ChangeSet@1.2236, 2005-04-03 20:03:53-07:00, torvalds@ppc970.osdl.org Merge bk://linux-sam.bkbits.net/kconfig into ppc970.osdl.org:/home/torvalds/v2.6/linux ChangeSet@1.2234, 2005-04-03 19:40:47-07:00, torvalds@ppc970.osdl.org Merge bk://bk.arm.linux.org.uk/linux-2.6-serial into ppc970.osdl.org:/home/torvalds/v2.6/linux ChangeSet@1.2181.17.4, 2005-04-03 20:16:16+02:00, zippel@linux-m68k.org kconfig: complete cpufreq Kconfig cleanup This completes the Kconfig cleanup for all other archs. CPU_FREQ_TABLE was moved to drivers/cpufreq/Kconfig and is selected as needed. Signed-off-by: Roman Zippel Signed-off-by: Sam Ravnborg ChangeSet@1.2224.7.3, 2005-04-03 10:18:11+01:00, pavel@cz.rmk.(none) [ARM] Fix u32 vs. pm_message_t in arm Patch from Pavel Machek This fixes u32 vs. pm_message_t confusion in arm. I was not able to even compile it, but it should not cause any problems. Please apply, ChangeSet@1.2224.1.127, 2005-04-02 16:09:12+02:00, bzolnier@trik.(none) [ide] fix via82cxxx resume failure With David Woodhouse . On resume from sleep, via_set_speed() doesn't reinstate the correct mode, because it thinks the drive is already configured correctly. Also kill redundant printk, ide_config_drive_speed() warns about errors. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.126, 2005-04-02 16:03:55+02:00, bzolnier@trik.(none) [ide] kill ide-default * add ide_drives list to list devices without a driver * add __ide_add_setting() and use it for adding no auto remove entries * kill ide-default pseudo-driver Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.125, 2005-04-02 16:00:50+02:00, bzolnier@trik.(none) [ide] get driver from rq->rq_disk->private_data * add ide_driver_t * to device drivers objects * set it to point at driver's ide_driver_t * store address of this entry in disk->private_data * fix ide_{cd,disk,floppy,tape,scsi}_g accordingly * use rq->rq_disk->private_data instead of drive->driver to obtain driver (this allows us to kill ide-default) ide_dma_intr() OOPS fixed by Tejun Heo . Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.124, 2005-04-02 15:49:01+02:00, bzolnier@trik.(none) [ide] kill ide_drive_t->disk * move ->disk from ide_drive_t to driver specific objects * make drivers allocate struct gendisk and setup rq->rq_disk (there is no need to do this for REQ_DRIVE_TASKFILE requests) * add ide_init_disk() helper and kill alloc_disks() in ide-probe.c * kill no longer needed ide_open() and ide_fops[] in ide.c ide_init_disk() fixed by Andrew Morton . Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.123, 2005-04-02 15:38:19+02:00, bzolnier@trik.(none) [ide] add ide_{un}register_region() Add ide_{un}register_region() and fix ide-{tape,scsi}.c to register block device number ranges. In ata_probe() only probe for modules. Behavior is unchanged because: * if driver is already loaded and attached to drive ata_probe() is not called et all * if driver is loaded by ata_probe() it will register new number range for a drive and this range will be found by kobj_lookup() If this is not clear please read http://lwn.net/Articles/25711/ and see drivers/base/map.c. This patch makes it possible to move drive->disk allocation from ide-probe.c to device drivers. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.122, 2005-04-02 15:33:42+02:00, bzolnier@trik.(none) [ide] destroy_proc_ide_device() cleanup When this function is called device is already unbinded from a driver so there are no driver /proc entries to remove. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.121, 2005-04-02 15:31:35+02:00, bzolnier@trik.(none) [ide] drive->dsc_overlap fix drive->dsc_overlap is supported only by ide-{cd,tape} drivers. Add missing clearing of ->dsc_overlap to ide_{cd,tape}_release() and move ->dsc_overlap setup from ide_register_subdriver() to ide_cdrom_setup() (ide-tape enables it unconditionally). Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.1.120, 2005-04-02 15:28:52+02:00, bzolnier@trik.(none) [ide] drive->nice1 fix It is drive's property independent of the driver being used so move drive->nice1 setup from ide_register_subdriver() to probe_hwif() in ide-probe.c. As a result changing a driver which controls the drive no longer affects this flag. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2224.6.2, 2005-04-01 20:02:28-08:00, gregkh@suse.de PCI: revert dumb SGI patch for resource freeing. Cset exclude: gregkh@suse.de|ChangeSet|20050317183046|30063 The patch doesn't do anything, and I'm not going to be trusting any more SGI pci hotplug patches for a long time... Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2226.2.1, 2005-04-01 14:35:01-08:00, davem@nuts.davemloft.net [SPARC64]: Store per-cpu pointer in IMMU TSB register. This trick requires that we PAGE_SIZE align the per-cpu areas, but that is fine. Also, reduce the __per_cpu_offsets[] array into just two values, a base and a shift. Signed-off-by: David S. Miller ChangeSet@1.2224.7.2, 2005-04-01 23:10:31+01:00, ben-linux@org.rmk.(none) [ARM PATCH] 2638/1: RX3715 - allow fclk as clock source Patch from Ben Dooks Since the RX3715 inits with fclk as the clock source, and to allow the system to generate the baud-rates for bluetooth control, this patch configures the platform data for "fclk" as a clock source Signed-off-by: Ben Dooks Signed-off-by: Russell King ChangeSet@1.2224.7.1, 2005-04-01 23:01:38+01:00, rpurdie@net.rmk.(none) [ARM PATCH] 2637/1: Combine code for Sharp SL series parameter area Patch from Richard Purdie The Sharp SL series bootloader puts a parameter structure into memory with important LCD parameters in it (amongst other things). The structure is common to collie, corgi, poodle, tosa and other models. This patch combines all the existing code into one place and simplifies access to the data. Signed-off-by: Richard Purdie Signed-off-by: Russell King ChangeSet@1.2224.1.119, 2005-04-01 13:30:25-08:00, torvalds@ppc970.osdl.org Merge proper selinux i_sock fix. ChangeSet@1.2224.1.118, 2005-04-01 12:44:10-08:00, trond.myklebust@fys.uio.no [PATCH] SELINUX: Fix i_sock reference. Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds ChangeSet@1.2224.6.1, 2005-04-01 12:36:30-08:00, gregkh@suse.de merge ChangeSet@1.2224.5.1, 2005-04-01 12:33:21-08:00, sds@tycho.nsa.gov [SELINUX]: Fix for removal of i_sock This patch against -bk eliminates the use of i_sock by SELinux as it appears to have been removed recently, breaking the build of SELinux in -bk. Simply replacing the i_sock test with an S_ISSOCK test would be unsafe in the SELinux code, as the latter will also return true for the inodes of socket files in the filesystem, not just the actual socket objects IIUC. Hence this patch reworks the SELinux code to avoid the need to apply such a test in the first place, part of which was obsoleted anyway by earlier changes to SELinux. Please apply. Signed-off-by: Stephen Smalley Signed-off-by: James Morris Signed-off-by: David S. Miller ChangeSet@1.2224.1.117, 2005-04-01 11:50:04-08:00, khali@linux-fr.org [PATCH] I2C: Fix a common race condition in hardware monitoring Grant Coady noticed that most hardware monitoring drivers are exposed to a race condition when one writes to the sysfs files they create. While the read calls properly request a lock on the internal data, write calls manipulate the internal data without proper locking. This big patch fixes that by adding locking wherever needed. Affected drivers: adm1021, adm1025, asb100, ds1621, fscher, fscpos, gl518sm, gl520sm, it87, lm63, lm75, lm77, lm78, lm80, lm83, lm87, lm90, lm92, max1619, pc87360, pcf8591, sis5595, smsc47m1, via686a, w83627hf and w83781d The adm1026, adm1031 and lm85 were already locking on write calls, but held the lock for code that did not require it, so they have been modified too. The smsc47b397 and w83l785ts drivers were not affected, because they are read-only. The patch should apply just fine on top of your stack, provided that you applied all previous patches in order (in particular, there is one lm87 indentation patch which is needed). Signed-off-by: Grant Coady Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.1.116, 2005-04-01 11:49:04-08:00, khali@linux-fr.org [PATCH] I2C: Move functionality handling from i2c-core to i2c.h So far, the functionality handling of i2c adapters was done in i2c-core by two exported functions: i2c_get_functionality and i2c_check_functionality. I found that both functions could be reduced to one line each, and propose that we turn them into inline function in the i2c.h header file, much like other i2c helper functions (e.g. i2c_get_clientdata, i2c_set_clientdata and i2c_clientname). The conversion of i2c_get_functionality suppresses a legacy check which shouldn't be needed anymore. Only one driver (s3c2410) was still relying on it, and was fixed some days ago. The conversion lets us get rid of two exports. Not only i2c-core gets smaller (by 458 bytes), but the client drivers using these functions get smaller too (typically by 48 bytes). And of course the new way is likely to be faster too, even if it wasn't my primary objective. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.1.115, 2005-04-01 11:48:28-08:00, khali@linux-fr.org [PATCH] I2C: pcf8574 doesn't need a lock While investigating the i2c chips drivers that were not properly locking, we found that the pcf8574 driver does the exact contrary. It uses a lock where it's not needed. While we were there, we did some additional cleanups to the driver: 1* Merge pcf8574_update_client() in show_read(), as it was the only user and the function became trivial once the locking was removed. 2* Add a validity check on values provided by user-space. Aurelien Jarno tested the modified code for confirmation and it worked just fine. Signed-off-by: Jean Delvare Signed-off-by: Aurelien Jarno Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.1.114, 2005-04-01 08:06:50-08:00, torvalds@ppc970.osdl.org Merge bk://kernel.bkbits.net/davem/net-2.6 into ppc970.osdl.org:/home/torvalds/v2.6/linux ChangeSet@1.2224.1.113, 2005-03-31 22:54:44-08:00, davem@sunset.davemloft.net [NBD]: Fix i_sock reference. Signed-off-by: David S. Miller ChangeSet@1.2224.1.112, 2005-03-31 22:35:09-08:00, tklauser@nuerscht.ch [NETFILTER]: ipt_hashlimit: Remove custom msecs_to_jiffies() macro Replace the MS2JIFFIES() macro with the msecs_to_jiffies() function provided in jiffies.h. The current macro is incorrect because HZ can have different values on different architectures. Signed-off-by: Tobias Klauser Signed-off-by: Harald Welte Signed-off-by: David S. Miller ChangeSet@1.2224.1.111, 2005-03-31 22:30:09-08:00, jix@bugmachine.ca [NETFILTER]: ipt_hashlimit: Fix bug introduced by hlist changes. Signed-off-by: Harald Welte Signed-off-by: David S. Miller ChangeSet@1.2224.1.110, 2005-03-31 22:28:03-08:00, bunk@stusta.de [ISDN]: Fix off-by-one errors in isdn_ppp.c This patch fixes several off by one errors found by the Coverity checker (ippp_table has ISDN_MAX_CHANNELS elements). Signed-off-by: Adrian Bunk Signed-off-by: David S. Miller ChangeSet@1.2224.1.109, 2005-03-31 22:27:11-08:00, herbert@gondor.apana.org.au [IPSEC]: Make IPCOMP more resilient. Since the IPCOMP header is left off when the payload is incompressible or too small, we can also do the same thing when we encounter an error during compression. In other words, we can let outbound IPCOMP always succeed. In the cases where it would currently fail we simply skip the IPCOMP transform. This makes IPCOMP slightly more resilient when memory is low and simplifies the code quite a bit. Signed-off-by: Herbert Xu Acked-by: James Morris Signed-off-by: David S. Miller ChangeSet@1.2224.1.108, 2005-03-31 22:25:26-08:00, matthew@wil.cx [NET]: Remove i_sock Remove i_sock from struct inode. Also remove some checks for SOCKET_I() returning NULL -- it can never return NULL for a valid inode. Signed-off-by: Matthew Wilcox Signed-off-by: David S. Miller ChangeSet@1.2224.1.107, 2005-03-31 22:24:38-08:00, herbert@gondor.apana.org.au [XFRM]: Simplify xfrm_policy_kill(). So here is a patch to simplify xfrm_policy_kill() by moving the GC linking after the write_unlock_bh(). Actually, as the code stands, xfrm_policy_kill() should/will never be called twice on the same policy. So we can add a warning to catch that. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2227, 2005-03-31 22:05:17-08:00, tcallawa@redhat.com [SPARC]: Implement pte_read() more cleanly. Signed-off-by: Tom 'spot' Callaway Signed-off-by: David S. Miller ChangeSet@1.2224.1.106, 2005-03-31 21:20:06-08:00, kaber@trash.net [IPSEC]: Check if SPI exists before creating acquire state. Signed-off-by: Patrick McHardy Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2224.1.105, 2005-03-31 21:19:27-08:00, kaber@trash.net [IPSEC]: Check SPI in xfrm_state_find() Signed-off-by: Patrick McHardy Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2224.1.104, 2005-03-31 21:18:16-08:00, herbert@gondor.apana.org.au [IPSEC]: Move xfrm_flush_bundles into xfrm_state GC Fixes ABBA deadlock noticed by Patrick McHardy. The locking in xfrm_state/xfrm_policy has always struck me as being an overkill. A lot of the locks should be replaced by rules that ensure the validity of most operations while a ref count is held. Now I have an excuse to do just that :) For 2.6.12 let's go for a simpler fix that breaks the dead lock. __xfrm_state_delete does not need to flush the bundles immediately. In fact, it is more efficient if we delay the flush to the GC worker since the flush is not dependent on any particular xfrm state. By delaying it we can do one single flush even when you're deleteing the entire xfrm state list. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2224.1.103, 2005-03-31 21:14:28-08:00, davem@sunset.davemloft.net Merge bk://kernel.bkbits.net/acme/net-2.6 into sunset.davemloft.net:/home/davem/src/BK/net-2.6 ChangeSet@1.2224.1.102, 2005-03-31 20:38:15-08:00, fubar@us.ibm.com [BONDING]: Do not drop non-VLAN traffic Change the bonding driver to not drop non-VLAN traffic when a VLAN is configured above it. Originally fixed by Olaf Kirch ; I changed his patch slightly to update comments. Signed-off-by: Jay Vosburgh Signed-off-by: David S. Miller ChangeSet@1.2224.1.101, 2005-03-31 20:36:08-08:00, util@deuroconsult.ro [PKT_SCHED]: Fix deadlock in sch_api.c While qdisc_create() is holding the rtnl_sem, it may try to load modules which in turn may try to register devices (teql is one such case), and it will then hang trying to retake the rtnl_sem. Signed-off-by: Catalin(ux aka Dino) BOIE Signed-off-by: David S. Miller ChangeSet@1.2224.1.100, 2005-03-31 20:32:53-08:00, shemminger@osdl.org [PKT_SCHED]: netem: Account for packets in delayed queue in qlen Netem has a private queue for delayed packets, and currently, packets in this queue are not accounted for in the qdisc qlen statistics. This is a problem if netem is used inside another qdisc doing rate control that peeks at the qlen. This patch changes the statistics to include the packets held but not ready to send. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller ChangeSet@1.2224.1.99, 2005-03-31 20:29:09-08:00, herbert@gondor.apana.org.au [NETLINK]: More complete fix for race. Unfortunately my patch only closed half the race. There is still a chunk of code between netlink_dump_start and netlink_dump that runs outside the cb lock which isn't protected by an sk reference. Here is a better patch which protects the entire netlink_dump function with a sk reference. The other call to netlink_dump by recvmsg is safe as the open file descriptor already holds a reference. As such the final sock_put in netlink_dump can be turned into a __sock_put since there is at least one reference held by the caller. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2224.4.5, 2005-03-31 16:24:49-08:00, akpm@osdl.org [IA64] CONFIG_NUMA requires CONFIG_ACPI_NUMA arch/ia64/kernel/smpboot.c:511:17: #error Fixme: Dunno how to build CPU-to-node map. Seems that CONFIG_NUMA requires CONFIG_ACPI_NUMA. Signed-off-by: Andrew Morton Signed-off-by: Tony Luck ChangeSet@1.2224.4.4, 2005-03-31 16:21:27-08:00, akpm@osdl.org [IA64] Andrew's fixes for warnings on ia64 build drivers/pci/msi.c: In function `msi_set_mask_bit': drivers/pci/msi.c:70: warning: cast from pointer to integer of different size drivers/pci/msi.c: In function `msi_capability_init': drivers/pci/msi.c:558: warning: cast to pointer from integer of different size Signed-off-by: Andrew Morton Signed-off-by: Tony Luck ChangeSet@1.2224.4.3, 2005-03-31 16:15:46-08:00, davidm@hpl.hp.com [IA64] Initialize ar.k7 to empty_zero_page early on. Without this initialization, early TLB misses to any user-regions will cause the TLB miss handlers to go down in flames. Normally, no such early TLB misses occur, but aggressive use of lfetch.fault can trigger it easily (e.g., when using lfetch.fault for the prefetch()/prefetchw() macros we get an early miss for address 0 due to a prefetch in find_pid()). Signed-off-by: David Mosberger-Tang Signed-off-by: Tony Luck ChangeSet@1.2224.4.2, 2005-03-31 16:10:45-08:00, bob.montgomery@hp.com [IA64] fix for unwind problem through dispatch_illegal_op_fault Here is the patch worked out with David Mosberger to fix the unwind info for dispatch_illegal_op_fault in ivt.S. Before adding this patch, an illegal op in the kernel would not show a complete stack trace in the oops message. Add PT_REGS_UNWIND_INFO to dispatch_illegal_op_fault to allow stack unwind to continue through that stack level when kernel code commits an illegal operation. Signed-off-by: Bob Montgomery Signed-off-by: Tony Luck ChangeSet@1.2226, 2005-03-31 14:56:22-08:00, davem@sunset.davemloft.net [SPARC64]: Put per-cpu area base into register g5. FINALLY, we can put the per-cpu base into register g5 on SMP. There are many simplifications and improvements now, but this is the base implementation. Thanks to Rusty and the IA64 folks for urging that I pursue this kind of scheme instead of locking stuff into the TLB at some fixed virtual address. Signed-off-by: David S. Miller ChangeSet@1.2224.3.30, 2005-03-31 14:33:14-08:00, mgreer@mvista.com [PATCH] i2c: i2c-mv64xxx - set adapter owner and class fields This patch adds the correct values for the 'owner' and 'class' fields of the adapter structure in the mv64xxx i2c bus driver. The missing class field caused some i2c chip drivers to refuse to attempt a probe on the mv64xxx i2c bus. Signed-off-by: Chris Elston Signed-off-by: Mark A. Greer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.29, 2005-03-31 14:32:55-08:00, grant_nospam@dodo.com.au [PATCH] I2C: Drop useless w83781d RT feature This patch removes useless RT feature from w83781d driver. Patch applies after the recent "I2C: Fix a common race condition in hardware monitoring" series. Signed-off-by: Grant Coady Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.28, 2005-03-31 14:32:36-08:00, khali@linux-fr.org [PATCH] I2C: Fix indentation of lm87 driver This trivial patch fixes indentation in the lm87 driver. I need this 'cause I'll soon post patches affecting these portions of code, and I'd like these patches to be easily readable. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.27, 2005-03-31 14:32:17-08:00, khali@linux-fr.org [PATCH] I2C: Fix broken force parameter handling I just noticed a nasty bug in the way the "force" parameter is handled for non-sensors i2c chip drivers. The "force" parameter supposedy is a list of adapter, address *pairs* where supported chips are unquestionably assumed to be. However, after handling one pair, the i2c core code searches for the next one *three* values later, not two. So with the current code, the second and third pairs wouldn't be properly handled. The fourth one would be, and so on. As a side note, this questions the need of an array parameter handling up to 24 of such pairs, when obviously nobody ever required more than one for the past 6 years. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.26, 2005-03-31 14:31:59-08:00, khali@linux-fr.org [PATCH] i2c: add adt7461 chip support to lm90 driver's Kconfig entry Hi Greg, James, all, > > > Attached is another version of my adt7461 patch, for inclusion in > > > the 2.6 tree. Reviewed by Jean. > > May we have an additional patch to Kconfig for this one? Here it finally comes. This simple patch adds a mention to the ADT7461 chip in Kconfig, now that the lm90 driver supports it. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.25, 2005-03-31 14:31:40-08:00, khali@linux-fr.org [PATCH] I2C: i2c-s3c2410 functionality and fixes This patch does the following updates to the i2c-s3c2410 bus driver: * Properly report the i2c functionality by adding to the `.functionality` field of the adapter * Change the dev_err() call on no-ack to an dev_dbg() to make it less noisy when the bus is being probed by i2cdetect, etc. * Add I2C_M_REV_DIR_ADDR to fully implement the I2C_FUNC_PROTOCOLO_MANGLING. * Ensure that the adapter owner field is set to THIS_MODULE Please apply, thanks. (Once this is applied, all i2c bus drivers will be properly reporting their functionality so I'll be able to go on with the i2c functionality core cleanups.) Signed-off-by: Ben Dooks Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.24, 2005-03-31 14:31:21-08:00, khali@linux-fr.org [PATCH] I2C: Fix race condition in it87 driver I noticed a race condition in the it87, affecting the IT8712F chip only. The VRM value is currently initialized *after* creating the vrm and vid sysfs files. This leaves a theorical room for reading from these files and get an invalid value. It's not critical, but let's still fix it. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.23, 2005-03-31 14:31:02-08:00, khali@linux-fr.org [PATCH] I2C: Delete useless instruction in it87 The IT8705F doesn't support VID, so it's quite pointless to give a value to it (and an arbitrary one at that). I think that this instruction was there for compatibility reasons some times ago, but the reasons went away while the instruction was left in place. We can safely delete it now. Thanks to Rudolf Marek for testing the patch (you never know). Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.22, 2005-03-31 14:30:43-08:00, rafael.espindola@gmail.com [PATCH] I2C: lsb in emc6d102 and adm1027 The attached patches add support for reading the lsb from the emc6d102 and change how they are read from the adm1027. The lm85_update_device function decodes the LSBs to temp_ext and in_ext. This strategy was suggested by Philip Pokorny. The patch also changes some macros to use the SCALE macro. I think that they become more readable this way. Signed-off-by: Rafael Ávila de Espíndola Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.21, 2005-03-31 14:30:24-08:00, ebrower@gmail.com [PATCH] I2C: lost arbitration detection for PCF8584 [PATCH] lost arbitration detection for PCF8584 algo driver Patch against a slightly-dated linux-2.6 BK tree This patch provides lost arbitration detection for the PCF8584 I2C algorithm driver. The PCF8584 LAB bit is set whenever lost arbitration is detected, so we check the bit in the wait_for_pin function and if LAB is detected we return -EINTR. The -EINTR value bubbles-up all the way to the master_xfer API call so callers may detect this condition explicitly. LAB could be checked more often, at the expense of code readability/maintainability. Signed-off-by: Eric Brower Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.20, 2005-03-31 14:30:05-08:00, R.Marek@sh.cvut.cz [PATCH] I2C: busses documentation update 2 of 2 Patch contains promised documentation update for i2c bus drivers. I would like to thank Jean Delvare and Aurelien Jarno for their comments. Signed-off-by: Rudolf Marek Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.19, 2005-03-31 14:29:46-08:00, R.Marek@sh.cvut.cz [PATCH] I2C: busses documentation update 1 of 2 This patch just moves i2c-parport file to busses directory. Patch for other busses documentation will follow. Signed-off-by: Rudolf Marek Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.18, 2005-03-31 14:29:23-08:00, khali@linux-fr.org [PATCH] I2C: Fix Vaio EEPROM detection This fixes a bug in the eeprom driver, which made all EEPROMs at location 0x57 be erroneously treated as Vaio EEPROMs. I have to say I'm quite ashamed that I introduced the bug in the first place, as this was a really stupid one. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.17, 2005-03-31 14:29:05-08:00, khali@linux-fr.org [PATCH] I2C: Recognize new revision of the ADT7463 chip This simple patch to the lm85 driver adds recognition of a new revision of the ADT7463 chip. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.16, 2005-03-31 14:28:46-08:00, khali@linux-fr.org [PATCH] I2C: Avoid repeated resets of i2c-viapro It was reported that the i2c-viapro SMBus driver sometimes has trouble on recent systems (VT8237 south bridge). The "Host Status" register has at least one additional bit used when compared with older south bridges of this family. The driver currently considers this additional bit as an error condition when it's set, causing repeated bus resets and sometimes read failures. This patch makes the driver ignore the bits of the "Host Status" register for which no definition is known. I wish I had a datasheet for the VIA VT8237, so that I could check what the additional bit is supposed to mean, but I don't. If someone has a datasheet or good contacts at VIA, please let me know. The patch was reported to fix the problem on a system with the VT8237, and was also tested not to break the driver on older VIA south bridges, so it seems to be safe. Thanks to Aurelien Jarno for the tests. Additionally, the patch makes the post-transaction bus reset slightly more efficient by sparing a few unneeded I/O operations. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.15, 2005-03-31 14:09:38-08:00, khali@linux-fr.org [PATCH] I2C: Kill outdated defines in i2c.h Some defines in i2c.h (I2C_CLIENT_MODPARM and friends) are now useless. They should have been removed when the i2c client parameters were converted from MODULE_PARAM to module_parm_array, but where not. This patch removes them now. Additionally, it moves the definition of I2C_CLIENT_MAX_OPTS next to where it is used rather than 220 lines before, which is preferable IMHO. As a side note, I think that there is a bug in the way these options are handled. The i2c code looks for I2C_CLIENT_END as a list terminator, but if the maximum number of parameters are actually provided, no terminator will be left. It's rather unlikely to happen because nobody will probably ever provide that many parameters, but this should probably be fixed. I'll address this issue later, since I plan to completely rewrite the way these parameters are handled anyway. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.14, 2005-03-31 14:09:20-08:00, khali@linux-fr.org [PATCH] I2C: Fix some i2c algorithm initialization While searching for i2c_algorithm declarations missing their .functionality member, I found three of them which were not properly initialized. i2c-algo-ite and i2c_sibyte_algo do not use the C99 initialization syntax, and i2c-ibm_iic.c explicitely initializes NULL members. Following patch puts some order in there. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.13, 2005-03-31 14:09:00-08:00, mgreer@mvista.com [PATCH] I2C: Fix breakage in m41t00 i2c rtc driver Remove setting of deleted i2c_client structure member. The latest include/linux/i2c.h:i2c_client structure no longer has an 'id' member. This patch removes the setting of that no longer existing member. Signed-off-by: Mark A. Greer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.12, 2005-03-31 14:08:41-08:00, frank.beesley@aeroflex.com [PATCH] I2C: Clean up of i2c-elektor.c build This patch changes the flags variable type from long to unsigned long in one function. This removes a couple of warnings from the compile messages for elektor i2c bus driver. Signed-off-by: Frank Beesley Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.11, 2005-03-31 14:08:21-08:00, jchapman@katalix.com [PATCH] i2c: add adt7461 chip support to lm90 driver i2c: add adt7461 chip support The Analog Devices ADT7461 temperature sensor chip is compatible with the lm90 device provided its extended temperature range is not enabled. The chip will be ignored if the boot firmware enables extended temperature range. Also, since the adt7461 treats temp values <0 as 0 and >127 as 127, the driver prevents temperature values outside the supported range from being set. Signed-off-by: James Chapman Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.10, 2005-03-31 14:08:02-08:00, jchapman@katalix.com [PATCH] i2c: new driver for ds1337 RTC Signed-off-by: James Chapman Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.9, 2005-03-31 14:07:43-08:00, grant_nospam@dodo.com.au [PATCH] I2C: group Intel on I2C Hardware Bus support From an end-user perspective it is easy to miss the third Intel PIIX entry on the menuconfig "I2C Hardware Bus support" screen. Also the Intel 801 menu item does not mention ICH. This trivial patch groups three Intel entries together, adds ICH to menu item, and ICH5/ICH5R to the help section. Includes suggestions from Jean Delvare. Signed-off-by: Grant Coady Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.8, 2005-03-31 14:07:24-08:00, khali@linux-fr.org [PATCH] I2C: Skip broken detection step in it87 One of the detection steps in the it87 chip driver was reported to be broken for some revisions of the IT8712F chip [1] [2]. This detection step is a legacy from the lm78 driver and the documentation available for the IT8705F and IT8712F chips does not mention it at all. For this reason, I propose to skip this detection step for Super-I/O chips. Super-I/O chips have already been identified when we reach this step, so it is redundant (additionally do being broken). This closes bug #4335. [1] http://bugzilla.kernel.org/show_bug.cgi?id=4335 [2] http://archives.andrew.net.au/lm-sensors/msg29962.html Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.7, 2005-03-31 14:07:05-08:00, khali@linux-fr.org [PATCH] I2C: Make master_xfer debug messages more useful While working on the recent saa7110 mess, I found that the debug message displayed when calling master_xfer wasn't as useful as it could be. Here is a patch improving this. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.6, 2005-03-31 14:06:47-08:00, khali@linux-fr.org [PATCH] I2C: Kill unused struct members in w83627hf driver I just noticed that the pwmenable struct members in the w83627hf driver are not used anywhere (and quite rightly so, as PWM cannot be disabled in these chips as far as I know). Let's just get rid of them and save some bytes of memory. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.5, 2005-03-31 14:06:28-08:00, khali@linux-fr.org [PATCH] I2C: Fix adm1021 alarms mask This patch fixes an incorrect bitmasking on the status register in the adm1021 driver, which was causing high alarm on remote temperature to be hidden. This bug was found and reported by Jayakrishnan: http://bugzilla.kernel.org/show_bug.cgi?id=4285 Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.4, 2005-03-31 14:06:09-08:00, khali@linux-fr.org [PATCH] I2C: Cleanup adm1021 unused defines While working on the adm1021 driver, I found that this driver has a number of unused (and useless) defines we could get rid of. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.3, 2005-03-31 14:05:50-08:00, khali@linux-fr.org [PATCH] I2C: New lm92 chip driver This is a new i2c chip driver named lm92. It supports the National Semiconductor LM92 and Maxim MAX6635 chips. This time I did not port the driver from the lm_sensors project but instead rewrote it. The reason is that the original driver has a different structure from the other i2c chip drivers, which would have made maintenance harder. I don't have a compatible chip myself but could test my code thanks to Mark Hoffman's i2c-stub fake bus driver. Later on, James Chapman tested it on a real chip, successfully. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.2, 2005-03-31 14:05:31-08:00, domen@coderock.org [PATCH] i2c/i2c-elektor: remove interruptible_sleep_on_timeout() usage Replace deprecated interruptible_sleep_on_timeout() with direct wait-queue usage. Patch is compile-tested. Signed-off-by: Nishanth Aravamudan Signed-off-by: Domen Puncer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.3.1, 2005-03-31 14:05:14-08:00, domen@coderock.org [PATCH] i2c/i2c-ite: remove interruptible_sleep_on_timeout() usage Replace deprecated interruptible_sleep_on_timeout() with direct wait-queue usage. Patch is compile-tested, sort of; the driver does not build in vanilla kernel either, but I don't seem to add any warnings.. Signed-off-by: Nishanth Aravamudan Signed-off-by: Domen Puncer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2224.2.1, 2005-03-31 11:27:34-08:00, dwmw2@infradead.org [PATCH] Fix incorrect bluetooth socket zapping On Tue, 2005-03-15 at 22:50 +0000, Linux Kernel Mailing List wrote: > - parent->sk_zapped = 1; > + sock_set_flag(sk, SOCK_ZAPPED); Er, no. That zapped the wrong socket. Signed-off-by: David Woodhouse Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.97, 2005-03-31 09:01:50-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: fix indentation in nfsd4_open OK, so it's trivial, but these misaligned comments have been bugging me forever.... Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.96, 2005-03-31 09:01:33-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: fix open returns for other claim types Switch on OPEN claim type instead of if-else. add remaining OPEN claim types. Return nfserr_notsupp instead of nfserr_inval for valid claim types. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.95, 2005-03-31 09:01:15-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: remove trailing whitespace from nfs4proc.c Trivial whitespace Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.94, 2005-03-31 09:00:57-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: rename release_delegation Release_delegation is responsible for removing the delegation's connection with the v4 state and with the lease subsystem. The word "unhash" seems closer than "release" as a description of this.... Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.93, 2005-03-31 09:00:38-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: acl error fix When a client sends an ACL that is a legal NFSv4 ACL, but that we don't support (because we can't represent it using a POSIX ACL), the correct error to return is NFS4ERR_ATTRNOTSUPP. (rfc3530, section 5.11.1: "If the server recieves a request to set an ACE that it cannot store, it MUST reject the request with NFS4ERR_ATTRNOTSUPP. If the server receives a request to set an ACE that it can store but cannot enforce, the server SHOULD reject the request with NFS4ERR_ATTRNOTSUPP"). Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.92, 2005-03-31 09:00:22-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: acl don't set named attrs We were expecting {read,write}_named_attributes to be set in sync with the read and write bits. But we don't actually support named attributes, and when we eventually do, it's unlikely that we'll want to use these ACL mode bits to control access to them. So let's just stop supporting these two bits. This also updates our mapping to the latest ietf draft. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.91, 2005-03-31 09:00:03-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: don't set WRITE_OWNER in either allow or deny bits WRITE_OWNER should be neither allowed nor denied. This improves interoperability with Solaris clients and represents our permissions more accurately. It breaks compatibility with clients using the old version of the NFSv4<->POSIX mapping. I believe the only code that did that was older versions of experimental client-side code available only from our website, so I prefer not to introduce a hack just to continue support the old mapping. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.90, 2005-03-31 08:59:47-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4_remove_unused_acl_function We're translating v4 ACL's to POSIX ACL's and relying on the filesystem to check those, so there's no need for a NFSv4 ACL-checking algorithm. Some day we'd like to find a way to implement the full v4 ACL protocol on the server side, but for now this is dead code. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.89, 2005-03-31 08:59:29-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: fix failure to truncate on some opens An OPEN upgrade on a file already open for WRITE will not upgrade the OPEN, but can still truncate the file. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.88, 2005-03-31 08:59:12-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: factor out common open_truncate code Factor out a bit of common code. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.87, 2005-03-31 08:58:55-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: allow read on open for write The rfc recommends allowing read using stateid's from opens that only requested write access, as clients may in some cases be unable to write without doing reads. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.86, 2005-03-31 08:58:38-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: fix use after put() in cb_recall nfs4_put_delegation() did a dprintk using a pointer that it had just called a put() on. This could cause a dereference of a pointer to freed memory in some situations. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.85, 2005-03-31 08:58:20-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: remove unneeded stateowner arguments Remove the unneeded stateowner argument from nfs4_check_open() and init_stateid. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.84, 2005-03-31 08:58:03-08:00, neilb@cse.unsw.edu.au [PATCH] nfsd4: fix share conflict tests In nfs4_check_open(): Move 'is_open_owner' check to be first. Remove continue so as to call test_share on OPENs with a previously seen open_owner as per rfc3530. Signed-off-by: Andy Adamson Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.83, 2005-03-31 08:57:45-08:00, neilb@cse.unsw.edu.au [PATCH] svcrpc: auth_domain documentation The use of auth_domains is somewhat confusing, in part because they were originally intended to be used in a more general way than they currently are. Update the documentation a little with an eye towards how it's currently used. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.82, 2005-03-31 08:50:03-08:00, corbet@lwn.net [PATCH] doc: where to find LDD3 Here's an only mildly self-serving patch so that interested people can find the free version of LDD3. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.81, 2005-03-31 08:49:44-08:00, benh@kernel.crashing.org [PATCH] Fix atyfb build on ppc This patch fixes a build problem with atyfb on ppc. It uses the stuff in macmodes.c, but doesn't trigger the build of it. So if no other driver using macmodes is built, the link will fail. Signed-off-by: David Woodhouse Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.80, 2005-03-31 08:49:28-08:00, benh@kernel.crashing.org [PATCH] radeonfb: Preserve TMDS setting radeonfb tends to override TMDS_TRANSMITTER_CNTL completely, which appears to be incorrect with some recent flat panels. This patch changes this so that we only tweak the bits we want in there and preserve the BIOS value for the rest. Along with the i2c fix, that patch fixes operations on the latest Apple cinema displays. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.79, 2005-03-31 08:49:09-08:00, benh@kernel.crashing.org [PATCH] radeonfb: Fix mode setting on CRT monitors Current radeonfb is a bit "anal" about accepting CRT modes, it basically only accepts modes that have the exact resolution, which tends to break with fbcon on console switches as it provides "approximate" modes. This patch fixes it by having the driver chose the closest possible mode instead of looking for an exact match. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.78, 2005-03-31 08:48:52-08:00, vandrove@vc.cvut.cz [PATCH] Fix matroxfb on big-endian hardware There was too much/too few byteswapping done by driver and hardware in matroxfb on big endian hardware. Change fixes mirrored/split/corrupted letters seen on screen when using accelerated matroxfb mode. Patch was tested on Mips (by Peter) and x86-64 (by Petr). Signed-off-by: Peter 'p2' De Schrijver Signed-off-by: Petr Vandrovec Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.77, 2005-03-31 08:48:34-08:00, juhl-lkml@dif.dk [PATCH] remove redundant NULL checks before kfree() in drivers/video/ Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.76, 2005-03-31 08:48:17-08:00, adaplas@hotpop.com [PATCH] neofb: Set hwaccel flags properly Not all neomagic chipsets are capable of hardware acceleration. Set the hwaccel flags only for chipsets that can do so. From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.75, 2005-03-31 08:48:00-08:00, adaplas@hotpop.com [PATCH] neofb: mmio fixes Changes in neo_map_mmio to make neofb working on NM2070, NM2090, NM2093. From: Marek Marczykowski Signed-off-by: Marek Marczykowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.74, 2005-03-31 08:47:43-08:00, adaplas@hotpop.com [PATCH] fbcon: Do not set palette if console is not visible A long standing bug in fbcon where switching/opening a vt will also set the palette even if the vt is in the background. This results in color corruption in the foreground vt. From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.73, 2005-03-31 08:47:26-08:00, adaplas@hotpop.com [PATCH] fbcon: Call set_par per fb_info once during init Currently, fbcon will unconditionally do set_par's on all info's mapped to each console. This results in repetetive hardware initialization when one is enough. Fix this by skipping all fbdev's that already underwent initialization. From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.72, 2005-03-31 08:47:09-08:00, adaplas@hotpop.com [PATCH] fbcon: Save var rotate field in struct display Add the rotate field among parameters saved in struct display (per console structure). Some out-of-tree drivers are possibly using this field. From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.71, 2005-03-31 08:46:52-08:00, varenet@parisc-linux.org [PATCH] s1d13xxxfb: Add support for Epson S1D13806 FB Add support for Epson S1D13806 framebuffer device. The driver is intended to be easily used with other S1D13xxx devices, hopefully by splitting the header file and changing a few defines. Since I haven't got the hardware to test that, though, I can only assert that it works with S1D13806. This driver has been succesfully tested on ARM embedded boards and reported working on SH architecture as well. Since this is my first framebuffer driver, I would welcome any suggestion/comment about it :) This driver has been built on top of some preliminary ARM specific work by Ben Dooks, and adapted from existing code (as stated in the header of s1d13xxxfb.c). Signed-off-by: Thibaut VARENE Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.70, 2005-03-31 08:46:34-08:00, adaplas@hotpop.com [PATCH] tridentfb: Clean up printk()'s This patch adds KERN_ constants to all of the printk()'s that need them in drivers/video/tridentfb.c Signed-off-by: James Nelson Signed-off-by: Domen Puncer Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.69, 2005-03-31 08:46:17-08:00, adaplas@hotpop.com [PATCH] sisfb: Trivial cleanups sis_main.c vfree() checking cleanups. convert from pci_module_init to pci_register_driver (from:http://kerneljanitors.org/TODO). Signed-off by: James Lamanna Signed-off-by: Domen Puncer Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.68, 2005-03-31 08:45:59-08:00, adaplas@hotpop.com [PATCH] fbdev: Convert drivers to pci_register_driver convert from pci_module_init to pci_register_driver (from:http://kerneljanitors.org/TODO). Signed-off-by: Christophe Lucas Signed-off-by: Domen Puncer Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.67, 2005-03-31 08:45:41-08:00, adaplas@hotpop.com [PATCH] fbdev: Kconfig fix for macmodes and PPC Fix link error for PPC-based drivers that also use functions in macmodes.c. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.66, 2005-03-31 08:45:24-08:00, adaplas@hotpop.com [PATCH] atyfb: Add boot/module option to override composite sync Some SPARC-based displays need the composite sync set to high. Since none of the entries in the default db has the composite sync set to high, using fb_find_mode() will produce a blank display. Signed-off-by: Antonino Daplas Acked-by: Alexander Kern Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.65, 2005-03-31 08:45:06-08:00, alex.kern@gmx.de [PATCH] Excessive atyfb debug messages Debug is no longer needed by default. Driver is pretty stable. Signed-off-by: Alexander Kern Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.64, 2005-03-31 08:44:48-08:00, adaplas@hotpop.com [PATCH] fbdev: Cleanups in drivers/video part 2 This patch contains cleanups under drivers/video/ including: - make some needlessly global code static - the following was needlessly EXPORT_SYMBOL'ed: - fbcon.c: fb_con - fbmon.c: get_EDID_from_firmware (completely unused) Signed-off-by: Adrian Bunk Signed-off-by: Antonino Daplas Signed-off-by: Benoit Boissinot Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.63, 2005-03-31 08:44:30-08:00, adaplas@hotpop.com [PATCH] pm2fb: X and VT switching crash fix From: Jim Hague This is a bug in pm2fb.c exposed by recent changes to fbcon. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.62, 2005-03-31 08:44:13-08:00, adaplas@hotpop.com [PATCH] nvidiafb: Delete i2c bus on driver unload The driver nvidiafb fails to delete the i2c bus on load failure or unload. Fix From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.61, 2005-03-31 08:43:56-08:00, adaplas@hotpop.com [PATCH] nvidiafb: Process boot options earlier A few options, such as noscale, are processed after hardware has been configured which render them useless. This manifest as failure of DVI displays to scale. Process them earlier. From: Antonino Daplas Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.60, 2005-03-31 08:43:39-08:00, adaplas@hotpop.com [PATCH] nvidiafb: Kconfig help text update for config FB_NVIDIA Tiny trivial patch to fix up the help text for config FB_NVIDIA. Signed-off-by: Jesper Juhl Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.59, 2005-03-31 08:43:22-08:00, rddunlap@osdl.org [PATCH] nvidiafb: fix section references Fix section usage in nvidiafb driver. Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.58, 2005-03-31 08:43:05-08:00, adaplas@hotpop.com [PATCH] nvidiafb: Maximize blit buffer capacity Currently, nvidiafb blit expands a rectangle a scanline at a time. This is very inefficient because the graphics pipeline is not maximized. It may also crash if the scanline is > 4096 (should not happen with fbcon though). Instead of blitting a scanline at a time, attempt to fill up buffer to maximum capacity. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.57, 2005-03-31 08:42:48-08:00, adaplas@hotpop.com [PATCH] fbcon: Stop framebuffer operations before hardware is properly initialized Accessing the hardware before it is properly initialized can lead to crashes or screen corruption. This happens when switching to X then back to console. When console comes back from X, the device is in an undefined state. During this window, accessing the hardware is disallowed. A new field in fbcon_par is added (graphics), which will be set to nonzero just before initialization of the framebuffer and when coming back from KD_GRAPHICS, then unset when an fb_set_var/fb_set_par is done. While this field is set, no accesses to the hardware is done. The consequence of this change is, hopefully, more robust switching between KD_GRAPHICS<-> KD_TEXT. An added benefit coming from this change is that the MODESWITCHLATE hack is not needed anymore and thus removed. This hack is used by savagefb, rivafb and nvidiafb. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.56, 2005-03-31 08:42:31-08:00, adaplas@hotpop.com [PATCH] fbdev: mvidia licensing clarification Mark said that the nVidia licensing is MIT-style, so it should be okay. I'm including a patch that adds a small licensing comment on the pertinent files and also adds the license notice in nv_accel.c, just to be on the safe side. Signed-off-by: Antonino Daplas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.55, 2005-03-31 08:42:14-08:00, benh@kernel.crashing.org [PATCH] radeonfb: DDC i2c fix The radeonfb code for DDC probing (like it's X.org counterpart) uses to leave the DDC clock & data lines asserted after the probing is complete. This causes problems with some Apple monitors like the new Cinema HD 23", who will turn themselves off when that happens. This fixes it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.54, 2005-03-31 08:41:57-08:00, benh@kernel.crashing.org [PATCH] radeonfb: Implement proper workarounds for PLL accesses After discussion with ATIs, it seems that the workarounds they initially gave me were not completely correct. This patch implements the proper ones, which includes sleeping in PLL accesses, and thus requires the previous patch to make sure we do not call unblank at interrupt time (unless oops_in_progress is set, in which case I use an mdelay). It also removes obsolete code that used to disable some power management features in the accel init code. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.53, 2005-03-31 08:41:40-08:00, benh@kernel.crashing.org [PATCH] ppc32: move powermac backlight stuff to a workqueue The powermac has a kernel-based driver for controlling the backlight from the keyboard that used to call into some fbdev's from interrupt contexts. This patch moves it to a workqueue (and additionally makes sure the console semaphore is taken and held). I hope I'll replace this by the new backlight framework in a future kernel version, but for now, this will fix the immediate issues with radeon. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.52, 2005-03-31 08:41:23-08:00, benh@kernel.crashing.org [PATCH] vt: don't call unblank at irq time This patch removes the call to unblank() from printk, and avoids calling unblank at irq() time _unless_ oops_in_progress is 1. I also export oops_in_progress() so drivers who care like radeonfb can test it and know what to do. I audited call sites of unblank_screen(), console_unblank(), etc... and I _hope_ I got them all, the patch includes a small patch to the s390 bust_spinlocks code that sets oops_in_progress back to 0 _after_ unblanking for example. I added a few might_sleep() to help us catch possible remaining callers. I'll soon write a document explaining fbdev locking. The current situation after this patch is that: - All callbacks have console_semaphore held (fbdev's are fully serialised). - Everything is called in schedule'able context, except the cfb_* rendering operations and cursor operations, with the special case of unblank who can be called at any time when "oops_in_progress" is true. A driver that needs to sleep in it's unblank implementation is welcome to test that variable and use a fallback path (or just do nothing if it's not simple). Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.51, 2005-03-31 08:41:06-08:00, rbultje@ronald.bitfreak.net [PATCH] zr36050 typo fix Fix a typo. It doesn't cause anything bad (at least not on my computer according to some tests), but who knows, maybe it fixes some obscure unfixeable bug for some people. Signed-off-by: Ronald S. Bultje Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.50, 2005-03-31 08:40:49-08:00, nickpiggin@yahoo.com.au [PATCH] sched: fix schedstats warning Quiet a warning when compiling without CONFIG_SMP Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.49, 2005-03-31 08:40:32-08:00, rml@novell.com [PATCH] iput() can sleep iput() can sleep. Document it. Signed-off-by: Robert Love Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.48, 2005-03-31 08:40:16-08:00, prasanna@in.ibm.com [PATCH] Kprobes: Allow/deny probes on int3/breakpoint instruction? Kprobes did an improper exit when a probe is inserted on an int3 instruction. In case of normal execution of int3/breakpoint instruction, it oops!. Probe on an int3 instruction was not handled properly by the kprobes, it generated faults after oops! doing an improper exit with holding the lock. This fix employes a bit different method to handle probe on an int3/breakpoint instruction. On execution of an int3/breakpoint instruction (placed by kprobe), kprobes_handler() is called which sets it for single stepping in-line(it does not matter whether we single step out-of-line/inline since the single stepping instruction is same). Now it single steps on int3/breakpoint instruction here, entering kprobes_handler() once again. Kprobes now check's the status that it is single stepping and avoids the recursion. It runs down through the trap handler and oops messages is seen on the console since it executed int3/breakpoint instruction. Here the kprobes single stepping handler never gets called. Is this behaviour acceptable ? Or should we avoid putting probes on an int3 /breakpoint instruction ? How should it handle such situations? Below is the patch to allow probes on an int3/breakpoint instruction. This patch fixes the above problem by doing a proper exit while avoiding recursion. Signed-off-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.47, 2005-03-31 08:39:59-08:00, ananth@in.ibm.com [PATCH] kprobe_handler should check pre_handler function The kprobe_handler() code currently does not check if a pre_handler is registered for the kprobe under process. This leads to a NULL pointer dereference in cases when a module does not define a pre_handler. The patch corrects the issue by explicitly checking that the pre_handler is not NULL before it is invoked. Signed-off-by: Ananth N Mavinakayanahalli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.46, 2005-03-31 08:39:42-08:00, rmk+lkml@arm.linux.org.uk [PATCH] paport oops fix It appears that the parport driver claims on-board superio devices without actually doing anything. When the driver is removed, we try to dereference non-existent driver data to unregister the ports. Since we didn't register anything, it's safe to ignore these devices in the remove function. Signed-off-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.45, 2005-03-31 08:39:25-08:00, ysato@users.sourceforge.jp [PATCH] nommu.c build error fix - Argument addition of security_file_mmap. - Because in_gate_area_no_task undefined with CONFIG_KALLSYM=y, add it. Signed-off-by: Yoshinori Sato Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.44, 2005-03-31 08:39:08-08:00, bunk@stusta.de [PATCH] kernel/rcupdate.c: make the exports EXPORT_SYMBOL_GPL Due to the patent situation at least in the USA, the exports of kernel/rcupdate.c should be EXPORT_SYMBOL_GPL. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.43, 2005-03-31 08:38:51-08:00, eike-kernel@sf-tec.de [PATCH] Kill stupid warning when compiling riocmd.c My gcc complains about RIOCommandRup(), this is because this one has no forward declaration and uses old style parameter declaration. This patch changes all function headers in riocmd.c to use their parameter types in function header directly. Signed-off-by: Rolf Eike Beer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.42, 2005-03-31 08:38:34-08:00, rlrevell@joe-job.com [PATCH] make Documentation/oops-tracing.txt relevant to 2.6 Here is a patch to finally bring oops-tracing.txt into the 2.6 era. Signed-Off-By: Lee Revell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.41, 2005-03-31 08:38:17-08:00, patrick@bitwizard.nl [PATCH] SX cli() conversion This patch converts all save_flags/restore_flags to the new spin_lock_irqsave/spin_unlock_irqrestore calls, as well as some other 2.6.X cleanups. This allows the "sx" driver to become SMP safe. Signed-off-by: Patrick vd Lageweg Signed-off-by: Rogier Wolff Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.40, 2005-03-31 08:38:00-08:00, sfr@canb.auug.org.au [PATCH] consolidate asm/ipc.h All the asm*/ipc.h files are basically the same (for things that are used) so I have consolidated them all into asm-generic/ipc.h Signed-off-by: Stephen Rothwell Acked-By: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.39, 2005-03-31 08:37:22-08:00, uwe.bugla@gmx.de [PATCH] cx24110 Conexant Frontend update With Peter Hettkamp I transformed Peter's attempt into a kernel-compatible patch. The dvb frontend cx24110 is given back a send burst function which is needed by the dvbstream-engine of MPlayer 1.0pre6a, for example. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.38, 2005-03-31 08:37:04-08:00, christian@leber.de [PATCH] make sysrq-F call oom_kill() Ky box (2.6.9-final) was yesterday completly stalled (mouse movable and stupid loadmeter was still working) after starting mutt and was swapping for half an hour until I sent SIGTERM to all processes. I suspect it was a 2 GB big galeon process that was the problem. I think sysrq needs a key to call oom_kill manually. From: Coywolf Qi Hunt Move it into a workqueue to avoid taking VM locks from IRQ context. Signed-off-by: Coywolf Qi Hunt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.37, 2005-03-31 08:36:46-08:00, rob@landley.net [PATCH] uml: Fix the console stuttering I spent far too much of the weekend tracking this sucker down through the guts of the tty code. The problem turns out to be that drivers/char/n_tty.c has a write_chan that does buffering and retransmitting data, and arch/um/drivers/chan_kern.c ALSO has a write_chan that buffers and retransmits data, and the first calls the second but the second doesn't always return correct status information for the -EAGAIN case. When they get confused, both of them try to buffer and retransmit data, hence the stuttering. The first fix is that if chan_kern's write_chan gets an -EAGAIN, it should NOT gratuitously change that to a 0 before returning. I don't know why that code is in there, but deleting those two lines makes 90% the stuttering go away. But not quite all of it. The second half of the fix is arch/um/drivers/line.c has a buffer_data() function that adds data to the buffer, tries to flush the buffer out to disk, gets -EAGAIN, and then returns -EAGAIN even though it successfully buffered all the data it was sent. So the upper layer resubmits the last chunk of data it sent when the console unblocks, even though the lower layer buffered it and sent it on by that point. With this patch, I can't get the UML console to stutter anymore by suspending the process it's writing to. (Add tee to the mix and you can still make it hang by suspending its xterm for a second or two, but I think that tee is hanging, not UML. Hangs with RHEL4 tee, but not busybox tee...) Signed-off-by: Rob Landley Acked-by: Jeff Dike Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.36, 2005-03-31 08:36:29-08:00, blaisorblade@yahoo.it [PATCH] uml: correct error message With "Anthony Brock" Replace the message with a more meaningful one. Noted by Anthony Brock. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.35, 2005-03-31 08:36:12-08:00, blaisorblade@yahoo.it [PATCH] uml: fix hostfs special perm handling When opening devices nodes on hostfs, it does not make sense to call access(), since we are not going to open the file on the host. If the device node is owned by root, the root user in UML should succeed in opening it, even if UML won't be able to open the file. As reported by Rob Landley, UML currently does not follow this, so here's a fix. Signed-off-by: Paolo 'Blaisorblade' Giarrusso CC: Rob Landley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.34, 2005-03-31 08:35:54-08:00, blaisorblade@yahoo.it [PATCH] uml: fixes a build failure with CONFIG_MODE_SKAS disabled This fixes a build failure (and also some warnings) when CONFIG_MODE_SKAS is disabled. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.33, 2005-03-31 08:35:37-08:00, blaisorblade@yahoo.it [PATCH] uml: quick fix syscall table *) Uml 2.6.11 does not compile with gcc 2.95.4 because some entries are duplicated, and that GCC does not accept this (unlike gcc 3). Plus various other bugs in the syscall table definitions: *) 223 is a syscall hole (i.e. ni_syscall) only on i386, on x86_64 it's a valid syscall (thus a duplicated one). *) __NR_vserver must be only once with sys_ni_syscall, and not multiple times with different values! *) syscalls duplicated in SUBARCHs and in common files (thus assigning twice to the same array entry and causing the GCC 2.95.4 failure mentioned above): sys_utimes, which is common, and sys_fadvise64_64, sys_statfs64, sys_fstatfs64, which exist only on i386. *) syscalls duplicated in each SUBARCH, to put in common files: sys_remap_file_pages, sys_utimes, sys_fadvise64 *) 285 is a syscall hole (i.e. ni_syscall) only on i386, on x86_64 the range does not arrive to that point. *) on x86_64, the macro name is __NR_kexec_load and not __NR_sys_kexec_load. Use the correct name in either case. Note: as you can see, part of the syscall table definition in UML is arch-independent (with everywhere defined syscalls), and part is arch-dependant. This has created confusion (some syscalls are listed in both places, some in the wrong one, some are wrong on one arch or another). Also, as add-ons: *) uses __va_copy instead of va_copy since some old versions of gcc (2.95.4 for instance) don't accept va_copy. *) some whitespace cleanups in the syscall table (if you don't like them, feel free to remove them). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.32, 2005-03-31 08:35:19-08:00, blaisorblade@yahoo.it [PATCH] uml: gprof depends on !TT CONFIG_GPROF depends on the fact that TT mode is disabled. I just verified this, and this dependency already exists in UML/2.4. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.31, 2005-03-31 08:35:02-08:00, blaisorblade@yahoo.it [PATCH] uml: fix sigio spinlock I just saw a "take twice spinlock" deadlock with the Spinlock debugging enabled on this lock, and static code analysis revealed this is the culprit: update_thread can take (in an error path) the sigio_lock, which is already held by all its callers (it's a static function, so it's easy to verify). Added some comments to mark where this function needs the lock, in case someone wants to reduce the locking here. Also clean an exitcall to mark the thread as killed (won't hurt, and could be useful if things go wrong). As a bonus, some CodingStyle cleanups. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.30, 2005-03-31 08:34:45-08:00, takata@linux-m32r.org [PATCH] m32r: build fix for CONFIG_DISCONTIGMEM This patch fixes build error for CONFIG_DISCONTIGMEM. * arch/m32r/mm/discontig.c: Fix build error for CONFIG_DISCONTIGMEM. * arch/m32r/kernel/setup.c: ditto. * arch/m32r/mm/discontig.c: - Add topology_init. - Cosmetics: change indentation of comments. Signed-off-by: Hayato Fujiwara Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.29, 2005-03-31 08:34:28-08:00, takata@linux-m32r.org [PATCH] m32r: Fix spinlock.h for CONFIG_DEBUG_SPINLOCK This patch is for fixing a build error of asm-m32r/spinlock.h for CONFIG_DEBUG_SPINLOCK. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.28, 2005-03-31 08:34:11-08:00, venkatesh.pallipadi@intel.com [PATCH] x86, x86_64: reading deterministic cache parameters and exporting it in /sysfs The attached patch adds support for using cpuid(4) instead of cpuid(2), to get CPU cache information in a deterministic way for Intel CPUs, whenever supported. The details of cpuid(4) can be found here IA-32 Intel Architecture Software Developer's Manual (vol 2a) (http://developer.intel.com/design/pentium4/manuals/index_new.htm#sdm_vol2a) and Prescott New Instructions (PNI) Technology: Software Developer's Guide (http://www.intel.com/cd/ids/developer/asmo-na/eng/events/43988.htm) The advantage of using the cpuid(4) ('Deterministic Cache Parameters Leaf') are: - It provides more information than the descriptors provided by cpuid(2) - It is not table based as cpuid(2). So, we will not need changes to the kernel to support new cache descriptors in the descriptor table (as is the case with cpuid(2)). The patch also adds a bunch of interfaces under /sys/devices/system/cpu/cpuX/cache, showing various information about the caches. Most useful field being shared_cpu_map, which says what caches are shared among which logical cpus. The patch adds support for both i386 and x86-64. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.27, 2005-03-31 08:33:54-08:00, roland@redhat.com [PATCH] x86-64 kprobes: handle %RIP-relative addressing mode The existing x86-64 kprobes implementation doesn't cope with the %RIP-relative addressing mode. Kprobes work by single-stepping a copy of an instruction overwritten by a breakpoint. When a probe is inserted on an instruction that uses the %RIP-relative data addressing mode, the copy run in a different location gets different data and so the presence of that probe causes the probed code to read or write the wrong memory location. Without this problem fixed, it is woefully unsafe to use the current kprobes code on x86-64 unless you are sure the instruction you instrument is not one that accesses global data using the %RIP addressing mode. This patch fixes the problem by recognizing the %RIP-relative addressing mode in an instruction when it's being copied to insert the kprobe, and adjusting its displacement so that it finds the right data. Taking this approach requires that the copied instruction's %RIP value be within 2GB of the virtual address of the data, i.e. the text/data areas of the kernel code and loaded modules. To satisfy this need the patch also replaces the use of vmalloc for getting instruction pages with lower-level calls to use a different part of the address space, the area at the top of the address space just above where modules are loaded. I left one page of red zone at the top, and the 1MB-4KB thus available allows for at most 69632 kprobes. (If we ever need to overcome that limit, we can change this to add a hook into the arch/x86_64/kernel/modules.c code and allocate pages inside the module area loading area instead.) Signed-off-by: Roland McGrath Acked-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.26, 2005-03-31 08:33:38-08:00, zwane@arm.linux.org.uk [PATCH] x86_64: reduce cacheline bouncing in cpu_idle_wait Andi noted that during normal runtime cpu_idle_map is bounced around a lot, and occassionally at a higher frequency than the timer interrupt wakeup which we normally exit pm_idle from. So switch to a percpu variable. I didn't move things to the slow path because it would involve adding scheduler code to wakeup the idle thread on the cpus we're waiting for. Signed-off-by: Zwane Mwaikambo Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.25, 2005-03-31 08:33:22-08:00, andrea@suse.de [PATCH] x86_64: avoid panic lockup This patch avoids a lockup with kernel.panic sysctl > 0, due an underflow of the number of the cpus. Later, Andrea said: While this still seems a needed fix, unfortunately this wasn't enough to fix the problem (the first positive report I got was because they did a mistake and they tested an UP kernel, oh well ;). Anyway the smp_stop_cpu executed by the cpu that invoked panic is wrong and it really causes an underflow of the number of cpus that can lead to further lockups. So I don't want to stop you from merging it since it seems a correct fix and it brings x86-64 in line with the x86 code too. But apparently we need something more than this too. Signed-off-by: Andrea Arcangeli Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.24, 2005-03-31 08:33:05-08:00, bunk@stusta.de [PATCH] arch/i386/kernel/smp.c: remove a pointless "inline" All callers of send_IPI_mask_sequence are in other files, so marking it "inline" is quite pointless. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.23, 2005-03-31 08:32:49-08:00, stsp@aknet.ru [PATCH] x86: fix ESP corruption CPU bug (take 2) This works around the corruption of the high word of the ESP register, which is the official bug of x86 CPUs. The bug triggers only when the one is using the 16bit stack segment, and is described here: http://www.intel.com/design/intarch/specupdt/27287402.PDF From: Oleg Nesterov I think that Stas tries to steal 1024 bytes from kernel's memory. Acked-by: Linus Torvalds Acked-by: Petr Vandrovec Acked-by: Pavel Machek Signed-off-by: Stas Sergeev Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.22, 2005-03-31 08:32:31-08:00, zwane@arm.linux.org.uk [PATCH] x86: reduce cacheline bouncing in cpu_idle_wait Andi noted that during normal runtime cpu_idle_map is bounced around a lot, and occassionally at a higher frequency than the timer interrupt wakeup which we normally exit pm_idle from. So switch to a percpu variable. I didn't move things to the slow path because it would involve adding scheduler code to wakeup the idle thread on the cpus we're waiting for. Signed-off-by: Zwane Mwaikambo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.21, 2005-03-31 08:32:14-08:00, michael@ellerman.id.au [PATCH] ppc64: Add mem=X boot command line option This patch adds the mem=X boot command line option for PPC64. On iSeries the user's mem=X value is aligned to PAGE_SIZE, on pSeries we align to 16 MB which is the size of a large page. The iSeries implementation is fairly straight forward, we declare mem=X as an early_param() and then in iSeries_init_early() we modify the systemcfg->physicalMemorySize based on that value. On pSeries the mem=X option is parsed in prom_init.c before the kernel proper starts, and is used to modify prom_init_mem()'s idea of memory. The mem=X value and computed tce_alloc_start/end values are saved by prom_init() into the device tree for later use by the kernel. The device tree properties are read by the kernel in early_dt_scan_chosen(), and used to modify the lmb structure in early_init_devtree(). That's the guts of it. On non-LPAR machines the tce_alloc_start/end values are read from the device tree and used in htab_initialize() to make sure the TCE table is mapped at the real top of RAM. If NUMA is enabled we also have to make changes in parse_numa_properties() and do_init_bootmem() to exclude memory regions above the memory limit, and truncate any region which stradles the limit. NB. This patch does not facilitate using mem=X to give drivers access to large regions of contiguous memory. Thanks to BenH, Anton, Olof, Stephen, Mike & Maneesh for their help. Signed-off-by: Michael Ellerman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.20, 2005-03-31 08:31:57-08:00, apw@us.ibm.com [PATCH] ppc64: fix zilog link error drivers/built-in.o(.text+0x2c640): In function `pmz_attach': /home/apw/devel/percs-head/linux-2.6/drivers/serial/pmac_zilog.c:1560: undefined reference to `.macio_request_resources' Signed-off-by: Amos Waterland Acked-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.19, 2005-03-31 08:31:40-08:00, andrea@cpushare.com [PATCH] seccomp for ppc64 This patch against 12-rc1 adds seccomp to the ppc64 arch. I tested it successfully with the seccomp_test. I didn't bother to change the syscall exit not to check for TIF_SECCOMP, in theory that bit could be optimized but it's an optimization in the slow path, and current code is a bit simpler. I also verified it still compiles and works fine on x86 and x86-64. Instead of the TIF_32BIT redefine, if you want to change x86-64 to use TIF_32BIT too (instead of TIF_IA32), let me know. Signed-off-by: Andrea Arcangeli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.18, 2005-03-31 08:31:22-08:00, Stefan.Nickl@kontron.com [PATCH] ppc32: MPC8555 CPM2 size/pointers for FCCs aka "All-ones problem" I've seen that MPC8555 support also made it into linux-2.6 by now. I don't have hardware and time to test, but it looks like this bug has made it over from the linuxppc-2.4 tree. The 8541/8555 has a smaller CPM DPRAM than the standard CPM2, just like the 8272. Unfortunately, the manual recommends you to use a segment of CPM DPRAM that is not implemented in the 8541/55. If the smaller DPRAM is not taken into account, fcc_enet.c will initialise the internal CPM buffer pointers (fcc_riptr/fcc_tiptr) to invalid offsets, with the result that all the FCC will ever send and receive are 0xff-en. See also "All-ones problem with FCC1 on MPC8541" ~October 2004 on linuxppc-embedded. Signed-off-by: Stefan Nickl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.17, 2005-03-31 08:31:04-08:00, galak@freescale.com [PATCH] ppc32: Fix MPC8555 & MPC8555E device lists (updated) Removed the FCC3 device from the lists of devices on MPC8555 & MPC8555E since it does not exist on these processors. Signed-off-by: Jason McMullan Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.16, 2005-03-31 08:30:47-08:00, galak@freescale.com [PATCH] ppc32: CPM2 PIC cleanup irq_to_siubit array Cleaned up irq_to_siubit array so we no longer need to do 1 << (31-bit), just 1 << bit. Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.15, 2005-03-31 08:30:30-08:00, galak@freescale.com [PATCH] ppc32: CPM2 PIC cleanup Cleaned up the CPM2 interrupt controller code: * Added the ability to offset the IRQs * Refactored common PIC init code out of platform files * Fixed IRQ offsets on MPC85xx so it can handle properly handled multiple interrupt controllers (i8259, CPM2 PIC, and OpenPIC) Signed-off-by: Jason McMullan Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.14, 2005-03-31 08:30:13-08:00, galak@freescale.com [PATCH] ppc32: cleanup of Book-E exception handling Cleaned up the Book-E exception handling code to remove saving/restoring registers that were not needed. Moved the register save/restore area onto the exception stacks instead of dedicated offsets. Additionally, this allows for proper SMP handling of the additional exception levels. Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.13, 2005-03-31 08:29:56-08:00, tnt@246tNt.com [PATCH] ppc32: Adds necessary cpu init to use USB on LITE5200 Platform To use external peripheral on MPC5200, some clocking registers and port-muxing must be done. Since this is platform specific, it's placed the platform support file. This particular patch is for USB support on the LITE5200. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.12, 2005-03-31 08:29:39-08:00, tnt@246tNt.com [PATCH] serial: Update mpc52xx_uart.c to use platform bus All Freescale MPC52xx related code now use new constants and the platform bus for it's driver. This patch makes this driver make use of that. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.11, 2005-03-31 08:29:22-08:00, tnt@246tNt.com [PATCH] ppc32: Use platform bus / ppc_sys model for Freescale MPC52xx This patch makes all platform based around the Freescale MPC52xx use the platform bus and more precisly the ppc_sys model put in place by Kumar Gala. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.10, 2005-03-31 08:29:04-08:00, tnt@246tNt.com [PATCH] ppc32: Change constants style in Freescale MPC52xx related code This patch changes the way the constants used for register block address are defined/used. This is a preparation for the use of the platform bus / ppc_sys model. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.9, 2005-03-31 08:28:47-08:00, tnt@246tNt.com [PATCH] ppc32: Remove the OCP system from the Freescale MPC52xx support We remove all usage of the OCP system as preparation to switch to the platform bus model / ppc_sys model. This is only for 'generic' support, drivers are adapted separatly, afterwards. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.8, 2005-03-31 08:28:30-08:00, tnt@246tNt.com [PATCH] ppc32: Remove unnecessary test in MPC52xx reset code That test is part of an old version of the code and erroneously made it to mainstream. Signed-off-by: Sylvain Munaut Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.7, 2005-03-31 08:28:13-08:00, rddunlap@osdl.org [PATCH] cpuset: make function decl. ANSI kernel/cpuset.c:1428:41: warning: non-ANSI function declaration Signed-off-by: Randy Dunlap Acked-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.6, 2005-03-31 08:27:57-08:00, pj@engr.sgi.com [PATCH] cpusets GFP_ATOMIC fix: tonedown panic comment This patch applies on top of my patch of March 26, entitled "cpusets special case GFP_ATOMIC allocs". It tones down my panic'y commentary. My commentary shouldn't imply that failed GFP_ATOMICs should lead to, or normally lead to, panics. Even though there are a few panic() calls following failed GFP_ATOMIC allocs, this is not the usual or desired result of a failed GFP_ATOMIC. The kernel will probably drop some detail on the floor and keep on working. Thanks to Nick Piggin for noticing (I hope this answers his point.) Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.5, 2005-03-31 08:27:39-08:00, pj@sgi.com [PATCH] cpusets: special-case GFP_ATOMIC allocs Stringent enforcement of cpuset memory placement could cause the kernel to fail a GFP_ATOMIC (!wait) memory allocation, even though memory was available elsewhere in the system. Relax the cpuset constraint, on the last zone loop in mm/page_alloc.c:__alloc_pages(), for ATOMIC requests. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.4, 2005-03-31 08:27:22-08:00, hirofumi@mail.parknet.co.jp [PATCH] read_kmem() fixes drivers/char/mem.c:289 if (p < PAGE_SIZE && read > 0) { [...] read -= tmp; count -= tmp; This part is losting the number of bytes which read. drivers/char/mem.c:302 sz = min_t(unsigned long, sz, count); This should use "read" instead of "count". drivers/char/mem.c:315 read -= sz; count -= sz; Also lost the number of bytes which read. In short, kmem returns incorrect number always if user is accessing the lowmem area. And also it doesn't handle the highmem boundary rightly. This patch uses "low_count" instead of "read", as the number of copy in lowmem area. And "read" is used as the number of bytes which read. Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.3, 2005-03-31 08:27:05-08:00, dhowells@redhat.com [PATCH] BDI: Provide backing device capability information [try #3] The attached patch replaces backing_dev_info::memory_backed with capabilitied bitmap. The capabilities available include: (*) BDI_CAP_NO_ACCT_DIRTY Set if the pages associated with this backing device should not be tracked by the dirty page accounting. (*) BDI_CAP_NO_WRITEBACK Set if dirty pages associated with this backing device should not have writepage() or writepages() invoked upon them to clean them. (*) Capability markers that indicate what a backing device is capable of with regard to memory mapping facilities. These flags indicate whether a device can be mapped directly, whether it can be copied for a mapping, and whether direct mappings can be read, written and/or executed. This information is primarily aimed at improving no-MMU private mapping support. The patch also provides convenience functions for determining the dirty-page capabilities available on backing devices directly or on the backing devices associated with a mapping. These are provided to keep line length down when checking for the capabilities. Signed-Off-By: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.2, 2005-03-31 08:26:47-08:00, trond.myklebust@fys.uio.no [PATCH] NFS: Fix typo in access caching code Currently we fail to clear the "invalid cache" flag when we've revalidated the cache. Signed-off-by: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2224.1.1, 2005-03-31 08:26:30-08:00, zam@namesys.com [PATCH] arm atomic_sub_and_test() Add a wrapper for atomic_sub_and_test into include/asm-arm/atomic.h Acked-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.51.8, 2005-03-31 14:34:31+02:00, tgraf@suug.ch Cset exclude: tgraf@suug.ch|ChangeSet|20050316221421|24742 ChangeSet@1.2181.1.16, 2005-03-30 20:44:11-08:00, davem@sunset.davemloft.net [SPARC64]: Kill final normal g5 register reference. This one was in strncpy_from_user(). Now we can finally put the per-cpu data area base into g5 on SMP. Signed-off-by: David S. Miller ChangeSet@1.2181.1.15, 2005-03-30 20:13:40-08:00, davem@sunset.davemloft.net [SPARC64]: Simplify checksumming code. The main impetus was to get rid of some of the remaining g5 register accesses. But this routine is about as fast as the older VIS stuff, and actually faster on UltraSPARC-III and later chips. Signed-off-by: David S. Miller ChangeSet@1.2181.53.2, 2005-03-30 20:37:02-05:00, fthain@telegraphics.com.au [PATCH] fix Jazzsonic driver build on m68k Signed-off-by: Finn Thain Signed-off-by: Jeff Garzik ChangeSet@1.2181.53.1, 2005-03-30 20:14:45-05:00, brazilnut@us.ibm.com [PATCH] pcnet32: 79C975 fiber fix From: "HARDY, Steven" I have found a bug in the pcnet32 driver (drivers/net/pcnet32.c) affecting all ethernet cards based on the AMD79C975 chip, using the fiber interface. It's a one line fix, where some config registers get corrupted during initialisation (which stops the Fiber interface working with this chip) This bug was introduced somewhere betweeen 2.4.17 and 2.6.x (noticed whilst upgrading to 2.6), and it may affect other chips too. I have checked all versions up to 2.6.11-bk6 and they are all broken. Signed-off-by: Andrew Morton Signed-off-by: Don Fry Signed-off-by: Jeff Garzik ChangeSet@1.2181.51.7, 2005-03-30 16:58:16-08:00, herbert@gondor.apana.org.au [NETLINK]: Fix sk_rmem_alloc assertion failure. In netlink_dump we're operating on sk after dropping the cb lock. This is racy because the owner of the socket could close it after we drop the cb lock. This is possible because netlink_dump isn't always called from the context of the process that owns the socket. For instance, if there is contention on rtnl then rtnetlink requests will be processed by the process that owns the rtnl. The solution is to hold a ref count on the socket before we drop the cb lock. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.52.3, 2005-03-30 19:53:35-05:00, ravinandan.arakali@neterion.com [PATCH] S2io: Changed copyright and added support for Xframe II 1. Change of Copyright to reflect S2io's new name "Neterion Inc." 2. Updated driver version number. 3. Add an additional PCI device id to support Neterion's new Xframe II. Some background info on Xframe-II, just in case - The NIC was announced back in January (see http://www.serverwatch.com/news/article.php/3464871) Xframe II is a PCI-X 2.0 DDR adapter designed to work in PCI-X 2.0 servers (it is also backwards compatible with current pci-x and pci slots); it's a first card to overcome pci-x 133 throughput bottleneck. Some of these pci-x 2.0 servers are available now and more are coming later in the year. Xframe II is completely backward compatible and hence the current driver will work as-is, except the device id change. There are some additional features/stateless offloads in Xframe-II as well; we plan to submit patches for these soon. Signed-off-by: Ravinandan Arakali Signed-off-by: Jeff Garzik ChangeSet@1.2181.52.2, 2005-03-30 19:53:21-05:00, ravinandan.arakali@neterion.com [PATCH] S2io: h/w initialization fixes Below patch addresses couple of issues w.r.t h/w initialization. 1. The SWAPPER_CTRL register is now initialized as per recommended procedure from h/w team. This register when initialized correctly ensures correct access to other registers on big-endian and little-endian systems. 2. Enabling the Continuous interrupt feature by default on Tx at a rate of 250 interrupts/sec. Signed-off-by: Ravinandan Arakali Signed-off-by: Jeff Garzik ChangeSet@1.2181.52.1, 2005-03-30 19:53:07-05:00, ravinandan.arakali@neterion.com [PATCH] S2io: Statistics fix Below patch fixes the statistics problem on big-endian systems(such as IBM PPC). Basically, the problem was "ethtool -s" was reporting huge values on counters like no. of Tx frames, no. of Rx packets etc. The fix has been tested on little-endian and big-endian systems. Signed-off-by: Ravinandan Arakali Signed-off-by: Jeff Garzik ChangeSet@1.2181.32.3, 2005-03-30 19:42:51-05:00, linville@tuxdriver.com [PATCH] e1000: add MODULE_VERSION Add MODULE_VERSION entry. Signed-off-by: John W. Linville Signed-off-by: Jeff Garzik ChangeSet@1.2181.51.5, 2005-03-30 16:35:41-08:00, yoshfuji@linux-ipv6.org [AF_UNIX]: unix_mkname comment Resurrect an old comment, explaining why this isn't an off-by-one error. Signed-off-by: Hideaki YOSHIFUJI Signed-off-by: Andrew Morton Signed-off-by: David S. Miller ChangeSet@1.2181.51.4, 2005-03-30 16:35:08-08:00, yoshfuji@linux-ipv6.org [IPV6]: Fix address/interface handling according to the scoping architecture I think this has been there for long time (maybe since 2.4...). With the following patch, I can connect local link-local address. - Change incoming interface according to the scoping architecture - Choose source address on appropriate interface, according to the scoping architecture. Signed-off-by: Hideaki YOSHIFUJI Signed-off-by: Andrew Morton Signed-off-by: David S. Miller ChangeSet@1.2181.51.3, 2005-03-30 16:34:22-08:00, herbert@gondor.apana.org.au [PKT_SCHED]: Memory leak in ipt.c Signed-off-by: Herbert Xu Signed-off-by: Andrew Morton Signed-off-by: David S. Miller ChangeSet@1.2181.51.2, 2005-03-30 16:33:25-08:00, rusty@rustcorp.com.au [NETFILTER]: Restore ports module parameter for ip_nat_{ftp,irq} There is no 'ports' parameter for the ip_nat_ftp and ip_nat_irc modules in 2.6.11: the ports parameter supplied to the ip_conntrack_ftp/ip_conntrack_irc module defines the ports. It was unfortunate that we were lazy in the original implementation, and forced the user to duplicate the arguments. Even more unfortunate, the removal of the parameter caused autoloading to break for various setups, with an 'Unknown parameter' message. The solution is to restore the parameter as a dummy, with a polite warning message that it is no longer neccessary. Signed-off-by: Rusty Russell Signed-off-by: Andrew Morton Signed-off-by: David S. Miller ChangeSet@1.2181.51.1, 2005-03-30 16:32:22-08:00, jan.kiszka@web.de [NET]: NULL pointer bug in netpoll.c It seems that there is a gremlin sleeping in net/core/netpoll.c:find_skb(). Even if no more buffers are available through skbs, skb is dereferenced anyway. The tiny patch should fix it. Signed-off-by: Andrew Morton Signed-off-by: David S. Miller ChangeSet@1.2220, 2005-03-30 15:41:44-08:00, torvalds@ppc970.osdl.org Sound/USB merge ChangeSet@1.2181.1.14, 2005-03-30 15:11:38-08:00, davem@sunset.davemloft.net [SPARC64]: Add UltraSPARC-IV cpu ids. Signed-off-by: David S. Miller ChangeSet@1.2181.1.13, 2005-03-30 13:49:13-08:00, davem@sunset.davemloft.net [SPARC64]: Simplified csum_partial() implementation. There is no need to make this thing use VIS et al. A simple straightforward prefetching integer version is fine. Actually, most of the time this routine is run to compute checksums of small header bits or similar. Signed-off-by: David S. Miller ChangeSet@1.2219, 2005-03-30 12:36:24-08:00, gregkh@suse.de Merge suse.de:/home/greg/linux/BK/bleed-2.6 into suse.de:/home/greg/linux/BK/usb-2.6 ChangeSet@1.2181.27.23, 2005-03-30 13:58:06-06:00, markh@osdl.org [PATCH] aacraid: endian cleanup An update to the patch from Adaptec with byte order changes: - drop byte swapping on all 0's and all 1's content - fix up missing swapping directives - ensure swapping on 16 bit values does not use 32 bit swap - Made hardware registers __le* types. - hopefully caught the style issues. Signed-off-by: Mark Haverkamp Signed-off-by: James Bottomley ChangeSet@1.2181.27.22, 2005-03-30 13:52:14-06:00, jejb@mulgrave.(none) Fix SCSI internal requests hang When a device is being torn down (mostly in USB ejection) we hit conditions where SCSI hangs in wait_for_completion(). The reason is that the scsi request prep function does a state check on the device and rejects it with BLKPREP_KILL if we're in the DEL (going away, no more I/O) state. Unfortunately, I/O submitted by scsi_wait_req() doesn't have its request->end_io initialised, so we're never notified of this (the submit path expects everything to come back via the scsi_request->sr_done() routine, which only happens if the I/O actually gets submitted. Note: almost everything that uses scsi_do_request() is wrong not only in this regard but also on other things, like ending tags. Problem Diagnosed with help from: Tejun Heo Signed-off-by: James Bottomley ChangeSet@1.2181.2.60, 2005-03-30 10:51:32-08:00, davem@cheetah.davemloft.net [SPARC64]: Missed some cases in U1memcpy register rework. Register o4 is referenced both with a preceeding percent sign and without (via macros), so make sure to replace all such cases. Signed-off-by: David S. Miller ChangeSet@1.2181.2.59, 2005-03-30 10:25:27-08:00, davem@cheetah.davemloft.net [SPARC64]: Kill unused header arch/sparc64/lib/VIS.h Signed-off-by: David S. Miller ChangeSet@1.2218, 2005-03-30 09:46:52-08:00, dhowells@redhat.com [PATCH] FRV: Fix kernel configuration This fixes the FRV configuration to work with 2.6.12-rc1. It does this by breaking out the kernel hacking menu into a separate file, in the same way this is done in other archs. Signed-Off-By: David Howells Signed-off-by: Linus Torvalds ChangeSet@1.2217, 2005-03-30 09:46:33-08:00, dhowells@redhat.com [PATCH] FRV: Cleanup unused variable This removes an unused variable from the FRV arch. Signed-Off-By: David Howells Signed-off-by: Linus Torvalds ChangeSet@1.2216, 2005-03-30 09:46:17-08:00, dhowells@redhat.com [PATCH] FRV: Fix TLB miss mapping cache flush This fixes the TLB miss mapping cache flush function. The flush was attempting to invalidate the coverage start virtual addresses for the cached page table mappings held in registers SCR0 and SCR1 by writing 0 into them. Unfortunately, 0x00000000-0x04000000 is itself a valid part of the virtual address range. This patches places -1 in there instead, thus specifying 0xfc000000-0xffffffff which is covered by a static I/O mapping, and so shouldn't ever be seen by the TLB-miss handler. Signed-Off-By: David Howells Signed-off-by: Linus Torvalds ChangeSet@1.2215, 2005-03-30 09:45:58-08:00, pavel@ucw.cz [PATCH] Remaining u32 vs. pm_message_t fixes This fixes three remaining places where we put u32 (or worse suspend_state_t) into pm_message_t-sized box. As a bonus, PCI_D0 is used instead of constant 0. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2214, 2005-03-30 09:45:42-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in AGP This should fix u32 vs. pm_message_t confusion in AGP. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2213, 2005-03-30 09:45:26-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t in macintosh This fixes u32 vs. pm_message_t confusion in macintosh. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2212, 2005-03-30 09:45:09-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in serials This fixes u32 vs. pm_message_t confusion in serials. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2211, 2005-03-30 09:44:51-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in MMC This fixes u32 vs. pm_message_t confusion in MMC layer. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2210, 2005-03-30 09:44:34-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in framebuffers This should fix u32 vs pm_message_t confusion in framebuffers, and do no code changes. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2209, 2005-03-30 09:44:16-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in PCMCIA This should fix part of u32 vs. pm_message_t confusion in pcmcia. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2208, 2005-03-30 09:43:58-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t confusion in OSS This fixes (part of) u32 vs. pm_message_t confusion in OSS. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2207, 2005-03-30 09:43:40-08:00, pavel@ucw.cz [PATCH] more pm_message_t fixes While fixing usb, I stomped on a few more fixes. Print error when some device fails to power down, and 2 is no longer valid state to pass in pm_message_t. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2206, 2005-03-30 09:43:23-08:00, pavel@suse.cz [PATCH] Fix u32 vs. pm_message_t in USB This fixes (part of) u32 vs. pm_message_t confusion in USB. It should cause no code changes. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2205, 2005-03-30 09:43:06-08:00, pavel@ucw.cz [PATCH] Fix pm_message_t in generic code This fixes u32 vs. pm_message_t in generic code. No code changes. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.50.4, 2005-03-30 18:34:03+01:00, vince@org.rmk.(none) [ARM PATCH] 2636/1: Missing include breaking cats build Patch from Vincent Sanders Fixes missing include which breaks footbridge_defconfig build Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.50.3, 2005-03-30 14:42:01+01:00, sascha@de.rmk.(none) [ARM PATCH] 2635/1: i.MX serial hardware handshaking support Patch from Sascha Hauer This patch adds support for hardwarehandshaking on i.MX soc Signed-off-by: Sascha Hauer Signed-off-by: Mike Lee Signed-off-by: Russell King ChangeSet@1.2181.50.2, 2005-03-30 14:24:57+01:00, nico@org.rmk.(none) [ARM PATCH] 2634/1: prevent the lack of any CPU and/or machine record at link time Patch from Nicolas Pitre Without this, someone could successfully link a kernel with, for example, no machine description structure at all. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King ChangeSet@1.2181.50.1, 2005-03-30 14:16:38+01:00, lucasvr@org.rmk.(none) [ARM PATCH] 2630/1: Fixes definition of GPB10 on S3C2410 Patch from Lucas Correia Villa Real This patch fixes the definition of the GPB10 bits on the S3C2410 by using the correct offset to access its functions. Signed-off-by: Lucas Correia Villa Real Signed-off-by: Ben Dooks Signed-off-by: Russell King ChangeSet@1.2181.49.2, 2005-03-30 12:22:42+01:00, rmk@flint.arm.linux.org.uk [SERIAL] Remove serial8250_late_console_init The serial core will try to register the console each time a port is registered with it. This makes serial8250_late_console_init redundant, so remove it. Signed-off-by: Russell King ChangeSet@1.2181.49.1, 2005-03-30 09:49:58+01:00, takata@org.rmk.(none) [SERIAL] m32r_sio driver update Patch from Hirokazu Takata m32r_sio driver updates: - Move m32r_sio specific description from asm-m32r/serial.h to driver/serial/m32r_sio.c. - Remove __register_m32r_sio, register_m32r_sio and unregister_m32r_sio from driver/serial/m32r_sio.c. Signed-off-by: Hirokazu Takata ChangeSet@1.2181.4.95, 2005-03-29 23:43:56-08:00, david-b@pacbell.net [PATCH] USB: ohci D3 resume fix This fixes a problem that cropped up resuming OHCI from PCI D3 on NForce2. Evidently the register controlling frame timing gets clobbered in D3, but not other registers ... a "by the book" reinit seems to solve this particular problem. (And ought to help a few startup glitches on other implementations. Linux never used that toggle bit before, and has had to struggle with glitchy init...) It also updates some diagnostics to be bit more useful, and tries to avoid a particular "wakeup while suspending" glitch that seems particular to one particular type of Australian mouse. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.94, 2005-03-29 22:47:47-08:00, gregkh@suse.de USB: fix up a lot of sparse warnings and bugs in the pwc driver. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.93, 2005-03-29 21:42:28-08:00, abbotti@mev.co.uk [PATCH] ftdi_sio: fix sysfs attribute permissions This patch changes the permissions of the 'event_char' and 'latency_timer' sysfs attributes for devices supported by the ftdi_sio driver. They should only be set by root. Signed-off-by: Ian Abbott Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.92, 2005-03-29 21:42:13-08:00, abbotti@mev.co.uk [PATCH] ftdi_sio: Support sysfs attributes for more chip A couple of sysfs attributes were added for the FT232BM (and FT245BM) chip type in 2.6.11. This patch extends there use to the FT8U232AM (and FT8U245AM) and FT2232C chip types, where applicable. Signed-off-by: Ian Abbott Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.91, 2005-03-29 21:41:57-08:00, abbotti@mev.co.uk [PATCH] ftdi_sio: add array to map chip type to a string This patch just adds an array to map the chip type to a string for use in debug messages for the ftdi_sio driver. I use it in the get_ftdi_divisor function and another patch. Signed-off-by: Ian Abbott Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.90, 2005-03-29 21:26:30-08:00, g.toth@e-biz.lu [PATCH] USB: rewrite the usblcd driver The contribution was created in whole or in part by me and I have the right to submit it under the open source license indicated in the file. Signed-off-by: Georges Toth Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.89, 2005-03-29 21:16:08-08:00, gregkh@suse.de [PATCH] USB: Put the Kconfig and Makefile back in proper order for the serial drivers Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.88, 2005-03-29 21:15:53-08:00, gregkh@suse.de [PATCH] USB: mark functions static in the cp2101 driver. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.87, 2005-03-29 21:15:36-08:00, craig@microtron.org.uk [PATCH] USB: add driver for CP2101/CP2102 RS232 adaptors This patch against 2.6.10 provides the driver for the CP2101 and CP2102 USB to RS232 adaptor chips which are found in some serial converter cables and usb mobile phone cables. It provides cp2101.c, and patches Kconfig and Makefile in drivers/usb/serial Since there is no technical data sheet freely available for these chips, much of the code in this driver has been based on protocol analysis, assumptions and guesswork. The cp210x family of devices have OEM programmable device IDs so updates will be required as and when new devices are found. The driver currently only supports configuration of baud rate, data bits, stop bits and parity mode. While this is enough to enable the use of a wide range of devices based on this chip, the driver is still incomplete in many areas such as flow control. Unfortunately without technical datasheets, implementing these features is made very difficult indeed. Signed-off-by: Craig Shelley craig@microtron.org.uk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.1.11, 2005-03-29 20:41:34-08:00, davem@sunset.davemloft.net Merge sunset.davemloft.net:/home/davem/src/BK/sparcwork-2.6 into sunset.davemloft.net:/home/davem/src/BK/sparc-2.6 ChangeSet@1.2181.45.3, 2005-03-30 01:34:30+02:00, tgraf@suug.ch [PKT_SCHED]: Fix action statistics dumping in compatibility mode Extends the action dumping function by a parameter to differ between regular calls and the one supposed to add the backward compatiblity bits for old userspace applications. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller ChangeSet@1.2181.27.21, 2005-03-29 17:20:00-06:00, michaelc@cs.wisc.edu [PATCH] fix fc class work queue usage According to this article http://lwn.net/Articles/125930/, "When cancel_delayed_work() returns zero, it means that the delayed work request was fired off before the call; it might, in fact, be running on another CPU when the cancel attempt is made". If it is successful, it returns a nonzero value. Tracing through cancel_delayed_work's timer usage would seem to confirm this. The fc class today though performs a flush_scheduled_work, when the return value is nonzero instead of zero. Also it appears the fc class will use flush_scheduled_work to flush the work from the shost_work_q when it should be using flush_workqueue(shost->work_q) (flush_scheduled_work() only flushes the default, keventd_wq, work queue). The attached patch adds a scsi_flush_work function for scsi_transport_fc to use and it fixes the cancel_delayed_work() test to detect when to flush the work queues correctly (it also only calls cancel_delayed_work when the work is queued as delayed (scan_work is not delayed). Signed-off-by: Mike Chrisite Signed-off-by: James Bottomley ChangeSet@1.2181.47.2, 2005-03-29 13:46:30-08:00, davem@sunset.davemloft.net [SPARC64]: More g5 register usage elimination. Use temporal stores and prefetches in memset/bzero while we are at it. Signed-off-by: David S. Miller ChangeSet@1.2181.48.1, 2005-03-29 15:48:28-05:00, jgarzik@pobox.com Merge pobox.com:/garz/repo/linux-2.6 into pobox.com:/garz/repo/net-drivers-2.6 ChangeSet@1.2181.1.9, 2005-03-29 15:42:20-05:00, jgarzik@pobox.com [libata sata_sil] Don't presume PCI cache-line-size reg is > 0 Some BIOSen are known to screw up the PCI cache-line-size register, so add a paranoia check. ChangeSet@1.2203, 2005-03-29 11:25:49-08:00, cifs.adm@bkbits.net Merge bk://linux.bkbits.net/linux-2.5 into bkbits.net:/repos/c/cifs/linux-2.5cifs ChangeSet@1.2181.27.20, 2005-03-29 11:06:06-06:00, jejb@mulgrave.(none) 3ware 9000 driver update From: adam radford - Add support for PAE mode. - Add lun support. - Fix twa_remove() to free irq handler/unregister_chrdev() before shutting down the card. - Change to new 'change_queue_depth' api. (From James) - Fix 'handled=1' ISR usage, remove bogus IRQ check (from Jeff Garzik) - Remove un-needed eh_abort handler. - Add support for embedded firmware error strings. Signed-off-by: Adam Radford Signed-off-by: James Bottomley ChangeSet@1.2181.27.19, 2005-03-29 10:55:43-06:00, rddunlap@osdl.org [PATCH] scsi_sysfs: use NULL instead of 0 drivers/scsi/scsi_sysfs.c:814:66: warning: Using plain integer as NULL pointer Signed-off-by: Randy Dunlap Signed-off-by: James Bottomley ChangeSet@1.2181.27.18, 2005-03-29 10:49:22-06:00, michaelc@cs.wisc.edu [PATCH] rm unused scan delay var Is FC_SCSI_SCAN_DELAY used by a FC driver that is not yet in mainline? This patch just deletes it if not since no one else is. Signed-off-by: Mike Chrisite Signed-off-by: James Bottomley ChangeSet@1.2181.1.7, 2005-03-29 07:43:57-08:00, christophe@saout.de [PATCH] x86-64: Fix preemption off of irq context with PREEMPT_BKL Fixing the interrupt tst for CONFIG_PREEMPT on x86-64 exposes another bug with CONFIG_PREEMPT_BKL. Calling schedule due to preemption releases the BKL which it shouldn't do. Call preempt_schedule_irq instead (like for i386). This fixes the easily reproducible filesystem errors I've seen (with reiserfs, which heavily relies on the BKL). Signed-off-by: Christophe Saout Signed-off-by: Andi Kleen Signed-off-by: Linus Torvalds ChangeSet@1.2181.16.26, 2005-03-28 22:29:40-08:00, gregkh@suse.de [PATCH] PCI: create PCI_DEBUG config option to make it easier for users to enable pci debugging Now you don't have to dig through a file to change a #define, it's a real config option. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.25, 2005-03-28 22:00:24-08:00, gregkh@suse.de PCI: clean up the dynamic id logic a little bit. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.1.6, 2005-03-28 21:24:58-08:00, torvalds@ppc970.osdl.org Merge bk://bk.arm.linux.org.uk/linux-2.6-rmk into ppc970.osdl.org:/home/torvalds/v2.6/linux ChangeSet@1.2202, 2005-03-28 22:50:11-06:00, stevef@smf-t23.(none) [CIFS] Return inode numbers (from server) more consistently on lookup and readdir to both types of servers (whether they support Unix extensions or not) when serverino mount parm specified. Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.43.7, 2005-03-28 20:43:20-08:00, torvalds@evo.osdl.org Merge whitespace and __nocast changes ChangeSet@1.2181.46.245, 2005-03-28 20:24:26-08:00, sct@redhat.com [PATCH] ext3: fix journal_unmap_buffer race There is a race condition in jbd between journal_unmap_buffer() and journal_commit_transaction(). This is leading to corruption of buffers on the transaction's t_locked_list, leading to a variety of symptoms usually involving an oops in kjournald. The problem is that various special-case exit paths in journal_unmap_buffer() call journal_put_journal_head without any locking. This is racing against a refiling of the same journal_head in journal_commit_transaction(): __journal_unfile_buffer(jh); __journal_file_buffer(jh, commit_transaction, BJ_Locked); The way these functions work, this leaves the jh temporarily with b_transaction==NULL; and if journal_unmap_buffer()'s call to journal_put_journal_head() hits this window, it sees the NULL transaction and frees the journal_head which is just about to get refiled on the locked list. The main exit path on journal_unmap_buffer() performs its journal_put_journal_head() before dropping the j_list_lock, so is not vulnerable to this race. The fix is to move the other similar calls on special-case exit branches in that function so that they also release the journal_head before dropping that lock. This is low-risk since the new order has already been tested as the normal exit path from this function. The change has had extensive testing and has been shown to fix the problem with no regressions found. Signed-off-by: Peter Keilty Signed-off-by: Nicholas Dokos Signed-off-by: Stephen Tweedie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.244, 2005-03-28 20:24:10-08:00, pbadari@us.ibm.com [PATCH] ext3 writeback "nobh" option Add a `nobh' mount option to ext3 in writeback mode: avoid attaching buffer_head to data pages, like ext2. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.243, 2005-03-28 20:23:53-08:00, pbadari@us.ibm.com [PATCH] ext3 writepages support for writeback mode Add writepages support for ext3 writeback mode. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.242, 2005-03-28 20:23:36-08:00, sct@redhat.com [PATCH] ext3/jbd race: releasing in-use journal_heads Fix destruction of in-use journal_head journal_put_journal_head() can destroy a journal_head at any time as long as the jh's b_jcount is zero and b_transaction is NULL. It has no locking protection against the rest of the journaling code, as the lock it uses to protect b_jcount and bh->b_private is not used elsewhere in jbd. However, there are small windows where b_transaction is getting set temporarily to NULL during normal operations; typically this is happening in __journal_unfile_buffer(jh); __journal_file_buffer(jh, ...); call pairs, as __journal_unfile_buffer() will set b_transaction to NULL and __journal_file_buffer() re-sets it afterwards. A truncate running in parallel can lead to journal_unmap_buffer() destroying the jh if it occurs between these two calls. Fix this by adding a variant of __journal_unfile_buffer() which is only used for these temporary jh unlinks, and which leaves the b_transaction field intact so that we never leave a window open where b_transaction is NULL. Additionally, trap this error if it does occur, by checking against jh->b_jlist being non-null when we destroy a jh. Signed-off-by: Stephen Tweedie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.241, 2005-03-28 20:23:19-08:00, js@linuxtv.org [PATCH] dvb: mt352: cleanups o remove s* from state, they are only used in read_status o remove casting of void* o remove FIXME in set_parameters, should be handled by dvb-core state machine o remove some unnecessary braces o remove #if 1 in read_status, and add note from Zarlink design manual o change read_signal_strength to read total AGC_GAIN in case some adapter turns on the RF_AGC loop. (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.240, 2005-03-28 20:23:03-08:00, js@linuxtv.org [PATCH] dvb: gcc 2.95 compile fixes Patch by Olaf Titz: gcc 2.95 compile fixes Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.239, 2005-03-28 20:22:46-08:00, js@linuxtv.org [PATCH] dvb: ttusb_dec: cleanup Cleanup patch from Peter Beutner: o unregister the input device on disconnect and move cleanup stuff of the RC to own function o keymap should be static not const o set up keymap correctly and completly for input device plus a more cosmetic one: o usb endpoints are only 4bit numbers[0...15], the 8th bit only specifies the direction and is set by the snd/rcvxxxpipe() macro Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.238, 2005-03-28 20:22:30-08:00, js@linuxtv.org [PATCH] dvb: support Nova-S rev 2.2 Support for Nova-S rev 2.2 (Gregor Kroesen) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.237, 2005-03-28 20:22:13-08:00, js@linuxtv.org [PATCH] dvb: sparse warnings on one-bit bitfields Remove some sparse warnings on one-bit bitfields. Signed-off-by: Peter Hagervall Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.236, 2005-03-28 20:21:57-08:00, js@linuxtv.org [PATCH] dvb: dibusb: support dtt200u (Yakumo/Typhoon/Hama) USB2.0 device o added native support for the dtt200u (Yakumo/Typhoon/Hama) USB2.0 device o URBs are now submitted, when the actual transfer began, not right from the device plugin (solves a lot of problems) o minor fixes in the dib3000-frontends (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.235, 2005-03-28 20:21:40-08:00, js@linuxtv.org [PATCH] dvb: convert from pci_module_init to pci_register_driver From: http://kerneljanitors.org/TODO o convert from pci_module_init to pci_register_driver Signed-off-by: Christophe Lucas Signed-off-by: Domen Puncer Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.234, 2005-03-28 20:21:23-08:00, js@linuxtv.org [PATCH] dvb: vfree() checking cleanups vfree() checking cleanups. Signed-off by: James Lamanna Signed-off-by: Domen Puncer Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.233, 2005-03-28 20:21:06-08:00, js@linuxtv.org [PATCH] dvb: dibusb: HanfTek UMT-010 fixes HanfTek UMT-010: adapted the pll-programming, the usb-ids and the firmware name to the new firmware (thanks to Sunny Liu from HanfTek) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.232, 2005-03-28 20:20:50-08:00, js@linuxtv.org [PATCH] dvb: support for Technotrend PCI DVB-T patch by Anssi Hannula: add support for Technotrend PCI DVB-T (0x13c2,0x0008, Grundig 29504-401 (LSI L64781 Based) frontend) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.231, 2005-03-28 20:20:33-08:00, js@linuxtv.org [PATCH] dvb: dibusb: remove useless ifdef removed useless ifdef: dvb_register_adapter always takes 3 parameters in this tree (Andreas Oberritter) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.230, 2005-03-28 20:20:17-08:00, js@linuxtv.org [PATCH] dvb: clear up confusion between ids and adapters clear up confusion between ids and adapters (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.229, 2005-03-28 20:19:59-08:00, js@linuxtv.org [PATCH] dvb: frontends: kfree() cleanup kfree(NULL) is safe (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.228, 2005-03-28 20:19:42-08:00, js@linuxtv.org [PATCH] dvb: fix CAMs on Typhoon DVB-S Fix for CAMs on Typhoon DVB-S, where it would constantly reset itself. (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.227, 2005-03-28 20:19:26-08:00, js@linuxtv.org [PATCH] dvb: saa7146: remove duplicate setgpio Remove duplicate setgpio (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.226, 2005-03-28 20:19:10-08:00, js@linuxtv.org [PATCH] dvb: tda10021: fix continuity errors Fix Continuity Errors with tda10021 (slickhenry, Robert Schlabbach) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.225, 2005-03-28 20:18:53-08:00, js@linuxtv.org [PATCH] dvb: dibusb: pll fix o fixed pll frequency calculation for channels > 700 MHz. (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.224, 2005-03-28 20:18:36-08:00, js@linuxtv.org [PATCH] dvb: ttusb_dec: IR support Add IR support added by Peter Beutner Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.223, 2005-03-28 20:18:20-08:00, js@linuxtv.org [PATCH] dvb: OREN or51211, or51132_qam and or51132_vsb firmware download info o add OREN or51211, or51132_qam and or51132_vsb firmware o correct some links Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.222, 2005-03-28 20:18:04-08:00, js@linuxtv.org [PATCH] dvb: dibusb: support nova-t usb ir o added native nova-t support, since only the nova-t firmware works with the nova-t remote control o added keys statically (maybe we should find something to sort this out... cinergyt2 is doing the same) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.221, 2005-03-28 20:17:47-08:00, js@linuxtv.org [PATCH] dvb: support pcHDTV HD2000 o DVB support for the pcHDTV HD2000 card, submitted by Rusty Scott o remove bttv dependency from frontend (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.220, 2005-03-28 20:17:30-08:00, js@linuxtv.org [PATCH] dvb: corrected links to firmware files corrected links to firmware files (reported by Stefan Frings) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.219, 2005-03-28 20:17:14-08:00, js@linuxtv.org [PATCH] dvb: av7110: error handling during attach Janitoring - error handling during attach o av7110_arm_sync(): small helper to factor out some code; o av7110_attach() does not check the status code returned by all the functions is uses; o balance the error path in av7110_attach and have it easy to check. Please check it; o if everything is correctly balanced, device_initialized is not needed anymore in struct av7110; o av7110_detach(): no need to cast a void * pointer; o av7110_detach(): die #ifdef, die ! o change the returned value of av7110_av_exit() as it can't fail; o change the returned value of av7110_ca_init() as it can fail. Removed extraneous casts while are it; o check for failure of vmalloc() in ci_ll_init(). o vfree(NULL) is safe. Signed-off-by: Francois Romieu Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.218, 2005-03-28 20:16:57-08:00, js@linuxtv.org [PATCH] dvb: saa7146: static initialization Static initialization. Signed-off-by: Francois Romieu Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.217, 2005-03-28 20:16:41-08:00, js@linuxtv.org [PATCH] dvb: av7110: fix Oops when av7110_ir_init() failed o don't call av7110_ir_init() if driver initialization failed already due to previous errors (resulted in Oops in out-of-memory conditions) (me) o don't do av7110_ir_exit if init was not done (Kenneth Aafloy) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.216, 2005-03-28 20:16:24-08:00, js@linuxtv.org [PATCH] dvb: ttusb-budget: s/usb_unlink_urb/usb_kill_urb/ patch by Colin Western: s/usb_unlink_urb/usb_kill_urb/ Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.215, 2005-03-28 20:16:07-08:00, js@linuxtv.org [PATCH] dvb: nxt2002: fix max frequency Patch by Taylor Jacob and Tom Dombrosky: There was a typo in the BBTI/B2C2 specs that stated the upper frequency of the air2pc/nxt2002 was 806Mhz, not 860Mhz. Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.214, 2005-03-28 20:15:50-08:00, js@linuxtv.org [PATCH] dvb: refactor sw pid filter to drop redundant code o added index field to struct dvb_demux_feed for having a unique feed id, which can be used for hardware pid filter tables o dibusb: adding the index to struct dvb_demux_feed makes dibusb-pid-filtering redundant o ttusb-budget: struct channel removed in favour of dvbdmxfeed->index (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.213, 2005-03-28 20:15:33-08:00, js@linuxtv.org [PATCH] dvb: cleanups, make stuff static This patch contains the following possible cleanups: o make needlessly global code static o remove the following EXPORT_SYMBOL'ed but unused function: - bt8xx/bt878.c: bt878_find_by_i2c_adap o remove the following unused global functions: - dvb-core/dvb_demux.c: dmx_get_demuxes - dvb-core/dvb_demux.c: dvb_set_crc32 o remove the following unneeded EXPORT_SYMBOL's: - dvb-core/dvb_demux.c: dvb_dmx_swfilter_packet - dvb-core/dvb_demux.c: dvbdmx_connect_frontend - dvb-core/dvb_demux.c: dvbdmx_disconnect_frontend - dvb-core/dvbdev.c: dvb_class Signed-off-by: Adrian Bunk Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.212, 2005-03-28 20:15:16-08:00, js@linuxtv.org [PATCH] dvb: support KWorld/ADSTech Instant DVB-T USB2.0 o added support for KWorld/ADSTech Instant DVB-T USB2.0 (DiB3000M-B) o added deactivation option of the pid parser for the DiB3000M-B (since there are USB2.0 devices and which now have the ability to deliver the complete Transport Stream) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.211, 2005-03-28 20:14:58-08:00, js@linuxtv.org [PATCH] dvb: skystar2: fix MAC address reading fixed MAC address reading (eeprom address to read, was not correctly set, respectively overwritten) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.210, 2005-03-28 20:14:42-08:00, js@linuxtv.org [PATCH] dvb: l64781: email address fix fix marko kohtala's mail address Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.209, 2005-03-28 20:14:26-08:00, js@linuxtv.org [PATCH] dvb: ttusb_dec: use alternative interface to save bandwidth Use alternative interface. Asks for less bandwidth and therefore works with OHCI as well as UHCI (Alex Woods) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.208, 2005-03-28 20:14:09-08:00, js@linuxtv.org [PATCH] dvb: dibusb: increased the number of urbs for usb1.1 devices increased the number of urbs for usb1.1 devices (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.207, 2005-03-28 20:13:53-08:00, js@linuxtv.org [PATCH] dvb: dibusb: debug changes o use own err,info,warn defines, driver description o remove warning about firmware bug (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.206, 2005-03-28 20:13:37-08:00, js@linuxtv.org [PATCH] dvb: dib3000: corrected device naming corrected device naming (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.205, 2005-03-28 20:13:20-08:00, js@linuxtv.org [PATCH] dvb: get_dvb_firmware: new unshield version patch by Mattias Holmlund: support new version of unshield for sp887x firmware extraction (changed cmdline parameters) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.204, 2005-03-28 20:13:04-08:00, js@linuxtv.org [PATCH] dvb: nxt2002: QAM64/256 support patch by Taylor Jacob: Add QAM64/256 Support Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.203, 2005-03-28 20:12:46-08:00, js@linuxtv.org [PATCH] dvb: dibusb: support Hauppauge WinTV NOVA-T USB2 o added support for Hauppauge WinTV NOVA-T USB2 (clone of MOD3000P by DiBcom) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.202, 2005-03-28 20:12:29-08:00, js@linuxtv.org [PATCH] dvb: dibusb readme update dibusb readme update (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.201, 2005-03-28 20:12:13-08:00, js@linuxtv.org [PATCH] dvb: ves1x93: invert_pwm fix fix unhandled invert_pwm option (needed on dbox2 hardware) submitted by Carsten Juttner Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.200, 2005-03-28 20:11:56-08:00, js@linuxtv.org [PATCH] dvb: skystar2: update email address Updated email address. Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.199, 2005-03-28 20:11:40-08:00, js@linuxtv.org [PATCH] dvb: support Activy Budget card support Activy Budget with ALPS BSRU6 tuner submitted by Andreas 'randy' Weinberger. Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.198, 2005-03-28 20:11:24-08:00, js@linuxtv.org [PATCH] dvb: mt352: Pinnacle 300i comments Comment preliminary Pinnacle 300i changes to the mt352 driver. Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.197, 2005-03-28 20:11:08-08:00, js@linuxtv.org [PATCH] dvb: skystar2: remove duplicate pci_release_region() remove duplicated pci_release_region() etc. Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.196, 2005-03-28 20:10:51-08:00, js@linuxtv.org [PATCH] dvb: dibusb: misc. fixes o worked around hw_sleep handling for usb1.1 devices o fixed oops when no frontend was attached (because of usb1.1 timeouts in my debugging sessions) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.195, 2005-03-28 20:10:35-08:00, js@linuxtv.org [PATCH] dvb: dibcom: frontend fixes o corrected the name in driver_desc o removed debug messages and some comments (see dib3000-watch) (Patrick Boettcher) Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.194, 2005-03-28 20:10:18-08:00, js@linuxtv.org [PATCH] dvb: clarify firmware upload messages clarify firmware upload messages Signed-off-by: Johannes Stezenbach Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.193, 2005-03-28 20:10:02-08:00, zippel@linux-m68k.org [PATCH] hfs: unicode decompose support HFS+ specificies that a number of unicode characters must be stored in its decomposed form on disk. This adds the support for the composition and decomposition of such HFS+ file names. There is also an option to turn it off again, in case one managed to create invalid names, which otherwise become inaccessible. Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.192, 2005-03-28 20:09:45-08:00, zippel@linux-m68k.org [PATCH] hfs: add nls support Add full nls support for HFS+. The default is still utf8, but that can be changed now via a mount option. Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.191, 2005-03-28 20:09:28-08:00, zippel@linux-m68k.org [PATCH] hfs: use parse library for mount options Make use of parser library to parse mount options. Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.190, 2005-03-28 20:09:12-08:00, zippel@linux-m68k.org [PATCH] hfs: fix sign problem in hfs_ext_keycmp The key values are unsigned, so comparing them using the difference doesn't always work. (Problem reported by k_guillaume@libertysurf.fr) Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.189, 2005-03-28 20:08:56-08:00, zippel@linux-m68k.org [PATCH] hfs: more bnode error checks Check for errors during reading of bnode pages and report them. Also improve error checks in case bnode validity checks failed. Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.188, 2005-03-28 20:08:39-08:00, zippel@linux-m68k.org [PATCH] hfs: fix umask behaviour This makes umask behaviour more consistent, so it's now also used for new files and the mount options behave like other umask users. (Problem reported by ) Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.187, 2005-03-28 20:08:23-08:00, zippel@linux-m68k.org [PATCH] hfs: free page buffers in releasepage Call try_to_free_buffers() when a page can be released. This behaviour changed during 2.5 and wasn't fixed in hfs while porting it to 2.6. Also remove an unnecessary PageActive() check, only the node ref count is really relevant. Signed-off-by: Roman Zippel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.186, 2005-03-28 20:08:06-08:00, patrick@bitwizard.nl [PATCH] Specialix/IO8 cli() conversion This patch converts all save_flags/restore_flags to the new spin_lock_irqsave/spin_unlock_irqrestore calls, as well as some other 2.6.X cleanups. This allows the "io8+" driver to become SMP safe. The large size of this patch comes mostly from the added debug features. Signed-off-by: Patrick vd Lageweg Signed-off-by: Rogier Wolff Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.185, 2005-03-28 20:07:50-08:00, stsp@aknet.ru [PATCH] au1x00_uart deadlock fix This driver also needs to drop the lock when calling back into the tty layer. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.184, 2005-03-28 20:07:34-08:00, bunk@stusta.de [PATCH] fs/attr.c: fix check after use This patch fixes a check after use found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.183, 2005-03-28 20:07:17-08:00, bunk@stusta.de [PATCH] drivers/telephony/ixj: fix a use after free This patch fixes a use after free found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.182, 2005-03-28 20:07:01-08:00, bunk@stusta.de [PATCH] drivers/block/DAC960.c: fix a use after free This patch fixes a use after free found by the Coverity checker. Controller is used in the Failure path. In the Failure patch, Controller will be freed in the end, so this kfree can simply be deleted. Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.181, 2005-03-28 20:06:44-08:00, gud@eth.net [PATCH] unified spinlock initialization - remove the usage of {SPIN,RW}_LOCK_UNLOCKED as far as possible - add a note to Documentation/spinlocks.txt about the deprecation of the macros {SPIN,RW}_LOCK_UNLOCKED Signed-off-by: Amit Gud Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.180, 2005-03-28 20:06:27-08:00, rbultje@ronald.bitfreak.net [PATCH] bt819 array indexing fix Signed-off-by: Ronald S. Bultje Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.179, 2005-03-28 20:06:09-08:00, bjorn.helgaas@hp.com [PATCH] Netmos parallel/serial/combo support There's a bugzilla entry for this here: http://bugzilla.kernel.org/show_bug.cgi?id=4334 This should fix all the problems I know about with Netmos combo cards: - 9735, 9835, and 9855 are not supported - combo cards with parallel are erroneously claimed by serial driver - serial and parport_serial blindly probe for ports parport_pc: Sort Netmos device IDs, no functional change. parport_serial: Previously supported 9735 and 9835. Add 9745, 9845, 9855, and add init hooks to discover how many serial/parallel ports are actually present (the boards are available in various configs). Add protection for overflow of static tables. quirks: Detect Netmos combo (parallel + serial) cards and change class from SERIAL to OTHER to prevent serial driver from claiming them. 8250: Add init hook to discover the number of serial ports present. This prevents us from poking at unused BARs. pci_ids: Add Netmos 9745, 9845, and sort. Signed-off-by: Bjorn Helgaas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.178, 2005-03-28 20:05:53-08:00, wendyx@us.ibm.com [PATCH] serial: Digi Neo driver This device driver is for the Digi Neo serial adapter. We got several rounds of detail comments from linux community and updated our earlier patches according to comments including Jeff Garzik and Greg KH's. Signed-off-by: Wen Xiong Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.177, 2005-03-28 20:05:36-08:00, patrick@bitwizard.nl [PATCH] generic-serial cli() conversion This patch converts all save_flags/restore_flags to the new spin_lock_irqsave/spin_unlock_irqrestore calls, as well as some other 2.6.X cleanups. This prepares the way for the "io8+", "sx" and "rio" drivers to become SMP safe. Patches for those drivers follow. Signed-off-by: Patrick vd Lageweg Signed-off-by: Rogier Wolff Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.176, 2005-03-28 20:05:21-08:00, rohit.seth@intel.com [PATCH] arch hook for notifying changes in PTE protections bits Recently on IA-64, we have found an issue where old data could be used by apps. The sequence of operations includes few mprotects from user space (glibc) goes like this: 1- The text region of an executable is mmaped using PROT_READ|PROT_EXEC. As a result, a shared page is allocated to user. 2- User then requests the text region to be mprotected with PROT_READ|PROT_WRITE. Kernel removes the execute permission and leave the read permission on the text region. 3- Subsequent write operation by user results in page fault and eventually resulting in COW break. User gets a new private copy of the page. At this point kernel marks the new page for defered flush. 4- User then request the text region to be mprotected back with PROT_READ|PROT_EXEC. mprotect suppport code in kernel, flushes the caches, updates the PTEs and then flushes the TLBs. Though after updating the PTEs with new permissions, we don't let the arch specific code know about the new mappings (through update_mmu_cache like routine). IA-64 typically uses update_mmu_cache to check for the defered flush flag (that got set in step 3) to maintain cache coherency lazily (The local I and D caches on IA-64 are incoherent). DavidM suggeested that we would need to add a hook in the function change_pte_range in mm/mprotect.c This would let the architecture specific code to look at the new ptes to decide if it needs to update any other architectual/kernel state based on the updated (new permissions) PTE values. We have added a new hook lazy_mmu_prot_update(pte_t) that gets called protection bits in PTEs change. This hook provides an opportunity to arch specific code to do needful. On IA-64 this will be used for lazily making the I and D caches coherent. Signed-off-by: David Mosberger Signed-off-by: Rohit Seth Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.175, 2005-03-28 20:05:04-08:00, relf@os2.ru [PATCH] fs/hpfs/*: fix HPFS support under 64-bit kernel The provided patch fixes HPFS filesystem support under 64-bit Linux kernel and closes the bugreport http://bugme.osdl.org/show_bug.cgi?id=4333 The problem is in 'time_t' size which is 8 bytes on 64-bit systems (comparing to 4 bytes on 32-bit systems). The patch introduces local 'time32_t' type of the fixed size 4 and uses it where required. Cc: Mikulas Patocka Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.174, 2005-03-28 20:04:48-08:00, damm@opensource.se [PATCH] module parameter fixes Crystal clear MODULE_PARM_DESC typo fixes for: bt878, edgeport, g_ether, budget, cifs, drm, eepro100 Signed-off-by: Magnus Damm Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.173, 2005-03-28 20:04:30-08:00, neilb@cse.unsw.edu.au [PATCH] nlm: fix f_count leak I can't see any reason for this file->f_count++. Removing it fixes a bug which leaves an exported filesystem busy if a callback for a lock held on that filesystem ever failed. Found by Terence Rokop. Signed-off-by: J. Bruce Fields Signed-off-by: Neil Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.172, 2005-03-28 20:04:14-08:00, mikukkon@gmail.com [PATCH] Fix compile warning in drivers/pnp/resource.c with !CONFIG_PCI drivers/pnp/resource.c:24: warning: `pnp_skip_pci_scan' defined but not used Signed-off-by: Mika Kukkonen (mikukkon@gmail.com) Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.171, 2005-03-28 20:03:58-08:00, hirofumi@mail.parknet.co.jp [PATCH] FAT: Fix msdos ->[ac]{date,time} MSDOS doesn't have ->adate and ->c{date,time}, those should be filled by zero. This fixes my recent changes. Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.170, 2005-03-28 20:03:42-08:00, hirofumi@mail.parknet.co.jp [PATCH] FAT: set MS_NOATIME to msdos MSDOS doesn't have atime, so this sets MS_NOATIME to msdos in order that we don't get unnecessary writes. Signed-off-by: Werner Almesberger Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.169, 2005-03-28 20:03:25-08:00, rth@twiddle.net [PATCH] alpha: elimitate two warnings from gcc4 Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.168, 2005-03-28 20:03:09-08:00, rth@twiddle.net [PATCH] small warning fix for gcc4 Eliminates "warning: type qualifiers ignored on function return type". Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.167, 2005-03-28 20:02:53-08:00, rmk+pcmcia@arm.linux.org.uk [PATCH] pcmcia: clean up suspend Where suspend is implemented, all socket drivers set their socket state to "dead_socket" as the first thing in their suspend ops method. Factor this out to the core PCMCIA code. This allows us to kill some suspend methods, and reduce the references to "dead_socket". Note that the suspend ops method is now optional. Signed-off-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.166, 2005-03-28 20:02:36-08:00, rmk@arm.linux.org.uk [PATCH] Fix PCMCIA resume with card inserted The following patch fixes a minor but annoying issue with PCMCIA resume. When a machine is suspended with an empty socket, and then resumed with a card in the socket, the card is ignored. The following patch adjusts the resume paths such that we detect the card and properly initialise it. Signed-off-by: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.165, 2005-03-28 20:02:19-08:00, jgarzik@pobox.com [PATCH] alpha build fixes Needed this to build Fedora rawhide kernel (2.6.12-rc1 + some patches) on alpha. This is the upstream portion of the build fixes. Signed-off-by: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.164, 2005-03-28 20:02:03-08:00, rddunlap@osdl.org [PATCH] io_remap_pfn_range: convert last callers io_remap_pfn_range() remaining callers: convert all remaining callers of io_remap_page_range() to io_remap_pfn_range(); add io_remap_page_range() to feature-removal-schedule.txt; Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.163, 2005-03-28 20:01:45-08:00, rddunlap@osdl.org [PATCH] io_remap_pfn_range: fix some callers for XEN (from Keir:) I have audited the drivers/ and sound/ directories. Most uses of remap_pfn_range are okay, but there are a small handful that are remapping device memory (mostly AGP and DRM drivers). Of particular driver is the HPET driver, whose mmap function is broken even for native (non-Xen) builds. If nothing else, vmalloc_to_phys should be used instead of __pa to convert an ioremapped virtual address to a valid physical address. The fix in this patch is to remember the original bus address as probed at boot time and to pass this to io_remap_pfn_range. Signed-off-by: Keir Fraser Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.162, 2005-03-28 20:01:29-08:00, rddunlap@osdl.org [PATCH] io_remap_pfn_range: convert sparc callers io_remap_pfn_range(): convert sparc32/64 callers of io_remap_page_range(with 6 args) to io_remap_pfn_range(with 5 args); Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.161, 2005-03-28 20:01:12-08:00, rddunlap@osdl.org [PATCH] io_remap_pfn_range: add for all arch-es This patch introduces a new interface function for mapping bus/device memory: io_remap_pfn_range. This accepts the same parameters as remap_pfn_range and io_remap_page_range but should be used in any situation where the caller is not simply remapping ordinary RAM. For example, when mapping device registers the new function should be used. The distinction between remapping device memory and ordinary RAM is critical for the Xen hypervisor. This patch series also cleans up the remaining users of io_remap_page_range (in particular, the several sparc-specific sections in various drivers that use a special form of io_remap_page_range: an extra argument for SPARC arch.) by converting them to use io_remap_pfn_range(), where io_remap_pfn_range() supports passing as part of the pfn argument. The sparc32 & sparc64 code needs live testing. (from Keir:) I have audited the drivers/ and sound/ directories. Most uses of remap_pfn_range are okay, but there are a small handful that are remapping device memory (mostly AGP and DRM drivers). Of particular driver is the HPET driver, whose mmap function is broken even for native (non-Xen) builds. If nothing else, vmalloc_to_phys should be used instead of __pa to convert an ioremapped virtual address to a valid physical address. The fix in this patch is to remember the original bus address as probed at boot time and to pass this to io_remap_pfn_range. io_remap_pfn_range(): add io_remap_pfn_range() for all arches; add MK_IOSPACE_PFN(), GET_IOSPACE(), and GET_PFN() for all arches but primarily for sparc32/64's extended IO space, sparc: kill the hack of using low bit of to mean write_combine or set side-effect (_PAGE_E) bit; (DaveM suggested that I kill it;) future: convert remaining callers of io_remap_page_range() to io_remap_pfn_range() and deprecate io_remap_page_range(); Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.160, 2005-03-28 20:00:54-08:00, jakub@redhat.com [PATCH] Futex: make futex_wait() atomic again Call get_futex_value_locked in futex_wait with futex hash bucket locked and only enqueue the futex if futex has the expected value. Simplify futex_requeue. Signed-off-by: Jakub Jelinek Acked-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.159, 2005-03-28 20:00:38-08:00, sct@redhat.com [PATCH] ext2/3 file limits to avoid overflowing i_blocks As discussed before, we can overflow i_blocks in ext2/ext3 inodes by growing a file up to 2TB. That gives us 2^32 sectors of data in the file; but once you add on the indirect tree and possible EA/ACL metadata, i_blocks will wrap beyond 2^32. Consensus seemed to be that the best way to avoid this was simply to stop files getting so large that this was a problem in the first place; anything else would lead to complications if a sparse file tried to overflow that 2^32 sector limit while filling in holes. I wrote a small program to calculate the total indirect tree overhead for any given file size, and 0x1ff7fffe000 turned out to be the largest file we can get without the total i_blocks overflowing 2^32. But in testing, that *just* wrapped --- we need to limit the file to be one page smaller than that to deal with the possibility of an EA/ACL block being accounted against i_blocks. So this patch has been tested, at least on ext3, by letting a file grow densely to its maximum size permitted by the kernel; at 0x1ff7fffe000, stat shows the file to have wrapped back exactly to 0 st_blocks, but with the limit at 0x1ff7fffd000, du shows it occupying the expected 2TB-blocksize bytes. Signed-off-by: Stephen Tweedie Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.158, 2005-03-28 20:00:22-08:00, schwidefsky@de.ibm.com [PATCH] posix-cpu-timers and cputime_t divisons. The posix cpu timers introduced code that will not work with an arbitrary type for cputime_t. In particular the division of two cputime_t values broke the s390 build because cputime_t is define as an unsigned long long. The first problem is the division of a cputime_t value by a number of threads. That is a cputime_t divided by an integer. The patch adds another macro cputime_div to the cputime macro regime which implements this type of division and replaces all occurences of a cputime / nthread in the posix cpu timer code. Next problem is bump_cpu_timer. This function is severly broken: 1) In the body of the first if statement a timer->it.cpu.incr.sched is used as the second argument of do_div. do_div expects an unsigned long as "base" parameter but timer->it.cpu.incr.sched is an unsigned long long. If the timer increment ever happens to be >= 2^32 the result is wrong and if the lower 32 bits are zero this even crashes with a fixed point divide exception. 2) The cputime_le(now.cpu, timer->it.cpu.expires.cpu) in the else if condition is wrong. The cputime_le() reads as "now.cpu <= timer->it.cpu.expires.cpu" and the subsequent cputime_ge() reads as "now.cpu >= timer.it.cpu.expires.cpu". That means that the two values needs to be equal to make the body of the second if to have any effect. The first cputime_le should be a cputime_ge. 3) timer->it.cpu.expires.cpu and delta in the else part of the if are of type cputime_t. A division of two cputime_t values is undefined (think of cputime_t as e.g. a struct timespec, that just doesn't work). We could add a primitive for this type of division but we'd end up with a 64 bit division or something even more complicated. The solution for bump_cpu_timer is to use the "slow" division algorithm that does shifts and subtracts. That adds yet another cputime macro, cputime_halve to do the right shift of a cputime value. The next problem is in arm_timer. The UPDATE_CLOCK macro does the wrong thing for it_prof_expires and it_virt_expires. Expanded the macro and added the cputime magic to it_prof/it_virt. The remaining problems are rather simple, timespec_to_jiffies instead of timespec_to_cputime and several cases where cputime_eq with cputime_zero needs to be used instead of "== 0". What still worries me a bit is to use "timer->it.cpu.incr.sched == 0" as check if the timer is armed at all. It should work but its not really clean. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.157, 2005-03-28 20:00:05-08:00, shaohua.li@intel.com [PATCH] Fix oops when inserting ipmi_si module In one of machines in our lab, spmi->addr.register_bit_width is 0 (so the returned address is invalid). Ignoring the check will cause inserting the module oops. Signed-off-by: Li Shaohua Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.156, 2005-03-28 19:59:49-08:00, george@mvista.com [PATCH] Fix POSIX timers expiring before their scheduled time This patch fixes the problem of POSIX timers returning too early due to not accounting for the time starting mid jiffie. Signed-off-by: George Anzinger Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.155, 2005-03-28 19:59:33-08:00, rddunlap@osdl.org [PATCH] cciss: range chcking fix Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.154, 2005-03-28 19:59:17-08:00, rddunlap@osdl.org [PATCH] mtrr: uaccess range checking fix mtrr: prevent copy_from_user(to, from, -1) or (if that should succeed somehow) write to line[-1] (on stack); Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.153, 2005-03-28 19:59:00-08:00, pj@sgi.com [PATCH] cpusets: alloc GFP_WAIT sleep fix The cpuset mems_allowed update code in alloc_pages_current could (in theory) put a task to sleep that didn't allow sleeping (did not have __GFP_WAIT flag set). In the rare circumstance that the current tasks mems_generation is outofdate compared to the tasks cpuset mems_generation, this mems_allowed update code needs to grap cpuset_sem, which can sleep. We avoid this by not trying to update mems_allowed here if we can't sleep (__GFP_WAIT not set). Thanks to Ray Bryant for noticing this. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.152, 2005-03-28 19:58:44-08:00, pj@sgi.com [PATCH] cpusets: mems generation deadlock fix The cpuset code to update mems_generation could (in theory) deadlock on cpuset_sem if it needed to allocate some memory while creating (mkdir) or removing (rmdir) a cpuset, so already held cpuset_sem. Some other process would have to mess with this tasks cpuset memory placement at the same time. We avoid this possible deadlock by always updating mems_generation after we grab cpuset_sem on such operations, before we risk any operations that might require memory allocation. Thanks to Jack Steiner for noticing this. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.151, 2005-03-28 19:58:26-08:00, randolph@tausq.org [PATCH] Missing set_fs() calls around kernel syscall Found by sparse... since we are passing kernel param to a syscall handler, we need to do the set_fs() wrappers. Signed-off-by: Randolph Chung Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.150, 2005-03-28 19:58:10-08:00, prasanna@in.ibm.com [PATCH] kprobes: incorrect spin_unlock_irqrestore() call in register_kprobe() register_kprobe() routine was calling spin_unlock_irqrestore() wrongly. This patch removes unwanted spin_unlock_irqrestore() call in register_kprobe() routine. Signed-off-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.149, 2005-03-28 19:57:54-08:00, pmeda@akamai.com [PATCH] pipe: save one pipe page Save one page in pipe writev without incuring additional cost (just that ampersand operator). Signed-off-by: Prasanna Meda Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.148, 2005-03-28 19:57:38-08:00, matthew@wil.cx [PATCH] New console flag: CON_BOOT CON_BOOT is like early printk in that it allows for output really early on. It's better than early printk because it unregisters automatically when a real console is initialised. So if you don't get consoles registering in console_init, there isn't a huge delay between the boot console unregistering and the real console starting. This is the case on PA-RISC where we have serial ports that aren't discovered until the PCI bus has been walked. I think all the current early printk users could be converted to this scheme with a minimal amount of effort. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.147, 2005-03-28 19:57:22-08:00, roland@topspin.com [PATCH] InfiniBand: remove unsafe use of in_atomic() Using in_atomic() to decide between GFP_KERNEL and GFP_ATOMIC is not safe (it doesn't work if CONFIG_PREEMPT=n). Change to just always allocating with GFP_ATOMIC, since we don't know if we can sleep or not. Signed-off-by: Roland Dreier Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.146, 2005-03-28 19:57:05-08:00, mikpe@user.it.uu.se [PATCH] drivers/char/isicom.c gcc4 fix Fix two array-of-incomplete-type errors from gcc4 in isicom.c. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.145, 2005-03-28 19:56:49-08:00, bunk@stusta.de [PATCH] kill drivers/cdrom/mcd.c As described in my patch that marked this obsolete driver as BROKEN, this patch completely removes it. Signed-off-by: Adrian Bunk Acked-by: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.144, 2005-03-28 19:56:31-08:00, christoph@lameter.com [PATCH] Per cpu irq stat The definition of the irq_stat as an array means that the individual elements of the irq_stat array are located on one NUMA node requiring internode traffic to access irq_stat from other nodes. This patch makes irq_stat a per_cpu variable which allows most accesses to be local. Signed-off-by: Christoph Lameter Signed-off-by: Shai Fultheim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.143, 2005-03-28 19:56:15-08:00, clameter@sgi.com [PATCH] Exports to enable clock driver modules The following exports are necessary to allow loadable modules to define new clocks. Without these the mmtimer driver cannot be build correctly as a module (there is another mmtimer specific fix necessary to get it to build properly but that will be a separate patch): Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.142, 2005-03-28 19:55:59-08:00, hong.liu@intel.com [PATCH] fix mmap() return value to conform to POSIX POSIX said: mmap() should return [EOVERFLOW] if the file is a regular file and the value of off + len exceeds the offset maximum established in the open file description associated with fildes. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.141, 2005-03-28 19:55:43-08:00, gordon.jin@intel.com [PATCH] fix mmap() return value to conform POSIX This patch fixes 2 return values in mmap() to conform POSIX spec: [EINVAL] The value of len is zero. [ENOMEM] MAP_FIXED was specified, and the range [addr,addr+len) exceeds that allowed for the address space of a process; or, if MAP_FIXED was not specified and there is insufficient room in the address space to effect the mapping. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.140, 2005-03-28 19:55:27-08:00, ashok.raj@intel.com [PATCH] Fix irq_affinity write from /proc for ia64 Made GENERIC_HARDIRQ mechanism work for ia64 and CPU hotplug. When write to /proc/irq is handled it is not appropriate to perform set_rte immediatly, since there is a race when the interrupt is asserted while the re-program is happening. Hence such programming is only safe when we do the re-program at the time of servicing an interrupt. This got broken when GENERIC_HARDIRQ got introduced for ia64. - added CONFIG_PENDING_IRQ so default /proc/irq write handler can do the right thing. TBD: We currently dont handle redirectable hint either in the display, or when we handle writes to /proc/irq/XX/smp_affinity. We need an arch specific way to account for the presence of "r" hint when we handle the proc write. Signed-off-by: Ashok Raj Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.139, 2005-03-28 19:55:10-08:00, joern@wohnheim.fh-wedel.de [PATCH] checkstack: fix sort misbehavior for long function names Fix sort behavior when long names are encountered. The previour regular expression depended on a tab to find the size in a string. For long names, this tab no longer exists, so it is smarter to check for a colon instead. For the kernel, this change shouldn't make a difference. But people started using the same script for c++ code in other projects. Signed-off-by: Jörn Engel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.138, 2005-03-28 19:54:54-08:00, gordon.jin@intel.com [PATCH] fix mprotect() with len=(size_t)(-1) to return -ENOMEM This patch fixes a corner case in sys_mprotect(): Case: len is so large that will overflow to 0 after page alignment. E.g. len=(size_t)(-1), i.e. 0xff...ff. Expected result: POSIX spec says it should return -ENOMEM. Current result: len is aligned to 0, then treated the same as len=0 and return success. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.137, 2005-03-28 19:54:38-08:00, janitor@sternwelten.at [PATCH] sedlbauer: eliminate bad section references Fix sedlbauer section references: convert __initdata to __devinitdata. Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 0000235f R_386_32 .init.data Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 0000236e R_386_32 .init.data Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 0000238d R_386_32 .init.data Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 000023b2 R_386_32 .init.data Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 00002417 R_386_32 .init.data Error: ./drivers/isdn/hisax/sedlbauer.o .text refers to 0000244c R_386_32 .init.data Signed-off-by: maximilian attems Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.136, 2005-03-28 19:54:21-08:00, janitor@sternwelten.at [PATCH] hfc_sx: eliminate bad section references Fix hfc_sx section references: convert __initdata to __devinitdata. Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 0000204d R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 0000205c R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 00002082 R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 0000209f R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 00002114 R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 0000211c R_386_32 .init.data Error: ./drivers/isdn/hisax/hfc_sx.o .text refers to 0000212e R_386_32 .init.data Signed-off-by: maximilian attems Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.135, 2005-03-28 19:54:05-08:00, janitor@sternwelten.at [PATCH] elsa eliminate bad section references Fix elsa section references: convert __initdata to __devinitdata. Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003d28 R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003d37 R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003d56 R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003d77 R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003ddb R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003e0e R_386_32 .init.data Error: ./drivers/isdn/hisax/elsa.o .text refers to 00003e20 R_386_32 .init.data Signed-off-by: maximilian attems Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.134, 2005-03-28 19:53:49-08:00, janitor@sternwelten.at [PATCH] teles3: eliminate bad section references Fix teles3 section references: convert __initdata to __devinitdata. Error: ./drivers/isdn/hisax/teles3.o .text refers to 000011ab R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 000011ba R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 000011e0 R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 000011fd R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 0000128c R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 00001294 R_386_32 .init.data Error: ./drivers/isdn/hisax/teles3.o .text refers to 000012a6 R_386_32 .init.data Signed-off-by: maximilian attems Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.133, 2005-03-28 19:53:31-08:00, janitor@sternwelten.at [PATCH] w6692: eliminate bad section references Fix w6692 section references: added __init to W6692Version(). Error: ./drivers/isdn/hisax/w6692.o .text refers to 0000002f R_386_32 .init.data Signed-off-by: maximilian attems Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.132, 2005-03-28 19:53:15-08:00, solar@openwall.com [PATCH] Enable gcc warnings for vsprintf/vsnprintf with "format" attribute Extend the gcc printk format-string checking. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.131, 2005-03-28 19:52:59-08:00, akpm@osdl.org [PATCH] revert recent gconfig changes Revert a 2.6.11 patch "make gconfig work with gtk-2.4". It causes the cute little pixmap buttons to not appear any more. Hopefully this will motivate someone to have another attempt at fixing gconf for gtk-2.4. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.130, 2005-03-28 19:52:42-08:00, domen@coderock.org [PATCH] cdrom/cdu31a: use wait_event Use wait_event instead of sleep_on. Also, remove two unused variables. Signed-off-by: Domen Puncer Acked-by: Ondrej Zary Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.129, 2005-03-28 19:52:26-08:00, domen@coderock.org [PATCH] cdrom/cdu31a: locking fixes Use semaphores instead of sleep_on*. Stolen from patch: http://lkml.org/lkml/2004/12/18/107 from Ondrej Zary Signed-off-by: Domen Puncer Acked-by: Ondrej Zary Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.128, 2005-03-28 19:52:09-08:00, domen@coderock.org [PATCH] cdrom/cdu31a: cleanups Pretty trivial cleanups: - reordered #includes - improved some printk's (note: this actually enabled some debug printk's) - removed ()'s from returns - removed SONY_POLL_EACH_BYTE, as grep doesn't find it anywhere else - removed panic() as it can't happen. Signed-off-by: Domen Puncer Acked-by: Ondrej Zary Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.127, 2005-03-28 19:51:53-08:00, mingo@elte.hu [PATCH] break_lock fix lock->break_lock is set when a lock is contended, but cleared only in cond_resched_lock. Users of need_lockbreak (journal_commit_transaction, copy_pte_range, unmap_vmas) don't necessarily use cond_resched_lock on it. So, if the lock has been contended at some time in the past, break_lock remains set thereafter, and the fastpath keeps dropping lock unnecessarily. Hanging the system if you make a change like I did, forever restarting a loop before making any progress. And even users of cond_resched_lock may well suffer an initial unnecessary lockbreak. There seems to be no point at which break_lock can be cleared when unlocking, any point being either too early or too late; but that's okay, it's only of interest while the lock is held. So clear it whenever the lock is acquired - and any waiting contenders will quickly set it again. Additional locking overhead? well, this is only when CONFIG_PREEMPT is on. Since cond_resched_lock's spin_lock clears break_lock, no need to clear it itself; and use need_lockbreak there too, preferring optimizer to #ifdefs. Signed-off-by: Hugh Dickins Signed-off-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.126, 2005-03-28 19:51:36-08:00, petero2@telia.com [PATCH] DVD-RAM support for pktcdvd This patch makes it possible to use the packet writing driver with DVD-RAM discs. The pktcdvd driver is not needed for writing to DVD-RAM discs but it can improve write performance. Polgár István reports: I wrote 178716Kb data to DVD-RAM without pktcdvd driver within 4.54 minutes. With pktcdvd driver it took me 2.33 minutes. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.125, 2005-03-28 19:51:19-08:00, petero2@telia.com [PATCH] Use __init and __exit in pktcdvd This patch adds __init and __exit annotations to the pktcdvd driver. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.124, 2005-03-28 19:51:02-08:00, dsw@gelato.unsw.edu.au [PATCH] Stallion driver module clean up These two patches continue the work that Wayne Meissner started and are against the current bk tree. These patches allow the stallion driver to be built-in and loaded at boot time, the current #ifdef MODULE only allows the init code to be included if compiled as a module. Tested for compile, boot and running on our console server as module and built-in. Signed-off-by Darren Williams Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.123, 2005-03-28 19:50:45-08:00, miklos@szeredi.hu [PATCH] Can't unmount bad inode This patch fixes a problem when a inode which is the root of a mount becomes bad (make_bad_inode()). In this case follow_link will return -EIO, so the name resolution fails, and umount won't work. The solution is just to remove the follow_link method from bad_inode_ops. Any filesystem operation (other than unmount) will still fail, since every other method returns -EIO. A test case for this is: 1) export an smbfs on host A and mount the share on host B 2) create directory X on A under the exported directory 3) bind mount X to Y on B (Y need not be under the share) 4) remove directory X, and create regular file X (same name) on A 5) stat X on B, this will make X a bad inode (file type changed) 6) umount Y Without the patch applied, umount won't succeed, and a reboot is necessary to get rid of the mount. With the patch applied, umount will succeed. The same is true for any filesystem which uses make_bad_inode() to mark an existing inode bad (NFS, SMBFS, FUSE, etc...). Signed-off-by: Miklos Szeredi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.122, 2005-03-28 19:50:29-08:00, richtera@us.ibm.com [PATCH] s390: claw network device driver Add support for claw network devices. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.121, 2005-03-28 19:50:10-08:00, pavlic@de.ibm.com [PATCH] s390: qeth tcp segmentation offload Add support for TCP Segmentation Offload to the qeth network driver. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.120, 2005-03-28 19:49:52-08:00, thoss@de.ibm.com [PATCH] s390: qeth blkt tuning Add support for qeth BLKT tuning, the OSA inbound blocking or packing algorithm. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.119, 2005-03-28 19:49:36-08:00, ptiedem@de.ibm.com [PATCH] s390: qeth 1920 device suppor cio/qeth changes: - Add new api to common i/o layer for qeth 1920 device support. - Add 1920 device support to qeth. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.118, 2005-03-28 19:49:20-08:00, pavlic@de.ibm.com [PATCH] s390: qeth layer2 fixes qeth network driver changes: - Clear qeth header struct in qeth_fill_header. - Drop broadcast packets in qeth_send_packet if the card can't do broadcasts. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.117, 2005-03-28 19:49:03-08:00, wein@de.ibm.com [PATCH] s390: dasd preferred path support Add code to the dasd driver to select the I/O path based on the path preferences of the device. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.116, 2005-03-28 19:48:46-08:00, bstroesser@fijitsu-siemens.com [PATCH] s390: signal stack bug If a signal handler is set to use the signal stack (SA_ONSTACK), but the signal stack is disabled, the signal frame should be written to the current stack without stack switching. The reason for the bug is get_sigframe() using on_sig_stack() instead of sas_ss_flags(), which would be ok. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.115, 2005-03-28 19:48:30-08:00, schwidefsky@de.ibm.com [PATCH] s390: kernel faults Correct check for user space faults. If the failing address space is in the secondary address space and uaccess has been switched to KERNEL_DS with set_fs, check_user_space() erroneously returns 1. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.114, 2005-03-28 19:48:13-08:00, vda@port.imtp.ilyichevsk.odessa.ua [PATCH] s390: swapped memset arguments. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.113, 2005-03-28 19:47:57-08:00, blaisorblade@yahoo.it [PATCH] uml: fix "cond. expr. as lvalues" warning Gcc 3.4.3 (and probably any 3.4) emits some "deprecation" warnings currently about usages of CHOOSE_MODE, since the below syntax has been deprecated: (a ? foo: bar) = foobar; which often results from expansion of: CHOOSE_MODE(foo, bar) = foobar; So add an additional __CHOOSE_MODE syntax for users which need to get a lvalue, which uses (a ? &foo : &bar) inside a function, casts the result to the correct type and dereference the pointer, and use it where needed (i.e. in ) The patch builds and runs correctly (this has been tested on i386 only, not on x86_64), and removes all the warnings. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.112, 2005-03-28 19:47:41-08:00, blaisorblade@yahoo.it [PATCH] uml: real fix for __gcov_init symbols Correctly export __gcov_init for cases where it's needed, by adding a weak definition for the case when GCC does not define this symbol and letting it being overriden by the real definition when GCC defines it (recent ones). Can't be implemented as a test on GCC version because SuSE has a crippled GCC, declared as 3.3.4 but having a lot of backported features. Also, since gcc 3.4.3 requires profiling options even during linking, add profiling options to final link stage. Signed-off-by: Paolo 'Blaisorblade' Giarrusso CC: Anton Altaparmakov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.111, 2005-03-28 19:47:24-08:00, blaisorblade@yahoo.it [PATCH] uml: add kconfig debug deps Add some needed dependencies for some debug options and hide the MAGIC_SYSRQ option for UML, since it displays this option in another menu. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.110, 2005-03-28 19:47:08-08:00, blaisorblade@yahoo.it [PATCH] uml - kbuild: link cmd Use the cmd_* syntax to define a command for symlink creations in UML (to reuse files from other archs). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.109, 2005-03-28 19:46:51-08:00, blaisorblade@yahoo.it [PATCH] uml: factor out common code in user-obj handling With Vadim Abrossimov *) Handle USER_OBJS through the general Kbuild infrastructure; the trick we use is to change c_flags only for USER_OBJS. This ain't at all worse than the previous kludgy solution, enables us to use a better dependency handling and to support MODVERSIONS. And it is UML-specific, as a bonus. So, no "it ain't clean enough" reasoning is allowed to hold this patch until you find a better solution. Leaving there the current broken code is not accepted. *) Move similar definitions from Makefiles to the newly created arch/um/scripts/Makefile.rules and include it everywhere needed. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.108, 2005-03-28 19:46:33-08:00, blaisorblade@yahoo.it [PATCH] Uml: little build fixes From: Al Viro Easier parts from "cross-build" (or "UML-kbuild") patch from Al Viro: *) abuses of host cc/ld/objcopy/paths are gone *) some #include path fixes *) other little abuses fixed *) remove LIBC_DIR var, ask gcc where libc.a is placed. This creates no problem so can be merged very likely. Note: only tested on i386, give a run on x86-64. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: "Vadim Abrossimov" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.107, 2005-03-28 19:46:15-08:00, blaisorblade@yahoo.it [PATCH] uml: disable more hardware kconfig opt and rename USERMODE to UML Disable some hardware-only configuration options when configuring for ARCH=um. By the way, we rename CONFIG_USERMODE to CONFIG_UML, as requested some time ago by the UML maintainer Jeff Dike. We also update defconfig as a consequence of all this. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.106, 2005-03-28 19:45:58-08:00, blaisorblade@yahoo.it [PATCH] uml: extend cmd line limits From: "Catalin(ux aka Dino) BOIE" , Paolo 'Blaisorblade' Giarrusso , Jeff Dike Increase UML command line size. And fix a crash from passing an overly-long command line to UML. XXX: check that init can handle 128 params and 128 env. var. The original patch set this limit to 256, but it seems me too much. Think! Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.105, 2005-03-28 19:45:42-08:00, blaisorblade@yahoo.it [PATCH] uml: cpu_relax fix Use rep_nop instead of barrier for cpu_relax, following $(SUBARCH)'s doing that (i.e. i386 and x86_64). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.104, 2005-03-28 19:45:26-08:00, blaisorblade@yahoo.it [PATCH] uml: fix compile Quick compile fix for i386-only change (which interacts with UML since we include headers from include/asm-$(SUBARCH)), which keeps the old behaviour and hence should cause no problems. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.103, 2005-03-28 19:45:09-08:00, blaisorblade@yahoo.it [PATCH] uml: cope with uml_net security fix Pass the interface to close as an fd besides that by name... passing it by name does not work with newer uml_net because that allows to ifconfig down arbitrary interfaces, while if you have an open fd to the SLIP interface it means you have full access to it (and thus can close it). Passing only by fd does not work with older utilities, so we do both things (which does not hurt). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.102, 2005-03-28 19:44:51-08:00, takata@linux-m32r.org [PATCH] m32r_sio driver update I made a patch to update m32r_sio driver. Compile checked and tested on a M32700UT eva board with SMP kernel. It looks working... __register_m32r_sio, register_m32r_sio, unregister_m32r_sio are still remained. I'm going to take a look them after. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.101, 2005-03-28 19:44:35-08:00, takata@linux-m32r.org [PATCH] m32r: Fix M32102 I-cache invalidation This patch fixes I-cache invalidation operation for M32102 chip, which is one of m32r MMU-less targets. Before this fix, the m32r kernel for M32102 chip missed I-cache invalidation operation and switched off I-cache unintentionally. This bug caused awful performance degradation. Signed-off-by: Mamoru Sakugawa Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.100, 2005-03-28 19:44:19-08:00, takata@linux-m32r.org [PATCH] m32r: Update MMU-less support #3 This patch is for updating m32r's MMU-less support. * arch/m32r/boot/compressed/Makefile: Use m32r-elf-gcc for MMU-less targets; change ELF object format from elf32-m32r-linux to elf32-m32r for !CONFIG_MMU. * arch/m32r/boot/compressed/head.S: Set up cache for M32102 chip. * arch/m32r/boot/setup.S: ditto. * arch/m32r/kernel/module.c: Module support for !CONFIG_MMU. Signed-off-by: Kazuhiro Inaoka Signed-off-by: Hayato Fujiwara Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.99, 2005-03-28 19:44:02-08:00, takata@linux-m32r.org [PATCH] m32r: Update MMU-less support #2 This patch is for updating m32r's MMU-less support. * arch/m32r/boot/compressed/m32r_sio.c: - Fix serial output routine * include/asm-m32r/mmu.h: - Update mm_context_t definition Signed-off-by: Yoshinori Sato Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.98, 2005-03-28 19:43:45-08:00, takata@linux-m32r.org [PATCH] m32r: Update MMU-less support #1 This patch is for updating m32r's MMU-less support. * arch/m32r/kernel/entry.s: - Fix syscall table for !CONFIG_MMU * arch/m32r/kernel/traps.c: - Fix EIT vector setup routine for !CONFIG_MMU Signed-off-by: Naoto Sugai Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.97, 2005-03-28 19:43:28-08:00, pavel@ucw.cz [PATCH] swsusp: kill swsusp_restore This kills swsusp_resume; it should be arch-neutral but some i386 code sneaked in. And arch-specific code is better done in assembly anyway. Plus it fixes memory leaks in error paths. Signed-off-by: Rafael J. Wysocki Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.96, 2005-03-28 19:43:12-08:00, pavel@ucw.cz [PATCH] swsusp: small updates This kills unused macro and write-only variable, and adds messages where something goes wrong with suspending devices. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.95, 2005-03-28 19:42:55-08:00, pavel@ucw.cz [PATCH] pm: remove obsolete pm_* from vt.c Those functions are useless these days, and should be gone. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.94, 2005-03-28 19:42:39-08:00, pavel@ucw.cz [PATCH] suspend-to-ram: update video.txt with more systems This updates video.txt documentation. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.93, 2005-03-28 19:42:22-08:00, pavel@ucw.cz [PATCH] swsusp: Add missing refrigerator calls This adds few more places where it is possible freeze kernel threads. From: Nigel Cunningham Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.92, 2005-03-28 19:42:04-08:00, rth@twiddle.net [PATCH] alpha spinlock.h update Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.91, 2005-03-28 19:41:48-08:00, rddunlap@osdl.org [PATCH] kernel-parameters: IA-32/X86-64 cleanups Convert "kernel-parameters.txt" to use IA-32 in place of x86 and X86-64 in place of x86_64, to be in line with other architecture documentation conventions. Add reference to Documentation/x86_64/boot-options.txt . Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.90, 2005-03-28 19:41:32-08:00, blaisorblade@yahoo.it [PATCH] x86-64: forgot asmlinkage on sys_mmap Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.89, 2005-03-28 19:41:15-08:00, kenneth.w.chen@intel.com [PATCH] x86_64: hugetlb fix x86-64 may have to allocate a bunch of upper levels of pagetables, and those allocations may fail. When they do, unmap_hugepage_range() needs to be able to clean up after them. Acked-by: William Lee Irwin III Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.88, 2005-03-28 19:40:59-08:00, blaisorblade@yahoo.it [PATCH] x86_64: remove old decl (trivial) vm_force_exec32 and friends were still alive on 2.6.9 release, but now (and in 2.6.10) they seem deleted. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.87, 2005-03-28 19:40:43-08:00, blaisorblade@yahoo.it [PATCH] x86-64: kconfig typo Trivial typo: default off instead of default n in kbuild. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.86, 2005-03-28 19:40:26-08:00, ak@suse.de [PATCH] x86_64: Clean up the IOMMU initialisation a bit Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.85, 2005-03-28 19:40:09-08:00, ak@suse.de [PATCH] x86_64: Fix missing delay when the TSC counter just overflowed Fix missing delay when the TSC counter just overflowed Noticed by Jan Beulich Signed-off-by: Andi Kleen Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.84, 2005-03-28 19:39:53-08:00, ak@suse.de [PATCH] x86_64: Fix gcc 3.4 warning in bitops.c Move include down to avoid: arch/x86_64/lib/bitops.c:18: warning: `find_first_zero_bit' declared inline after being called Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.83, 2005-03-28 19:39:37-08:00, ak@suse.de [PATCH] x86_64: Fix exception stack detection during backtraces The test in in_exception_stack was done the wrong way round, which lead to incorrect exception stack detection in the kernel backtracer. Also fix a off by one in the test. Noticed by Jan Beulich Signed-off-by: Andi Kleen Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.82, 2005-03-28 19:39:20-08:00, ak@suse.de [PATCH] x86_64: Fix CONFIG_PREEMPT Fix a pretty bad bug in the x86-64 preempt support. The test for interrupts off was done the wrong way round, which would lead to kernel preemption only when interrupts were off. Found by Jan Beulich. Signed-off-by: Andi Kleen Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.81, 2005-03-28 19:39:05-08:00, ak@suse.de [PATCH] x86_64: Fix SMP fallback to UP When falling back to UP on a SMP kernel make sure cpu_sibling_map is always initialized. Otherwise scheduler_tick eventually runs into incompletely initialized sched domains and oopses. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.80, 2005-03-28 19:38:48-08:00, ak@suse.de [PATCH] x86_64: Don't enable interrupts in oopses unconditionally The original reason for it was to allow backwards scrolling on the console, but that didn't work for other reasons anyways. And in some cases the early interrupt enabling can lead to a nested oops. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.79, 2005-03-28 19:38:32-08:00, ak@suse.de [PATCH] x86_64: Only free PMDs and PUDs after other CPUs have been flushed This avoids a race on AMD systems where other CPUs could speculatively follow an already freed page table. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.78, 2005-03-28 19:38:15-08:00, ak@suse.de [PATCH] x86_64: Change the y2069 bug in the RTC timer code to be a y2100 bug. That's ok since x86-64 systems only exist since 2001. More is unfortunately not possible because the RTC CMOS only supports two decimal digits. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.77, 2005-03-28 19:37:59-08:00, ak@suse.de [PATCH] x86_64: Fix LDT descriptor Fix bug introduced with the TLS system calls in 2.5. The LDT descriptor needs two entries, not one. It would overlap into the TLS range, which means setting an LDT would corrupt the first TLS descriptor. Noticed by Jan Beulich Signed-off-by: Andi Kleen Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.76, 2005-03-28 19:37:42-08:00, ak@suse.de [PATCH] x86_64: Always reload CR3 completely when a lazy MM thread drops a MM. Always reload CR3 completely when a lazy MM thread drops a MM. This avoids keeping stale mappings around in the TLB that could be run into by the CPU by itself (e.g. during prefetches). Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.75, 2005-03-28 19:37:27-08:00, ak@suse.de [PATCH] x86_64: Minor fix to TLB flush IPI Minor fix to TLB flush IPI Fix minor harmless bug. When doing a flush IPI we used 0xffffffff as the catch all wildcard, which is a valid address in a 64bit process. When this page was flushed the IPI handler would do a full flush by mistake instead of just flushing the page. Use -1ULL instead. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.74, 2005-03-28 19:37:10-08:00, ak@suse.de [PATCH] x86_64: Fix NMI RTC access race Fix NMI RTC access race This patch fixes a race between the CMOS clock setting and the NMI code. The NMI code indiscriminatly sets index registers and values in the same place the CMOS clock is set. If you are setting the CMOS clock and an NMI occurs, Bad values could be written to or read from the CMOS RAM, or the NMI operation might not occur correctly. Resetting the NMI is not required on x86_64 (in fact, it should not be done according to the ICH5 documentation). This patch simply removes the useless code. Signed-off-by: Corey Minyard Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.73, 2005-03-28 19:36:54-08:00, ak@suse.de [PATCH] x86_64: resume PIT for x86_64 resume PIT for x86_64 Signed-off-by: Luming Yu Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.72, 2005-03-28 19:36:37-08:00, ak@suse.de [PATCH] x86_64: mem=XXX will now limit kernel memory to XXX instead of XXX+1MB mem=XXX will now limit kernel memory to XXX instead of XXX+1MB Signed-off-by: Suresh Siddha Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.71, 2005-03-28 19:36:20-08:00, ak@suse.de [PATCH] x86_64: Fix formatting and white space in signal code Fix formatting and white space in signal code No functional changes. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.70, 2005-03-28 19:36:04-08:00, ak@suse.de [PATCH] x86_64: Fix special ISA case in iounmap() Fix special ISA case in iounmap(), from i386. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.69, 2005-03-28 19:35:48-08:00, ak@suse.de [PATCH] x86_64: Cleanups in new backtrace code in oprofile Remove 4/4 support code. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.68, 2005-03-28 19:35:31-08:00, ak@suse.de [PATCH] x86_64: Include PCI-Express configuration Include PCI-Express configuration Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.67, 2005-03-28 19:35:15-08:00, ak@suse.de [PATCH] x86_64: Work around Tyan BIOS MTRR initialization bug. Work around Tyan BIOS MTRR initialization bug. Some Tyan AMD BIOS don't initialize the first fixed range MTRR, which causes it to contain random bogus values. When the MTRR tries to duplicate the MTRR state to other CPUs at startup it oopses because of this. This patch works around this by catching exception while setting MTRRs. It would be better to validate all fixed range MTRRs and fix them, but that would be very complicated code. This simple hack seems to work too (except that the first 64k of physical memory are likely uncached). A BIOS update fixes that. Signed-off-by: Andi Kleen Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.66, 2005-03-28 19:34:59-08:00, ak@suse.de [PATCH] x86_64: Give out of line get_user better calling conventions Give out of line get_user better calling conventions Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.65, 2005-03-28 19:34:42-08:00, ak@suse.de [PATCH] x86_64: Move put_user out of line Move put_user out of line. Generates smaller code. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.64, 2005-03-28 19:34:25-08:00, ak@suse.de [PATCH] x86_64: Remove stale unused file Remove stale unused file Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.63, 2005-03-28 19:34:09-08:00, ak@suse.de [PATCH] x86_64: Remove unused and broken code in io.h Remove unused and broken code in io.h Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.62, 2005-03-28 19:33:53-08:00, ak@suse.de [PATCH] x86_64: Remove noisy printk in K8 bus detection code Remove noisy printk in K8 bus detection code Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.61, 2005-03-28 19:33:37-08:00, ak@suse.de [PATCH] x86_64: Remove obsolete comments in vsyscall.c and fix some others. Remove obsolete comments in vsyscall.c and fix some others. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.60, 2005-03-28 19:33:20-08:00, mikpe@csd.uu.se [PATCH] x86_64: fix vsyscall.c syntax error arch/x86_64/kernel/vsyscall.c:193: error: syntax error before 'vsyscall_sysctl_change' Signed-off-by: Mikael Pettersson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.59, 2005-03-28 19:33:04-08:00, ak@suse.de [PATCH] x86_64: Nop out system call instruction in vsyscall page when not needed Nop out system call instruction in vsyscall page when not needed. This may make some attacks more difficult. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.58, 2005-03-28 19:32:48-08:00, ak@suse.de [PATCH] x86_64: Fix indentation in vsyscall.c. No functional changes. Fix indentation in vsyscall.c. No functional changes. Needed for followup patches. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.57, 2005-03-28 19:32:32-08:00, ak@suse.de [PATCH] x86_64: Remove never used obsolete file Remove never used obsolete file. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.56, 2005-03-28 19:32:15-08:00, ak@suse.de [PATCH] x86_64: Move HPET selection into processor specific options Move HPET selection into processor specific options. Pointed out by Pavel Roskin Fix the help text up a bit. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.55, 2005-03-28 19:31:58-08:00, ak@suse.de [PATCH] x86_64: Fix compilation with CONFIG_PROC_FS=n Fix compilation with CONFIG_PROC_FS=n Pointed out by Randy Dunlap Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.54, 2005-03-28 19:31:42-08:00, ak@suse.de [PATCH] x86_64: Busses array is only indexed with a 8bit value, doesn't make sense Busses array is only indexed with a 8bit value, doesn't make sense to make it bigger. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.53, 2005-03-28 19:31:26-08:00, ak@suse.de [PATCH] x86_64: Add an 64bit entry path for exec Add an 64bit entry path for exec. From: Also fix up the entry address in vmlinux. >> I have moved most of what was code duplication until after we are in 64bit mode. And I have reduced the amount of code in the 32bit entry path a little bit. And to keep things maintainable we now always execute the full 64bit entry point. << Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.52, 2005-03-28 19:31:09-08:00, ak@suse.de [PATCH] x86_64: Add new AMD cpuid flags to cpuinfo Add new AMD cpuid flags to cpuinfo Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.51, 2005-03-28 19:30:53-08:00, ak@suse.de [PATCH] x86_64: Update defconfig Update defconfig Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.50, 2005-03-28 19:30:36-08:00, juhl-lkml@dif.dk [PATCH] rename FPU_*verify_area to FPU_*access_ok Since verify_area is going the way of the Dodo soon it seems resonable to no longer refer to it in wrapper functions/macros. FPU_verify_area and FPU_code_verify_area have already been converted to call access_ok so now seems a good time to rename them. This patch makes no functional changes at all. Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.49, 2005-03-28 19:30:20-08:00, rddunlap@osdl.org [PATCH] i386: add kstack=N option (from x86_64) Add "kstack=N" boot option for IA-32 (from x86_64). Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.48, 2005-03-28 19:30:04-08:00, pazke@donpac.ru [PATCH] es7000 dmi cleanup This moves es7000_plat global variable out of DMI code. Signed-off-by: Andrey Panin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.47, 2005-03-28 19:29:48-08:00, jason@rightthere.net [PATCH] ES7000 Legacy Mappings Update This update only affects Unisys' ES7000 machines. The patch reflects a change needed to determine which generation of ES7000 is currently running. The next generation of ES7000s will have conventional legacy support so the patch accommodates for this. This patch has been tested and verified on both an authentic 5xx ES7000 box and the next generation ES7000 box. Signed-off-by: Natalie Protasevich Signed-off-by: Jason Davis Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.46, 2005-03-28 19:29:31-08:00, manfred@colorfullife.com [PATCH] fix put_user for 80386 Linus noticed that put_user doesn't to the manual page table lookup that is required for cpus without a working WP flag. The solution is simple: if CONFIG_X86_WP_WORKS_OK is not set, then the put_user macros must call __copy_to_user_ll(). That function contains the required checks. This is already implemented for __put_user_size(), somehow I overlooked __put_user_{1,2,4,8,X}. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.45, 2005-03-28 19:29:15-08:00, venkatesh.pallipadi@intel.com [PATCH] rtc_lock is irq-safe rtc_lock is held during timer interrupts. So, we should block interrupts while holding it. Signed-off-by: Venkatesh Pallipadi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.44, 2005-03-28 19:28:58-08:00, zwane@arm.linux.org.uk [PATCH] APM: fix interrupts enabled in device_power_up APM was calling device_power_down and device_power_up with interrupts enabled, resulting in a few calls to get_cmos_time being done with interrupts enabled (rtc_lock needs to be acquired with interrupts disabled). Signed-off-by: Zwane Mwaikambo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.43, 2005-03-28 19:28:42-08:00, george@mvista.com [PATCH] x86: CMOS time update optimisation This patch changes the update of the cmos clock to be timer driven rather than poll driven by the timer interrupt function. If the clock is not being synced to an outside source the timer is removed and thus system overhead is nill in that case. The update frequency is still ~11 minutes and missing the update window still causes a retry in 60 seconds. We want the calls to sync_cmos_clock() to be made in a consistent environment. This was not true when calling it directly from the NTP call code. The change means that sync_cmos_clock() is ALWAYS called from run_timers(), i.e. as a timer call back function. Also, call the timer code only through the timer interface (set a short timer to do it from the ntp call). Signed-off-by: George Anzinger Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.42, 2005-03-28 19:28:26-08:00, yuasa@hh.iij4u.or.jp [PATCH] mips: update VR41xx RTC support This patch updates NEC VR4100 series RTC support. Signed-off-by: Yoichi Yuasa Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.41, 2005-03-28 19:27:38-08:00, akpm@osdl.org [PATCH] mips linkage fix Jim Gifford The iomap.o functions aren't dragged into the kernel if the build is fully modular. But modules need them. So link iomap.o into the kernel unconditionally. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.40, 2005-03-28 19:27:21-08:00, michael@ellerman.id.au [PATCH] ppc64: numa: Remove redundant and broken overlap check The numa code used to have to handle the fact that memory regions (as reported by OF) had been coallesced in the lmb struct and so might overlap node boundaries. Since Mike's patch went in this doesn't happen anymore, because we iterate over the memory regions from OF directly. This patch simply removes a check that catered for the overlapping case, which now "can't happen". The condition also happens to be broken in the current code, but in a painless way, so it's as good as removed already. Signed-off-by: Michael Ellerman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.39, 2005-03-28 19:27:06-08:00, olof@austin.ibm.com [PATCH] PPC64: Fix LPAR IOMMU setup code for p630 Here's a fix to deal with p630 systems in LPAR mode. They're to date the only system that in some cases might lack a dma-window property for the bus, but contain an overriding property in the device node for the specific adapter/slot. This makes the device setup code a bit more complex since it needs to do some of the things that the bus setup code has already done. Signed-off-by: Olof Johansson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.38, 2005-03-28 19:26:49-08:00, benh@kernel.crashing.org [PATCH] ppc64: Fix ethernet PHY reset on iMac G5 On iMac G5, when netbooting (or causing any other ethernet activity from within the Open Firmware environment), the PHY is put into a low power state before booting the OS. The result is that Linux doesn't see it and networking doesn't work. This patch adds the ethernet PHY reset platform hook to pmac_feature.c on ppc64 (it already is commonly used on ppc32 as lots of macs have this same problem, so the hook definition is already there and sungem is already calling it). Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.37, 2005-03-28 19:26:33-08:00, michael@ellerman.id.au [PATCH] ppc64: Make numa=off command line argument work again Mike's patch "ppc64: NUMA memory fixup (another try)" broke the numa code when "numa=off" is specified on the kernel command line. The fix is to pretend everything is in node 0 when numa is disabled. Boot tested on pSeries LPAR with numa=off and numa=debug (ie. on). Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.36, 2005-03-28 19:26:17-08:00, anton@samba.org [PATCH] ppc64: fix pseries hcall stubs Fix a number of bugs in our pseries hcall stubs: - store parameters in the area specified by the ABI, no need to create stack frames. - plpar_hcall_4out would corrupt r14 - merge multiple HVSC definitions Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.35, 2005-03-28 19:26:01-08:00, anton@samba.org [PATCH] ppc64: fix semtimedop compat syscall As with sparc64, the ppc64 version of semtimedop was incorrect - the timeout is in the fifth argument. I got caught copying again :) Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.34, 2005-03-28 19:25:45-08:00, anton@samba.org [PATCH] ppc64: fix linkage error on G5 Move the ppc64 specific cond_syscall(ppc_rtas) into sys_ni.c so that it takes effect. With this fixed we can remove the #define hack. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.33, 2005-03-28 19:25:28-08:00, mikpe@csd.uu.se [PATCH] ppc64: fix compile error in prom.c arch/ppc64/kernel/prom.c:1691: error: syntax error before 'prom_reconfig_notifier' arch/ppc64/kernel/prom.c:1692: error: field name not in record or union initializer arch/ppc64/kernel/prom.c:1692: error: (near initialization for 'prom_reconfig_nb') arch/ppc64/kernel/prom.c:1692: warning: initialization makes pointer from integer without a cast Signed-off-by: Mikael Pettersson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.32, 2005-03-28 19:25:12-08:00, mikpe@csd.uu.se [PATCH] ppc64: fix gcc4 compile error in paca.h In file included from include/asm/spinlock.h:20, from include/linux/spinlock.h:43, from include/linux/signal.h:5, from arch/ppc64/kernel/asm-offsets.c:17: include/asm/paca.h:25: error: array type has incomplete element type Signed-off-by: Mikael Pettersson Acked-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.31, 2005-03-28 19:24:56-08:00, paulus@samba.org [PATCH] ppc64: pSeries_iommu.c: use pSeries reconfig notifier Use the pSeries_reconfig notifier chain for tearing down the iommu table when a device node is removed. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.30, 2005-03-28 19:24:40-08:00, paulus@samba.org [PATCH] ppc64: pci_dn.c: use pSeries reconfig notifier Use the pSeries_reconfig notifier list to handle newly added pci device nodes. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.29, 2005-03-28 19:24:23-08:00, trini@kernel.crashing.org [PATCH] ppc32: 8xx typo fix The following patch fixes a typo which was introduced in the great SPRN_ renaming. Signed-off-by: Pantelis Antoniou Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.28, 2005-03-28 19:24:07-08:00, waite@skycomputers.com [PATCH] ppc32: fix broken compile on Sky Computers HDPU platform This fixes some compile errors in Sky Computers HDPU platform that were created by a bad patch. Signed-off-by: Brian Waite Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.27, 2005-03-28 19:23:50-08:00, benh@kernel.crashing.org [PATCH] ppc32/64: Map prefetchable PCI without guarded bit While experimenting with framebuffer access performances, we noticed a very significant improvement in write access to it when not setting the "guarded" bit on the MMU mappings. This bit basically says that reads and writes won't have side effects (it allows speculation). It appears that it also disables write combining. This patch implements a new phys_mem_access_prot() arch callback for use by /dev/mem and fbdev when available, implements it for ppc32 and ppc64, and modifies /dev/mem and fbdev to use it, respectively when available or on ppc. I didn't change fbdev to use it on all archs when available because there is already a whole lot of arch specific mess in there (more than in /dev/mem !) that I didn't feel like messing with, but archs maintainers are welcome to give it a go). The old mechanism in /dev/mem is still there, but arch maintainers should probably switch to this once which is more consistent imho. Finally, the ppc32 and ppc64 implementation of this and of the PCI mmap calls (used by /proc and /sys) are modified to check if the mapping happens in a prefetchable PCI resource, in which case, the guarded bit is not set for the pgprot. In fact, ppc32 implementation of this code is updated to be identical to ppc64. This improves framebuffer write performance on a simple test paul wrote from about 50Mb/sec to 200Mb/sec on my M9 based laptop and on a G5. The new hook will automatically catch up Xfree mmap's from /dev/mem, so it will work out of the box with existing X servers. Kernel fbdev accesses aren't improved yet as ioremap doesn't use that mecanism. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.26, 2005-03-28 19:23:33-08:00, mgreer@mvista.com [PATCH] ppc32: Fix Sandpoint Soft Reboot This patch allows the Freescale Sandpoint to perform soft reboots. A write of 0x00 to the Winbond's Chip Select Control Register was clearing the Upper BIOS Chip Select Enable bit which unmaps the boot flash. The comment associated with the write noted that it was enabling the RTC and Keyboard address locations, but the bits in question (1 and 0) are reserved when the Winbond chip is in PowerPC mode. Also, the bits are 1 for enable, 0 for disable, therefore the original code was actually disabling the address locations. The patch also corrects errors in the definitions of 2 configuration bits in the Tundra Tsi107 bridge's MAPB Options register. Signed-off-by Randy Vinson Signed-off-by Mark A. Greer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.25, 2005-03-28 19:23:16-08:00, perchrh@pvv.org [PATCH] ppc32: dmasound compilation fix sound/built-in.o(.init.text+0xb68): In function `dmasound_awacs_init': : undefined reference to `pmac_xpram_read' Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.24, 2005-03-28 19:22:59-08:00, galak@freescale.com [PATCH] ppc32: Report chipset version in common /proc/cpuinfo handling Moved reporting of chipset revision from board specific to common handing of /proc/cpuinfo. In light of numerous PPC system-on-chip devices, we report both the cpu version (reflects the core version) and the chipset version (reflects the system-on-chip or bridge version). Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.23, 2005-03-28 19:22:43-08:00, galak@freescale.com [PATCH] ppc32: typo fix in load/store string emulation Patch fixes a typo in the emulation of load/store string emulations pointed out by Segher Boessenkool. Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.22, 2005-03-28 19:22:27-08:00, galak@freescale.com [PATCH] ppc32: Fix CONFIG_SERIAL_TEXT_DEBUG support on 83xx The uart initialization for CONFIG_SERIAL_TEXT_DEBUG on 83xx was passing in physical addresses instead of effective. Additional, fix the Kconfig support to be for all 83xx devices, not just the MPC834x SYS board. Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.21, 2005-03-28 19:22:11-08:00, galak@freescale.com [PATCH] ppc32: Move 83xx & 85xx device and system description files This patch moves the 83xx & 85xx device and system description files out of the platform directory (used for board code) and into the syslib directory (used for common system code). Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.20, 2005-03-28 19:21:42-08:00, mgreer@mvista.com [PATCH] ppc32: Fix mv64x60 internal SRAM size ppc32: Fix wrong size for mv64[34]60's internal SRAM. - Fix incorrect SRAM size - Minor Kconfig cleanups for mv64x60 platforms Signed-off-by: Mark A. Greer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.19, 2005-03-28 19:21:26-08:00, sds@tycho.nsa.gov [PATCH] SELinux: add name_connect permission check This patch adds a name_connect permission check to SELinux to provide control over outbound TCP connections to particular ports distinct from the general controls over sending and receiving packets. Signed-off-by: Stephen Smalley Signed-off-by: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.18, 2005-03-28 19:21:10-08:00, sds@tycho.nsa.gov [PATCH] SELinux: audit unrecognized netlink messages This patch changes SELinux to audit any unrecognized netlink messages in controlled classes rather than silently rejecting them, and to allow them if in permissive mode. Please apply. Signed-off-by: Stephen Smalley Signed-off-by: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.47.1, 2005-03-28 19:20:57-08:00, davem@sunset.davemloft.net [SPARC64]: Some more cheetah+ patches needed for fptraps. Signed-off-by: David S. Miller ChangeSet@1.2181.46.17, 2005-03-28 19:20:54-08:00, sds@tycho.nsa.gov [PATCH] SELinux: allow mounting of filesystems with invalid root inode context This patch alters the SELinux handling of inodes with invalid security contexts so that a filesystem with a root inode that has an invalid security context can still be mounted for administrative recovery without disabling SELinux altogether. Signed-off-by: Stephen Smalley Signed-off-by: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.16, 2005-03-28 19:20:37-08:00, sds@tycho.nsa.gov [PATCH] SELinux: make code static and remove unused code This patch from Adrian Bunk makes needlessly global code static and removes a number of unused global and static functions from SELinux. Please apply. Author: Adrian Bunk Signed-off-by: Stephen Smalley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.15, 2005-03-28 19:20:20-08:00, jmoyer@redhat.com [PATCH] unused 'size' assignment in filemap_nopage filemap_nopage has the following code: retry_all: size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; if (pgoff >= size) goto outside_data_content; /* If we don't want any read-ahead, don't bother */ if (VM_RandomReadHint(area)) goto no_cached_page; /* * The "size" of the file, as far as mmap is concerned, isn't bigger * than the mapping */ if (size > endoff) size = endoff; After this, size is not referenced. So, either this potential reassignment of size is superfluous, or we are missing some other code later on in the function. If it is the former, I've attached a patch which will remove the code. (akpm: and endoff can go away too. That was the unused variable which gcc has been warning about for ages. Also, gratuitous whitespace fiddling). Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.14, 2005-03-28 19:20:04-08:00, ak@muc.de [PATCH] Fix mmap of /dev/kmem Fix mmap of /dev/kmem. It cannot ever have worked before. vmalloc is still not supported because that would be more complicated. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.13, 2005-03-28 19:19:48-08:00, herbert@13thfloor.at [PATCH] include cleanup in pgalloc.h This patch cleans up asm-*/pgalloc.h by removing the generous includes which are obsoleted (duplicated) by including linux/mm.h (and friends) They are double checked and verified by the PLM cross compiling service (the patched kernel gives the same warnings/errors as the unpatched) http://osdl.org/plm-cgi/plm?module=patch_info&patch_id=4313 Signed-off-by: Herbert Pötzl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.12, 2005-03-28 19:19:32-08:00, akpm@osdl.org [PATCH] slab shrinkers: use vfs_cache_pressure Fix some bugs spotted by Andrea Arcangeli - When we added /proc/sys/vm/vfs_cache_pressure we forgot to allow it to tune the dquot and mbcache slabs as well. - Reduce lock contention in shrink_dqcache_memory(). - Use dqstats.free_dquots in shrink_dqcache_memory(): this is the count of reclaimable objects. - Export sysctl_vfs_cache_pressure to GPL modules for mbcache.o Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.11, 2005-03-28 19:19:14-08:00, clameter@sgi.com [PATCH] mm counter operations through macros This patch extracts all the operations on counters protected by the page table lock (currently rss and anon_rss) into definitions in include/linux/sched.h. All rss operations are performed through the following macros: get_mm_counter(mm, member) -> Obtain the value of a counter set_mm_counter(mm, member, value) -> Set the value of a counter update_mm_counter(mm, member, value) -> Add to a counter inc_mm_counter(mm, member) -> Increment a counter dec_mm_counter(mm, member) -> Decrement a counter With this patch it becomes easier to add new counters and it is possible to redefine the method of counter handling. The counters are an issue for scalability since they are used in frequently used code paths and may cause cache line bouncing. F.e. One may not use counters at all and count the pages when needed, switch to atomic operations if the mm_struct locking changes or split the rss into counters that can be locally incremented. The relevant fields of the task_struct are renamed with a leading underscore to catch out people who are not using the acceessor macros. Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.10, 2005-03-28 19:18:57-08:00, mort@sgi.com [PATCH] vmscan: move code to isolate LRU pages into separate function I noticed that the loop to pull pages out of the LRU lists for processing occurred twice. This just sticks that code into a separate function to improve readability. Signed-Off-By: Martin Hicks Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.9, 2005-03-28 19:18:41-08:00, manfred@colorfullife.com [PATCH] slab: 64-bit fix Use kmem_bufctl_t instead of int for the loop counters in check_slabp. Signed-of-by: Manfred Spraul Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.8, 2005-03-28 19:18:25-08:00, akpm@osdl.org [PATCH] slab: kfree(null) is unlikely - mark kfree(NULL) as being unlikely - Coding-style consistency for function definitions Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.7, 2005-03-28 19:18:09-08:00, manfred@colorfullife.com [PATCH] slab.[ch]: kmalloc() cleanups I found a new idea for optimizing the loop in kmalloc - now the cleanup both reduces the size of the .text segment in mm/slab.o and speeds up largs kmallocs a bit. I still loose 2 cycles for kmalloc(32,GFP_KERNEL) compared to the current code, but I haven't figured out how I could optimize further. - inline kmem_find_general_cachep() - optimize kmem_find_general_cachep() and document the optimization. - remove duplicate code from __kmalloc(): call kmem_find_general_cachep() directly. Based on a patch from Renaud Lienhart Signed-off-by: Manfred Spraul Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.6, 2005-03-28 19:17:52-08:00, akpm@osdl.org [PATCH] tty overrun time fix moreau francis points out that we need to initialise overrun_time, else the overrun reporting screws up when jiffies is "negative". So initialise tty->overrun_time apropriately, and handle the case where the tty has been open for so long that the overrun time has become "greater than" jiffies. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.5, 2005-03-28 19:17:36-08:00, jonsmirl@gmail.com [PATCH] handle multiple video cards on the same bus When detecting the boot video device, allow for the case of multiple cards on the same bus. Check each candidate to make sure that the card is active. Signed-off-by: Jon Smirl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.4, 2005-03-28 19:17:20-08:00, daniel@osdl.org [PATCH] ppc64: fix AIO panic on PPC64 caused by is_hugepage_only_range() When testing AIO on PPC64 (a power5 machine) running 2.6.11 with CONFIG_HUGETLB_PAGE=y, I ran into a kernel panic when a process exits that has done AIO (io_queue_init()) but has not done the io_queue_release(). The exit_aio() code is cleaning up and panicking when trying to free the aio ring buffer. I tracked this down to is_hugepage_only_range() (include/asm-ppc64/page.h) which is doing a touches_hugepage_low_range() which is checking current->mm->context.htlb_segs. The problem is that exit_mm() cleared tsk->mm before doing the mmput() which leads to the exit_aio() and then the panic. Looks like is_hugepage_only_range() is only used in ia64 and ppc64. Fix is to change is_hugepage_only_range() to take an 'mm' as a parameter as well as 'addr' and 'len' and then the ppc64 code could change to use 'mm'. It looks like it has been broken for quite a while. Signed-off-by: Daniel McNeil Acked-by: David Gibson Acked-by: "Luck, Tony" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.3, 2005-03-28 19:17:04-08:00, linux@dominikbrodowski.net [PATCH] pcmcia: don't send eject request events to userspace Here follow PCMCIA-related patches which allow for kernel-based matching between 16-bit PCMCIA devices and their drivers. If you're using this (and using a special startup script or small tool), you don't need to run "cardmgr" any longer. cardmgr still works, though, allowing for a gradual transition. These patches apply on top of what's in 2.6.11-rc4-mm1; for proper operation pcmcia-bridge-resource-management-fix.patch should be removed first, though. Andrew, please apply them to your next -mm release. 1. Theory of operation I: setting up the sockets ================================================ - the appropriate socket driver is initialized (modprobe or built-in) - a hotplug event (CLASS=pcmcia_socket) is generated - if it's a socket which doesn't map the resources statically, resources available for use by PCMCIA cards need to be marked as "allowed". This is done using either a small shell script[*] or pcmcia-socket-startup[**]. The latter parses /etc/pcmcia/config.opts -- the same file cardmgr currently uses for this purpose. [*] e.g. #!/bin/sh echo -n "+ 0x00000100 - 0x000004cf" > /sys/class/pcmcia_socket/pcmcia_socket1/available_resources_io echo -n "+ 0x60000000 - 0x60ffffff" > /sys/class/pcmcia_socket/pcmcia_socket1/available_resources_mem echo -n "1" > /sys/class/pcmcia_socket/pcmcia_socket1/available_resources_setup_done [**] note: for pcmcia-socket-startup to work properly you need a patch for libsysfs (it is already accepted by the maintainer of the package). See below for details. 2. Theory of operation II: setting up a device ============================================== - a pcmcia device is discovered in the socket - device information is determined and exported to sysfs - a hotplug event is created, passing -- among others -- a complete MODNAME string to userspace. Available match flags are the product ID strings [in userspace, only the hashes are matched -- you can't use strings in modules.alias], manufactor and card ID, the multifunction device number, the pseudo-multifunction device number [this is what e.g. bind "serial_cs", bind "serial_cs" was before], whether it needs a CIS override, and the function ID [this match is only used if it is allowed by userspace -- it should be avoided and will be removed in future, after existing users have been converted to matches of other types]. - modprobe $MODNAME [*] - the device is matched against drivers. function ID matches are ignored at this stage. If the device needs a CIS override, it is requested from userspace using the firmware helpers. It is recommended to have a symlink either from /lib/firmware/cis/ to /etc/pcmcia/cis/ or the other way to let both (old) cardmgr and (new) firmware.agent access the override CIS data. - if no driver was found, an userspace helper checks whether the card is one of two(!) which can only be detected using cardmgr's tuple directive. If this is the case, a CIS override is done using /sys/class/pcmcia_socket/pcmcia_socket%n/cis , loading a CIS with proper identification values. - if this doesn't lead to a driver being bound to the device, function ID-based matches are allowed. [*] note: for this to work properly you need a patch to be written which loads _all_ matching modules, not just one] 3. patches ========== I'll not send the patches which add the proper device identification tables to this list, but only (and a bit later) to Andrew and the appropriate maintainers. You can find them already at http://www.kernel.org/pub/linux/kernel/people/brodo/patches/2.6.11-rc5/ -- they start with pcmcia-25- I used all identification information I could find in pcmcia-cs, the kernel, and some web resources and, but probably missed some. If I did, please send me either the appropriate device info from the cardmgr config file or the output of "pcmcia-modalias" (see below) -- or a PCMCIA_MATCH patch. 4. userspace tools ================== Several GPL[*] userspace tools are required or useful for this (new) method of operation. I'll send them to this list for review, and they're also available _temporarily_ at http://www.kernel.org/pub/linux/kernel/people/brodo/temp-pcmcia-tools.tar.gz I'm still undecided on what to do with these userspace tools. IMO they could either be merged into pcmcia-cs or distributed in addition to pcmcia-cs. Also, I'd be very glad if somebody else could maintain and improve these tools. pcmcia-socket-startup This tool is based on "cardmgr" by Dave Hinds, but most of it was removed. The remains parse the /etc/pcmcia/config file for allowed ioport and iomem ranges and for disallowed IRQs, and sets these using libsysfs. Due to a bug in libsysfs, a patch is needed for this to work properly. It takes one parameter, the socket number to be set up. pcmcia-modalias This tool reads device information and generates a MODALIAS string just like the kernel exports using hotplug for this device. It takes one paramter, the device's bus_id string (e.g. "pcmcia-modalias 1.0" for the first device on the second PCMCIA socket) pcmcia-check-broken-cis This checks whether the CIS is so broken that only the "tuple"-based match is possible. It takes one parameter, the socket the PCMCIA card to be tested is inserted to, and returns nothing (if no CIS override is needed), or the CIS file name. The CIS parsing is done using a much simplified version of what is to be found in drivers/pcmcia/cistpl.c, so this is based on Dave Hinds work as well. hotplug/pcmcia* The hotplug scripts are quite self-explanatory and also allow for "coldplug" (i.e. they work around "missed" events during boot). pccardctl This previously announced tool uses sysfs instead of an ioctl for commands like "eject", "insert", "info" and "ident". [*] reason: they're partly based on other GPL-licensed tools. This patch: Don't inform cardmgr of ejection requests so that "cardctl eject" behaves _exactly_ the same as a physical ejection from an userspace point of view. Removing this allows us to clean up some other code which is otherwise unused. Signed-off-by: Dominik Brodowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.2, 2005-03-28 19:16:47-08:00, linux@dominikbrodowski.net [PATCH] pcmcia: don't lock up in rsrc_nonstatic pcmcia_validate_mem struct socket_data wasn't zeroed, so pcmcia_validate_mem() didn't get called. If it is called, though, one possible code-path already holds skt_sem, so lockups occur. Therefore, change calling conventions to pcmcia_validate_mem(). Signed-off-by: Dominik Brodowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.46.1, 2005-03-28 19:16:31-08:00, linux@dominikbrodowski.net [PATCH] pcmcia: properly bail out on MTD-related ioctl invocation Properly set the return value for DS_BIND_MTD, DS_GET_FIRST_REGION and DS_GET_NEXT_REGION. Else trying to bind MTD devices the old (deprecated) way may cause an endless loop in cardmgr. This also happens if the resources weren't made available properly in /etc/pcmcia/config.opts. Signed-off-by: Dominik Brodowski Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.27.17, 2005-03-28 19:58:58-06:00, stern@rowland.harvard.edu [PATCH] Add a scsi_device flag for RETRY_HWERROR It turns out that a bunch of USB-IDE converters make the mistake of returning SK = 04 (Hardware Error) whenever the IDE device signals any sort of error, without bothering to distinguish recoverable from non-recoverable errors. The best way to handle this is for usb-storage to set a per-device flag indicating that these errors should always be retried. The current scheme (blacklist flag but no per-device flag) isn't well suited for this situation. This patch adds the per-device flag and sets it initially based on the blacklist setting. Once this has been merged, a separate patch will be submitted to Matt Dharm adding the corresponding support to usb-storage. Signed-off-by: Alan Stern Signed-off-by: James Bottomley ChangeSet@1.2181.43.6, 2005-03-28 17:42:49-08:00, torvalds@ppc970.osdl.org Mark "gfp" masks as "unsigned int" and use __nocast to find violations. This makes it hard(er) to mix argument orders by mistake for things like kmalloc() and friends, since silent integer promotion is now caught by sparse. ChangeSet@1.2181.45.2, 2005-03-29 02:55:11+02:00, tgraf@suug.ch [NET]: Make primary TLV type optional Allows the use of the gnet_stats API for backward compatiblity cases where no "modern" TLV structure is needed. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller ChangeSet@1.2181.45.1, 2005-03-29 02:45:52+02:00, tgraf@suug.ch Cset exclude: hadi@cyberus.ca|ChangeSet|20050325173452|50562 ChangeSet@1.2181.2.56, 2005-03-28 16:19:59-08:00, davem@sunset.davemloft.net [SPARC64]: FPU disabled trap needs context register patching. It writes the nucleus context into the secondary context register so that it can do block stores and loads from nucleus context. So, we need to patch on cheetah+. Signed-off-by: David S. Miller ChangeSet@1.2181.2.55, 2005-03-28 15:29:13-08:00, davem@sunset.davemloft.net [SPARC64]: Support >=cheetah+ dual-dtlbs properly. UltraSPARC-III+ (aka. cheetah+) and later chips have three D-TLB units. One is fully assosciative and has 16 entries, the other two have 512 entries each and are 2 way subblocked and then indexed by the PAGE_MASK bits. This implies that the two 512 entry TLBs need to know the page size in order to index things properly. The page size to use is encoded in page size fields within the TLB context registers. Up until this point we just left them at zero, which meant that both 512 entry TLBs were used for 8K translations only. The rest (including the 4MB kernel mappings) went into the 16 entry fully assosciative TLB. Now, for the kernel, we use the first 512 entries for 4MB mappings and the second 512 entries for PAGE_SIZE mappings (ie. for vmalloc and modules). For the user, we use the first 512 entries for PAGE_SIZE and the second 512 entries for HPAGE_SIZE if the user maps any hugetlb pages else we use the second half for user PAGE_SIZE stuff as well. Most of this changeset is clerical. We move most of the mm->context layout defines into asm/mmu.h, we abstract the mm->context type so that it is now much easier to see and trap accesses to the context values. In particular, one can find all context handling by grepping the sparc64 sources for CTX_* and {PRIMARY,SECONDAY}_CONTEXT. Happily, the fast paths of the kernel are mostly untouched by this new stuff. We add 2 cycles to trap entry, 1 cycle to trap exit, and 2 cycles to the window trap fixup code. Signed-off-by: David S. Miller ChangeSet@1.2181.16.24, 2005-03-28 15:20:34-08:00, bunk@stusta.de [PATCH] drivers/pci/msi.c: fix a check after use This patch fixes a check after use found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.23, 2005-03-28 15:19:00-08:00, eike-kernel@sf-tec.de [PATCH] PCI: remove pci_find_device usage from pci sysfs code. Greg KH wrote: > On Sun, Mar 20, 2005 at 03:53:58PM +0100, Rolf Eike Beer wrote: > > Greg KH wrote: > > > ChangeSet 1.1998.11.23, 2005/02/25 08:26:11-08:00, gregkh@suse.de > > > > > > PCI: remove pci_find_device usage from pci sysfs code. > > Any reasons why you are not using "for_each_pci_dev(pdev)" here? > > Nope, I forgot it was there :) Patch is against 2.6.12-rc1-bk1 and does the same think like your one, except it uses for_each_pci_dev() Signed-off-by: Rolf Eike Beer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.22, 2005-03-28 15:10:57-08:00, akpm@osdl.org [PATCH] PCI: handle multiple video cards on the same bus From: Jon Smirl When detecting the boot video device, allow for the case of multiple cards on the same bus. Check each candidate to make sure that the card is active. Signed-off-by: Jon Smirl Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.21, 2005-03-28 15:10:34-08:00, eike-kernel@sf-tec.de [PATCH] PCI: shrink drivers/pci/proc.c::pci_seq_start() this patch shrinks pci_seq_start by using for_each_pci_dev() macro instead of explicitely using a loop and avoiding a goto. Signed-off-by: Rolf Eike Beer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.20, 2005-03-28 15:10:15-08:00, bunk@stusta.de [PATCH] drivers/pci/hotplug/cpqphp_core.c: fix a check after use This patch fixes a check after use found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.19, 2005-03-28 15:09:51-08:00, khali@linux-fr.org [PATCH] PCI: Quirk for Asus M5N One more Asus laptop which requires a PCI quirk to unhide the SMBus. Contributed by Matthias Hensler through bugzilla (#4391). Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.16.18, 2005-03-28 15:09:31-08:00, prarit@sgi.com [PATCH] PCI Hotplug: add documentation about release pointer. Adds "release" func pointer comments to nano-doc. Signed-off-by: Prarit Bhargava Index: hp/drivers/pci/hotplug/pci_hotplug.h =================================================================== RCS file: /usr/local/src/cvsroot/bk/linux-2.5/drivers/pci/hotplug/pci_hotplug.h,v retrieving revision 1.1.1.1 ChangeSet@1.2181.4.86, 2005-03-28 12:36:04-08:00, nacc@us.ibm.com [PATCH] sound/usbmidi: change parameters of usb_bulk_msg() to msecs Use milliseconds at the timeout parameter in the call to usb_bulk_msg() to match the converted interface. Signed-off-by: Nishanth Aravamudan Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.85, 2005-03-28 12:35:42-08:00, nacc@us.ibm.com [PATCH] sound/usbaudio: change parameters of snd_usb_ctl_msg() to msecs Use milliseconds at the timeout parameter in the call to snd_usb_ctl_msg(), a wrapper to usb_control_msg(), to match the converted interface. Signed-off-by: Nishanth Aravamudan Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.84, 2005-03-28 12:35:20-08:00, nacc@us.ibm.com [PATCH] usb/kl5kusb105: change parameters of usb_control_msg() to msecs Use milliseconds at the timeout parameter in the call to usb_control_msg() to match the converted interface. Signed-off-by: Nishanth Aravamudan Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.83, 2005-03-28 12:34:56-08:00, nacc@us.ibm.com [PATCH] usb/pwc-ctrl: change parameters of usb_control_msg() to msecs Use milliseconds at the timeout parameter in the call to usb_control_msg() to match the converted interface. Signed-off-by: Nishanth Aravamudan Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.82, 2005-03-28 12:34:31-08:00, colin@colino.net [PATCH] USB: fix shared key auth in zd1201 It's currently impossible to associate with a shared-key-only access point using the zd1201 driver. The attached patch fixes it. The reason was (probably) a typo in the definitions of the authentification types. I found that they should be (1,2) instead of (0,1) by looking at the old linux-wlan-ng driver by Zydas. Signed-off-by: Colin Leroy Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.81, 2005-03-28 12:34:11-08:00, bunk@stusta.de [PATCH] MAINTAINERS: remove obsolete HPUSBSCSI entry It seems I forgot MAINTAINERS in my patch that removed this driver... Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.80, 2005-03-28 12:33:49-08:00, bunk@stusta.de [PATCH] drivers/usb/media/usbvideo.c: fix a check after use This patch fixes a check after use found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.79, 2005-03-28 12:33:26-08:00, zaitcev@redhat.com [PATCH] USB: fix for ub for sleeping function called from invalid context at kernel/workqueue.c:264 From: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.78, 2005-03-28 12:33:04-08:00, bunk@stusta.de [PATCH] drivers/usb/misc/usbtest.c: fix a NULL dereference This patch fixes a NULL dereference found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.77, 2005-03-28 12:32:43-08:00, bunk@stusta.de [PATCH] drivers/usb/class/usb-midi.c: remove dead code This patch removes some obviously dead code found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.44.54, 2005-03-28 21:29:59+01:00, buytenh@org.rmk.(none) [ARM PATCH] 2583/1: add several registers to arch-ixp2000/ixp2000-regs.h Patch from Lennert Buytenhek These registers ({RAW_STATUS,ENABLE_SET,ENABLE_CLEAR}_[AB]_[0123]) are needed for checking the status of, acking, unmasking and masking the 256 individual thread interrupt sources. This is used by the microengine thread interrupt driver. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2181.4.76, 2005-03-28 12:11:14-08:00, gregkh@suse.de USB: add fossil watch ids to the visor driver. Thanks to Gunter Ohrner for the information. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.44.53, 2005-03-28 21:04:52+01:00, buytenh@org.rmk.(none) [ARM PATCH] 2582/1: correct thread interrupt comments in arch-ixp2000/irqs.h Patch from Lennert Buytenhek Some comments in arch-ixp2000/irqs.h say that thread interrupts 64-127 are IXP2800-only, but this is not true: IXP2400 has 0-31 and 64-95, and 32-63 and 96-127 are IXP2800-only. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2181.44.52, 2005-03-28 20:38:21+01:00, rmk@flint.arm.linux.org.uk [ARM] Fix ARM TLB shootdown code We missed flushing the TLB when we're unmapping only reserved pages. Since the kernel is now better at passing vmas which correspond solely with the region to unmap, we can use tlb_*_vma() to do "just enough" flushing, both for the TLB and cache. We can avoid all flushes in tlb_*_vma() for the full-MM case, and just do a flush_tlb_mm(). We can omit the cache flushing because this thread will die soon. Signed-off-by: Russell King ChangeSet@1.2181.44.51, 2005-03-28 14:32:20+01:00, vince@org.rmk.(none) [ARM PATCH] 2629/1: Update shark_defconfig default config Patch from Vincent Sanders Updates shark_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.50, 2005-03-28 14:09:29+01:00, vince@org.rmk.(none) [ARM PATCH] 2628/1: Update simpad_defconfig default config Patch from Vincent Sanders Updates simpad_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.49, 2005-03-28 13:47:55+01:00, vince@org.rmk.(none) [ARM PATCH] 2627/1: Update lusl7200_defconfig default config Patch from Vincent Sanders Updates lusl7200_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.48, 2005-03-28 13:39:24+01:00, vince@org.rmk.(none) [ARM PATCH] 2626/1: Update versatile_defconfig default config Patch from Vincent Sanders Updates versatile_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.47, 2005-03-28 13:31:48+01:00, vince@org.rmk.(none) [ARM PATCH] 2625/1: Update lubbock_defconfig default config Patch from Vincent Sanders Updates lubbock_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.46, 2005-03-28 13:22:45+01:00, vince@org.rmk.(none) [ARM PATCH] 2624/1: Update lpd7a404_defconfig default config Patch from Vincent Sanders Updates lpd7a404_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.45, 2005-03-28 13:15:10+01:00, vince@org.rmk.(none) [ARM PATCH] 2623/1: Update lpd7a400_defconfig default config Patch from Vincent Sanders Updates lpd7a400_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.44, 2005-03-28 13:07:48+01:00, vince@org.rmk.(none) [ARM PATCH] 2622/1: Update lart_defconfig default config Patch from Vincent Sanders Updates lart_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.43, 2005-03-28 12:59:29+01:00, vince@org.rmk.(none) [ARM PATCH] 2621/1: Update hackkit_defconfig default config Patch from Vincent Sanders Updates hackkit_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.42, 2005-03-28 12:51:11+01:00, vince@org.rmk.(none) [ARM PATCH] 2620/1: Update h7202_defconfig default config Patch from Vincent Sanders Updates h7202_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.41, 2005-03-28 12:42:46+01:00, vince@org.rmk.(none) [ARM PATCH] 2619/1: Update h7201_defconfig default config Patch from Vincent Sanders Updates h7201_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.40, 2005-03-28 12:35:10+01:00, vince@org.rmk.(none) [ARM PATCH] 2618/1: Update h3600_defconfig default config Patch from Vincent Sanders Updates h3600_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.39, 2005-03-28 12:27:36+01:00, vince@org.rmk.(none) [ARM PATCH] 2617/1: Update fortunet_defconfig default config Patch from Vincent Sanders Updates fortunet_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.38, 2005-03-28 12:19:02+01:00, vince@org.rmk.(none) [ARM PATCH] 2616/1: Update smdk2410_defconfig default config Patch from Vincent Sanders Updates smdk2410_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.37, 2005-03-28 12:11:19+01:00, vince@org.rmk.(none) [ARM PATCH] 2615/1: Update shannon_defconfig default config Patch from Vincent Sanders Updates shannon_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.36, 2005-03-28 12:03:00+01:00, vince@org.rmk.(none) [ARM PATCH] 2614/1: Update jornada720_defconfig default config Patch from Vincent Sanders Updates jornada720_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.35, 2005-03-28 11:54:31+01:00, vince@org.rmk.(none) [ARM PATCH] 2613/1: Update ixp4xx_defconfig default config Patch from Vincent Sanders Updates ixp4xx_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.34, 2005-03-28 11:47:29+01:00, vince@org.rmk.(none) [ARM PATCH] 2612/1: Update ixdp2801_defconfig default config Patch from Vincent Sanders Updates ixdp2801_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.33, 2005-03-28 11:39:28+01:00, vince@org.rmk.(none) [ARM PATCH] 2611/1: Update footbridge_defconfig default config Patch from Vincent Sanders Updates footbridge_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.32, 2005-03-28 11:31:18+01:00, vince@org.rmk.(none) [ARM PATCH] 2610/1: Update epxa10db_defconfig default config Patch from Vincent Sanders Updates epxa10db_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.31, 2005-03-28 11:22:33+01:00, vince@org.rmk.(none) [ARM PATCH] 2609/1: Update ep80219_defconfig default config Patch from Vincent Sanders Updates ep80219_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.30, 2005-03-28 11:14:40+01:00, vince@org.rmk.(none) [ARM PATCH] 2608/1: Update pxa255-idp_defconfig default config Patch from Vincent Sanders Updates pxa255-idp_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.29, 2005-03-28 11:07:06+01:00, vince@org.rmk.(none) [ARM PATCH] 2607/1: Update pleb_defconfig default config Patch from Vincent Sanders Updates pleb_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.28, 2005-03-28 10:58:14+01:00, vince@org.rmk.(none) [ARM PATCH] 2606/1: Update omnimeter_defconfig default config Patch from Vincent Sanders Updates omnimeter_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.27, 2005-03-28 10:49:23+01:00, vince@org.rmk.(none) [ARM PATCH] 2605/1: Update ixdp2800_defconfig default config Patch from Vincent Sanders Updates ixdp2800_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.26, 2005-03-28 10:40:47+01:00, vince@org.rmk.(none) [ARM PATCH] 2604/1: Update ixdp2401_defconfig default config Patch from Vincent Sanders Updates ixdp2401_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.25, 2005-03-28 10:09:55+01:00, vince@org.rmk.(none) [ARM PATCH] 2603/1: Update ixdp2400_defconfig default config Patch from Vincent Sanders Updates ixdp2400_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.24, 2005-03-28 10:02:38+01:00, vince@org.rmk.(none) [ARM PATCH] 2602/1: Update integrator_defconfig default config Patch from Vincent Sanders Updates integrator_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.23, 2005-03-28 09:53:34+01:00, vince@org.rmk.(none) [ARM PATCH] 2601/1: Update enp2611_defconfig default config Patch from Vincent Sanders Updates enp2611_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.22, 2005-03-28 09:44:30+01:00, vince@org.rmk.(none) [ARM PATCH] 2600/1: Update edb7211_defconfig default config Patch from Vincent Sanders Updates edb7211_defconfig to take account of the Kconfig changes this change is slightly more invasive than otehrs in this series because this configuration has not been updated in a *long* time. It appears to build with this config now. Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.2.54, 2005-03-27 21:43:19-08:00, davem@sunset.davemloft.net [SPARC64]: Do not use magic constant in mmu_context.h Signed-off-by: David S. Miller ChangeSet@1.2181.2.53, 2005-03-27 20:46:20-08:00, davem@sunset.davemloft.net [SPARC64]: Make PAGE_SIZE configurable. Signed-off-by: David S. Miller ChangeSet@1.2201, 2005-03-27 22:24:51-06:00, stevef@smf-t23.(none) [CIFS] various code formatting cleanup Signed-off-by: Jesper Juhl (juhl-lkml@dif.dk) Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.44.21, 2005-03-28 01:57:43+01:00, vince@org.rmk.(none) [ARM PATCH] 2599/1: Update s3c2410_defconfig default config Patch from Vincent Sanders Updates s3c2410_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.20, 2005-03-28 01:35:40+01:00, vince@org.rmk.(none) [ARM PATCH] 2598/1: Update omap_h2_1610_defconfig default config Patch from Vincent Sanders Updates omap_h2_1610_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.43.5, 2005-03-27 16:25:11-08:00, torvalds@ppc970.osdl.org Add '__nocast' sparse annotation to allow people to mark places where implicit casts are not appropriate. This can be used to mark unsigned integers as being uncastable to signed, for example. Or enums as not degrading to integers and vice versa. ChangeSet@1.2181.44.19, 2005-03-28 01:13:42+01:00, vince@org.rmk.(none) [ARM PATCH] 2597/1: Update netwinder_defconfig default config Patch from Vincent Sanders Updates netwinder_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.18, 2005-03-28 00:56:04+01:00, vince@org.rmk.(none) [ARM PATCH] 2596/1: Update mx1ads_defconfig default config Patch from Vincent Sanders Updates mx1ads_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.17, 2005-03-28 00:31:05+01:00, vince@org.rmk.(none) [ARM PATCH] 2595/1: Update mainstone_defconfig default config Patch from Vincent Sanders Updates mainstone_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.16, 2005-03-28 00:08:19+01:00, vince@org.rmk.(none) [ARM PATCH] 2594/1: Update iq80332_defconfig default config Patch from Vincent Sanders Updates iq80332_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.15, 2005-03-27 23:47:51+01:00, vince@org.rmk.(none) [ARM PATCH] 2593/1: Update iq80331_defconfig default config Patch from Vincent Sanders Updates iq80331_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.14, 2005-03-27 23:23:01+01:00, vince@org.rmk.(none) [ARM PATCH] 2592/1: Update iq80321_defconfig default config Patch from Vincent Sanders Updates iq80321_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.13, 2005-03-27 23:01:32+01:00, vince@org.rmk.(none) [ARM PATCH] 2591/1: Update iq31244_defconfig default config Patch from Vincent Sanders Updates iq31244_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.12, 2005-03-27 22:39:04+01:00, vince@org.rmk.(none) [ARM PATCH] 2590/1: Update ebsa110_defconfig default config Patch from Vincent Sanders Updates ebsa110_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.11, 2005-03-27 22:14:50+01:00, vince@org.rmk.(none) [ARM PATCH] 2589/1: Update cerfcube_defconfig default config Patch from Vincent Sanders Updates cerfcube_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.10, 2005-03-27 22:07:05+01:00, vince@org.rmk.(none) [ARM PATCH] 2588/1: Update bast_defconfig default config Patch from Vincent Sanders Updates bast_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.9, 2005-03-27 21:59:21+01:00, vince@org.rmk.(none) [ARM PATCH] 2587/1: Update badge4_defconfig default config Patch from Vincent Sanders Updates badge4_defconfig to take account of the Kconfig changes Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.8, 2005-03-27 21:51:56+01:00, vince@org.rmk.(none) [ARM PATCH] 2586/1: Update clps7500_defconfig default config Patch from Vincent Sanders Updates clps7500_defconfig to take account of the Kconfig changes this depends on 2585/1 Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.7, 2005-03-27 20:09:22+01:00, vince@org.rmk.(none) [ARM PATCH] 2585/1: missing ARCH_CLPS7500 depends in video Kconfig Patch from Vince Sanders While updating default configurations it was noted that the FB_ACORN video config option was missing a dependancy on ARCH_CLPS7500 Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.6, 2005-03-27 20:01:58+01:00, vince@org.rmk.(none) [ARM PATCH] 2584/1: cpufreq Kconfig menu tidyup Patch from Vincent Sanders While updating default configurations it was noted that CONFIG_SA1100 seems to enable the cpufreq entries however the sourced drivers/cpufreq/Kconfig doesnt do a menu section so it looks awful, this tiny patch fixes it. Signed-off-by: Vincent Sanders Signed-off-by: Russell King ChangeSet@1.2181.44.5, 2005-03-27 18:48:13+01:00, dsaxena@net.rmk.(none) [ARM PATCH] 2576/1: Fix LDRD and LDRSB (Thumb) abort handling Patch from Deepak Saxena The ARM LDRD (v5+) and Thumb LDRSB instructions use bit 20/11 (ARM/Thumb) differently than every other instruction, so it is set to 0 (write) even though the instructions are read instructions. This means that during an abort the instructions will be treated as a write and the handler will raise a signal from unwriteable locations if they fault. We have to specifically check for these instructions from the abort handlers to treat them properly. EABI toolchains emit LDRD instructions and even those not using EABI might have handcoded ASM that uses these instructions. This patch creates an abort-macro.S file that is included by the abort-ev* files that need the special cases and changes those handlers to use the macros. ARMv6 does not need the special case handling as the HW takes care of setting the FSR bit appropriately. Signed-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2181.44.4, 2005-03-27 18:32:03+01:00, buytenh@org.rmk.(none) [ARM PATCH] 2581/1: two more ixp2000 typo fixes Patch from Lennert Buytenhek Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King ChangeSet@1.2181.44.3, 2005-03-27 18:24:18+01:00, buytenh@org.rmk.(none) [ARM PATCH] 2580/1: remove nonsensical comment from arch-ixp2000/io.h Patch from Lennert Buytenhek Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King ChangeSet@1.2181.44.2, 2005-03-27 18:17:05+01:00, nico@org.rmk.(none) [ARM PATCH] 2579/1: make early boot failure more verbose Patch from Nicolas Pitre It is pointless for cases where one tries to boot a kernel on the wrong ARM subarchitecture, but for those cases where the architecture is similar enough so the low level routines do work then something a bit more informative than "Error: a" might be handy, especially when the bootloader is the culprit. All extra strings are compiled in only if CONFIG_DEBUG_LL is selected of course. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King ChangeSet@1.2181.44.1, 2005-03-27 18:09:23+01:00, nico@org.rmk.(none) [ARM PATCH] 2578/1: unsigned compare in processor and machine list walking Patch from Nicolas Pitre Signed-off-by: Nicolas Pitre Signed-off-by: Russell King ChangeSet@1.2181.43.4, 2005-03-27 17:56:44+10:00, airlied@starflyer.(none) drm: free kbuf if copy from user fails.. From: Eric Anholt Signed-off-by: Dave Airlie ChangeSet@1.2181.43.3, 2005-03-27 17:53:53+10:00, airlied@starflyer.(none) drm: radeon idct defines Add some type 3 idct packets for reference. Signed-off-by: Dave Airlie ChangeSet@1.2181.43.2, 2005-03-27 17:52:15+10:00, airlied@starflyer.(none) drm: radeon driver update 1.16 add R200_EMIT_PP_TRI_PERF_CNTL packet to support brilinear filtering on r200 fix a bug in the 1.15 merge also. From: Roland Scheidegger Signed-off-by: Dave Airlie ChangeSet@1.2181.38.8, 2005-03-27 17:17:28+10:00, airlied@starflyer.(none) drm: change DRIVER_ to CORE_ Change some defines to better naming. Signed-off-by: Dave Airlie ChangeSet@1.2181.38.7, 2005-03-27 16:49:13+10:00, airlied@starflyer.(none) Fix sparse NULL/0 warning: drivers/char/drm/radeon_state.c:1845:15: warning: Using plain integer as NULL pointer Signed-off-by: Randy Dunlap Signed-off-by: Dave Airlie ChangeSet@1.2181.38.6, 2005-03-27 16:47:24+10:00, airlied@starflyer.(none) drm: Remove incorrect "drm_"-prefix from parameter description. Signed-off-by: Magnus Damm Signed-off-by: Dave Airlie ChangeSet@1.2181.2.52, 2005-03-26 21:23:28-08:00, davem@nuts.davemloft.net [SPARC64]: Kill stray reference to pgdcache_size. Signed-off-by: David S. Miller ChangeSet@1.2181.2.51, 2005-03-26 21:37:09-08:00, davem@sunset.davemloft.net [SPARC64]: Handle non-8K PAGE_SIZE better in TLB miss handlers. The existing code mostly worked, but only for 32-bit processes. This caught a serious bug too, pgd_index() was masking with plain PTRS_PER_PGD instead of (PTRS_PER_PGD - 1). That has been with us for several years. It is a wonder that things work at all some times :-) Signed-off-by: David S. Miller ChangeSet@1.2181.2.50, 2005-03-26 19:48:01-08:00, davem@nuts.davemloft.net [SPARC64]: Make *_LOCKED_TLBENT and L1DCACHE_SIZE asm visible. Signed-off-by: David S. Miller ChangeSet@1.2181.2.49, 2005-03-26 19:31:26-08:00, davem@sunset.davemloft.net [SPARC64]: Create and use new macro, DCACHE_ALIASING_POSSIBLE. It determines whether D-cache aliasing issues are possible, and if not elide all of the D-cache flushing code and logic. Signed-off-by: David S. Miller ChangeSet@1.2181.42.2, 2005-03-26 20:04:49-03:00, acme@toy.ghostprotocols.net [NET] make all protos partially use sk_prot sk_alloc_slab becomes proto_register, that receives a struct proto not necessarily completely filled, but at least with the proto name, owner and obj_size (aka proto specific sock size), with this we can remove the struct sock sk_owner and sk_slab, using sk->sk_prot->{owner,slab} instead. This patch also makes sk_set_owner not necessary anymore, as at sk_alloc time we have now access to the struct proto onwer and slab members, so we can bump the module refcount exactly at sock allocation time. Another nice "side effect" is that this patch removes the generic sk_cachep slab cache, making the only last two protocols that used it use just kmalloc, informing a struct proto obj_size equal to sizeof(struct sock). Ah, almost forgot that with this patch it is very easy to use a slab cache, as it is now created at proto_register time, and all protocols need to use proto_register, so its just a matter of switching the second parameter of proto_register to '1', heck, this can be done even at module load time with some small additional patch. Another optimization that will be possible in the future is to move the sk_protocol and sk_type struct sock members to struct proto, but this has to wait for all protocols to move completely to sk_prot. This changeset also introduces /proc/net/protocols, that lists the registered protocols details, some may seem excessive, but I'd like to keep them while working on further struct sock hierarchy work and also to realize which protocols are old ones, i.e. that still use struct proto_ops, etc, yeah, this is a bit of an exaggeration, as all protos still use struct proto_ops, but in time the idea is to move all to use sk->sk_prot and make the proto_ops infrastructure be shared among all protos, reducing one level of indirection. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller ChangeSet@1.2181.24.8, 2005-03-26 15:03:00-08:00, davem@nuts.davemloft.net [SPARC64]: Eliminate g5 register usage from ultra.S Signed-off-by: David S. Miller ChangeSet@1.2181.42.1, 2005-03-26 13:47:42-08:00, davem@sunset.davemloft.net Merge sunset.davemloft.net:/home/davem/src/BK/network-2.6 into sunset.davemloft.net:/home/davem/src/BK/net-2.6 ChangeSet@1.2181.2.45, 2005-03-26 13:45:09-08:00, geert@linux-m68k.org [PATCH] 3dfx DRM depends on PCI 3dfx DRM depends on PCI Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.44, 2005-03-26 13:44:52-08:00, geert@linux-m68k.org [PATCH] TPM depends on PCI TPM depends on PCI. Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.43, 2005-03-26 13:44:35-08:00, geert@linux-m68k.org [PATCH] M68k: Add missing pieces of thread info TIF_MEMDIE support M68k: Add missing pieces of thread info TIF_MEMDIE support (introduced in 2.6.11-rc3) Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.42, 2005-03-26 13:44:17-08:00, geert@linux-m68k.org [PATCH] Mac NCR5380 SCSI: Fix bus error Mac NCR5380 SCSI: Fix bus error by passing the correct instance pointer to request_irq() Signed-off-by: Finn Thain Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.41, 2005-03-26 13:44:00-08:00, geert@linux-m68k.org [PATCH] M68k: IP checksum updates M68k: IP checksum updates: - ip_fast_csum() needs a "memory" constraint with new gcc versions - Do not always use d0 in ip_fast_csum(), leave register allocation to gcc - Fixed constraints of csum_fold() - Fixed constraints of csum_tcpudp_nofold() - Moved comment for csum_tcpudp_magic() to the right place Signed-off-by: Kars de Jong Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.40, 2005-03-26 13:43:42-08:00, geert@linux-m68k.org [PATCH] Sun-3/3x: Enable Sun partition tables support by default Sun-3/3x: Enable Sun partition tables support by default Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.39, 2005-03-26 13:43:26-08:00, geert@linux-m68k.org [PATCH] Zorro: replace printk() with pr_info() in drivers/zorro/zorro.c Zorro: This fixes the only printk() in drivers/zorro that has no KERN_* constant. Signed-off-by: James Nelson Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.38, 2005-03-26 13:43:09-08:00, geert@linux-m68k.org [PATCH] M68k/stdma: Replace sleep_on() with wait_event() M68k/stdma: Use wait_event() instead of the deprecated sleep_on() function. Since wait_event() expects the condition passed in to be the stopping condition, negate the current one. Signed-off-by: Nishanth Aravamudan Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.37, 2005-03-26 13:42:52-08:00, geert@linux-m68k.org [PATCH] M68k: Update signal delivery handling M68k: Update signal delivery handling, which was broken by the removal of notify_parent() in 2.6.9-rc2 Signed-off-by: Roman Zippel Signed-off-by: Geert Uytterhoeven Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.36, 2005-03-26 13:31:00-08:00, gerg@snapgear.com [PATCH] m68knommu: remove unused variables in mcfserial.c Removed unused/unrefernced mcfrs_tmp_buf array in mcfserial.c, and mutex associated with it. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.35, 2005-03-26 13:30:45-08:00, gerg@snapgear.com [PATCH] m68knommu: cleanup ColdFire specific trap handling asm code A couple of fixes to the ColdFire specific trap handling code: . do not clear the stack alignment bits in exception frame (in Lreturn) breaks return stack badly if applications work with stack un-aligned. . use generated asm-offsets instead of local offset definitions . use THREAD_SIZE definition instead of hard coding stack size (this is needed to support 4k stacks) . some instruction optimization (uses instructions that encode using ledd bytes) Patches originally from Philippe De Muyter Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.34, 2005-03-26 13:30:30-08:00, gerg@snapgear.com [PATCH] m68knommu: use generated asm-offsets in trap handlers Use generated asm-offsets in trap handling code, not local definitions of the offsets. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.33, 2005-03-26 13:30:15-08:00, gerg@snapgear.com [PATCH] m68knommu: move LED variable definitions for 5307 The LED specific setup code that used to be in platform specific start up code moved to board level code. This is specificaly for the Motorola/Freescale 5307 family. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.32, 2005-03-26 13:30:01-08:00, gerg@snapgear.com [PATCH] m68knommu: generate asm-offsets for thread_info struct Offsets into the thread_info structure should be auto generated, instead of using hard coded offset values. This adds entries to the asm-offsets generated for the m68knommu varients. Original patch submitted by Philippe De Muyter Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.31, 2005-03-26 13:29:46-08:00, gerg@snapgear.com [PATCH] m68knommu: move LED variable definitions for 5272 The LED specific setup code that used to be in platform specific start up code moved to board level code. This is specificaly for the Motorola/Freescale 5272 family. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.30, 2005-03-26 13:29:30-08:00, gerg@snapgear.com [PATCH] m68knommu: update MAINTAINERS entry Update entry in MAINTAINERS file to use my generic uclinux.org email address. And also clarify who is support the m68knommu architecture branch. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.29, 2005-03-26 13:29:14-08:00, gerg@snapgear.com [PATCH] m68knommu: 4k stack support Support for 4k kernel stacks on m68knommu. This cleans up the use of THREAD_SIZE, so it is used instead of hard coded size. Also remove local offset definitions, these are generated in asm-offsets. Patches original from Patch submitted by Philippe De Muyter Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.28, 2005-03-26 13:28:59-08:00, gerg@snapgear.com [PATCH] m68knommu: remove duplicate definition of THREAD_SIZE Remove definition of THREAD_SIZE in page.h for m68knommu. It is defined in thread_info.h Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.27, 2005-03-26 13:28:43-08:00, gerg@snapgear.com [PATCH] m68knommu: fix spelling mistakes in mafcache.h Fix some spelling mistakes in mcfcache.h Patch originally submitted by Geert Uytterhoeven Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.26, 2005-03-26 13:28:27-08:00, gerg@snapgear.com [PATCH] m68knommu: add missing KM_ enums Bring the km_type enum definitions into like with other architectures. Patch originally submitted by Leon Woestenberg Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.2.25, 2005-03-26 13:28:10-08:00, gerg@snapgear.com [PATCH] m68knommu: optimize trap handling asm code Optimize the trap handling asm code for m68knommu targets. Mostly this is using instructions that use fewer bytes to encode. Also remove local definitions for constants that are automatically generated in the asm-offsets header. Original patch came from Philippe De Muyter Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.41.1, 2005-03-25 21:05:43-08:00, gregkh@suse.de Merge suse.de:/home/greg/linux/BK/bleed-2.6 into suse.de:/home/greg/linux/BK/foo ChangeSet@1.2181.15.2, 2005-03-25 20:57:47-08:00, gregkh@suse.de merge mainline with 2.6.11.6 and 2.6.11.5 ChangeSet@1.2181.28.43, 2005-03-25 20:50:14-08:00, davem@sunset.davemloft.net [NET]: Forgot to remove doc file when I killed ethertap. Noticed by Chris Wright. Signed-off-by: David S. Miller ChangeSet@1.2181.40.5, 2005-03-25 19:07:07-08:00, torvalds@ppc970.osdl.org Undo VIA AGP TLB flush low-bits-zero patch It causes problems for people, and as DaveJ says: "Yes, it's clearly completely broken". Cset exclude: davej@delerium.kernelslacker.org|ChangeSet|20050223022752|60156 ChangeSet@1.2181.40.4, 2005-03-25 18:14:44-08:00, paulus@samba.org [PATCH] ppc64: prom.c: use pSeries reconfig notifier Use the pSeries_reconfig notifier list to fix up a device node which is about to be added. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.40.3, 2005-03-25 18:14:28-08:00, paulus@samba.org [PATCH] ppc64: introduce pSeries_reconfig.[ch] Move as much pSeries-specific DLPAR/hotplug code as possible into its own file, which is built only when pSeries support is enabled in the config. This new file is intended to contain support code for the "Dynamic Reconfiguration" option in the RISC Platform Architecture, which encompasses both PCI hotplug and dynamic logical partitioning (DLPAR). This patch mostly just moves code around, but the device node addition and removal API is slightly modified. In this way, of_add_node and of_remove_node are now responsible only for safely updating the device tree and global list, without all the other stuff like proc entries etc. of_add_node and of_remove_node have been renamed to of_attach_node and of_detach_node, respectively. This also adds the definitions and api for a notifier chain which is meant to be used by code that must act upon device node addition or removal. Patches to migrate code to the notifier api follow in this series. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.23.2, 2005-03-25 18:11:48-08:00, davidm@hpl.hp.com [IA64] minstate.h: fix stray backslash Trivial fix to drop a stray backslash in the MINSTATE_START_SAVE_MIN_VIRT macro. Signed-off-by: David Mosberger-Tang Signed-off-by: Tony Luck ChangeSet@1.2181.40.2, 2005-03-25 18:04:55-08:00, paulus@samba.org [PATCH] ppc64: make OF node fixup code usable at runtime At boot we recurse through the device tree "fixing up" various fields and properties in the device nodes. Long ago, to support DLPAR and hotplug, we largely duplicated some of this fixup code, the main data structures which are attached to the new device nodes. This patch introduces a helper function (prom_alloc) for handling allocations at both boot and runtime, kills most of the duplicated code, and makes finish_node, finish_node_interrupts, and interpret_pci_props suitable for use at runtime by converting them to use prom_alloc. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.40.1, 2005-03-25 18:04:37-08:00, paulus@samba.org [PATCH] ppc64: preliminary changes to OF fixup functions Preliminary modifications to support using some of the interpret_func family of functions at runtime. Changes the mem_start argument to be passed by reference, and the return type to int for error handling to be implemented in following patches. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.1982.168.38, 2005-03-25 18:00:30-08:00, chrisw@osdl.org Linux 2.6.11.6 TAG: v2.6.11.6 ChangeSet@1.1982.168.37, 2005-03-25 17:50:52-08:00, akpm@osdl.org [PATCH] Potential DOS in load_elf_library From: Herbert Xu Yichen Xie points out that load_elf_library can modify `elf_phdata' before freeing it. CAN-2005-0749 is assigned to this issue. Signed-off-by: Andrew Morton Signed-off-by: Chris Wright ChangeSet@1.1982.168.36, 2005-03-25 17:50:01-08:00, mlafon@arkoon.net [PATCH] Suspected information leak (mem pages) in ext2 From: "Mathieu Lafon" I think I have discovered a potential security problem in ext2: when a new directory is created, the ext2 block written to disk is not initialized. Included is a proposed patch for Linux 2.6 (ext2_make_empty() function): CAN-2005-0400 is assigned to this issue. Signed-off-by: Chris Wright ChangeSet@1.1982.168.35, 2005-03-25 17:49:05-08:00, meissner@suse.de [PATCH] Fix signedness problem at socket creation From: Marcel Holtmann CAN-2005-0750 is assigned to this issue ilja discovered potential local root exploit in bluetooth socket creation. This patch fixes a small signedness problem when creating the socket. Signed-off-by: Marcel Holtmann Signed-off-by: Chris Wright ChangeSet@1.1982.168.34, 2005-03-25 17:46:03-08:00, chrisw@osdl.org [PATCH] isofs: more defensive checks against corrupt isofs images Michal Zalewski discovers range checking flaws in iso9660 filesystem. http://marc.theaimsgroup.com/?l=bugtraq&m=111110067304783&w=2 CAN-2005-0815 is assigned to this issue. Some more defensive checks to keep corrupt isofs images from corrupting memory or causing Oops. Signed-off-by: Chris Wright ===== fs/isofs/rock.c 1.23 vs edited ===== ChangeSet@1.1982.168.33, 2005-03-25 17:45:24-08:00, torvalds@osdl.org [PATCH] isofs: more "corrupted iso image" error cases Michal Zalewski discovers range checking flaws in iso9660 filesystem. http://marc.theaimsgroup.com/?l=bugtraq&m=111110067304783&w=2 CAN-2005-0815 is assigned to this issue. From: Linus Torvalds isofs: more "corrupted iso image" error cases Thanks to Michal Zalewski for testing. Signed-off-by: Chris Wright ChangeSet@1.1982.168.32, 2005-03-25 17:44:34-08:00, torvalds@osdl.org [PATCH] isofs: Handle corupted rock-ridge info slightly better Michal Zalewski discovers range checking flaws in iso9660 filesystem. http://marc.theaimsgroup.com/?l=bugtraq&m=111110067304783&w=2 CAN-2005-0815 is assigned to this issue. From: Linus Torvalds isofs: Handle corupted rock-ridge info slightly better. Keyword here being 'slightly'. The code is a mess. Signed-off-by: Chris Wright ChangeSet@1.2181.38.5, 2005-03-26 11:16:56+11:00, airlied@starflyer.(none) drm: fixup pci ids Add new ATI PCI ID, and fixup i915GM one... Signed-off-by: Dave Airlie ChangeSet@1.2181.18.35, 2005-03-25 22:43:46+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2577/1: more ixp2000 comment work (typo fixes and annotations) Patch from Lennert Buytenhek Fix two more ixp2000 typos, and place some comments in ixp2000-regs.h to remind the developer that the virtual addresses of some of the io mappings have been hardcoded in various assembly files. (I've been bitten by this too many times now.) Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King ChangeSet@1.2181.18.34, 2005-03-25 21:06:29+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2507/1: work around ixp2400 erratum #66 Patch from Lennert Buytenhek Intel's ixp2400 can occasionally corrupt writes to its on-chip registers, leading to system crashes. The suggested workaround for this is to map in those registers using XCB=101 instead of XCB=000. Patch 2491/1 makes all mappings of ixp2000 on-chip I/O regions use section mappings. This patch makes all iotable section mappings use XCB=101. This patch depends on 2491/1. Kernel iotable mappings that are not 1MB-aligned, ioremap(), and userspace mappings using mmap() on /dev/mem will still generate XCB=000 mappings, which can trigger erratum #66 quite easily. For this, an extra patch (to arch/arm/mm/proc-xscale.S) is needed, which is maintained out-of-tree because it has been rejected in the past. See http://ixp2xxx.sourceforge.net/kernel/ and look for 'ixp2000-erratum66-ptes.diff'. 2.6.11-rc4 with 2491/1 plus this patch survives a week of heavy testing on my ENP-2611, while 2.6.11-rc4 vanilla doesn't stand a chance and crashes within 30 minutes under the same workload. Signed-off-by: Lennert BuytenhekSigned-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2198, 2005-03-25 14:58:46-06:00, stevef@smf-t23.(none) [CIFS] cleanup unnecessary casts, and redundant null pointer checks Suggested by Jesper Juhl, although slightly modified. Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2197, 2005-03-25 14:12:19-06:00, stevef@smf-t23.(none) [CIFS] add generic readv/writev and aio support. Suggested by Christoph Hellwig Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.28.42, 2005-03-25 09:34:52-08:00, hadi@cyberus.ca [PKT_SCHED]: action stats double dip Following patch fixes a scenario where we have forward compatibility but action stats being double dipped. Signed-off-by: David S. Miller ChangeSet@1.2181.28.41, 2005-03-25 09:25:23-08:00, mingo@elte.hu [XFRM]: xfrm_policy destructor fix the patch below fixes a bug that i encountered while running a PREEMPT_RT kernel, but i believe it should be fixed in the generic kernel too. xfrm_policy_kill() queues a destroyed policy structure to the GC list, and unlocks the policy->lock spinlock _after_ that point. This created a scenario where GC processing got to the new structure first, and kfree()d it - then the write_unlock_bh() was done on the already kfreed structure. There is no guarantee that GC processing will be done after policy->lock has been dropped and softirq processing has been enabled. Signed-off-by: Ingo Molnar Signed-off-by: David S. Miller ChangeSet@1.2181.24.7, 2005-03-24 20:48:11-08:00, davem@nuts.davemloft.net [SPARC64]: Eliminate g5 register usage from switch_to(). Signed-off-by: David S. Miller ChangeSet@1.2181.33.3, 2005-03-24 23:32:42-05:00, Carlos.Pardo@siliconimage.com [PATCH] sata_sil: Fix FIFO PCI Bus Arbitration This patch set default values for the FIFO PCI Bus Arbitration to avoid data corruption. The root cause is due to our PCI bus master handling mismatch with the chipset PCI bridge during DMA xfer (write data to the device). The patch is to setup the DMA fifo threshold so that there is no chance for the DMA engine to change protocol. We have seen this problem only on one motherboard. Signed-off-by: Silicon Image Corporation Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.12, 2005-03-24 22:56:42-05:00, bunk@stusta.de [PATCH] drivers/net/wireless/airo.c: correct a wrong check The Coverity checker correctly noted that this condition can't ever be fulfilled. This patch changes it to what it should have been. Signed-off-by: Adrian Bunk Signed-off-by: Jeff Garzik ChangeSet@1.2181.24.6, 2005-03-24 17:59:43-08:00, davem@nuts.davemloft.net [SPARC64]: Move rwsem helpers into asm file. They were all purely inline asm statements anyways. Signed-off-by: David S. Miller ChangeSet@1.2196, 2005-03-24 19:33:28-06:00, stevef@smf-t23.(none) [CIFS] Check if cifs demultiplex thread valid (not exited, or exiting) before we wake it on unmount (otherwise can cause oops in send_sig). Pointed out by Ameet Paranjape Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.4.75, 2005-03-24 16:11:28-08:00, phil@ipom.com [PATCH] USB Storage: Remove dup in unusual_devs Matthew Dharm pointed out that your BK tree somehow got 2 versions of an unusual_devs entry in it. One was from a patch in January and the other was from an automerge. I dug around in bk - which I find to be the most counter-intuitive program in the world - and generated a patch against your latest BK tree, and I've attached it. It just removes the dup. Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.74, 2005-03-24 16:11:11-08:00, colin@colino.net [PATCH] USB: fix harmful typos in zd1201.c I was looking around to see why monitor mode wouldn't work using the zd1201 driver, and spotted these obvious typos. I think you can safely apply :) Signed-off-by: Colin Leroy Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.73, 2005-03-24 15:31:46-08:00, david-b@pacbell.net [PATCH] USB: ehci split ISO fixes (full speed audio etc) This contains patches to the EHCI driver for the full speed isochronous transfer support: - The sitd->hw_buf[1] and sitd->hw_buf_hi[1] fields were not correctly initialized, affecting transfers which crossed 4K boundaries. Helps resolve some "buzz" in audio playback. - Correctly rounds 188-byte OUT transfers to fit into a single packet ... no split transfers needed. Resolves other "buzz" in audio OUT streams. - Fixes a cut'n'paste error in the logic to check for microframe scheduling collisions with a given transaction translator. The error caused full speed iso tds to be treated in part as if they were high speed iso tds, which could oops. - The split transaction state flag doesn't indicate an error, and it should be ignored when deciding if (IN) transfers had errors. - Remove some code rejecting full speed iso IN transfers; it's not fully working yet though. Plus a some minor cleanups: reporting the iso start frame in the relevant units, making a debug message more consistent. Signed-off-by: Karsten Wiese Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.72, 2005-03-24 15:31:29-08:00, david-b@pacbell.net [PATCH] USB: usbnet uses netif_msg_*() ethtool filtering This converts most of the usbnet code to actually use the ethtool message flags. The ASIX code is left untouched, since there are a bunch of patches pending there ... that's where the remaining handful of "sparse -Wbitwise" warnings come from. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.71, 2005-03-24 15:31:13-08:00, david-b@pacbell.net [PATCH] USB: usbnet minor bugfixes Two bugfixes to usbnet. The important one is that it's OK when the minidriver doesn't want to use a status endpoint. That shouldn't be treated as an error ... errors prevent probe() from succeeding! The minor fix is that software driven interface shutdown should neither be accounted as an error, nor generate a diagnostic. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.70, 2005-03-24 15:30:57-08:00, david-b@pacbell.net [PATCH] USB: pegasus uses netif_msg_*() filters This updates the messaging for the pegasus driver: - Use driver model diagnostics or printk using the interface name for all diagnostic messages ... not dbg()/warn()/err(). - Almost everywhere, use the netif_msg_XXX() macros to check the message control bitmask maintained by ethtool. The default mask is initialized using a new "message_level" module parameter. Also: - Removes the needless PEGASUS_RUNNING flag, replacing it with the standard netdevice mechanism. - Cleaner access for unaligned values. Not all processors spend silicon to support them like x86 does. - Adds a few "sparse" fixes. - Saves the return values for the requests that manipulate chip registers ... doesn't yet check them, but at least anyone looking at the code (e.g. to find out why the link check task is wedged...) will see where those failure modes are ignored. Currently the errors may be reported by printk, but the netif_msg_*() filters make that an even worse alert mechanism. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.69, 2005-03-24 15:30:41-08:00, david-b@pacbell.net [PATCH] USB: usb rndis gadget sparse fixes [4/5] This is a bunch of "sparse" fixups for the RNDIS code, saying it's little-endian on the wire and swapping cpu_to_le32() calls with more-correct le32_to_cpu() ones. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.68, 2005-03-24 15:30:24-08:00, david-b@pacbell.net [PATCH] USB: gadget zero sparse fixes [5/5] Some "sparse" updates for the Gadget Zero driver, preparing it to expect SETUP packets to be little endian. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.67, 2005-03-24 15:30:07-08:00, david-b@pacbell.net [PATCH] USB: usb gadgetfs sparse fixes [3/5] Many "sparse" updates for GadgetFS driver, preparing it to expect SETUP packets to be little endian. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.66, 2005-03-24 15:29:50-08:00, stern@rowland.harvard.edu [PATCH] USB: fix usb file_storage gadget sparse fixes [2/5] On Sun, 20 Mar 2005, David Brownell wrote: > > - gfile-0319 ... gets rid of most byteorder warnings from the > > file_storage gadget driver (except for SETUP packet fields) This patch causes an undeclared variable error when verbose debugging is enabled. Please add the following on top of it. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.65, 2005-03-24 15:29:34-08:00, david-b@pacbell.net [PATCH] USB: usb file_storage gadget sparse fixes [2/5] Some "sparse" updates for the File-backed Storage Gadget driver, preparing it to expect SETUP packets to be little endian. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.64, 2005-03-24 15:29:17-08:00, david-b@pacbell.net [PATCH] USB: usb gadget misc sparse fixes [1/5] This lets the file_storage gadget appear again in Kconfig, and fixes some minor "sparse" warnings. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.26.4, 2005-03-25 00:26:46+01:00, marcel@holtmann.org [Bluetooth] Fix signedness problem at socket creation This patch fixes a small signedness problem when creating the socket. Signed-off-by: Marcel Holtmann ChangeSet@1.2181.4.63, 2005-03-24 15:14:10-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: remove RW_DETECT from being a config option This patch started life as as454b from Alan Stern. It has been rediffed against the tip, including updates for the way unusual_devs flags are defined. This patch removes the Kconfig option USB_STORAGE_RW_DETECT. That option was used to enable/supress the attempt to detect if a device was write protected. It seems that the vast majority of devices properly respond to the latest algorithm for making that determination. So now we move to excluding only those devices that can't handle it. We accomplish this via the unusual_devs.h list. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.62, 2005-03-24 15:13:53-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: combine waitqueues This patch started life as as476, and has been rediffed against the tip. However, that was a few days ago. This patch combines the two separate waitqueue heads used by the scsi-scanning thread and the device-reset routine into one. After all, until the scanning thread is through waiting there will be no SCSI devices and hence no device resets. Once the scanning thread is done waiting, the waitqueue can be used by the reset logic -- so even if the act of scanning produces resets, we're fine. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.61, 2005-03-24 15:13:36-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: allow disconnect to complete faster This patch started life as as476 from Alan Stern. It has been rediffed against the tip, tho that was a few days ago. This patch makes the disconnect() routine not wait for the control and scanning threads to exit. This may not seem important now, but it will become important later: We would end up with a deadlock if disconnect() (which is called with the device locked) was waiting for the control thread to exit, while the control thread was waiting to lock the device so it could do an autosuspend. It's necessary to make sure that the host and us_data structures aren't deallocated before the control and scanning threads are through with them. This is done by calling scsi_host_get and scsi_host_put at the start and end of each thread, before signalling that the threads are running. Since the probe() and disconnect() routines cannot run concurrently (guaranteed to us by the USB core), this method will guarantee the structures are not deallocated too soon. While there's nothing wrong with leaving the threads alive after disconnect() returns, there would be a real problem if the threads were still alive when usb_stor_exit returned! So now usb_stor_exit has to wait to make sure all the threads have died. Apparently the only safe way for one thread to signal another while exiting is to use complete_and_exit, which we've been doing. So the patch adds a new driver-wide struct completion, named threads_gone, and each thread signals it while exiting. usb_stor_exit must call wait_for_completion the appropriate number of times, and that number is stored in a new counter named total_threads. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.60, 2005-03-24 15:13:20-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: exit control thread immediately upon disconnect This patch started life as as475 from Alan Stern. It has been rediffed against the tip, tho that was several days ago. This patch causes the main control thread to exit as soon as possible, i.e., as soon as the DISCONNECTING flag is set. It no longer waits for an explicit exit command, one with srb == NULL. There won't be any bad implications for our interaction with the SCSI midlayer, because once the DISCONNECTING flag is set we fail every submitted command immediately in the queuecommand routine. And if a command manages to squeeze through the crack (submitted and accepted before we disconnect but not yet processed), the SCSI midlayer will cancel it automatically when we remove the host. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.59, 2005-03-24 15:13:02-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: make usb-storage structures refcounted by SCSI This patch started life as as474 from Alan Stern. It's been rediffed against the tip, tho that is now several days old. This patch changes the way our private struct us_data is allocated; now it gets stored at the end of the Scsi_Host rather than separately. That's what the hostdata field is intended for, and this is how other low-level host drivers operate. In order to convert between us_data and the corresponding Scsi_Host I added two new inline routines: us_to_host and host_to_us. (The conversion actually should be quicker than before by a microscopic amount, because now it only involves adding an offset whereas before it involved dereferencing a pointer.) The main advantage is that the host is refcounted, so now our us_data automatically is too. Although that doesn't matter at the moment, it will matter later on when the control thread may need to outlive the disconnect callback. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.58, 2005-03-24 14:45:05-08:00, bunk@stusta.de [PATCH] drivers/usb/core/devices.c: small corrections total_written is used at places where it can't have any value different from 0. This patch is partially based on findings of the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.24.5, 2005-03-24 14:44:07-08:00, davem@nuts.davemloft.net [SPARC64]: Eliminate g5 register usage in rwsem. Also, semaphore stuff needs a g5 register clobber until we move over to actually passing -ffixed-g5 to the compiler. Signed-off-by: David S. Miller ChangeSet@1.2181.4.57, 2005-03-24 14:34:40-08:00, rkagan@mail.ru [PATCH] drivers/usb/core/usb.c: add MODALIAS env var to hotplug The patch below adds MODALIAS environment variable to usb hotplug callout, allowing for its straightforward use with modprobe. Signed-off-by: Roman Kagan Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.56, 2005-03-24 14:29:06-08:00, david-b@pacbell.net [PATCH] USB: pxa25x udc updates, mostly PM This has various updates to the PXA 21x/25x/26x UDC driver. - Implement the "new" pullup() and vbus_session() methods, and use them to keep the UDC 48 MHz clock off much of the time. * Reworked that ugly Lubbock VBUS IRQ code. Claim both IRQs, enable only one at a time; clock the UDC only when VBUS is present. (And get rid of rude runtime messages.) * Implement driver model suspend() and resume() calls. When this device suspends, it clocks off the UDC. On boards that support it (including Zaurus clamshells, but not Lubbock) the D+ pullup is disabled, so the host won't see the device. - Hmm, the "latest" errata defined some "Must Be One" bits. OK. - Change the LED support for debugging. It stopped compiling for Lubbock a while back. This switches to the standard LED calls (so it can work on non-Lubbock hardware), removes the EP0 calls (not very useful any more), and for Lubbock now initializes the hex leds (U-Boot doesn't enable them, BLOB did). - "sparse" updates, and get rid of a warning that's pointless unless someone's working on DMA support; Tested on Lubbock (VBUS sensing but no pullup) and some Zaurus clamshells (pullup but no VBUS). Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.55, 2005-03-24 14:28:49-08:00, david-b@pacbell.net [PATCH] USB: ohci-omap update (mostly clock gating) This syncs the OMAP OHCI code with the latest from the OMAP tree. The main changes are updated clock gating support (goes with some clock tree updates, submitted separately) and two minor cleanups: the platform device resources use physical addressing, not the static i/o mapped addresses; and there's no longer a need for a (noexistent in mainstream) separate header file. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.54, 2005-03-24 14:28:28-08:00, colin@colino.net [PATCH] USB: fix missing hunk in drivers/usb/Makefile I see there's been a driver for zd1201 added in drivers/usb/net/. There's a hunk missing in drivers/usb/Makefile, the driver doesn't get built in nothing else in drivers/usb/net is configured in : Signed-off-by: Colin Leroy Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.11.88, 2005-03-24 21:10:51+00:00, rmk@flint.arm.linux.org.uk [SERIAL] Remove SERIAL_INLINE, and move debug macro to 8250_pci.c This removes the unnecessary SERIAL_INLINE macro, and moves SERIAL_DEBUG_PCI to 8250_pci.c Signed-off-by: Russell King ChangeSet@1.2181.28.39, 2005-03-24 10:29:25-08:00, davem@sunset.davemloft.net [TG3]: Update driver version and reldate. Signed-off-by: David S. Miller ChangeSet@1.2195, 2005-03-24 12:13:35-06:00, stevef@smfhome.smfdom [CIFS] Display pool sizes in cifs stats Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.18.33, 2005-03-24 18:07:49+00:00, rmk@flint.arm.linux.org.uk [ARM] Don't call force_sig_info() for kernel-mode exceptions. If an exception happens during boot, we may call force_sig_info(), which tries to allocate memory. If the memory subsystem has not been setup, we oops. However, we have lost the real cause of the problem (the original exception.) Avoid this problem by not calling force_sig_info() for kernel mode exceptions. Signed-off-by: Russell King ChangeSet@1.2181.18.32, 2005-03-24 17:49:23+00:00, rmk@flint.arm.linux.org.uk [ARM] Move alignment_trap/zero_fp macros into usr_entry macro. Signed-off-by: Russell King ChangeSet@1.2181.18.31, 2005-03-24 17:34:28+00:00, rmk@flint.arm.linux.org.uk [ARM] mach-types update Signed-off-by: Russell King ChangeSet@1.2181.11.87, 2005-03-24 16:22:25+00:00, rmk@flint.arm.linux.org.uk [SERIAL] Add UART_CAP_UUE Rather than relying on port->type == PORT_XSCALE, determine when we need to set UUE via a capability. Signed-off-by: Russell King ChangeSet@1.2181.13.3, 2005-03-24 08:06:44-06:00, shaggy@austin.ibm.com JFS: remove aops from directory inodes jfs no longer uses a directory inode's address space. Clean up the code by removing aops for directories altogether Signed-off-by: Dave Kleikamp ChangeSet@1.2181.11.86, 2005-03-24 12:18:28+00:00, rmk@flint.arm.linux.org.uk [SERIAL] au1x00_uart: remove duplicate serial registration functions au1x00_uart doesn't need to provide duplicated 8250 registration functions. Remove them. Signed-off-by: Russell King ChangeSet@1.2181.11.85, 2005-03-24 11:58:57+00:00, rmk@flint.arm.linux.org.uk [SERIAL] Set port.dev to PCMCIA device. Since serial_cs provides power management callbacks, we should not use the fallback methods in 8250.c. We tell 8250.c about this when we associate a struct device with the port, and, since we now have such a beast, use it. This also makes sysfs indicate which port is associated with which PCMCIA device. Signed-off-by: Russell King ChangeSet@1.2181.11.84, 2005-03-24 11:34:14+00:00, rmk@flint.arm.linux.org.uk [SERIAL] Allow drivers to use uart_match_port This removes the duplicate of uart_match_port in 8250.c Signed-off-by: Russell King ChangeSet@1.2181.31.54, 2005-03-24 12:13:42+01:00, perex@suse.cz ALSA 1.0.9rc2 ChangeSet@1.2181.31.53, 2005-03-24 12:00:49+01:00, perex@suse.cz ALSA CVS update ALSA Version release: 1.0.9rc2 Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.52, 2005-03-24 12:00:05+01:00, perex@suse.cz [ALSA] seq - fix local variable initialization ALSA sequencer This patch re-adds the initialization of callbacks and pcallbacks that was accidentally removed in the last revision. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.51, 2005-03-24 11:59:21+01:00, perex@suse.cz [ALSA] usb - change timeout of USB control/bulk msg functions to msecs USB generic driver This changes the timeout in the remaining (indirect) calls to usb_control/bulk_msg from jiffies to msecs. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.50, 2005-03-24 11:58:40+01:00, perex@suse.cz [ALSA] cs4281 - fix typos in the case gameport is disabled CS4281 driver This patch fixes the wrong names of the dummy gameport functions used when CONFIG_GAMEPORT isn't set. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.49, 2005-03-24 11:57:55+01:00, perex@suse.cz [ALSA] Add AC97_SCAP_NO_SPDIF flag AC97 Codec,ATIIXP driver,au88x0 driver,EMU10K1/EMU10K2 driver Added a new flag, AC97_SCAP_NO_SPDIF, to prevent to build the SPDIF-related controls on ac97 codec. This flag is used when the sound chip has its native SPDIF support and it conflicts with the one of AC97 codec. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.48, 2005-03-24 11:57:09+01:00, perex@suse.cz [ALSA] Fix EFX voice allocation/preparation EMU10K1/EMU10K2 driver Fixed a bug (possibly Oops) in allocation/preparation of EFX voices The invalid voice pointer was accessed when voices are allocated over the voice table boundary. The patch includes a small clean-up & optimization. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.47, 2005-03-24 11:56:28+01:00, perex@suse.cz [ALSA] Fix Oops in snd_emu10k1_add_controls EMU10K1/EMU10K2 driver Fixed Oops in snd_emu101k_add_controls (introduced in the last patch for stack usage reduction). Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.46, 2005-03-24 11:55:42+01:00, perex@suse.cz [ALSA] Clean up the chip detection EMU10K1/EMU10K2 driver Minor clean-ups of the chip detectoin code. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.45, 2005-03-24 11:54:58+01:00, perex@suse.cz [ALSA] Add Mono volume controls for ALC260 HDA Codec driver Added Mono volume controls for ALC260 codec. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.44, 2005-03-24 11:54:18+01:00, perex@suse.cz [ALSA] Add AD1986A support HDA generic driver,HDA Codec driver Added the patch for Analog Device AD1986A codec. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.43, 2005-03-24 11:53:32+01:00, perex@suse.cz [ALSA] correct comment for setting widget output HDA Codec driver This patch has no real logical change, it simply correct the comment. Signed-off-by: ChenLi Tien Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.42, 2005-03-24 11:52:53+01:00, perex@suse.cz [ALSA] Fixes AC3 output on Audigy2 sound cards EMU10K1/EMU10K2 driver This patch adds a DSP patch to fix an spdif_bug on some Audigy2 cards. Signed-off-by: James Courtier-Dutton Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.41, 2005-03-24 11:52:11+01:00, perex@suse.cz [ALSA] Add framework for better audigy sound card capabilities selection EMU10K1/EMU10K2 driver This patch adds more options to help identify all the many different creative sound cards. It will eventually be used to control features more finely. Signed-off-by: James Courtier-Dutton Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.40, 2005-03-24 11:51:23+01:00, perex@suse.cz [ALSA] Replace with macros for gameport initialization ALSA Core,ALS4000 driver,AZT3328 driver,CMIPCI driver,CS4281 driver ENS1370/1+ driver,ES1938 driver,ES1968 driver,SonicVibes driver VIA82xx driver,au88x0 driver,CS46xx driver,Trident driver,YMFPCI driver Use some macros for gameport initialization. This makes much easier to write the compatible layer for the old gameport API in alsa-driver tree. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.39, 2005-03-24 11:50:38+01:00, perex@suse.cz [ALSA] Fix Oops with joystick support YMFPCI driver Fix Oops when joystick is enabled. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.38, 2005-03-24 11:49:58+01:00, perex@suse.cz [ALSA] Fix Oops with joystick support ES1968 driver Fix Oops when joystick is enabled. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.37, 2005-03-24 11:48:53+01:00, perex@suse.cz [ALSA] Use vprintk() ALSA Core Use vprintk() instead of printk with a temporary line buffer. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.36, 2005-03-24 11:46:07+01:00, perex@suse.cz [ALSA] Reduce stack usage Control Midlevel,ALSA Core,PCM Midlevel,Timer Midlevel ALSA<-OSS emulation,ALSA sequencer,ALSA<-OSS sequencer AK4XXX AD/DA converters,GUS Library,Wavefront drivers EMU10K1/EMU10K2 driver,HDA generic driver,MIXART driver PDAudioCF driver,USB generic driver Reduce the stack usage, mostly by replacing large structs with kmalloc'ed instances. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.35, 2005-03-24 11:44:51+01:00, perex@suse.cz [ALSA] Remove unnecessary ac97 init code VIA82xx driver,VIA82xx-modem driver Removed unnecessary ac97 init code in snd_via82xx_chip_init(). This reduces eventually the big stack usage, too. Signed-off-by: Takashi Iwai ChangeSet@1.2181.38.4, 2005-03-24 21:18:38+11:00, airlied@starflyer.(none) agp: export agp_find_bridge for drm Signed-off-by: Dave Airlie ChangeSet@1.2181.38.3, 2005-03-24 21:15:09+11:00, airlied@starflyer.(none) drm: fix issue where agp is acquired before agp_init With integrated chipsets ala i865 the X server acquires the bridge for 2D operations then the DRM acquires it for 3D kaboom.. Based on patch from Brice Goglin but I think this patch is safer if it can't find a bridge it acquires it. Tested on i865 (i830/i915) and Radeon on Xorg CVS and XFree86 4.3.0 Signed-off-by: Dave Airlie ChangeSet@1.2181.4.53, 2005-03-24 00:35:43-08:00, gregkh@suse.de USB: mark usb-serial interface GPL only This is done because you have to use a GPL only licensed .h file in order to write a usb-serial driver. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.38.2, 2005-03-24 17:40:57+11:00, airlied@starflyer.(none) drm: issue with unique for XFree86 4.3 backwards compatibility This got broken at some stage not sure when exactly... but it caused XFree86 4.3 issues and I had to install sarge... Signed-off-by: Dave Airlie ChangeSet@1.2181.38.1, 2005-03-24 17:27:39+11:00, airlied@starflyer.(none) verify_area is deprecated, replaced by access_ok. Seems I missed this one when I did the big overall conversion. Signed-off-by: Jesper Juhl Signed-off-by: Dave Airlie ChangeSet@1.2181.28.38, 2005-03-23 21:18:39-08:00, hadi@cyberus.ca [PKT_SCHED]: Use proper attritbute for action stats. Action stats were being sent on the wrong TLV. Patch attached. Thanks to Andy Furniss for finding and helping debug this. Signed-off-by: David S. Miller ChangeSet@1.2181.4.52, 2005-03-23 20:34:15-08:00, zaitcev@redhat.com [PATCH] USB: Add myself to MAINTAINERS A Jan Kasprzak asked a few days ago to have a MAINTAINERS entry for ub. This patch updates my entries in MAINTAINERS (ub & ymfpci). Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.51, 2005-03-23 20:33:55-08:00, zaitcev@redhat.com [PATCH] USB: usbmon - document and kill pipe from API Someone at linux-usb-devel observed that usbmon should not leak the concept of pipe to the user space. This patch makes a compromise between having too many words in the line and the pipe. It breaks API, but I doubt anyone has any usbmon tools in circulation at this time, so I decided not to move to a different filename. Also, I decided to add a document which describes what is going on. I was reluctant to do it because I doubt my ability to keep it synchronized and updated, but there were requests to have one. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.50, 2005-03-23 20:33:34-08:00, zaitcev@redhat.com [PATCH] USB: Fix baud selection in mct_u232 This is a patch by Hansjoerg Lipp to fix coding error in divisor calculation. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.49, 2005-03-23 20:33:13-08:00, zaitcev@redhat.com [PATCH] USB: ub static patch I'm pretty sure this comes from Bunk originally, but I didn't keep the original, so only one signed-off-by is present. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.48, 2005-03-23 15:26:40-08:00, gregkh@suse.de USB: fix bug in visor driver with throttle/unthrottle causing oopses. Thanks to Mark Lord for reporting this and helping with testing. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.13.2, 2005-03-23 16:43:50-06:00, shaggy@austin.ibm.com JFS: Don't allow xtLookup to run against directory with inline data Signed-off-by: Dave Kleikamp ChangeSet@1.2181.27.16, 2005-03-23 16:32:55-06:00, jejb@mulgrave.(none) 3ware driver update - Increase max ioctl buffer size to 512 sectors. - Make tw_scsi_queue() return 0 for 'Unknown scsi opcode'. - Fix tw_remove() to free irq handler/unregister_chrdev() before shutting down the card. - Change to new 'change_queue_depth' api (from James). - Fix 'handled=1' ISR usage, remove bogus IRQ check (from Jeff). Signed-off-by: Adam Radford Signed-off-by: James Bottomley ChangeSet@1.2181.28.37, 2005-03-23 12:32:50-08:00, davem@sunset.davemloft.net Merge bk://kernel.bkbits.net/acme/net-2.6 into sunset.davemloft.net:/home/davem/src/BK/acme-2.6 ChangeSet@1.2181.28.36, 2005-03-23 12:27:01-08:00, davem@sunset.davemloft.net [TG3]: Missing counter bump in tigon3_4gb_hwbug_workaround(). Reported by Adrian Bunk. Signed-off-by: David S. Miller ChangeSet@1.2181.28.35, 2005-03-23 12:19:14-08:00, bunk@stusta.de [EQL]: Kill dead code This patch removes some obviously dead code found by the Coverity checker. Signed-off-by: Adrian Bunk Signed-off-by: David S. Miller ChangeSet@1.2181.28.34, 2005-03-23 12:17:47-08:00, herbert@gondor.apana.org.au [IPV4]: Clear DF bit in ip_fragment fast path It is possible for ip_fragment() to send out head fragments with both DF and MF set for packets with local_df set to true. This is because the fast path tries to only modify the MF bit of the head fragment. Since the offset is always zero for the head fragment, and we know that DF should be cleared in case of local_df, we can change |= to a straight assignment. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.28.33, 2005-03-23 12:16:53-08:00, herbert@gondor.apana.org.au [IPV4]: Check mtu instead of frag_list in ip_push_pending_frames() I still didn't like the fact that ip_append_data was the only user of dst_pmtu :) So I went looking for bugs in the surrounding code. I managed to find something in ip_push_pending_frames. When dst_mtu < dst_pmtu - IPsec overhead (which can be caused by PMTU discovery within an IPsec tunnel), and we transmit a packet that's longer than dst_mtu but shorter than dst_pmtu - IPsec overhead, then the DF bit will be incorrectly set in the inner IP header. This will cause the packet to be dropped when it hits the router that generated the original PMTU event. Unfortunately the ICMP packet coming back doesn't tell us anything new so the next time we send a packet we will do exactly the same thing. The fix is similar to what we did in ip_output. Instead of checking whether frag_list is empty, we check the condition skb->len <= dst_mtu directly and set the DF bit based on that. We can enumerate all the possibilities to see that this is correct. If skb->len <= dst_mtu and frag_list is empty then this does the samething as before and is obviously correct. If skb->len <= dst_mtu and frag_list is non-empty then it implies that dst_pmtu has increased since the fragments were constructed as dst_pmtu = dst_mtu + IPsec overhead. So the skb will now fit within a single fragment which means that setting DF is correct. The fragments will be merged by skb_linearise in dev_queue_xmit. If skb->len > dst_mtu and frag_list is non-empty then again this maintains the status quo. If skb->len > dst_mtu and frag_list is empty then we will leave the DF bit clear as the packet will need to be fragmented between the remote IPsec gateway and the final destination. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.28.32, 2005-03-23 12:13:40-08:00, herbert@gondor.apana.org.au [CRYPTO]: Remap when walk_out crosses page in crypt() This is needed so that we can keep the in_place assignment outside the inner loop. Without this in pathalogical situations we can start out having walk_out being different from walk_in, but when walk_out crosses a page it may converge with walk_in. Signed-off-by: David S. Miller ChangeSet@1.2181.28.31, 2005-03-23 12:12:46-08:00, herbert@gondor.apana.org.au [CRYPTO]: Split cbc_process into encrypt/decrypt Rather than taking a branch on the fast path, we might as well split cbc_process into encrypt and decrypt since they don't share anything in common. We can get rid of the cryptfn argument too. I'll do that next. Signed-off-by: David S. Miller ChangeSet@1.2181.28.30, 2005-03-23 12:11:58-08:00, herbert@gondor.apana.org.au [CRYPTO]: Kill obsolete iv check in cbc_process() Here's some more optimisations plus a bug fix for a pathological case where in_place might not be set correctly which can't happen with any of the current users. Here is the first one: We have long since stopped using a null cit_iv as a means of doing null encryption. In fact it doesn't work here anyway since we need to copy src into dst to achieve null encryption. No user of cbc_encrypt_iv/cbc_decrypt_iv does this either so let's just get rid of this check which is sitting in the fast path. Signed-off-by: David S. Miller ChangeSet@1.2181.28.29, 2005-03-23 12:10:18-08:00, herbert@gondor.apana.org.au [CRYPTO]: Fix walk->data handling The problem is that walk->data wasn't being incremented anymore after my last change. This patch should fix it up. Signed-off-by: David S. Miller ChangeSet@1.2181.28.28, 2005-03-23 12:09:27-08:00, herbert@gondor.apana.org.au [CRYPTO]: Optimise kmap calls in crypt() Perform kmap once (or twice if the buffer is not aligned correctly) per page in crypt() instead of the current code which does it once per block. Consequently it will yield once per page instead of once per block. Signed-off-by: David S. Miller ChangeSet@1.2181.28.27, 2005-03-23 12:08:48-08:00, herbert@gondor.apana.org.au [CRYPTO]: Eliminate most calls to scatterwalk_copychunks from crypt() Only call scatterwalk_copychunks when the block straddles a page boundary. This allows crypt() to skip the out-of-line call most of the time. Signed-off-by: David S. Miller ChangeSet@1.2181.28.26, 2005-03-23 12:08:10-08:00, herbert@gondor.apana.org.au [CRYPTO]: Split src/dst handling out from crypt() Move src/dst handling from crypt() into the helpers prepare_src, prepare_dst, complete_src and complete_dst. complete_src doesn't actually do anything at the moment but is included for completeness. This sets the stage for further optimisations down the track without polluting crypt() itself. These helpers don't belong in scatterwalk.[ch] since they only help the particular way that crypt() is walking the scatter lists. Signed-off-by: David S. Miller ChangeSet@1.2181.28.25, 2005-03-23 12:07:31-08:00, herbert@gondor.apana.org.au [CRYPTO]: Handle in_place flag in crypt() Move the handling of in_place into crypt() itself. This means that we only need two temporary buffers instead of three. It also allows us to simplify the check in scatterwalk_samebuf. Signed-off-by: David S. Miller ChangeSet@1.2181.28.24, 2005-03-23 12:06:51-08:00, herbert@gondor.apana.org.au [CRYPTO]: Do scatterwalk_whichbuf inline. scatterwalk_whichbuf is called once for each block which could be as small as 8/16 bytes. So it makes sense to do that work inline. It's also a bit inflexible since we may want to use the temporary buffer even if the block doesn't cross page boundaries. In particular, we want to do that when the source and destination are the same. So let's replace it with scatterwalk_across_pages. I've also simplified the check in scatterwalk_across_pages. It is sufficient to only check len_this_page. Signed-off-by: David S. Miller ChangeSet@1.2181.28.23, 2005-03-23 11:52:39-08:00, davem@sunset.davemloft.net [NETROM]: net/netrom.h now needs net/sock. Signed-off-by: David S. Miller ChangeSet@1.2181.28.22, 2005-03-23 11:44:24-08:00, davem@sunset.davemloft.net [TG3]: Add missing CHIPREV_5750_{A,B}X defines. Signed-off-by: David S. Miller ChangeSet@1.2181.28.21, 2005-03-23 11:16:37-08:00, ralf@linux-mips.org [ROSE]: Get rid of sk_protinfo use Below patch puts struct sock into rose_cb to get rid of the need for the use of sk_protinfo in rose_sk(). While we're touching the data structure convert it from a typedef into a struct. Signed-off-by: David S. Miller ChangeSet@1.2181.28.20, 2005-03-23 11:15:42-08:00, ralf@linux-mips.org [NETROM]: Get rid of sk_protinfo use Below patch puts struct sock into nr_cb to get rid of the need for the use of sk_protinfo in nr_sk(). While we're touching the data structure convert it from a typedef into a struct. Signed-off-by: David S. Miller ChangeSet@1.2181.28.19, 2005-03-23 11:14:06-08:00, mchan@broadcom.com [TG3]: Add Broadcom copyright. Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.18, 2005-03-23 11:12:04-08:00, mchan@broadcom.com [TG3]: Fix ethtool set functions Fix all relevant ethtool set functions to properly handle the not-netif_running() case. In most cases, the new settings are accepted without setting the hardware if not-netif_running(). The new settings will take effect when the device is subsequently brought up. tg3_nway_reset() is the exception where it will return -EAGAIN if not-netif_running(). Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.17, 2005-03-23 11:09:49-08:00, mchan@broadcom.com [TG3]: Fix jumbo frames phy settings Fix jumbo frame settings on all copper phys that support jumbo frames by setting the fifo elasticity bit. This setting is for the phy's tx fifo to properly handle jumbo frames. Note that a similar jumbo frame fix for the phy's rx fifo was made to tg3 in the past. Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.16, 2005-03-23 11:07:37-08:00, mchan@broadcom.com [TG3]: Add unstable PLL workaround for 5750 Add unstable PLL clock workaround for 5750 Ax and Bx devices. The workaround code is run just before entering D3hot state. Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.15, 2005-03-23 11:05:03-08:00, mchan@broadcom.com [TG3]: Flush status block in tg3_interrupt() Add register read of PCI state register in tg3_interrupt() if status block's updated bit is not set. This will flush the status block and confirm whether the interrupt is ours or not. PCI ordering rules allow the interrupt to arrive at the CPU ahead of the status block that may be posted at the chipset. Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.14, 2005-03-23 11:01:22-08:00, mchan@broadcom.com [TG3]: Add 5705_plus flag Add a 5705_plus flag to indicate the device is 5705, 5750, or future chips that all share the same basic architecture. This makes it easier to add support for future devices. Signed-off-by: Michael Chan ACKed-by: Jeff Garzik Signed-off-by: David S. Miller ChangeSet@1.2181.28.13, 2005-03-23 10:58:41-08:00, chas@cmf.nrl.navy.mil [ATM]: assorted cleanups Signed-off-by: Adrian Bunk Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.28.12, 2005-03-23 10:57:53-08:00, chas@cmf.nrl.navy.mil [ATM]: [lanai] alpha build fixes Signed-off-by: Jeff Garzik Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.28.11, 2005-03-23 10:57:06-08:00, chas@cmf.nrl.navy.mil [ATM]: [ambassador] fix sparse warnings Signed-off-by: Alexey Dobriyan Signed-off-by: Domen Puncer Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.28.10, 2005-03-23 10:53:53-08:00, chas@cmf.nrl.navy.mil [ATM]: [nicstar] fix some sparse warnings Signed-off-by: Alexey Dobriyan Signed-off-by: Domen Puncer Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.28.9, 2005-03-23 10:52:51-08:00, chas@cmf.nrl.navy.mil [ATM]: [zatm] fix sparse warning Signed-off-by: Alexey Dobriyan Signed-off-by: Domen Puncer Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.28.8, 2005-03-23 10:51:57-08:00, chas@cmf.nrl.navy.mil [ATM]: Remove bridge/lec interdependency Signed-off-by: Chas Williams Signed-off-by: David S. Miller ChangeSet@1.2181.27.15, 2005-03-23 11:13:21-06:00, jejb@mulgrave.(none) Make Fusion-MPT much faster as module From: Moore, Eric Dean Between the 3.01.16 and 3.01.18, we introduced new method to passing command line options to the driver. Some of the command line options are used for fine tuning dv(domain validation) in the driver. By accident, these command line options were wrapped around #ifdef MODULE in the 3.01.18 version of the driver. What this meant is when the driver is compiled built-in the kernel, the optimal settings for dv were ignored, thus poor performance. There was actually a fix for this when I submitted SAS drivers to the mailing list back in November, however the SAS drivers was rejected, and later on I overlooked submitting a single patch to solve this. Signed-off-by: Eric Moore Signed-off-by: James Bottomley ChangeSet@1.2181.27.14, 2005-03-23 11:05:01-06:00, bunk@stusta.de [PATCH] drivers/scsi/osst.c: remove unused code Thanks to both the Coverity checker and GNU gcc, it was found that this variable is completely unused. Signed-off-by: Adrian Bunk Signed-off-by: Willem Riede Signed-off-by: James Bottomley ChangeSet@1.2181.27.13, 2005-03-23 10:59:06-06:00, bunk@stusta.de [PATCH] drivers/scsi/osst.c: make code static This patch makes needlessly global code static. Signed-off-by: Adrian Bunk Signed-off-by: Willem Riede Signed-off-by: James Bottomley ChangeSet@1.2194, 2005-03-22 22:51:15-06:00, stevef@smfhome.smfdom [CIFS] clean up source code formatting Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.33.2, 2005-03-22 23:24:12-05:00, russb@emc.com [PATCH] libata: support descriptor sense in ctrl page libata must support the descriptor format sense blocks as they are required to properly report results of ATA pass through commands as well as other SCSI commands reporting 48b LBAs. This patch adjusts the control mode page to properly report this. Signed-off-by: Brett Russ Signed-off-by: Jeff Garzik ChangeSet@1.2181.28.7, 2005-03-22 20:07:41-08:00, davej@redhat.com [IPV4]: Fix swapped memset args in multipath_wrandom.c Signed-off-by: Dave Jones Signed-off-by: David S. Miller ChangeSet@1.2181.28.6, 2005-03-22 20:06:35-08:00, romieu@fr.zoreil.com [IPV4]: Fix early use of inline in route.c Signed-off-by: Francois Romieu Signed-off-by: David S. Miller ChangeSet@1.2181.20.36, 2005-03-22 19:37:26-08:00, davem@sunset.davemloft.net [IRDA]: Squash warnings introduced by DEBUG cleanups. Signed-off-by: David S. Miller ChangeSet@1.2181.20.35, 2005-03-22 19:23:10-08:00, jt@hpl.hp.com [IRDA]: DEBUG macro fixes o [CRITICA] Eliminate all 'ASSERT(..., break;)' -> use goto; That would compile differenty with/without CONFIG_IRDA_DEBUG o [CORRECT] Add '()' to IRDA_DEBUG macro to avoid side effects o [CORRECT] Add 'do {} while(0)' to IRDA_ASSERT to avoid side effects o [FEATURE] Rename ASSERT to IRDA_ASSERT (namespace pollution) o [FEATURE] Rename MESSAGE to IRDA_MESSAGE (namespace pollution) o [FEATURE] Rename ERROR to IRDA_ERROR (namespace pollution) o [FEATURE] Disable IRDA_ASSERT when no CONFIG_IRDA_DEBUG -> footprint Signed-off-by: Jean Tourrilhes Signed-off-by: David S. Miller ' ChangeSet@1.2181.20.34, 2005-03-22 19:20:39-08:00, davem@sunset.davemloft.net [NET]: Kill NETLINK_DEV and its only user, ethertap. This stuff has been scheduled to die for 2 years. Signed-off-by: David S. Miller ChangeSet@1.2181.20.33, 2005-03-22 19:17:22-08:00, chrisw@osdl.org [NETLINK]: Remove unused netlink NL_EMULATE_DEV code Now that netlink_attach() has been removed, the NL_EMULATE_DEV handler functions can't ever be set. So let's rip them out too, because what's left behind can't be used at all. Signed-off-by: Chris Wright Signed-off-by: David S. Miller ChangeSet@1.2181.20.32, 2005-03-22 19:09:27-08:00, davem@sunset.davemloft.net [NETPOLL]: netpoll_queue needs to be exported to modules Signed-off-by: David S. Miller ChangeSet@1.2181.20.31, 2005-03-22 19:05:52-08:00, shemminger@osdl.org [TCP]: BIC not binary searching correctly While redoing BIC for the split up version, I discovered that the existing 2.6.11 code doesn't really do binary search. It ends up being just a slightly modified version of Reno. See attached graphs to see the effect over simulated 1mbit environment. The problem is that BIC is supposed to reset the cwnd to the last loss value rather than ssthresh when loss is detected. The correct code (from the BIC TCP code for Web100) is in this patch. Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller ChangeSet@1.2181.20.30, 2005-03-22 19:04:16-08:00, davem@sunset.davemloft.net [NETPOLL]: Do not use __smp_processor_id(). It is not necessarily available in all configurations. Signed-off-by: David S. Miller ChangeSet@1.2181.20.29, 2005-03-22 18:46:25-08:00, juhl-lkml@dif.dk [NET]: Remove redundant NULL pointer check before kfree in socket.c Signed-off-by: Jesper Juhl Signed-off-by: David S. Miller ChangeSet@1.2181.20.28, 2005-03-22 18:44:45-08:00, akepner@sgi.com [BONDING]: Use NETIF_F_LLTX in bonding device Lock contention on the bonding device's xmit_lock can become a bottleneck when 3 or more gige links are aggregated. And it looks like it's unnecessary too, so use the NETIF_F_LLTX flag to avoid grabbing this lock. Signed-off-by: Signed-off-by: David S. Miller ChangeSet@1.2181.20.27, 2005-03-22 18:39:35-08:00, mpm@selenic.com [NETPOLL]: Fix racy dev->flags usage Put ndev->flags usage under the lock. Spotted by Patrick McHardy. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.26, 2005-03-22 18:38:55-08:00, mpm@selenic.com [NETPOLL]: Carrier clarification Clarify the flaky carrier detect code and use msleep(). Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.25, 2005-03-22 18:33:48-08:00, mpm@selenic.com [NETPOLL]: Avoid kfree_skb() on packets with destructor Packets that have destructors should not be zapped here as that might produce additional printk warnings via netconsole. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.24, 2005-03-22 18:31:52-08:00, mpm@selenic.com [NETPOLL]: Handle xmit_lock recursion similarly Handle possible recursion on xmit_lock while we're at it. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.23, 2005-03-22 18:31:14-08:00, mpm@selenic.com [NETPOLL]: Add optional dropping and queueing support This adds a callback for packets we can't deliver immediately and a helper function for clients to queue such packets to the device post-interrupt. Netconsole is modified to use the queueing function for best-effort delivery. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.22, 2005-03-22 18:30:38-08:00, mpm@selenic.com [NETPOLL]: Fix ->poll() locking Introduce a per-client poll lock and flag. The lock assures we never have more than one caller in dev->poll(). The flag provides recursion avoidance on UP where the lock disappears. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.21, 2005-03-22 18:30:01-08:00, mpm@selenic.com [NETPOLL]: Add netpoll pointer to net_device Add struct netpoll pointer to struct netdevice Move netpoll rx flags to netpoll struct Stop traversing rx_list and get np pointer from skb->dev->np Remove now unneeded rx_list Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.20, 2005-03-22 18:29:23-08:00, mpm@selenic.com [NETPOLL]: Filter inlines Add netpoll rx helpers Move skb_free for rx into __netpoll_rx Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.19, 2005-03-22 18:28:36-08:00, mpm@selenic.com [NETPOLL]: Shorten carrier detect timeout. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller ChangeSet@1.2181.20.18, 2005-03-22 18:27:00-08:00, davem@sunset.davemloft.net [IPV4]: The multipath select_route method must be implemented. Signed-off-by: David S. Miller ChangeSet@1.2181.36.1, 2005-03-22 22:40:53-03:00, acme@toy.ghostprotocols.net [NET] use sk_acceptq_is_full Replacing the open coded equivalent. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller ChangeSet@1.2181.34.11, 2005-03-22 19:00:15-05:00, domen@coderock.org [PATCH] net/sk98lin: remove duplicate delay Remove an unnecessary second (and identical) delay. schedule_timeout() does not need to be called, as msleep_interruptible() already delayed the task. Signed-off-by: Nishanth Aravamudan Signed-off-by: Domen Puncer Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.10, 2005-03-22 18:57:37-05:00, komurojun-mbn@nifty.com [PATCH] net/Kconfig: remove unsupported network adapter names 3c569B (98) and RE1000Plus(C-Bus) is not supported at all by the kernel 2.6.11. Signed-off-by: Jun Komuro Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.9, 2005-03-22 18:24:56-05:00, dale@farnsworth.org [PATCH] mii: GigE support bug fixes Signed-off-by: Dale Farnsworth Acked-by: James Chapman Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.8, 2005-03-22 18:14:13-05:00, venza@brownhat.org [PATCH] Maintainer change for the sis900 driver Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.7, 2005-03-22 17:54:26-05:00, linville@tuxdriver.com [PATCH] bonding: avoid tx balance for IGMP (alb/tlb mode) Add special case to bond_alb_xmit() to avoid tx balance for IGMP. Signed-off-by: John W. Linville Signed-off-by: Jeff Garzik ChangeSet@1.2181.35.1, 2005-03-22 17:53:22-05:00, akpm@osdl.org [PATCH] b44: allocate tx bounce bufs as needed From: "John W. Linville" The b44 hardware has a DMA mask that only covers 1GB. On x86, a DMA mask <4GB results in allocations using GFP_DMA. The GFP_DMA pool (16MB) gets exhausted very quickly in some configurations. The b44 driver has been pre-allocating bounce buffers in a single large (~750k) contiguous block. On boxes w/ limited GFP_DMA memory, this allocation can fail. Such failure results in the driver being unable to load and function. The solution here is to check each tx skb against the DMA mask. If it is outside the allowable range, a single buffer is allocated from the GFP_DMA range and discarded after the tx completes. This behaviour mimics what is done for bounce buffers on the rx side. The pre-allocation of tx bounce buffers is, of course, removed. Acked-by: Pekka Pietikäinen Signed-off-by: John W. Linville Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.6, 2005-03-22 17:50:47-05:00, mikpe@user.it.uu.se [PATCH] drivers/net/depca.c gcc4 fix Fix drivers/net/depca.c: In function 'load_packet': drivers/net/depca.c:1829: warning: operation on 'i' may be undefined warning from gcc4 in depca.c. Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.5, 2005-03-22 17:50:35-05:00, mikpe@user.it.uu.se [PATCH] drivers/net/arcnet/arcnet.c gcc4 fixes Fix drivers/net/arcnet/arcnet.c: In function 'release_arcbuf': drivers/net/arcnet/arcnet.c:256: warning: operation on 'i' may be undefined drivers/net/arcnet/arcnet.c: In function 'get_arcbuf': drivers/net/arcnet/arcnet.c:292: warning: operation on 'i' may be undefined warnings from gcc4 in arcnet.c. Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.4, 2005-03-22 17:48:09-05:00, akpm@osdl.org [PATCH] drivers/net/sis900.c: fix a warning From: Adrian Bunk drivers/net/sis900.c:199: warning: 'sis900_poll' declared `static' but never defined Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.3, 2005-03-22 17:47:57-05:00, akpm@osdl.org [PATCH] fix pci_disable_device in 8139too From: Andres Salomon http://linux.bkbits.net:8080/linux-2.6/cset@418391928THbmFKdJ5UCOhnFPMYbOA added an unconditional pci_disable_device() to __rtl8139_cleanup_dev(). That's fine for rtl8139_remove_one and rtl8139_init_one; however, for rtl8139_init_board, it ends up being called in the error path. That is, if pci_enable_device or pci_request_regions fails, err_out calls __rtl8139_cleanup_dev, which calls pci_disable_device. Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.2, 2005-03-22 17:45:58-05:00, akpm@osdl.org [PATCH] bonding needs inet The bonding driver needs CONFIG_INET, for arp_create(), arp_send(), arp_xmit(). Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik ChangeSet@1.2181.34.1, 2005-03-22 17:33:07-05:00, akpm@osdl.org [PATCH] Fix suspend/resume on via-velocity From: Pavel Machek This fixes suspend-resume on via-velocity. It was confused w.r.t. pointers... Now uses netdev_priv(). [Well, someone should run sed over that driver, there are many more dev->priv]. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik ChangeSet@1.2193, 2005-03-22 16:17:42-06:00, stevef@smfhome.smfdom [CIFS] whitespace/formatting cleanup Signed-off-by: Jesper Juhl (juhl-lkml@dif.dk) Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.33.1, 2005-03-22 15:41:49-05:00, jason.d.gaston@intel.com [PATCH] SATA AHCI correction Intel ICH7R This patch removes an invalid DID for Intel ICH7R from the ahci.c SATA AHCI driver. Signed-off-by: Jason Gaston Signed-off-by: Jeff Garzik ChangeSet@1.2181.32.2, 2005-03-22 15:37:06-05:00, linville@tuxdriver.com [PATCH] e1000: flush work queues on remove Flush work queues in ->remove() for e1000. Acked-by: Ganesh Venkatesan Signed-off-by: John W. Linville Signed-off-by: Jeff Garzik ChangeSet@1.2181.32.1, 2005-03-22 15:36:54-05:00, linville@tuxdriver.com [PATCH] e1000: avoid sleeping in watchdog timer context Move bulk of e1000_watchdog to a workqueue to make it safe to call functions which can sleep. Signed-off-by: John W. Linville Signed-off-by: Jeff Garzik ChangeSet@1.2181.30.3, 2005-03-22 09:45:15-08:00, davem@sunset.davemloft.net [SPARC64]: Handle straddling VA space hole correctly. Noticed by Hugh Dickins. Signed-off-by: David S. Miller ChangeSet@1.2181.18.30, 2005-03-22 09:36:50+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2575/1: pass -mbig-endian/-mlittle-endian to invocations of cpp Patch from Lennert Buytenhek We currently pass -mbig-endian/-mlittle-endian to the compiler and the assembler, but we do not pass it to cpp, even though it also needs it. So, add this flag to CPPFLAGS, and since CPPFLAGS is also passed to the compiler and the assembler, we can then then remove the flag from CFLAGS and AFLAGS. The symptom of this problem is the kernel hanging after "Memory: ......KB available (....K code, ...K data, ...K init)" when compiling a big-endian kernel with a little-endian toolchain or vice versa, and this happens because 'jiffies' points to the wrong half of 'jiffies_64' (due to vmlinux.lds having been generated with the wrong preprocessor flags), so calibrate_delay() takes forever to complete. Derived from an earlier patch by Krzysztof Halasa. Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King ChangeSet@1.2181.31.34, 2005-03-22 09:17:29+01:00, perex@suse.cz [ALSA] Increase buffer sizes for the CA0106 driver CA0106 driver This patch increases the buffer size for the ca0106 driver, so this might help prevent over/underruns. Signed-off-by: James Courtier-Dutton Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.33, 2005-03-22 09:16:48+01:00, perex@suse.cz [ALSA] Fix 96000 SPDIF out from Audigy 2 P16V EMU10K1/EMU10K2 driver This allows one to output at 96000 to the SPDIF using the P16V chip. Note: The sample phase is wrong when using the P16V chip, but at least no resampling is done. Signed-off-by: James Courtier-Dutton Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.32, 2005-03-22 09:16:05+01:00, perex@suse.cz [ALSA] emu10k1 external tram size EMU10K1/EMU10K2 driver This patch fixes wrong size reported by driver for external tram. It reports size in bytes and should report it in samples as for internal tram. Signed-off-by: Peter Zubaj Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.31, 2005-03-22 09:15:23+01:00, perex@suse.cz [ALSA] use amp capabilities from afg if amp override not set HDA Codec driver Fix by Matt : Some HDA codec nodes contain an amp, but do not provide local amp capabilities. In these cases, AC_WCAP_AMP_OVRD is not set so we should query the AFG nid in order to get the general amp capabilities. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.30, 2005-03-22 09:14:43+01:00, perex@suse.cz [ALSA] fix bug with pci hotplug mode MIXART driver Fix the Oops with hotplug fw loader. (Theis fix is missing in the last commit to mixart.c accidentally.) Signed-off-by: Markus Bollinger Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.29, 2005-03-22 09:13:58+01:00, perex@suse.cz [ALSA] add HPET support Timer Midlevel,ALSA Core add a wrapper for the HPET driver Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.28, 2005-03-22 09:13:15+01:00, perex@suse.cz [ALSA] remove unneeded interrupt locks in rawmidi drivers Generic drivers,MPU401 UART,CS4281 driver,ENS1370/1+ driver CS46xx driver,EMU10K1/EMU10K2 driver Now that the output trigger callback is called from a softirq instead of an hardirq, we don't need anymore to disable interrupts in our interrupt handlers. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.27, 2005-03-22 09:12:31+01:00, perex@suse.cz [ALSA] rawmidi - move output trigger into a tasklet Documentation,RawMidi Midlevel Calling the output trigger callback from another interrupt handler can lead to unintuitive locking requirements (i.e., spin_lock_irqsave) in the sound card interrupt handler. Moving the call to the callback into a tasklet cures this, and has the added benefit that the callback is called only once if more that one sequencer event has been delivered in one timer interrupt tick. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.26, 2005-03-22 09:11:47+01:00, perex@suse.cz [ALSA] rawmidi - fix locking in drop_output and drain_input RawMidi Midlevel The comments in snd_rawmidi_drop_output and snd_rawmidi_drain_input are wrong -- interrupts _are_ enabled, and spinlocks _are_ required. So remove the comments and add spinlocks. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.25, 2005-03-22 09:11:04+01:00, perex@suse.cz [ALSA] Replace '/' with commas in ac97 codec names AC97 Codec Replaced '/' letters with commas in ac97 codec names, so that they can be used for sysfs entries in ac97_bus. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.24, 2005-03-22 09:10:24+01:00, perex@suse.cz [ALSA] Fix SPDIF output on CMI9880 HDA Codec driver There is mute control on 9880's spdif (IEC958) out, so it needs to be turned on/off in mixer. The patch is for CVS version and I think it can be patched to azx too. Hope this also fix the 9880 SPDIF-out bug. Signed-off-by: ChenLi Tien Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.23, 2005-03-22 09:09:39+01:00, perex@suse.cz [ALSA] fix bug with pci hotplug mode MIXART driver There is a bug with mixart driver, when using hotplug: accessing NULL pointer -> segmentation fault Signed-off-by: Markus Bollinger Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.22, 2005-03-22 09:08:57+01:00, perex@suse.cz [ALSA] rme32 - remove superfluous spin_lock_irqsave call RME32 driver In the PCM trigger callback, replace spin_lock_irqsave() with spin_lock() because interrupts are already guaranteed to be disabled. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.21, 2005-03-22 09:08:16+01:00, perex@suse.cz [ALSA] Fix typos ALSA sequencer,ALSA Core Fix typos in alsa-kernel code for MIDI sostenuto. Signed-off-by: William Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.20, 2005-03-22 09:07:30+01:00, perex@suse.cz [ALSA] fix misc oopses EMU10K1/EMU10K2 driver Fix Oops with Multi-channel (EFX) mixer controls. Signed-off-by: Arnaud Patard Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.19, 2005-03-22 09:06:46+01:00, perex@suse.cz [ALSA] fix P16V breakage for non Audigy2 cards EMU10K1/EMU10K2 driver The P16V patch unconditionally checks the IPR2 register in the interrupt handler resulting in infinite loop and system lockup on any non Audigy2 cards. I really hate checking emu->is_audigy and emu->revision in a fast path like the IRQ handler but I don't see another way. Also, don't bother allocating/freeing the DMA buffer for P16V unless it's really present. This is a critical fix and should trigger an immediate rc2 release IMO. Currently any emu10k1 users other than Audigy 2 will lock up hard as soon as they play any sound. Signed-off-by: Lee Revell Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.18, 2005-03-22 09:06:04+01:00, perex@suse.cz [ALSA] remove superfluous spin_lock_irqsave calls MPU401 UART,CS4281 driver,ENS1370/1+ driver,CS46xx driver RME HDSP driver In PCM trigger and pointer callbacks, replace spin_lock_irqsave() with spin_lock() because interrupts are already guaranteed to be disabled. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.17, 2005-03-22 09:05:21+01:00, perex@suse.cz [ALSA] documentation - clarify information about atomic callbacks Documentation Document that the ack callback is atomic, too, and that the atomic callbacks are called with disabled interrupts. Additionally, clarify the description of the rawmidi trigger callback. Signed-off-by: Clemens Ladisch ChangeSet@1.2181.31.16, 2005-03-22 09:04:37+01:00, perex@suse.cz [ALSA] Add new C-Media 9880 codec ID HDA Codec driver Following change need to be added for newer C-Media 9880 codec ID. Signed-off-by: ChenLi Tien Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.15, 2005-03-22 09:03:57+01:00, perex@suse.cz [ALSA] Use full-digital model as default for CMI9880 HDA Codec driver Use full-digital model as default for CMI9880 rather than the minimal model. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.14, 2005-03-22 09:03:14+01:00, perex@suse.cz [ALSA] ak4114 (Juli@) - increased delay between statistics update & rate check AK4114 receiver Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.13, 2005-03-22 09:02:33+01:00, perex@suse.cz [ALSA] Wake up polls and signals at timer notification Timer Midlevel Wake up polls and signals at timer notification (TREAD mode only). Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.12, 2005-03-22 09:01:52+01:00, perex@suse.cz [ALSA] Fix resume of es1968 ES1968 driver Fixed the resume of es1968. - restore the running PCM set up properly in resume - ignore spurious hw-vol irqs during resume Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.11, 2005-03-22 09:01:11+01:00, perex@suse.cz [ALSA] Fix Oops with timer notifying Timer Midlevel Fixed Oops with timer notifying after TIMER_TREAD ioctl. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.10, 2005-03-22 09:00:28+01:00, perex@suse.cz [ALSA] Fix suspend/resume with ATIIXP ATIIXP driver Fixed the suspend/resume with ATIIXP driver. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.9, 2005-03-22 08:59:49+01:00, perex@suse.cz [ALSA] Add proper spin/irq locks to suspend PCM Midlevel Add the proper spin/irq locks to PCM suspend functions so that PCM trigger and pointer callbacks can be called safely without irqsave. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.8, 2005-03-22 08:59:10+01:00, perex@suse.cz [ALSA] isa/Kconfig - added SND_AD1848_LIB and SND_CS4231_LIB tristates ISA This patch fixes problem with missing SND_GENERIC_PM for isa cards using ad1848 and cs4231 library modules. Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.7, 2005-03-22 08:58:26+01:00, perex@suse.cz [ALSA] emu10k1 - add support for p16v chip (24-bit playback) EMU10K1/EMU10K2 driver Add 24bit, 96khz support for multichannel playback on the Audigy 2 sound cards. Signed-off-by: James Courtier-Dutton Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.6, 2005-03-22 08:57:43+01:00, perex@suse.cz [ALSA] emu10k1 - copyright additions/fixes EMU10K1/EMU10K2 driver The next two patches add my copyright for adding the multichannel PCM support to emupcm.c and emumixer.c. The final patch adds Clemens Ladisch to the copyright on timer.c, rather than just saying 'Copied from similar code by CL'. This has been bugging me for a while, since I just copied and pasted from the ymfpci driver & changed the registers. Signed-off-by: Lee Revell Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.5, 2005-03-22 08:57:02+01:00, perex@suse.cz [ALSA] emu10k1 - Silence the 'BUG (or not enough voices)' message EMU10K1/EMU10K2 driver Silence the 'BUG (or not enough voices)' message. This does not in fact represent a bug; it's a normal ocurrence when the wavetable synth is in use. Signed-off-by: Lee Revell Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.4, 2005-03-22 08:56:19+01:00, perex@suse.cz [ALSA] emu10k1 - give the subdevices descriptive names EMU10K1/EMU10K2 driver Give the subdevices descriptive names, like 'ADC Capture/Standard PCM Playback' instead of 'EMU10K1' for hw:x,0 and 'Multichannel Capture/PT Playback' instead of 'EMU10K1 EFX' for hw:x,2. Now that qjackctl enumerates the devices automatically, this is a significant usability improvement. Signed-off-by: Lee Revell Signed-off-by: Jaroslav Kysela ChangeSet@1.2181.31.3, 2005-03-22 08:55:39+01:00, perex@suse.cz [ALSA] Fix voice allocation corruption EMU10K1/EMU10K2 driver Fixed the corrupted voice allocation in snd_emu10k1_pcm_channel_alloc(). Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.2, 2005-03-22 08:54:56+01:00, perex@suse.cz [ALSA] Add DXS support for MSI K8T Neo2-FI VIA82xx driver Added the DXS entry for MSI K8T Neo2-FI. Signed-off-by: Takashi Iwai ChangeSet@1.2181.31.1, 2005-03-22 08:54:10+01:00, perex@suse.cz [ALSA] Fix ALC655/658/850 SPDIF playback route AC97 Codec Fix ALC655/658/850 IEC958 (SPDIF) playback route control. Signed-off-by: Takashi Iwai ChangeSet@1.2181.16.17, 2005-03-21 22:55:26-08:00, gregkh@suse.de PCI Hotplug: enforce the rule that a hotplug slot needs a release function. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.47, 2005-03-21 22:36:57-08:00, stern@rowland.harvard.edu [PATCH] UHCI updates This is the fifth of five updates to the uhci-hcd driver: Separate out the part of the driver responsible for scanning the schedule and doing delayed processing. Put it in a new routine which can be called as needed (such as when the controller is suspended) from several places in addition to the usual IRQ handler. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.46, 2005-03-21 22:36:17-08:00, stern@rowland.harvard.edu [PATCH] UHCI updates This is the fourth of five updates to the uhci-hcd driver: Reimplement the get_current_frame routines so that when the controller isn't running they return a cached value. Also add a flag to track whether the controller is running and allow critical data structure updates to occur immediately if the controller is stopped. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.45, 2005-03-21 22:30:56-08:00, stern@rowland.harvard.edu [PATCH] UHCI updates This is the third of five updates to the uhci-hcd driver: Change the first argument to uhci_finish_completion from hcd to uhci and expand the spinlock-protected region to include the entire timer callback routine plus the suspend and resume routines. It's a little awkward that the reset routine must run without the spinlock, but that whole pathway will vanish before long. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.44, 2005-03-21 22:27:09-08:00, stern@rowland.harvard.edu [PATCH] UHCI updates This is the second of five updates to the uhci-hcd driver: Reimplement the port reset function by splitting it into two parts like the other HC drivers do, where the second part is triggered by a port status request when the reset finishes. Now the entire hub_control routine can run without sleeping and can hold the device spinlock. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.43, 2005-03-21 22:26:12-08:00, stern@rowland.harvard.edu [PATCH] UHCI updates This is the first of five updates to the uhci-hcd driver: Rename the uhci->schedule_lock to just plain uhci->lock. Originally I had thought of adding a separate device-management lock, but now it seems better to use a single lock for both purposes. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.42, 2005-03-21 22:24:51-08:00, stern@rowland.harvard.edu [PATCH] USBcore updates This is the fifth of five updates to usbcore: Add some extra debugging messages to the hub driver and make it set the power state for root hubs like it does for external hubs. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2192, 2005-03-21 22:32:17-06:00, stevef@smf-t23.(none) [CIFS] Fix NT4 attribute setting Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2191, 2005-03-21 20:09:44-06:00, stevef@smf-t23.(none) [CIFS] add new retry on failure to legacy servers such as NT4 of delete of readonly files. Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.4.41, 2005-03-21 15:04:00-08:00, stern@rowland.harvard.edu [PATCH] USBcore updates This is the fourth of five updates to usbcore: Add a new usb_hcd_resume_root_hub function for use by HCDs in processing resume requests from their root hubs. No calls to this new routine are in the patch but they will be added in due course. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.40, 2005-03-21 14:58:02-08:00, stern@rowland.harvard.edu [PATCH] USBcore updates This is the third of five updates to usbcore: Adjust the usb_hc_died routine to eliminate races with root-hub registration/deregistration and have it tell khubd to remove all devices below the root hub. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.39, 2005-03-21 14:43:56-08:00, stern@rowland.harvard.edu [PATCH] USBcore and HCD updates This is the second of five updates to usbcore: Rename the hcd->state constants so that they all begin with HC_STATE. Right now some of them start with HCD_STATE and others with USB_STATE, which can be very confusing. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.18.29, 2005-03-21 21:26:20+00:00, icampbell@com.rmk.(none) [ARM PATCH] 2574/1: PXA2xx: Save CCLKCFG over sleep Patch from Ian Campbell As discussed on the ARM kernel list, the CCLKCFG[TURBO] bit is documented as being cleared when sleep mode exits so we need to save and restore it ourselves. Signed-off-by: Ian Campbell Signed-off-by: Russell King ChangeSet@1.2181.18.28, 2005-03-21 21:17:05+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2573/1: simplify align[bw]() in ixp2000's io.h and update comments Patch from Lennert Buytenhek We currently have a workaround for the fact that early IXP2400s assert byte lanes for PCI I/O transactions the other way round. Newer silicon has a bit that can be set to 1 to get the 'proper' behavior, but since that is not available on older silicon we choose to keep working around this issue by hand. The workaround can be done in an easier way than is currently the case, though -- instead of substracting the lower two bits of the I/O address from 3 and putting the result back into the lower two bits, we can just toggle the two lowest bits using a simple XOR with 3. While we are at it, the outb() macro does not seem to parenthesize its first argument, while outw() and outl() do, so fix this up. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2181.18.27, 2005-03-21 20:57:28+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2572/1: remove ifdefs from enp2611.c Patch from Lennert Buytenhek If enp2611.c is compiled in, it implies that CONFIG_ARCH_ENP2611 is defined, so we don't need the ifdefs. Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King ChangeSet@1.2181.18.26, 2005-03-21 20:47:35+00:00, buytenh@org.rmk.(none) [ARM PATCH] 2571/1: minor time-keeping fixes for ixp2000 Patch from Lennert Buytenhek There are two very minor issues with time-keeping on the ixp2000: 1. It turns out that if T#_CLD (timer reload value) is set to X, it means the timer will spend a full tick on 0, then a full tick on X, and then start counting down towards 0, which then effectively gives the timer a period of X+1 instead of the intended X. Solve by writing X-1 instead of X to T#_CLD. This is relatively harmless, though -- assuming a 50MHz system clock (like on the ENP-2611) and HZ=100, there will be 500000 timer interrupts generated per 500001 real-world jiffies, but that last timer interrupt will call timer_tick twice, taking the system time from 499999 jiffies plus 20000 usec to 500001 jiffies plus 0 usec. Every 500000th jiffy will 'not exist', and every other jiffy will last 10002us instead of 10000us. 2. next_jiffy_time is initialised as '0xffffffff - ticks_per_jiffy', but timer 4 starts ticking from 0xffffffff downwards, which will cause the time offset in the first 1/HZ sec after timer int init to be negative, and the first timer interrupt ever not to call timer_tick(). Neither of these issues can cause clock drift or backwards clock motion, but it's good to fix them nevertheless. Signed-off-by: Lennert Buytenhek Signed-off-by: Deepak Saxena Signed-off-by: Russell King ChangeSet@1.2181.30.2, 2005-03-21 09:56:38-08:00, davem@sunset.davemloft.net [SPARC64]: Fix fifth arg pointer check for SEMTIMEDOP. Noticed by Anton Blanchard. Signed-off-by: David S. Miller ChangeSet@1.2181.30.1, 2005-03-21 09:28:06-08:00, davem@sunset.davemloft.net [PARISC]: Fix type in unaligned.h header. Meant to include asm-generic/unaligned.h, not make this file include itself :-) Noticed by Mika Kukkonen. Signed-off-by: David S. Miller ChangeSet@1.2181.27.12, 2005-03-20 21:05:57-06:00, jejb@mulgrave.(none) 53c700: Alter interrupt assignment This change makes interrupt assignment the job of the glue driver. The upshot being that the glue driver can now set the name of the interrupt and, in the case of the NCR_D700 driver, make the interrupt routing more efficient. Signed-off-by: James Bottomley ChangeSet@1.2181.2.22, 2005-03-21 12:16:10+11:00, nathans@sgi.com [XFS] Make trivial extension to sync flag to implement dirsync, instead of silently ignoring it. SGI-PV: 932297 SGI-Modid: xfs-linux-melb:xfs-kern:21888a Signed-off-by: Nathan Scott ChangeSet@1.2181.2.21, 2005-03-21 12:03:51+11:00, nathans@sgi.com [XFS] Fix sync mount option to also do metadata updates synchronously. SGI-PV: 932297 SGI-Modid: xfs-linux-melb:xfs-kern:21887a Signed-off-by: Nathan Scott ChangeSet@1.2190, 2005-03-20 17:09:22-06:00, stevef@smf-t23.(none) [CIFS] streamlining cifs open with various helper functions Signed-off-by: Jesper Juhl (juhl-lkml@dif.dk) Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.26.3, 2005-03-20 18:12:44+01:00, marcel@holtmann.org [Bluetooth] Fix race condition in virtual HCI driver This patch fixes a race condition in the hci_vhci_chr_close() routine of the virtual HCI driver. Signed-off-by: Imre Deak Signed-off-by: Marcel Holtmann ChangeSet@1.2181.27.11, 2005-03-20 10:12:34-06:00, jejb@mulgrave.(none) SCSI: Re-export code incorrectly made static The interrupt routine in 53c700 is designed to be exported for a future code change Signed-off-by: James Bottomley ChangeSet@1.2181.18.25, 2005-03-20 09:46:11+00:00, rmk@flint.arm.linux.org.uk [ARM] Add vserver syscall allocation Signed-off-by: Russell King ChangeSet@1.2181.28.4, 2005-03-20 20:36:54+11:00, domen@coderock.org [CRYPTO]: Fix sparse warnings in tea Signed-off-by: Domen Puncer Signed-off-by: Herbert Xu ChangeSet@1.2181.28.3, 2005-03-20 20:30:42+11:00, domen@coderock.org [CRYPTO]: Fix sparse warnings in blowfish Signed-off-by: Domen Puncer Signed-off-by: Herbert Xu ChangeSet@1.2181.28.2, 2005-03-20 20:24:16+11:00, domen@coderock.org [CRYPTO]: Fix sparse warning in sha512 Signed-off-by: Domen Puncer Signed-off-by: Herbert Xu ChangeSet@1.2181.28.1, 2005-03-20 20:21:42+11:00, domen@coderock.org [CRYPTO]: Fix sparse warning in sha256 Signed-off-by: Domen Puncer Signed-off-by: Herbert Xu ChangeSet@1.2189, 2005-03-19 23:16:23-06:00, stevef@smf-t23.(none) [CIFS] code cleanup, rearranging of large function Mostly following suggestions of Jesper Juhl Signed-of-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.24.4, 2005-03-19 20:27:17-08:00, davem@nuts.davemloft.net [SPARC64]: Eliminate g5 register usage from bitops assembly. Signed-off-by: David S. Miller ChangeSet@1.2181.27.10, 2005-03-19 20:36:15-06:00, jejb@mulgrave.(none) ncr53c8xx: Fix small problem with initial negotiation The driver produces messages list this: ncr53c720-1: ID 7, Fast-10, Parity Checking scsi1 : ncr53c8xx-3.4.3g target1:0:1: target did not report SYNC. Vendor: TOSHIBA Model: CD-ROM XM-5401TA Rev: 3605 Type: CD-ROM ANSI SCSI revision: 02 target1:0:1: Beginning Domain Validation target1:0:1: asynchronous. target1:0:1: Domain Validation skipping write tests target1:0:1: FAST-5 SCSI 4.2 MB/s ST (236 ns, offset 8) target1:0:1: Ending Domain Validation Although everything is fine. The initial "target did not report SYNC" is because the driver is trying to negotiate over the initial inquiry (so, naturally we don't know if it supports sync yet). The solution is to rip out the last remnants of the spontaneous attempts at negotiation to force it to wait for the DV negotiation in slave_configure. Signed-off-by: James Bottomley ChangeSet@1.2181.20.16, 2005-03-19 16:46:45-08:00, yoshfuji@linux-ipv6.org [IPV4]: Size ip_mp_alg_table[] correctly. Signed-off-by: Hideaki YOSHIFUJI Signed-off-by: David S. Miller ChangeSet@1.2181.20.15, 2005-03-19 16:45:16-08:00, davem@sunset.davemloft.net [IPV4]: Check multipath ops func pointers against NULL. Multipath algorithms are not required to implement all of the methods. Just the ones they need. Signed-off-by: David S. Miller ChangeSet@1.2181.27.9, 2005-03-19 15:43:33-06:00, jejb@titanic.il.steeleye.com Q720: fix compile warning Signed-off-by: James Bottomley ChangeSet@1.2181.18.24, 2005-03-19 21:20:40+00:00, tony@com.rmk.(none) [ARM PATCH] 2540/1: OMAP update 10/10: Add boards VoiceBlue and NetStar Patch from Tony Lindgren This patch by Ladislav Michl adds support for boards VoiceBlue and NetStar Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.18.23, 2005-03-19 21:12:49+00:00, tony@com.rmk.(none) [ARM PATCH] 2542/1: OMAP update 9/10: Board specific updates Patch from Tony Lindgren This patch updates the board specific files for OMAP. The highlights of this patch are: - Change flash memory to use device model style init by Ladislav Michl - USB related resource fixes by David Brownell - Resource fixes for smc91x by Tony Lindgren Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.18.22, 2005-03-19 21:04:42+00:00, tony@com.rmk.(none) [ARM PATCH] 2541/1: OMAP update 8/10: Leds related changes Patch from Tony Lindgren Better leds support by David Brownell Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.27.8, 2005-03-19 14:58:03-06:00, James.Bottomley@steeleye.com [PATCH] fix breakage in the SCSI generic tag code When I redid the target allocation routines, I inadvertently broke the generic tag code. The problem is that the device is added to the host list after slave_configure in the new code. This is too late since if any command disconnects during the initial inquiry phase, we now cannot find the device for the reconnect. The fix is to add the device to the lists earlier. Signed-off-by: James Bottomley ChangeSet@1.2181.18.21, 2005-03-19 20:57:37+00:00, tony@com.rmk.(none) [ARM PATCH] 2547/1: Update OMAP 7/10: USB low-level init Patch from Tony Lindgren This patch by David Brownell updates the USB low-level init on OMAP. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.27.7, 2005-03-19 14:55:41-06:00, willy@parisc-linux.org [PATCH] Fix small bug in scsi_transport_spi Fix small bug in scsi_transport_spi "reserved" was being printed without a trailing \n. Signed-off-by: Matthew Wilcox Signed-off-by: James Bottomley ChangeSet@1.2181.27.6, 2005-03-19 14:54:23-06:00, willy@parisc-linux.org [PATCH] ncr53c8xx update ncr53c8xx update ncr-3.4.3g: - set the starget in slave_alloc instead of slave_configure to avoid problems with drives sending negotiation messages before we try to configure them. - Implement ->get_signalling. - Pass the scsi_cmnd to ncr_get_ccb() instead of the target & lun numbers - All ncr_show_msg() callers changed to use ncr_print_msg() - Combine ncr_show_msg() into ncr_print_msg() - Move ncr_print_msg() to near the top of the file and eliminate prototype - Change PRINT_ADDR() to take a fmt string - Make PRINT_ADDR use dev_info() - Eliminate PRINT_LUN -- all callers can use PRINT_ADDR or dev_info - Replace PRINT_TARGET() with a direct call to dev_info() - Whitespace cleanups - Get rid of superfluous ; after } of if/switch statements - Use IDENTIFY() instead of M_IDENTIFY | ... - Convert to use spi_display_xfer_agreement() Signed-off-by: Matthew Wilcox Signed-off-by: James Bottomley ChangeSet@1.2181.27.5, 2005-03-19 14:53:04-06:00, willy@parisc-linux.org [PATCH] Zalon updates Zalon updates - Add KERN_INFO to printk Signed-off-by: Andrew McGregor Signed-off-by: Domen Puncer Signed-off-by: Matthew Wilcox - Remove iomem-related warnings - Improve printk message - Use cpu_relax() Signed-off-by: Matthew Wilcox Signed-off-by: James Bottomley ChangeSet@1.2181.18.20, 2005-03-19 20:49:36+00:00, tony@com.rmk.(none) [ARM PATCH] 2544/1: OMAP update 6/10: Change OCPI to use clock framework Patch from Tony Lindgren This patch by Paul Mundt and Tony Lindgren changes cleans up the OCPI bus code and changes it to use clock framework. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.27.4, 2005-03-19 14:47:37-06:00, willy@parisc-linux.org [PATCH] Misc Lasi 700 fixes Misc Lasi 700 fixes - Use the DMA_32BIT_MASK constants when calling dma_set_mask() Signed-off-by: Tobias Klauser Signed-off-by: Matthew Wilcox - ioremap fixes Signed-off-by: Matthew Wilcox Signed-off-by: James Bottomley ChangeSet@1.2181.27.3, 2005-03-19 14:46:29-06:00, axboe@suse.de [PATCH] queue <-> sdev reference counting problem There is a problem with the way sdev is freed currently. The reason is really that there is a circular referencing problem: the sdev needs to hold on to the queue, but the queue (through the request function) also needs to hold on to the sdev. The easiest way to work-around this problem is to kill the sdev reference in the queue when the sdev is freed. On invocation of scsi_request_fn(), kill io to this device. Signed-off-by: Jens Axboe Signed-off-by: James Bottomley ChangeSet@1.2181.27.2, 2005-03-19 14:45:29-06:00, jejb@titanic.il.steeleye.com atp870u DMA mask fix From: Alan Cox Noted by James Bottomley Signed-off-by: Alan Cox Signed-off-by: James Bottomley ChangeSet@1.2181.18.19, 2005-03-19 20:43:23+00:00, tony@com.rmk.(none) [ARM PATCH] 2546/1: OMAP update 5/10: GPIO interrupt changes Patch from Tony Lindgren This patch fixes problems with missed GPIO interrupts. FPGA interrupt handling is also simplified based on the GPIO changes. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.27.1, 2005-03-19 14:39:40-06:00, jejb@titanic.il.steeleye.com atp870u: Re-merge cleanups From: Alan Cox This restores the Adrian Bunk and Al Viro cleanups that got trashed in the driver update. It also fixes a few formatting glitches and adds cpu_relax() calls to the polls spinning on the controller/bus. Signed-off-by: Alan Cox Signed-off-by: James Bottomley ChangeSet@1.2181.18.18, 2005-03-19 20:35:11+00:00, tony@com.rmk.(none) [ARM PATCH] 2564/1: OMAP update 4/10: Pin multiplexing updates, take 2 Patch from Tony Lindgren This patch from David Brownell and Nishant Menon adds new pin configurations. A fix for the mux spinlock is also included. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.18.17, 2005-03-19 20:28:36+00:00, tony@com.rmk.(none) [ARM PATCH] 2565/1: OMAP update 3/10: Clock changes, take 2 Patch from Tony Lindgren This patch by Tuukka Tikkanen and David Brownell updates OMAP clocks and removes old clocks.h. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.18.16, 2005-03-19 20:12:56+00:00, tony@com.rmk.(none) [ARM PATCH] 2548/1: OMAP update 2/10: Include files Patch from Tony Lindgren This patch syncs the mainline kernel with the linux-omap tree. The patch mostly contains minor fixes and extra register definitions by various OMAP developers. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.18.15, 2005-03-19 19:54:31+00:00, tony@com.rmk.(none) [ARM PATCH] 2539/1: OMAP update 1/10: Arch files Patch from Tony Lindgren This patch syncs the mainline kernel with the linux-omap tree. The patch contains various minor fixes by various OMAP developers. Signed-off-by: Tony Lindgren Signed-off-by: Russell King ChangeSet@1.2181.26.2, 2005-03-19 20:40:25+01:00, marcel@holtmann.org [Bluetooth] Make another variable static This patch makes another needlessly global variable static. Signed-off-by: Adrian Bunk Signed-off-by: Marcel Holtmann ChangeSet@1.2181.26.1, 2005-03-19 20:37:30+01:00, marcel@holtmann.org [Bluetooth] Remove now unneeded references to sk_protinfo Now that sk_protinfo is not used anymore for storing private protocol information, we can safely remove the code in the socket destruct path to free it. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller Signed-off-by: Marcel Holtmann ChangeSet@1.2181.11.83, 2005-03-19 18:26:40+00:00, dick@com.rmk.(none) [SERIAL] sealevel 8 port RS-232/RS-422/RS-485 board Patch from Dick Hollenbeck Vendor Sealevel suggested these changes for its new board. Tried them, they work with the card. Please apply the patch below, which was made from 2.6.10 but can be applied to 2.6.11.2 without errors. ChangeSet@1.2181.18.14, 2005-03-19 16:54:46+00:00, ben-linux@org.rmk.(none) [ARM PATCH] 2557/1: S3C2410 - fix otom/nexcoder buiilds due to sparse fixes Patch from Ben Dooks These two machines where missed out when updating for sparse fixes, return them to building correctly. Signed-off-by: Ben Dooks Signed-off-by: Russell King ChangeSet@1.2181.25.1, 2005-03-19 16:39:44+00:00, rmk@flint.arm.linux.org.uk [MMC] SD support : protocol Patch from Pierre Ossman Protocol definitions. The basic commands needed for the later patches. The R1_APP_CMD seems to be misdefined in protocol.h so this patch changes it. ChangeSet@1.2181.11.82, 2005-03-19 16:22:50+00:00, afong@org.rmk.(none) [SERIAL] 8250/sbc8560 bug/fix Patch from Amy Fong On the sbc8560 platform, changes were introduced as of 2.6.10-rc2 which caused the sbc8560 to splutter noise all over the serial console. It seems that ALPHA_KLUDGE_MCR was mulitply and differently defined in both drivers/serial/8250.h and include/linux/serialP.h Below is a patch that clones the definition of ALPHA_KLUDGE_MCR from serialP.h, it has been tested and works on the sbc8560. ChangeSet@1.2181.11.81, 2005-03-19 16:11:36+00:00, alex.williamson@com.rmk.(none) [SERIAL] new hp diva console port Patch from Alex Williamson The patch below adds IDs and setup for a new PCI Diva console port. This device provides a single UART described by PCI Bar 1. ID already submitted to pciids.sf.net. Signed-off-by: Alex Williamson ChangeSet@1.2181.18.13, 2005-03-19 09:23:19+00:00, ben-linux@org.rmk.(none) [ARM PATCH] 2563/1: RiscPC - update IOMEM annotations Patch from Ben Dooks Fixes `iomem` annotations on a number of parts of the arch-rpc build, including IO_BASE, IOC_BASE and IOMD_BASE. Changed the include/asm-arm/arch-rpc/io.h and hardware.h to reflect this, and fixed-up the build problems Signed-off-by: Ben Dooks Signed-off-by: Russell King ChangeSet@1.2181.18.12, 2005-03-19 09:16:10+00:00, nico@org.rmk.(none) [ARM PATCH] 2552/2: woops Patch from Nicolas Pitre argh! Signed-off-by: Nicolas Pitre Signed-off-by: Russell King ChangeSet@1.2181.18.11, 2005-03-19 09:07:26+00:00, ben-linux@org.rmk.(none) [ARM PATCH] 2562/2: CL7500 - iomem fixes Patch from Ben Dooks Fixes for iomem annotations in include/asm-arm/arch-cl7500/hardware.h Signed-off-by: Ben Dooks Signed-off-by: Russell King ChangeSet@1.2187, 2005-03-19 01:32:29-06:00, stevef@smf-t23.(none) [CIFS] remove redundant null pointer checks before kfrees Signed-off-by: Jesper Juhl (juhl-lkml@dif.dk) Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2186, 2005-03-19 01:15:34-06:00, stevef@smf-t23.(none) [CIFS] whitespace cleanups and source formatting improvements Signed-off-by: Jesper Juhl (juhl-lkml@dif.dk) Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2185, 2005-03-19 00:50:27-06:00, stevef@smf-t23.(none) [CIFS] remove sparse warnings Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.24.3, 2005-03-18 22:43:10-08:00, davem@nuts.davemloft.net [SPARC64]: Kill g5 register usage from rtrap.S Signed-off-by: David S. Miller ChangeSet@1.2181.24.2, 2005-03-18 22:30:06-08:00, davem@nuts.davemloft.net [SPARC64]: Kill all smp_tune_scheduling(), totally unused. Kill off cheetah_tune_scheduling() as well. Adrian Bunk's changes made this all apparent. Signed-off-by: David S. Miller ChangeSet@1.2181.24.1, 2005-03-18 22:22:06-08:00, davem@nuts.davemloft.net [SPARC64]: Eliminate g5 register usage in semaphore support code. Signed-off-by: David S. Miller ChangeSet@1.1982.168.31, 2005-03-18 22:11:36-08:00, gregkh@suse.de Linux 2.6.11.5 TAG: v2.6.11.5 ChangeSet@1.1982.168.30, 2005-03-18 21:47:12-08:00, khc@pm.waw.pl [PATCH] Fix kernel panic on receive with WAN Hitachi SCA HD6457x Another patch for 2.6.11.x: already in main tree, fixes kernel panic on receive with WAN cards based on Hitachi SCA/SCA-II: N2, C101, PCI200SYN. The attached patch fixes NULL pointer dereference on RX. Signed-off-by: Krzysztof Halasa Acked-by: Jeff Garzik Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.29, 2005-03-18 21:43:33-08:00, dilinger@debian.org [PATCH] Possible VIA-Rhine free irq issue It seems to me that in the VIA Rhine device driver the requested irq might not be freed in case the alloc_ring() function fails. alloc_ring() can fail with a ENOMEM return value because of possible pci_alloc_consistent() failures. Updated to CodingStyle. Signed-off-by: Jeff Garzik Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.28, 2005-03-18 21:41:52-08:00, hugh@veritas.com [PATCH] tasklist left locked On 4-way SMP, about one reboot in twenty hangs while killing processes: exit needs exclusive tasklist_lock, but something still holds read_lock. do_signal_stop race case misses unlock, and fixing it fixes the symptom. Signed-off-by: Hugh Dickins Acked-by: Roland McGrath Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.27, 2005-03-18 21:40:56-08:00, shemminger@osdl.org [PATCH] : Fix check for underflow http://bugme.osdl.org/show_bug.cgi?id=4279 Summary: When I try to start vpnc the net/core/skbuff.c:91 crash This check is wrong, gcc optimizes it away: if ((len -= sizeof(pi)) > len) return -EINVAL; This could be responsible for the BUG. If len is 2 or 3 and TUN_NO_PI isn't set it underflows. alloc_skb() allocates len + 2, which is 0 or 1 byte. skb_reserve tries to reserve 2 bytes and things explode in skb_put. [TUN]: Fix check for underflow Signed-off-by: Patrick McHardy Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.26, 2005-03-18 21:39:03-08:00, davem@davemloft.net [PATCH] [IPSEC]: Fix __xfrm_find_acq_byseq() This function, as it's name implies, is supposed to only return IPSEC objects which are in the XFRM_STATE_ACQ ("acquire") state. But it returns any object with the matching sequence number. This is wrong and confuses IPSEC daemons to no end. [XFRM]: xfrm_find_acq_byseq should only return XFRM_STATE_ACQ states. Signed-off-by: Patrick McHardy Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.25, 2005-03-18 21:37:15-08:00, kaber@trash.net [PATCH] Fix crash while reading /proc/net/route [IPV4]: Fix crash while reading /proc/net/route caused by stale pointers Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.24, 2005-03-18 21:32:44-08:00, dilinger@debian.org [PATCH] Possible AMD8111e free irq issue It seems to me that if in the amd8111e_open() fuction dev->irq isn't zero and the irq request succeeds it might not get released anymore. Specifically, on failure of the amd8111e_restart() call the function returns -ENOMEM without releasing the irq. The amd8111e_restart() function can fail because of various pci_alloc_consistent() and dev_alloc_skb() calls in amd8111e_init_ring() which is being called by amd8111e_restart. 1374 if(dev->irq ==0 || request_irq(dev->irq, amd8111e_interrupt, SA_SHIRQ, 1375 dev->name, dev)) 1376 return -EAGAIN; Signed-off-by: Jeff Garzik Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.23, 2005-03-18 21:32:08-08:00, dsd@gentoo.org [PATCH] Fix stereo mutes on Surround volume control As of 2.6.11, I have no output out of the rear right speaker of my 4.1 surround sound setup. I am using snd-intel8x0 based on a Realtek ALC650F chip on an nvidia motherboard. A gentoo user with completely different hardware also ran into this: http://bugs.gentoo.org/84276 2.6.11-mm3 fixes this problem and I've identified the specific fix, which is already in the ALSA development tree. An ALSA developer asked me to submit the fix for 2.6.11.x when I'd found it, so here it is :) -- AC97 Codec Fix stereo mutes on Surround volume control. Signed-off-by: Takashi Iwai Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.1982.168.22, 2005-03-18 21:30:08-08:00, ralf@linux-mips.org [PATCH] NetROM locking Fix deadlock in NetROM due to double locking. I was sent the patch by Alan and have doublechecked it. This bug hits Net/ROM users really hard. It's accepted by DaveM - but just too late to make it into 2.6.11. Signed-off-by: Ralf Baechle Signed-off-by: Chris Wright Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2184, 2005-03-18 22:52:54-06:00, stevef@smf-t23.(none) [CIFS] handle passwords with multiple commas in them Signed-off-by: Steve French (sfrench@us.ibm.com) ChangeSet@1.2181.22.1, 2005-03-18 15:17:04-08:00, tony.luck@intel.com [IA64] Another fix for pgd_addr_end (last one was wrong). I confused the hole in the middle of a region when an implementation of the Itanium architecture doesn't implement all the virtual address bits with the actual layout in Linux. Linux doesn't put a hole in the middle of the region, it stacks all the address space that the page tables can reference at the start of the region. Thanks to Dave Miller (again) and Peter Chubb. Signed-off-by: Tony Luck ChangeSet@1.2181.4.38, 2005-03-18 15:08:47-08:00, stern@rowland.harvard.edu [PATCH] USBcore updates This is the first of five updates to usbcore: Merge the hcd_register_root macro with the usb_register_root_hub function. This is part of the ongoing campaign to flatten out the hcd glue layer. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.37, 2005-03-18 13:56:40-08:00, olh@suse.de [PATCH] USB: another broken usb floppy Signed-off-by: Olaf Hering Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.36, 2005-03-18 13:51:56-08:00, phil@ipom.com [PATCH] USB unusual_devs: add another datafab device The following was reported by Josef Reisinger . This entry allows the device to be found and prevents it from showing up twice. Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.35, 2005-03-18 13:51:36-08:00, phil@ipom.com [PATCH] USB unusual_devs: Add another Tekom entry This patch adds support for Another "Tekom 300" camera - or more specifically a BenQ DC5330. Versions of this patch were submitted by both Frank Copeland and Manuel Fombuena - but neither were quite right, so I've tweaked it accordingly. Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.34, 2005-03-18 13:51:17-08:00, gmenguez@usuarios.retecal.es [PATCH] USB: Support for new ipod mini (and possibly others) + usb Just a little patch to unusual_devs.h in usb-storage in order to support new ipods mini (ie. the new 6 gig model) and possibly other new big models reported to have problems through usb in linux 2.6. From: Guillermo Menguez Alvarez Signed-off-by: Phil Dibowitz Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.11.79, 2005-03-18 13:46:48-08:00, schwidefsky@de.ibm.com [PATCH] s390: oprofile support Remove experimental tag from the s390 oprofile support. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.78, 2005-03-18 13:46:29-08:00, ptiedem@de.ibm.com [PATCH] s390: ctc buffer size ctc network driver changes: - Allow to change the ctc buffer size while ctc is offline. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.77, 2005-03-18 13:46:10-08:00, schwidefsky@de.ibm.com [PATCH] s390: missing timer ticks s390 core changes: - Fix missing timer ticks. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.76, 2005-03-18 13:45:52-08:00, holzheu@de.ibm.com [PATCH] s390: s390dbf permissions Use more specific permissions for the procfiles if s390dbf. Read only views should have read permission, write only views should have write permission. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.75, 2005-03-18 13:45:31-08:00, cohuck@de.ibm.com [PATCH] s390: device unregistering Common i/o layer changes: - Don't unregister devices from ccw_device_{on,off}line_notoper directly, but put the unregister on the ccw_device_work workqueue (as it is done for all other unregisters). Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.74, 2005-03-18 13:45:12-08:00, schwidefsky@de.ibm.com [PATCH] s390: add run_posix_cpu_timers to account_user_vtime The posix-timers patch introduces a call to run_posix_cpu_timers in update_process_times. The same call is required in the s390 private account_user_vtime function as well. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.73, 2005-03-18 13:44:55-08:00, schwidefsky@de.ibm.com [PATCH] s390: define atomic_sub_return Add missing atomic_sub_return for skb_release_data. Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.72, 2005-03-18 13:44:36-08:00, schwidefsky@de.ibm.com [PATCH] s390: system calls s390 system call fixes: - Add missing waitid and remap_file_pages system calls to s390. - Keep consistent naming scheme xxx_wrapper in compat_wrapper.S. - Remove #undef of __NR_getdents64 for 64 bit. The system call is present for 64 bit (linux_getdents and linux_getdents64 differ). Signed-off-by: Martin Schwidefsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.71, 2005-03-18 13:44:18-08:00, juhl-lkml@dif.dk [PATCH] mips: convert a remaining verify_area to access_ok Signed-off-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.70, 2005-03-18 13:44:00-08:00, paulus@samba.org [PATCH] ppc64: allow xmon=on,off,early allow 'xmon' or 'xmon=early' to enter xmon very early during boot. allow 'xmon=on' to just enable it, or 'xmon=off' to disable it. Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.69, 2005-03-18 13:43:42-08:00, paulus@samba.org [PATCH] ppc64: fix error cases in nvram partition scan The NVRAM on pSeries machines and powermacs is structured as a series of partitions, each of which has a header containing its length (including the header) and a header checksum. When the checksum was wrong or the length was zero, we would previously keep trying to scan through the partitions, possibly looping forever if the length was zero. This patch changes the behaviour to terminate the scan when a corrupted partition is found. Signed-off-by: Utz Bacher Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.68, 2005-03-18 13:43:21-08:00, paulus@samba.org [PATCH] ppc64: remove unnecessary ISA ioports During boot, pSeries_request_regions() should only request I/O ports for legacy ISA in the case that ISA exists on the system. Add a check for this. This patch was suggested by Anton. Signed-off-by: John Rose Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.67, 2005-03-18 13:43:04-08:00, paulus@samba.org [PATCH] ppc64: make cpu hotplug play well with maxcpus and smt-enabled This patch allows you to boot a pSeries system with maxcpus=x or smt-enabled=off (or both) and bring up the offline cpus later from userspace, assuming the kernel was built with CONFIG_HOTPLUG_CPU=y. - Record cpus which were started from OF in a cpu map and use that instead of system_state to decide how to start a cpu in smp_startup_cpu. - Change the smp bootup logic slightly so that the path for bringing up secondary threads is exactly the same as hotplugging a cpu later from userspace. - Add a new function to smp_ops - cpu_bootable. This is implemented only by pSeries to filter out secondary threads during boot with smt-enabled=off. Another way this could be done is to change the kick_cpu member to return int and we can check for this case in smp_pSeries_kick_cpu. - Remove the games we play with cpu_present_map and the hard_smp_processor_id to handle smt-enabled=off, since they're now unnecessary. - Remove find_physical_cpu_to_start; assigning threads to logical slots should be done at bootup and at DLPAR time, not during a cpu online operation. One caveat: you need up-to-date firmware on Power5 for the maxcpus option to work on systems with more than one processor. Otherwise interrupts get misrouted, typically resulting in hangs or "unable to find root filesystem" problems. Tested on Power5 with and without CONFIG_HOTPLUG_CPU and with various combinations of the maxcpus= and smt-enabled= parameters. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.66, 2005-03-18 13:42:41-08:00, paulus@samba.org [PATCH] ppc64: use pSeries reconfig notifier for cpu DLPAR Use the pSeries_reconfig notifier API to handle processor addition and removal on pSeries LPAR. This is the "right" way to do it, as opposed to setting cpu_present_map = cpu_possible_map at boot (this is fixed in a following patch). Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.65, 2005-03-18 13:42:16-08:00, olh@suse.de [PATCH] ppc64: missing newline/carrige return in zImage Some eyecandy for zImage and zImage.initrd. Most OF implementations do not print a newline after their last line of output, so the "zImage starting..." appears right after the last number or netboot output. A zImage.initrd misses a carrige return to avoid a staircase effect. Tested on JS20. Signed-off-by: Olaf Hering Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.64, 2005-03-18 13:41:56-08:00, benh@kernel.crashing.org [PATCH] ppc64: thermal control for Xserve This patch adds support for Xserve G5 to the thermal control driver. It also adds a few updates to the desktop G5 code based from changes Apple did to their own drivers. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.63, 2005-03-18 13:41:39-08:00, kravetz@us.ibm.com [PATCH] ppc64: NUMA memory fixup (another try) Below is a new version of the patch that allows holes within nodes on ppc64 NUMA. I would appreciate it if someone familiar with OF device tree parsing could take a look at this part of the code. So far, I've gotten this wrong twice. Patch was tested in various configurations on a G5 and OpenPower 720. Signed-off-by: Mike Kravetz Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.62, 2005-03-18 13:41:20-08:00, sfr@canb.auug.org.au [PATCH] ppc64 iSeries: cleanup iSeries_setup This patch does some trivial cleanups on iSeries_setup.[ch]: - eliminiate warning about iommu_init_early_iSeries not being declared - remove trailing whitespace - change some functions to static - remove defunct function declarations Built and booted on iSeries. Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.61, 2005-03-18 13:41:02-08:00, sfr@canb.auug.org.au [PATCH] ppc64 iSeries: cleanup viopath I figured I might as well do some simple cleanups. This patch does: - single bit int bitfields are a bit suspect and Anndrew pointed out recently that they are probably slower to access than ints - get rid of some more studly caps - define the semaphore and the atomic in struct alloc_parms rather than pointers to them since we just allocate them on the stack anyway. - one small white space cleanup - use the HvLpIndexInvalid constant instead of ita value Built and booted on iSeries (which is the only place it is used). Signed-off-by: Stephen Rothwell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.60, 2005-03-18 13:40:44-08:00, paulus@samba.org [PATCH] ppc64: delete unused file iSeries_fixup.h This patch is from Domen Puncer . Remove nowhere referenced file. (egrep "filename\." didn't find anything) Signed-off-by: Domen Puncer Acked-by: Stephen Rothwell Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.59, 2005-03-18 13:40:25-08:00, paulus@samba.org [PATCH] ppc64: delete unused file no_initrd.c This patch is from Domen Puncer . Remove nowhere referenced file. (egrep "filename\." didn't find anything) Signed-off-by: Domen Puncer Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.4.33, 2005-03-18 13:40:24-08:00, clemens@ladisch.de [PATCH] emi26: add another product ID for the Emi2|6/A26 This adds the product ID for another hardware revision of the Emi 2|6. Except for the ID, this device behaves the same and uses the same firmware. Signed-off-by: Clemens Ladisch Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.11.58, 2005-03-18 13:40:06-08:00, paulus@samba.org [PATCH] ppc64: make RTAS code usable on non-pSeries machines This patch is from Arnd Bergmann . RTAS is not actually pSeries specific, but some PPC64 code that relies on RTAS is currently protected by CONFIG_PPC_PSERIES. This introduces a generic configuration option PPC_RTAS that can be used by other subarchitectures as well. The existing option with the same name is renamed to the more specific RTAS_PROC. Signed-off-by: Arnd Bergmann Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.57, 2005-03-18 13:39:47-08:00, paulus@samba.org [PATCH] ppc64: kill might_sleep() warnings in __copy_*_user_inatomic This patch is from Arnd Bergmann and Olof Johansson. This implements the __copy_{to,from}_user_inatomic() functions on ppc64. The only difference between the inatomic and regular version is that inatomic does not call might_sleep() to detect possible faults while holding locks/elevated preempt counts. Signed-off-by: Arnd Bergmann Acked-by: Olof Johansson Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.56, 2005-03-18 13:39:27-08:00, paulus@samba.org [PATCH] ppc64: fix kprobes calling smp_processor_id when preemptible When booting with kprobes and preemption both enabled and CONFIG_DEBUG_PREEMPT=y, I get lots of warnings about smp_processor_id being called in preemptible code, from kprobe_exceptions_notify. On ppc64, interrupts and preemption are not disabled in the handlers for most synchronous exceptions such as breakpoints and page faults (interrupts are disabled in the very early exception entry code but are reenabled before calling the C handler). This patch adds a preempt_disable/enable pair to kprobe_exceptions_notify, and moves the preempt_disable() in kprobe_handler() to be done only in the case where we are about to single-step an instruction. This eliminates the bug warnings. From: Ananth N Mavinakayanahalli The patch is fine, but it seems to break jprobes - we have an unbalanced preempt_enable/disable path while handling jprobes. Patch below is against 2.6.11-mm4 and fixes the issue. Signed-off-by: Paul Mackerras Signed-off-by: Ananth N Mavinakayanahalli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.55, 2005-03-18 13:39:10-08:00, waite@skycomputers.com [PATCH] ppc32: add support for Sky Computers HDPU Compute blade enhanced features The Sky Computers HDPU compute blade is an embedded compute platform for the Sky Computers SMARTpac 600/1200 series. The blade consists of ppc7445/ppc7447A UP or SMP, Marvell Discovery II, 100 BaseT ethernet, a single PCI/PCI-X slot, and a PCI-X on-board Infiniband device. This patch contains drivers for enhanced features of the HDPU compute blade. Namely, 2 drivers for registers used by the Health monitoring co-processor. The cpustate register is a write only register used to convey critical states to the health monitor and the nexus registers provides a read-only interface from the health monitoring system to provide location based information. Signed-off-by: Brian Waite Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.54, 2005-03-18 13:38:51-08:00, ntl@pobox.com [PATCH] ppc64: rtasd shouldn't hold cpucontrol while sleeping The rtasd thread should not hold the cpucontrol semaphore while sleeping between event scans in its first pass; it needlessly delays boot by one second per cpu when CONFIG_HOTPLUG_CPU=y. Signed-off-by: Nathan Lynch Signed-off-by: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.53, 2005-03-18 13:38:22-08:00, waite@skycomputers.com [PATCH] ppc32: add support for Sky Computers HDPU Compute blade The Sky Computers HDPU compute blade is an embedded compute platform for the Sky Computers SMARTpac 600/1200 series. A blade consists of ppc7445/ppc7447A, Marvell Discovery II, 100 BaseT ethernet, a single PCI/PCI-X slot, and a PCI-X on-board Infiniband device. This patch contains the base platform support required for booting Linux. Signed-off-by: Brian Waite Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.52, 2005-03-18 13:37:48-08:00, tnt@246tNt.com [PATCH] ppc32: sparse clean ups for the Freescale MPC52xx related code These clean-ups are mainly missing __iomem qualifier, unnecessary (and 'wrong') casting, missing static. Signed-off-by: Sylvain Munaut Acked-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.51, 2005-03-18 13:37:25-08:00, benh@kernel.crashing.org [PATCH] ppc32: Add pegasos ethernet support This patch declares the necessary platform device on the pegasos platform so that the mv643xx eth driver can be used on it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Sven Luther Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.50, 2005-03-18 13:37:05-08:00, tnt@246tNt.com [PATCH] ppc32: Add PCI bus support for Freescale MPC52xx Note that this support has "known" problem but theses are believed to be due to hardware issues. Signed-off-by: Sylvain Munaut Acked-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.49, 2005-03-18 13:36:44-08:00, galak@freescale.com [PATCH] ppc32: Fix FEC ethernet intialization on MPC8540 ADS board The PHY interrupt for the DM9121 PHY connected to the FEC ethernet port does not work on the MPC8540 ADS board. If we tell the driver that the PHY does not have an interrupt the FEC works properly on the MPC8540 ADS board. Signed-off-by: Roy Zang Signed-off-by: Kumar Gala Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.48, 2005-03-18 13:36:24-08:00, mgreer@mvista.com [PATCH] ppc32: Clean up mv64x60 bootwrapper support This patch removes the call to mv64x60_init() in arch/ppc/boot/simple/head.S and now uses the 'load_kernel()' hook to call to have mv64x60-specific code called. This means that the mv64x60 code will be called after the bootwrapper has relocated itself. The platforms affected by this change are updated by this patch as well. Signed-off-by: Mark A. Greer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.47, 2005-03-18 13:35:33-08:00, mgreer@mvista.com [PATCH] ppc32: update Radstone ppc7d platform - Recent mv643xx #define name changes broke the PPC7D platform compile. Fixed by this patch. - Change default platform config to add mv643xx_eth and mv64xxx-i2c config options. - Add i2c platform data and update to cope with recent platform device name change. Signed-off-by: James Chapman Signed-off-by: Mark A. Greer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.46, 2005-03-18 13:35:15-08:00, mgreer@mvista.com [PATCH] ppc32: Patch for changed dev->bus_id format - Recent changes to drivers/base/platform.c:platform_device_register() changed the format of dev->bus_id (there is now a '.' between the name & instance (e.g., the old mpsc0 is now mpsc.0)). This field is used by some platform's platform_notify() routine to identify the dev entry. This patch updates the bus_id value compared to include the dot. - Fix an bad macro name change by a previous patch. - Some coding style fixups, etc. Signed-off-by: Mark A. Greer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.45, 2005-03-18 13:34:57-08:00, trini@kernel.crashing.org [PATCH] ppc32: Update 8260_io/fcc_enet.c to function again There's too many things in here that've sat too long (I'd been hoping to just delete the driver, but that hasn't happened yet, so). A cobbled together list of changes is: - Update MDIO support for workqueues. - Make use of - Add RPX6 support. - Comment out set_multicast_list (broken). - Rework tx_ring stuff so we have tx_free, not tx_Full/n_pkts. - Other PHY updates/fixes. - Leo Li: Rework FCC clock configuration, make it easier. - 2.4 : VLAN header room, other misc bits. - Kill MII_REG_NNN in favor of defines from - DM9161 PHY support (2.4, Myself & alebas@televes.com) - PQ2ADS and PQ2FADS support bits (Myself & alebas@televes.com From: Leo Li Signed-off-by: Tom Rini Signed-off-by: Alexandre Bastos Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.44, 2005-03-18 13:34:37-08:00, trini@kernel.crashing.org [PATCH] ppc32: Fix a typo on 8260 This fixes a lingering typo in arch/ppc/boot/simple/m8260_tty.c Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.43, 2005-03-18 13:34:08-08:00, trini@kernel.crashing.org [PATCH] ppc32: Serial fix for PAL4 Add PAL4's bit to Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.42, 2005-03-18 13:33:38-08:00, trini@kernel.crashing.org [PATCH] ppc32: Better comment arch/ppc/syslib/cpc700.h This adds better comments to arch/ppc/syslib/cpc700.h Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.41, 2005-03-18 13:33:17-08:00, trini@kernel.crashing.org [PATCH] ppc32: Lindent include/asm-ppc/dma.h This originally came from Paul, back in July of 2003. Run Lindent over include/asm-ppc/dma.h Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.40, 2005-03-18 13:32:53-08:00, trini@kernel.crashing.org [PATCH] ppc32: Delete arch/ppc/syslib/ppc4xx_serial.c arch/ppc/syslib/ppc4xx_serial.c is unused cruft, delete. Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.39, 2005-03-18 13:32:35-08:00, trini@kernel.crashing.org [PATCH] ppc32: Fix a warning in planb video driver [ aside: This has been sitting in the linuxppc-2.5 bk tree for I don't know how long. And the driver is still horribly broken. ] The following patch moves overlay_is_active to before its first use. It was originally written when gcc wouldn't complain, but now does, about not having the definition before usage. Signed-off-by: Tom Rini Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.38, 2005-03-18 13:32:14-08:00, benh@kernel.crashing.org [PATCH] ppc32: Add virtual DMA support to legacy floppy driver This patch adds support for pseudo-dma transfers on ppc32 for the legacy floppy driver. It is useful on some machines like pegasos where the legacy DMA doesn't seem to work properly (possibly to the lack of a "legacy" DMA zone on ppc32). Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Pavel Fedin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.37, 2005-03-18 13:31:55-08:00, benh@kernel.crashing.org [PATCH] ppc32: Update PowerMac models table This patch updates the table of PowerMac models, adding the Mac mini, a few missing ones in older slots too, and sorts it in a more logical way. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.36, 2005-03-18 13:31:36-08:00, benh@kernel.crashing.org [PATCH] ppc32: Fix overflow in cpuinfo freq. display The CPU frequency in /proc/cpuinfo would overflow because of a signed/unsigned bug. This fixes it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.11.35, 2005-03-18 13:31:16-08:00, benh@kernel.crashing.org [PATCH] ppc32: Fix PowerMac cpufreq for newer machines This patch fixes the cpufreq support for newer machines, including latest Apple laptops using the 7447A CPU. With this patch, it should now propertly detect that the CPU is booting low speed on some models, and let you switch it to full speed (previously, /proc/cpuinfo would display the frequency of the full speed CPU but it was really running low speed). Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds ChangeSet@1.2181.20.14, 2005-03-18 10:45:06-08:00, herbert@gondor.apana.org.au [NET]: Make dst_allfrag use dst instead of dst->path > BTW, shouldn't dst_allfrag be called dst_path_allfrag? Rather than doing that, let's make the path usage explicit in the one place that it's needed (the same place where we use dst_mtu(dst->path)). Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.13, 2005-03-18 10:43:40-08:00, herbert@gondor.apana.org.au [NET]: Kill dst_pmtu/dst_path_metric This would have been the patch that removed dst->path. But ip_append_data got in the way :) However, using dst->path is only needed rarely and should always require a bit of deliberation. So let's get rid of dst_pmtu and dst_path_metric and use dst_mtu and dst_metric directly. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.12, 2005-03-18 10:42:32-08:00, herbert@gondor.apana.org.au [NET]: Kill unnecessary uses of dst_path_metric This gets rid of the last unnecessary use of dst_path_metric. In decnet dst->path is always equal to dst. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.11, 2005-03-18 10:41:41-08:00, herbert@gondor.apana.org.au [IPSEC]: Get ttl from child instead of path Now that dst_pmtu is almost gone let's do the same to dst_path_metric. I've only found one legitimate use of it and that's the one that was recently added in dst_allfrag. This patch makes xfrm4_encap/xfrm6_encap use dst->child instead of dst->path so that we choose the correct route to get the hoplimit from when nested tunnels are present. For simple tunnels dst->child == dst->path so there is no change. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.10, 2005-03-18 10:39:12-08:00, herbert@gondor.apana.org.au [IPV4]: Kill remaining unnecessary uses of dst_pmtu Once again here is a couple of trivial dst_pmtu/dst_mtu replacements. In both locations, we can only have simple dst entries which means that dst == dst->path. BTW, this is the rule that we should apply in future for uses of dst_mtu/dst_pmtu (or other metrics on dst). If the only dst's that can appear are simple dst's (dst == dst->path), then we should use dst_mtu or dst_metric. If dst != dst->path, then whoever is writing the code will need to think about which of dst or dst->path is the right one. In most instances dst will be the one. However, as we have seen in ip_append_data, dst->path may be needed rarely. In particular, if we're doing fragmentation immediately after IPsec, then you may need it. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.9, 2005-03-18 10:37:41-08:00, herbert@gondor.apana.org.au [NETFILTER]: Use correct IPSEC MTU in TCPMSS This patch makes ipt_TCPMSS use the correct MTU value for clamping. This is a bit tricky actually since TCPMSS can be used in FORWARD, LOCAL_OUT as well as POST_ROUTING. In the first two cases we haven't performed IPsec yet so dst_mtu obviously does the right thing. As it is, POST_ROUTING is performed after xfrm_output so MSS clamping is useless there. With Patrick's IPsec netfilter stuff, there will be a POST_ROUTING processing before IPsec processing, in which case dst_mtu also returns exactly what we want. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.8, 2005-03-18 10:36:23-08:00, herbert@gondor.apana.org.au [IPV4]: Fix MTU check in ipmr_queue_xmit Somehow I missed four files with dst_pmtu usages in them. I'm going to split them along the sames lines I did before: bug fixes and then the trivial changes. Here is a patch that replaces dst_pmtu with dst_pmtu in ipmr.c since this is straight IPIP tunneling equivalent to what we have in ipip.c. As it is we may send ICMP packets when IPsec is present which is exactly what the comment says that we shouldn't do. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller ChangeSet@1.2181.20.7, 2005-03-18 10:34:53-08:00, shenkel@gmail.com [TUN]: Align only ethernet packets to NET_IP_ALIGN. Signed-off-by: Sven Henkel Signed-off-by: David S. Miller ChangeSet@1.2181.20.6, 2005-03-18 10:30:20-08:00, shenkel@gmail.com [NETPOLL]: Align UDP packets to NET_IP_ALIGN. This avoids unnecessary alignment traps on some platforms. Signed-off-by: Sven Henkel Signed-off-by: David S. Miller ChangeSet@1.2181.21.11, 2005-03-18 08:01:17-08:00, tali@admingilde.org [PATCH] docbook: fix escaping of kernel-doc This fixes a bug I introduced with the last patches of the DocBook generation. Signed-off-by: Martin Waitz Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.10, 2005-03-18 08:01:00-08:00, aia21@cam.ac.uk [PATCH] uml: Fix compilation due to mismerge. The recent slew of UML updates that appeared in BK seems to have gone wrong somewhere. The file "arch/um/kernel/syscall_user.c" contains identical content twice over, thus breaking compilation. Below is a patch to fix this. Please apply. Signed-off-by: Anton Altaparmakov Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.9, 2005-03-18 07:52:44-08:00, gerg@snapgear.com [PATCH] m68k-nommu: remove vendor/board specific startup code Remove all board specific startup code files. Code has been merged into a much smaller set of common code for each CPU class type, so that is 3 types now. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.8, 2005-03-18 07:52:22-08:00, gerg@snapgear.com [PATCH] m68k-nommu: move PILOT platform startup code Create more common 68328 startup code for the PILOT platform. It requires some unique setup from boot, that no other platform requires. This is part of the re-organization of the start up code to make more if it common. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.7, 2005-03-18 07:52:05-08:00, gerg@snapgear.com [PATCH] m68k-nommu: create common 68328 RAM based startup code Create common 68328 startup code for the ram based platforms. This is part of the re-organization of the start up code to make more if it common. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.6, 2005-03-18 07:51:48-08:00, gerg@snapgear.com [PATCH] m68k-nommu: remove nowhere referenced file semp3.h Remove nowhere referenced file. (egrep "filename\." didn't find anything) Signed-off-by: Domen Puncer Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.5, 2005-03-18 07:51:32-08:00, gerg@snapgear.com [PATCH] m68k-nommu: create common 68328 ROM based startup code Create common 68328 startup code for the ROM based platforms. This is part of the re-organization of the start up code to make more if it common. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.4, 2005-03-18 07:51:14-08:00, gerg@snapgear.com [PATCH] m68k-nommu: fix broken GET_MEM_SIZE macro in ColdFire head code Fix GET_MEM_SIZE macro, name was wrong. It is only used on the 5272 platform, so got overlooked in original changes. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.3, 2005-03-18 07:50:57-08:00, gerg@snapgear.com [PATCH] m68k-nommu: change build process to use common head code Update m68knommu Makefiles to build just the common start up head code now. Remove all referneces to the older vendor/board specific start codes. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.2, 2005-03-18 07:50:40-08:00, gerg@snapgear.com [PATCH] m68k-nommu: use vma list in nommu mmap support Modify the mm_context struct to keep a list of vma's instead of the uClinux specific mm_block struct's that used to be used. This reflects the changes made to mm/nommu.c in 2.6.11. Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.21.1, 2005-03-18 07:50:25-08:00, gerg@snapgear.com [PATCH] m68k-nommu: remove nowhere referenced file io_hw_swap.h Remove nowhere referenced file. (egrep "filename\." didn't find anything) Signed-off-by: Domen Puncer Signed-off-by: Greg Ungerer Signed-off-by: Linus Torvalds ChangeSet@1.2181.20.3, 2005-03-18 15:10:40+01:00, marcel@holtmann.org [Bluetooth] Kill bt_sock_alloc() and its usage Kill bt_sock_alloc() function and make the derived socks have a struct bt_sock as its first member, so that the _pi() functions can just cast the struct sock pointer to its respective types, taking advantage of the fact that sk_alloc() now use kmalloc() when no slab is passed. This is another step, close to the final one, to kill sk_protinfo and make further planned changes possible. Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: David S. Miller Signed-off-by: Marcel Holtmann ChangeSet@1.2181.20.2, 2005-03-18 13:25:42+01:00, marcel@holtmann.org [Bluetooth] Fix session reference counting for RFCOMM When an incoming connection terminates, the signal DLC is never closed and thus the underlaying L2CAP connection stays open. This problem doesn't show up often, because most times the other side takes care of terminating the signal DLC. Signed-off-by: Marcel Holtmann ChangeSet@1.2181.20.1, 2005-03-18 13:21:52+01:00, marcel@holtmann.org [Bluetooth] Support HCI Extensions in BCSP driver To support the vendor specific HCI commands and events the BCSP drivers needs to convert these to BCSP packets for the correct channel. Signed-off-by: Marcel Holtmann ChangeSet@1.2181.11.31, 2005-03-18 12:06:33+01:00, bzolnier@trik.(none) [ide] ide-tape: fix character device ->open() vs ->cleanup() race Similar to the same race but for the block device. * store pointer to struct ide_tape_obj in idetape_chrdevs[] * rename idetape_chrdevs[] to idetape_devs[] and kill idetape_chrdev_t * add ide_tape_chrdev_get() for getting reference to the tape * store tape pointer in file->private_data and fix all users of it * fix idetape_chrdev_{open,release}() to get/put reference to the tape Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.30, 2005-03-18 12:05:37+01:00, bzolnier@trik.(none) [ide] ide-scsi: add basic refcounting * pointers to a SCSI host and a drive are added to idescsi_scsi_t * pointer to the SCSI host is stored in disk->private_data * ide_scsi_{get,put}() is used to {get,put} reference to the SCSI host Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.29, 2005-03-18 12:01:47+01:00, bzolnier@trik.(none) [ide] ide-tape: add basic refcounting Similar changes as for ide-cd.c. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.28, 2005-03-18 11:59:51+01:00, bzolnier@trik.(none) [ide] ide-floppy: add basic refcounting Similar changes as for ide-cd.c. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.27, 2005-03-18 11:57:25+01:00, bzolnier@trik.(none) [ide] ide-disk: add basic refcounting Similar changes as for ide-cd.c (except that struct ide_disk_obj is added). Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.26, 2005-03-18 11:54:40+01:00, bzolnier@trik.(none) [ide] ide-cd: add basic refcounting * based on reference counting in drivers/scsi/{sd,sr}.c * fixes race between ->open() and ->cleanup() (ide_unregister_subdriver() tests for drive->usage != 0 but there is no protection against new users) * struct kref and pointer to a drive are added to struct ide_cdrom_info * pointer to drive's struct ide_cdrom_info is stored in disk->private_data * ide_cd_{get,put}() is used to {get,put} reference to struct ide_cdrom_info * ide_cd_release() is a release method for struct ide_cdrom_info Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.25, 2005-03-18 11:51:25+01:00, bzolnier@trik.(none) [ide] make ide_generic_ioctl() take ide_drive_t * as an argument As a result disk->private_data can be used by device drivers now. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.24, 2005-03-18 11:45:52+01:00, htejun@gmail.com [ide] hdio.txt update This patch updates Documentation/ioctl/hdio.txt to include more detailed descriptions about HDIO_DRIVE_{CMD|TASK|TASKFILE} ioctls. Signed-off-by: Tejun Heo Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.23, 2005-03-18 11:42:11+01:00, tklauser@nuerscht.ch [ide] drivers/ide/cs5520.c: use the DMA_{64,32}BIT_MASK constants Description: Use the DMA_{64,32}BIT_MASK constants from dma-mapping.h when calling pci_set_dma_mask() or pci_set_consistent_dma_mask() See http://marc.theaimsgroup.com/?t=108001993000001&r=1&w=2 for details Signed-off-by: Tobias Klauser Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.11.22, 2005-03-18 11:36:11+01:00, Jason.d.gaston@intel.com [ide] pci_ids.h correction for Intel ICH7R This patch removes an incorrect ICH7R DID in pci_ids.h. Signed-off-by: Jason Gaston Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.2181.9.15, 2005-03-18 00:00:20-05:00, len.brown@intel.com merge ChangeSet@1.2181.9.14, 2005-03-17 23:26:54-05:00, len.brown@intel.com [ACPI] build fix in acpi_pci_irq_disable() bk-acpi-acpi_pci_irq_disable-build-fix.patch Signed-off-by: Andrew Morton Signed-off-by: Len Brown ChangeSet@1.2181.12.23, 2005-03-17 20:00:13-08:00, davem@sunset.davemloft.net [IPV4]: Make multipath algs into true drivers. This also makes them configurable on a per-route basis via rtnetlink route attributes. Based upon suggestions from Thomas Graf and Alexey Kuznetsov. Signed-off-by: David S. Miller ChangeSet@1.2181.4.32, 2005-03-17 18:15:50-08:00, gregkh@suse.de [PATCH] USB Storage: remove unneeded unusual_devs.h entry. Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.31, 2005-03-17 18:15:30-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: change how unusual_devs.h flags are defined This patch started life as as473 from Alan Stern, and has been rediffed against the current tip. This patch changes the way the flags bits are defined, so that the proc_info routine can automatically list all the flags that are set. As a side effect it will also include the NEED_OVERRIDE flag; that shouldn't really matter. (As another side effect, the definitions grew too long to fit on a single 80-character line so I had to wrap the comments.) This patch also takes the opportunity to remove a couple of obsolete flags and to renumber the bit values. The patch also removes an unusual_devs entry whose only purpose was to set one of those obsolete flags; apparently it got left behind when all the other uses of those flags were removed. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.30, 2005-03-17 18:15:11-08:00, rddunlap@osdl.org [PATCH] pwc: fix printk arg types Fix gcc printk arg type warnings: drivers/usb/media/pwc/pwc-if.c:325: warning: int format, different type arg (arg 2) drivers/usb/media/pwc/pwc-if.c:1182: warning: int format, different type arg (arg 4) Signed-off-by: Randy Dunlap Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.29, 2005-03-17 18:14:52-08:00, rddunlap@osdl.org [PATCH] sisusb: fix arg. types Fix gcc printk arg type and other function parameter warnings: drivers/usb/misc/sisusbvga/sisusb.c: In function `sisusb_send_packet': drivers/usb/misc/sisusbvga/sisusb.c:583: warning: passing arg 7 of `sisusb_send_bulk_msg' from incompatible pointer type drivers/usb/misc/sisusbvga/sisusb.c:591: warning: passing arg 6 of `sisusb_recv_bulk_msg' from incompatible pointer type drivers/usb/misc/sisusbvga/sisusb.c: In function `sisusb_send_bridge_packet': drivers/usb/misc/sisusbvga/sisusb.c:620: warning: passing arg 7 of `sisusb_send_bulk_msg' from incompatible pointer type drivers/usb/misc/sisusbvga/sisusb.c:628: warning: passing arg 6 of `sisusb_recv_bulk_msg' from incompatible pointer type drivers/usb/misc/sisusbvga/sisusb.c: In function `sisusb_write_mem_bulk': drivers/usb/misc/sisusbvga/sisusb.c:989: warning: int format, different type arg (arg 3) drivers/usb/misc/sisusbvga/sisusb.c:989: warning: int format, different type arg (arg 3) drivers/usb/misc/sisusbvga/sisusb.c: In function `sisusb_clear_vram': drivers/usb/misc/sisusbvga/sisusb.c:1407: warning: passing arg 7 of `sisusb_write_mem_bulk' from incompatible pointer type Signed-off-by: Randy Dunlap Signed-off-by: Thomas Winischhofer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.28, 2005-03-17 18:14:33-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: remove unneeded NULL tests This patch started life as as472 from Alan Stern, and has been rediffed against the current tip. This patch simply removes some unnecessary NULL checking before kfree() calls. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.27, 2005-03-17 18:14:14-08:00, mdharm-usb@one-eyed-alien.net [PATCH] USB Storage: Header reorganization This patch started life as as471 from Alan Stern, and has been regenerated against the current tip. This patch cleans up the use of header files. Primarily it makes sure that "usb.h" is included before any of the other local headers. It also removes some unnecessary declarations of struct us_data and struct scsi_cmnd, and it moves the inclusion of to usb.h where it will be needed by a later patch. Signed-off-by: Alan Stern Signed-off-by: Matthew Dharm Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.26, 2005-03-17 18:13:54-08:00, okir@suse.de [PATCH] USB: fix uhci irq 10: nobody cared! error From: Olaf Kirch Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.25, 2005-03-17 18:13:34-08:00, domen@coderock.org [PATCH] USB: compile warning cleanup compile warning cleanup - handle error return from scsi_add_host Signed-off-by: Stephen Biggs Signed-off-by: Domen Puncer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.24, 2005-03-17 18:13:15-08:00, domen@coderock.org [PATCH] usb/digi_acceleport: remove interruptible_sleep_on_timeout() usage Replace deprecated interruptible_sleep_on_timeout() with direct wait-queue usage. Also replace some rather odd wait-queue usage with the existent macros. Also adjusted the wake_up_interruptible() call appropriately, as I changed all the states to TASK_UNINTERRUPTIBLE (signals were not be checked in the current code). Patch is compile-tested. Signed-off-by: Nishanth Aravamudan Signed-off-by: Domen Puncer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.23, 2005-03-17 18:12:55-08:00, domen@coderock.org [PATCH] usb/rio500: remove interruptible_sleep_on_timeout() usage Replace deprecated interruptible_sleep_on_timeout() with direct wait-queue usage. Patch is compile-tested. Signed-off-by: Nishanth Aravamudan Signed-off-by: Domen Puncer Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.22, 2005-03-17 17:58:07-08:00, oliver@neukum.org [PATCH] USB: removal of obsolete error code from kaweth this patch from David removes an obsolete error code from kaweth. Signed-off-by: Oliver Neukum Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.21, 2005-03-17 17:57:47-08:00, stern@rowland.harvard.edu [PATCH] USB: Prevent hub driver interference during port reset This patch causes the hub driver to ignore ports that are being reset by another driver. Without this protection khubd will sometimes notice that the port is behaving funny while the reset is taking place and will unilaterally disconnect it, which is not what we want. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.20, 2005-03-17 17:57:27-08:00, stern@rowland.harvard.edu [PATCH] g_file_storage: add configuration and interface strings This patch adds iConfiguration and iInterface strings to the g_file_storage driver (just to spiff it up and help in testing software that looks at those strings). It also enlarges the space reserved for the Manufacturer string, since combinations of long kernel names and long UDC driver names were sometimes getting truncated. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.19, 2005-03-17 17:57:07-08:00, stern@rowland.harvard.edu [PATCH] usb-midi: fix arguments to usb_maxpacket() The usb-midi driver uses an incorrect value for the is_out argument to usb_maxpacket(), triggering a WARN_ON. This patch fixes the mistake. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.18, 2005-03-17 17:56:48-08:00, akpm@osdl.org [PATCH] usb hcd u64 warning fix drivers/usb/core/hcd.c:1689: warning: long long unsigned int format, u64 arg (arg 6) drivers/usb/core/hcd.c:1695: warning: long long unsigned int format, u64 arg (arg 5) We must not assume that u64 is implemented as `unsigned long long'. On ppc64 (for example) it is `unsigned long'. Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.17, 2005-03-17 17:49:09-08:00, zaitcev@redhat.com [PATCH] USB: Patch for ub to fix oops after disconnect This patch was developed by Glenn Maynard for his oops back in December, but for some reason I forgot about it and had it reimplemented when I stepped on this problem myself. In my case, there was no oops, but a warning about slab corruption. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.16, 2005-03-17 17:35:24-08:00, david-b@pacbell.net [PATCH] USB: ethernet/rndis gadget driver updates Various fixes to the Ethernet/RNDIS gadget core code: - Pre-allocate the request used to transfer status back to the host. Used initially for CDC Ethernet; RNDIS will change later. This resolves a longstanding FIXME, elimininating fault modes. - Use larger packets for those status reports, 16 bytes not 8; this eliminates some fault modes, without losing hardware support. - Streamline endpoint configuration, just save the endpoints during driver binding. The previous scheme was a complex leftover from before the endpoint autoselection library code existed, and this bit of cleanup prepares for more simplifications later. - Implement a basic outgoing packet filter, for CDC Ethernet and RNDIS but not the CDC subset. This improves conformance to both specs. - Correct the bit rate reports for CDC Ethernet and RNDIS to match the peak bulk transfer rates, not the raw signaling rates. This still doesn't issue CDC or RNDIS link status change notifications to the host as often as it should, but that'll be easier now. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.15, 2005-03-17 17:35:05-08:00, david-b@pacbell.net [PATCH] USB: net2280 reports correct dequeue status Minor bugfix to net2280: don't return incorrect dequeue() status. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.14, 2005-03-17 17:34:45-08:00, david-b@pacbell.net [PATCH] USB: usbnet fix for Zaurus C-860 This patch resolves a recent problem with the Zaurus C-860 support. A change to correct handling of Zaurii that are lying about their support for the "CDC Ethernet" class specification broke the C-860, which tells an entirely different lie (that it supports "CDC MDLM", providing access to a cell phone modem). The code expecting it to be telling a lie about CDC Ethernet support naturally misbehaved. (Sharp should straighten out its story. The 2.6 OpenZaurus kernels don't have any such issues...) The fix is just to recognize this bogus MDLM stuff and ignore it. This patch also includes the two MDLM descriptors in although they're not currently used. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.13, 2005-03-17 17:34:24-08:00, david-b@pacbell.net [PATCH] USB: usbnet gets status polling, uses for CDC Ethernet This adds status/interrupt transfer infrastructure to "usbnet", and uses it for CDC Ethernet support. It can be used with other devices that define an interrupt-IN endpoint (quite a few!), so long as the meaning of the events is documented (erm, not so many). Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.12, 2005-03-17 17:34:04-08:00, david-b@pacbell.net [PATCH] USB: ehci and short in-bulk transfers with 20KB+ urbs This changes handling of short bulk IN transfers with URB buffers that require two or more transfer descriptors. The case is rare in most systems, since few drivers use such large buffers (bigger than 20KB, assuming 4K-aligned). The existing code misbehave on at least NEC's EHCI 0.95 silicon, where it seems to hit a new variant of a silicon quirk relating to dummy TDs. Symptom of the misbehavior is that the host stops polling on the IN endpoint, and the URB queue no longer progresses. This fix uses simpler logic to detect those short reads, dodging that quirk. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.11, 2005-03-17 17:33:44-08:00, david-b@pacbell.net [PATCH] USB: ohci zero length control IN transfers This fixes a longstanding bug in the OHCI driver, inherited from the 2.4 code. It also fixes a related comment in the EHCI driver, which came when the EHCI code was first derived from OHCI. (The EHCI code doesn't have that bug; the comment was effectively "FIXME add bug".) The bug: control-IN transfers with no data stages have status stages with an IN packet (like a control-OUT transfer), instead of using an OUT packet. See figure 8-37 of the USB spec. The current code makes trouble with some peripheral hardware and zero length IN transfers (scarce in the Real World). Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.10, 2005-03-17 17:33:24-08:00, david-b@pacbell.net [PATCH] USB: usb gadget kconfig tweaks This makes two small changes to the gadget Kconfig. - It removes SA-1100 support ... if anyone gets around to finishing that controller driver, it could go back, but until then there's no real point. - The "dummy_hcd" moves to the end, with a comment to please keep it there. Moving that Kconfig entry helps keep "real hardware" as the default on systems that have it ... otherwise Kconfig will surprise people with a curious default. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.9, 2005-03-17 17:33:03-08:00, david-b@pacbell.net [PATCH] USB: add at91_udc recognition Add declarations to recognize the AT91 USB peripheral controller, as used in various ARMv4T chips including the mmu-ful at91rm9200. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.8, 2005-03-17 17:13:54-08:00, bunk@stusta.de [PATCH] remove drivers/usb/image/hpusbscsi.c USB_HPUSBSCSI was marked as BROKEN in 2.6.11 since libsane is the preferred way to access these devices. Unless someone plans to resurrect this driver, I'm therefore proposing this patch to completely remove it. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.7, 2005-03-17 17:09:47-08:00, bunk@stusta.de [PATCH] drivers/usb/net/pegasus.c: make some code static This patch makes some needlessly global code static. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.6, 2005-03-17 17:09:18-08:00, bunk@stusta.de [PATCH] drivers/usb/storage/: cleanups This patch contains the following cleanups: - make needlessly global code static - scsiglue.c: remove the unused usb_stor_sense_notready Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.5, 2005-03-17 17:07:38-08:00, bunk@stusta.de [PATCH] drivers/usb/serial/: make some functions static This patch makes some needlessly global functions static. Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.4.4, 2005-03-17 17:04:11-08:00, bunk@stusta.de [PATCH] USB: possible cleanups Before I'm getting flamed to death: This patch contains possible cleanups. If parts of this patch conflict with pending changes these parts of my patch have to be dropped. This patch contains the following possible cleanups: - make needlessly global code static - #if 0 the following unused global functions: - core/usb.c: usb_buffer_map - core/usb.c: usb_buffer_unmap - remove the following unneeded EXPORT_SYMBOL's: - core/hcd.c: usb_bus_init - core/hcd.c: usb_alloc_bus - core/hcd.c: usb_register_bus - core/hcd.c: usb_deregister_bus - core/hcd.c: usb_hcd_irq - core/usb.c: usb_buffer_map - core/usb.c: usb_buffer_unmap - core/buffer.c: hcd_buffer_create - core/buffer.c: hcd_buffer_destroy Signed-off-by: Adrian Bunk Signed-off-by: Greg Kroah-Hartman ChangeSet@1.2181.11.21, 2005-03-17 16:52:58-08:00, torvalds@ppc970.osdl.org Linux 2.6.12-rc1 TAG: v2.6.12-rc1