ChangeSet@1.1058, 2003-08-04 18:08:39-03:00, marcelo@logos.cnet Cset exclude: bunk@fs.tum.de|ChangeSet|20030804201535|32414 ChangeSet@1.1052.1.3, 2003-08-04 17:15:35-03:00, bunk@fs.tum.de [PATCH] fix a compile warning in acpi/system.c On Fri, Aug 01, 2003 at 01:19:11PM -0300, Marcelo Tosatti wrote: >... > Summary of changes from v2.4.22-pre9 to v2.4.22-pre10 > ============================================ >... > Willy Tarreau: > o ACPI poweroff fix >... This patch causes the following compile warnings: <-- snip --> ... gcc -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.22-pre10-full/include - Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=k6 -Os -DACPI_DEBUG_OUTPUT -nostdinc -iwithprefix include -DKBUILD_BASENAME=system -c -o system.o system.c system.c: In function `acpi_power_off': system.c:93: warning: implicit declaration of function `acpi_suspend' system.c: At top level: system.c:303: warning: type mismatch with previous implicit declaration system.c:93: warning: previous implicit declaration of `acpi_suspend' system.c:303: warning: `acpi_suspend' was previously implicitly declared to return `int' ... <-- snip --> The following patch removes these warnings: ChangeSet@1.1052.1.2, 2003-08-04 16:45:36-03:00, bunk@fs.tum.de [PATCH] fix a compile warning in acpi/system.c On Fri, Aug 01, 2003 at 01:19:11PM -0300, Marcelo Tosatti wrote: >... > Summary of changes from v2.4.22-pre9 to v2.4.22-pre10 > ============================================ >... > Willy Tarreau: > o ACPI poweroff fix >... This patch causes the following compile warnings: <-- snip --> ... gcc -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.22-pre10-full/include - Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=k6 -Os -DACPI_DEBUG_OUTPUT -nostdinc -iwithprefix include -DKBUILD_BASENAME=system -c -o system.o system.c system.c: In function `acpi_power_off': system.c:93: warning: implicit declaration of function `acpi_suspend' system.c: At top level: system.c:303: warning: type mismatch with previous implicit declaration system.c:93: warning: previous implicit declaration of `acpi_suspend' system.c:303: warning: `acpi_suspend' was previously implicitly declared to return `int' ... <-- snip --> The following patch removes these warnings: ChangeSet@1.1052.1.1, 2003-08-04 14:52:36-03:00, marcelo@logos.cnet Delete: fs/noquot.c ChangeSet@1.1056, 2003-08-02 10:11:36-04:00, lethal@unusual.internal.linux-sh.org sh: Further support for SecureEdge5410 and SH7751R This adds in further support and general fixes for the SH7751R PCI host bridge, as well as the SnapGear SH7751R-based SecureEdge5410 boards. ChangeSet@1.1055, 2003-08-02 10:06:00-04:00, lethal@unusual.internal.linux-sh.org sh: Add pgprot_nocached() definition. Reported and fixed by Alex Bennee: As far as I can tell this only affect user-mode access that mmap /dev/mem to access hardware. If pgprot_nonchached is not set in drivers/char/mem.c it defines it own which funnily enough does nothing for the SH arch. This little patch fixes that (and my 2.4.21 problems). It was included in one of the older versions of the ST kernels but not currently. ChangeSet@1.1054, 2003-08-02 10:01:27-04:00, lethal@unusual.internal.linux-sh.org sh: Fix single stepping from looping As reported by Sugioka-san: I found that single stepping infinitely loops when delayed jump instruction jumps into it's own delay slot. For example, bra 1f 1: nop In this case, when single stepping start from 'bra' instruction, UBC set to 'nop' instruction, and then breaks before executing 'bra' instruction, saved PC register holds the address of 'bra' instruction. So next single stepping will start from 'bra' instruction again. Yes, it seems strange and non-sense instruction sequence, but gcc-3.2.3 generates that kind of sequence and there are in /lib/libc-2.2.5.so. Solution, When single stepping, if instruction has delay slot and calculated next pc points it's own delay slot, add 2 to the next pc. patch from SUGIOKA Toshinobu. ChangeSet@1.1053, 2003-08-02 09:49:40-04:00, lethal@unusual.internal.linux-sh.org sh: Define __flush_icache_all() for SH-3 Add a dummy __flush_icache_all() definition for SH-3, which is needed for proper compilation of some of the new cache flushing code that was merged from the ST tree. ChangeSet@1.1052, 2003-08-01 10:24:28-03:00, marcelo@logos.cnet NMI watchdog documentation for x86-64 TAG: v2.4.22-pre10 ChangeSet@1.1051, 2003-07-31 15:09:42-04:00, fubar@us.ibm.com [netdrvr bonding] fix ifenslave ia64 build ChangeSet@1.1050, 2003-07-31 14:52:02-04:00, willy@w.ods.org [netdrvr bonding] fix kernel panic when optional feature used - now the last one fixes a kernel panic due to a cheap hack which was introduced to determine the source IP address to use with ARP checks. It takes the first address of the first slave, and puts a lock on it. If there's no address, its ip_ptr is NULL, and the kernel panics while trying to get the lock. You can reproduce it easily this way : # modprobe eth0 # modprobe bonding mode=active-backup miimon=1000 # ip link set bond0 up # ifenslave bond0 eth0 => kernel panic ! ChangeSet@1.1049, 2003-07-31 14:50:38-04:00, willy@w.ods.org [netdrvr bonding] fix a typo in the MODULE_PARM_DESC ChangeSet@1.1048, 2003-07-31 14:58:03-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre10 ChangeSet@1.1047, 2003-07-31 14:55:11-03:00, m.c.p@wolk-project.de [PATCH] Add missing -EFAULT for sysctl Hi Marcelo, attached patch does the following things: 1. Add two missing -EFAULT checks for sysctl 2. Code cleanup ChangeSet@1.1046, 2003-07-31 13:42:27-03:00, m.c.p@wolk-project.de [PATCH] Fix /proc/self security issue [ The following text is in the "iso-8859-15" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] Hi Marcelo, CAN-2003-0501: The /proc filesystem in Linux allows local users to obtain sensitive information by opening various entries in /proc/self before executing a setuid program. This causes the program to fail to change the ownership and permissions of already opened entries. In -ac and -wolk for some time. ChangeSet@1.1041.1.12, 2003-07-31 13:18:27-03:00, adam@nmt.edu [PATCH] 3ware driver update ChangeSet@1.1041.3.1, 2003-07-31 10:36:50-05:00, shaggy@shaggy.austin.ibm.com JFS: write_super_lockfs should mark superblock clean LVM and EVMS snapshots of JFS filesystems were not mountable because write_super_lockfs was not marking the volume as clean. The volume was otherwise in a consistent state. ChangeSet@1.1041.1.10, 2003-07-31 12:36:00-03:00, m.c.p@wolk-project.de [PATCH] MXCSR Handler Unspecified Vulnerability Hi Marcelo, the Linux Kernel MXCSR handler code has been reported prone to an unspecified vulnerability. The issue presents itself when low-level MXCSR kernel code encounters a malformed address. It has been reported that the MXCSR code fails to sufficiently handle malformed address data and will leave garbage in the CPU state registers. Although speculative, it has been conjectured that this issue may allow an attacker to trigger a denial of service condition. Although unconfirmed other attacks may also be possible. In all known vendor kernels, -ac, -aa and -wolk for some time. ChangeSet@1.1041.2.5, 2003-07-31 12:14:59-03:00, mike.miller@hp.com [PATCH] cciss update: Fix problem with shared IRQs This patch patch built & tested using the 2.4.21 kernel with the 2.4.22-pre9 patch applied. It is intended for inclusion in the 2.4.22 kernel. Problem scenario: The system is running in Full-Table Mapped APIC mode and a cciss device is sharing an IRQ with another device that gets a lot of interrupts. If the driver is unloaded and reloaded during runtime there a period of time during intialization in which we are polling. We send down an inquiry to see whats attached to the controller. If the command completes BEFORE our next poll AND an interrupt occurs on that line the OS will call the interrupt handler for all devices on that line. We see that we have a completed command and attempt to service it before all of our data structures have been initialized. Nasty. Bug fix. Changes: 1. Added a flag to the ctlr_info struct to track if interrupts are enabled or disabled. 2. Modified our condition for an interrupt. Now we check if anything is outstanding AND if are interrupts enabled. ChangeSet@1.1041.2.4, 2003-07-31 12:14:41-03:00, mike.miller@hp.com [PATCH] cciss update: author change This patch was built & tested using kernel 2.4.21 with the 2.4.22-pre9 patch applied. It is intended for inclusion in the 2.4.22 kernel. Changes: 1. Changes the author from Charles White to Hewlett-Packard Company. Requirement from management. Sorry, Charles. ChangeSet@1.1041.2.3, 2003-07-31 12:09:37-03:00, m.c.p@wolk-project.de [PATCH] Fix irq handling of IO-APIC edge IRQs on UP Hi Marcelo, send_IPI_self is needed to resend irqs with IRQ_PENDING status when enabled. Checked with Ingo, and it's in 2.4-ac for some time, also in 2.5 since 2 weeks, also in 2.4-WOLK. See here: http://linux.bkbits.net:8080/linux-2.5/cset@1.1455.1.9?nav=index.html|ChangeSet@-12w This should fix ide lost interrupts on UP with IO-APIC. The needed ide-fix is already in .22-pre. Ide trigers it this way: - disable_irq - do stuff that triggers IRQ. - irq is IRQ_PENDING - enable_irq - IRQ is lost, needs to be resend. ACKed by Alan some minutes ago. ChangeSet@1.1041.2.2, 2003-07-31 11:46:10-03:00, m.c.p@wolk-project.de [PATCH] Fix AGPGART problem with 4GB RAM I forward it to you since Marcelo E. Magallon forgot to CC you. Verified by me and works! - Please include it before -pre10. Thanks. ChangeSet@1.1041.2.1, 2003-07-31 11:38:15-03:00, herbert@13thfloor.at [PATCH] ROOT NFS fixes I'm sure you're a busy man, but please have a look at the attached patch, which addresses the root=/dev/wossname issue, when ROOT_NFS is configured. IMHO this is a NFS only issue, and fixes an erroneous behaviour. so to clarify this patch once again, beside doing some documentational stuff, it does _not_ produce the same code, because otherwise I would not have bothered in the first palce ... - if (MAJOR(ROOT_DEV) == UNNAMED_MAJOR) { + if (MAJOR(ROOT_DEV) == NFS_MAJOR + && MINOR(ROOT_DEV) == NFS_MINOR) { this trivial code change, as stated several times, changes the behaviour regarding unknown root= devices at boot time root_dev_setup() is called with "root=" and, in turn calls name_to_kdev_t() which searches root_dev_names[] for the given device names, which if not found results in (0:0) this is handled by the current code like dev=/dev/nfs, which IMHO is the wrong behaviour ... ChangeSet@1.1041.1.8, 2003-07-31 17:01:50+04:00, green@angband.namesys.com reiserfs: fix savelinks for bigendian arches. Remove unneeded cpu->disk order conversion for savelink data. ChangeSet@1.1041.1.7, 2003-07-30 17:55:34-03:00, benh@kernel.crashing.org [PATCH] ppc32: export hash_table_lock on SMP for MacOnLinux Hi Marcelo ! The MacOnLinux "virtual machine" needs hash_table_lock symbol to be exported. This patch fixes that. Please apply. (That used to be in my tree for ages). Cheers, Ben. ===== arch/ppc/kernel/ppc_ksyms.c 1.30 vs edited ===== ChangeSet@1.1041.1.6, 2003-07-30 17:28:07-03:00, jaharkes@cs.cmu.edu [PATCH] Coda fixes Hi Marcello, Here are some Coda related patches. - In coda_readdir, don't use vfs_readdir as we are already holding the inode semaphore. - Remove an unlock_kernel() in a place where we don't hold the BKL. - Avoid an oops when the cache manager passes back a bad fd. The last part of the patch is against fs/namei.c, and avoids dentry revalidation on all names that starts with '.', instead of only revalidating '.' and '..' lookups. Here is Trond's response, http://marc.theaimsgroup.com/?l=linux-kernel&m=105768272826912&w=2 Jan ChangeSet@1.1041.1.5, 2003-07-30 12:46:16-03:00, bunk@fs.tum.de [PATCH] fix IPMI build error #if CONFIG_ACPI_HT_ONLY I got the following build error in 2.6.0-test2: <-- snip --> ... LD .tmp_vmlinux1 ... drivers/built-in.o(.init.text+0xdff5): In function `init_ipmi_kcs': : undefined reference to `acpi_find_bmc' make: *** [.tmp_vmlinux1] Error 1 $ grep ACPI .config # Power management options (ACPI, APM) # ACPI Support CONFIG_ACPI=y CONFIG_ACPI_HT_ONLY=y $ <-- snip --> acpi_find_bmc is only available #ifdef CONFIG_ACPI_INTERPRETER. The following patch fixes it (the same problem does exist in 2.4.22-pre8, the patch applies with a few lines offset): ChangeSet@1.1041.1.4, 2003-07-30 11:09:37-03:00, willy@w.ods.org [PATCH] ACPI poweroff fix On Thu, Jul 24, 2003 at 06:02:15PM -0300, Marcelo Tosatti wrote: > > Great. I`ll apply it to the 2.4 tree later and it will be present in > -pre9. Hi Marcelo, it seems you forgot the patch in -pre9. Never mind, I've just rediffed it, here it is. Cheers, Willy ChangeSet@1.1041.1.3, 2003-07-30 11:03:02-03:00, vandrove@vc.cvut.cz [PATCH] matroxfb: extended support for mplayer Hi, Arpi, one of mplayer developers, asked me if I could submit this patch to you, as it appears stable in the wild, and its main portion (hooking matroxfb's interrupt) is used in 2.5.x kernels for about 6 months. Fix invalid behavior when FBIO_PAN_DISPLAY is invoked on fbdev without attached virtual terminal. Add frame count reporting to the matroxfb. Add FBIO_WAITFORSYNC for output frame synchronization. Add FB_ACTIVATE_VBL support to get rid of tearing effects without slowing down apps. drivers/video/matrox/matroxfb_DAC1064.c | 2 drivers/video/matrox/matroxfb_Ti3026.c | 1 drivers/video/matrox/matroxfb_base.c | 259 ++++++++++++++++++++++++++++---- drivers/video/matrox/matroxfb_base.h | 30 ++- drivers/video/matrox/matroxfb_crtc2.c | 48 ++++- drivers/video/matrox/matroxfb_misc.c | 13 - include/linux/matroxfb.h | 2 7 files changed, 303 insertions(+), 52 deletions(-) Thanks, Petr Vandrovec vandrove@vc.cvut.cz ChangeSet@1.1041.1.2, 2003-07-30 10:59:22-03:00, vandrove@vc.cvut.cz [PATCH] ncpfs: Support for clustered NetWare volumes Hi Marcelo, clustered volumes are numbered from 255 down to 0. This makes them unreachable on Linux, as it accepts only volume numbers 0 to 63. Fortunately fix is really simple, just changing maximum from 64 to 256. Thanks, Petr Vandrovec vandrove@vc.cvut.cz ChangeSet@1.1041.1.1, 2003-07-29 21:22:32-03:00, bjorn.helgaas@hp.com [PATCH] trivial 2.4 HCDP documentation/config patch Hi Marcelo, This just adds help for CONFIG_SERIAL_HCDP and fixes indentation. #### AUTHOR bjorn.helgaas@hp.com #### COMMENT START ### Comments for ChangeSet CONFIG_SERIAL_HCDP documentation and indent fix. ### Comments for Documentation/Configure.help Add CONFIG_SERIAL_HCDP documentation. ### Comments for drivers/char/Config.in Fix CONFIG_SERIAL_HCDP indentation. #### COMMENT END # This is a BitKeeper generated patch for the following project: # Project Name: Linux kernel tree # This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas: # ChangeSet 1.1039 -> 1.1040 # drivers/char/Config.in 1.46 -> 1.47 # Documentation/Configure.help 1.168 -> 1.169 # # The following is the BitKeeper ChangeSet Log # -------------------------------------------- # 03/07/29 bjorn.helgaas@hp.com 1.1040 # CONFIG_SERIAL_HCDP documentation and indent fix. # -------------------------------------------- # ChangeSet@1.1019.1.22, 2003-07-30 01:54:20+02:00, trond.myklebust@fys.uio.no If xdr_kmap() fails, we need to ensure that it unmaps all the pages, and returns 0. We don't want to be sending partial RPC requests to the server. ChangeSet@1.1019.1.21, 2003-07-30 01:52:27+02:00, trond.myklebust@fys.uio.no If an RPC request has to be resent due to a timeout, it turns out that call_encode() may cause rq_rcv_buf to be reset despite the fact that a reply might be delivered at any moment by a softirq. This typically results in 'NFS: server cheating in read reply' error messages. The following patch adds rq_private_buf, which is a mirror of rq_rcv_buf for the benefit of the softirqs. ChangeSet@1.1040, 2003-07-29 15:34:12-04:00, shmulik.hen@intel.com [netdrvr bonding] fix ARP monitoring bug ChangeSet@1.1019.1.20, 2003-07-29 14:46:58-03:00, marcelo@freak.distro.conectiva Merge bk://ppc.bkbits.net/for-marcelo-ppc into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: v2.4.22-pre9 TAG: v2.422-pre9 ChangeSet@1.1003.42.2, 2003-07-29 10:27:45-07:00, trini@kernel.crashing.org PPC32: Correctly set intfreq / busfreq on the Motorola 860FADS. Patch from Wade Farnsworth ChangeSet@1.1019.4.1, 2003-07-29 10:06:20-07:00, trini@kernel.crashing.org Merge kernel.crashing.org:/home/trini/work/kernel/pristine/linux-2.4-misc into kernel.crashing.org:/home/trini/work/kernel/pristine/for-marcelo-ppc ChangeSet@1.1039, 2003-07-29 11:59:57-04:00, jgarzik@redhat.com [netdrvr] add new broadcom 440x net driver, "b44" By David Miller, with many fixes from Pekka Pietikainen. ChangeSet@1.1003.42.1, 2003-07-29 08:55:59-07:00, trini@kernel.crashing.org PPC32: Allow eth0 and eth1 to work on MPC8xx boards with QS6612 PHYs. Fix from Wade Farnsworth . ChangeSet@1.1038, 2003-07-29 11:49:49-04:00, achirica@telefonica.net [wireless airo] adds support for noise level reporting (if available) ChangeSet@1.1037, 2003-07-29 11:48:35-04:00, achirica@telefonica.net [wireless airo] makes the card passive when entering monitor mode ChangeSet@1.1036, 2003-07-29 11:46:31-04:00, achirica@telefonica.net [wireless airo] eliminate infinite loop makes sure a possible (never happened, but just in case) infinite loop in the transmission code terminates. ChangeSet@1.1035, 2003-07-29 11:44:58-04:00, achirica@telefonica.net [wireless airo] safer shutdown sequence changes the card shutdown sequence to a safer one ChangeSet@1.1034, 2003-07-29 11:33:59-04:00, achirica@telefonica.net [wireless airo] fix Tx race ChangeSet@1.1019.1.19, 2003-07-29 10:06:40-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/davem/net-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 ChangeSet@1.1019.3.10, 2003-07-28 17:25:49-07:00, olof@austin.ibm.com [RANDOM]: Fix SMP deadlock in __check_and_rekey(). ChangeSet@1.1019.3.9, 2003-07-28 12:25:16-07:00, jones@ingate.com [IGMP]: linux/igmp.h needs asm/byteorder.h ChangeSet@1.1019.1.18, 2003-07-28 11:52:39-03:00, vherva@niksula.hut.fi [PATCH] NMI watchdog documentation This was posted to lkml earlier. No objections from anyone. People have complained about this before, so there is some demand for this. It is only a documentation fix, so it should be safe. Please apply. -- v -- v@iki.fi -- Documentation/nmi-watchdoc.txt doesn't actually tell what options need to be enabled in kernel config in order to use NMI watchdog. I for one found it confusing. ChangeSet@1.1019.1.17, 2003-07-28 11:42:05-03:00, bunk@fs.tum.de [PATCH] MTD Configure.help cleanups The patch below changes the following in Configure.help: - remove entry for the CONFIG_MTD_BOOTLDR_PARTS option that is no longer present - document the following options (all help texts stolen from 2.6): - CONFIG_MTD_CMDLINE_PARTS - CONFIG_MTD_CFI_B8 - CONFIG_MTD_CFI_I8 - CONFIG_MTD_CFI_STAA ChangeSet@1.1019.1.16, 2003-07-28 10:20:13-03:00, neilb@cse.unsw.edu.au [PATCH] md - Resolve problem with refcounting of md arrays. Because an md array may not exist yet when the device it opened, the ->active count, which is incremented on opening if the array exists, is not 100% reliable. This patch changes md to test inode->i_bdev->bd_openers rather than mddev->active to test if an array is still in used before stopping it. (patch against 2.4.22-pre9) ChangeSet@1.1019.1.15, 2003-07-28 10:18:58-03:00, neilb@cse.unsw.edu.au [PATCH] knfsd - BKL is missing in once place in knfsd Knfsd still uses the BKL in many places, and the exp_*lock() routines for locking the export table depend on it. However reading /proc/fs/nfs/exports uses those locks without first grabing the BKL. This patch is reported to fix a hang that was being experienced on an SMP nfs server on which /proc/fs/nfs/exports was being read frequently. (patch against 2.4.22-pre9) ChangeSet@1.1019.1.14, 2003-07-28 10:13:18-03:00, neilb@cse.unsw.edu.au [PATCH] knfsd - Convert error code to nfserror code in nfsd_symlink As Andreas Gruenbacher noticed, without this as wrong error code can leak out. (patch against 2.4.22-pre9) ChangeSet@1.1019.1.13, 2003-07-27 18:25:00-03:00, bcollins@debian.org [PATCH] Interim IEEE-1394 fixes Select fixes from the ieee1394-linux-2.4 repo: IEEE1394 : - Replace tightly scheduled timeout loop with a 50ms timer. - Better return and usage of BUG_ON() in dma handler. - Calculate timeout expiry interval at change instead of at every timeout check. Index: linux-2.4/drivers/ieee1394/ieee1394_core.c =================================================================== ChangeSet@1.1019.1.12, 2003-07-27 18:23:47-03:00, bcollins@debian.org [PATCH] Include param.h for HZ in ieee1394 This got missed with the select patches I just sent. Causes a compiler error in certain cases. ChangeSet@1.1019.3.8, 2003-07-25 14:16:20-07:00, solt@dns.toxicfilms.tv [NETFILTER]: Make REJECT target compliant with RFC 1812. Add support for iptables --reject-with-admin-prohib option of the REJECT target, making it compliant with RFC 1812. ChangeSet@1.1019.3.7, 2003-07-25 14:14:36-07:00, kaber@trash.net [NETFILTER]: Drop reference to conntrack after removing confirmed expectation. ChangeSet@1.1019.3.6, 2003-07-25 14:13:46-07:00, bunk@fs.tum.de [NETFILTER]: Add missing Configure.help entry for ipt_recent. With help from Harald Welte ChangeSet@1.1019.3.5, 2003-07-25 14:12:45-07:00, martin.bene@icomedias.com [NETFILTER]: Add missing include to ip_conntrack_core.h ChangeSet@1.1019.3.4, 2003-07-25 14:11:36-07:00, kaber@trash.net [NETFILTER]: Fix locking of ipt_helper. ChangeSet@1.1019.3.3, 2003-07-25 14:10:48-07:00, laforge@netfilter.org [NETFILTER]: Fix a bug in the IRC DCC command parser of ip_conntrack_irc. ChangeSet@1.1019.3.2, 2003-07-25 14:09:45-07:00, kaber@trash.net [NETFILTER]: Fix issues with iptables REJECT and MIRROR targets wrt. policy routing. ChangeSet@1.1019.3.1, 2003-07-25 14:08:27-07:00, kaber@trash.net [NETFILTER]: Fix problems with iptables MIRROR target - check TTL before rewriting so icmp_send gets clean packet - skb_copy_expand(skb) for tcpdump and asymmetric routing - inline some function - remove unneccessary struct in_device declaration - remove RTO_CONN ChangeSet@1.1019.1.11, 2003-07-25 16:38:23-03:00, pp@netppl.fi [PATCH] Avoid annoying "can't emulate rawmode" messages with logitech cordless mice [ The following text is in the "iso-8859-15" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] Hi Could you please apply this patch to 2.4. It fixes a very annoying problem with logitech wireless keyboard/mouse combinations. Every time a mouse button is pressed the message keyboard.c: can't emulate rawmode for keycode 272 appears on the console, which is somewhat annoying especially if you want to use gpm. For 2.5 this has been fixed by the patch in http://www.cs.helsinki.fi/linux/linux-kernel/2003-06/0754.html, Here's a similar patch for 2.4, which fixes the problem for me. ChangeSet@1.1019.1.10, 2003-07-25 16:25:01-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre9 ChangeSet@1.1019.1.9, 2003-07-25 16:24:21-03:00, agruen@suse.de [PATCH] Fix warning in fs/binfmt_elf.c Hello Geert and Marcelo, it turns oout that the warning is bogus, and trivially fixed. Marcelo, you may want to apply the attached patch. ChangeSet@1.1019.1.8, 2003-07-25 09:53:06-07:00, kaber@trash.net [NET]: Fix signnedness test in socket filter code. ChangeSet@1.1019.1.7, 2003-07-24 14:21:29-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre8 TAG: v2.4.22-pre8 ChangeSet@1.1019.1.6, 2003-07-24 14:18:57-03:00, agruen@suse.de [PATCH] unshare-files fix breaks file locks Hello Marcelo, as already mentioned in private mail, the unshare-files patch in 2.4.22-pre7 breaks POSIX file lock semantics after an execve(2): unshare_files() gives the process a new task_struct->files struct. Existing POSIX file locks continue to refer to the old files struct. POSIX requires that the locks migrate to the exec'ed process. (The LSB checks that.) The fix is to steal the locks from the old file struct after unshare_files(), when it is known that the new, unshared files struct will be used. The fixes introduce changes in behavior for processes that share a common files struct, a case which can be constructed using clone(2). This use of clone is pathological. Before the unshare-files fix, the clones would still share the same files struct. With unshare files, the exec'ed process would lose the locks. We should ensure that the exec'ed process finally holds the locks. Index: linux-2.4.22-pre7.orig/fs/binfmt_elf.c =================================================================== ChangeSet@1.1019.1.5, 2003-07-22 09:07:38-03:00, mark.fasheh@oracle.com [PATCH] Fix deadlock in journal_create Marcelo, I sent this initially against 2.4.21-rc6 and it didn't make it in -- even though it got Stephen's OK. Here's a resend -- I've identified that the bug still exists and 2.4.22-pre7. The patch didn't need to be changed as it still applies cleanly. I ran across a deadlock when trying to do a journal_create at mount time. The problem is that journal_create does a sync_dev which eventually tries to do a get_super which does: down_read(&s->s_umount); The problem arises if I call journal_create from my read_super method in which case get_sb_bdev has already done: down_write(&s->s_umount); from alloc_super. Replacing the sync_dev call with an fsync_no_super seems to have fixed the deadlock. It you want to test this out using ext3 (I have verified it with my own filesystem), simply follow these steps: 1) create an ext2 filesystem on a device 2) mount that new partition and make a "journal file" on it (using dd). get the inode number of that file. 3) unmount it and remount it as an ext3 filesystem using the option: journal=inode_number where inode number is the inode number of the journal file you just created. This will get ext3's read_super method to call ext3_create_journal which will hang during a journal_create. A trivial patch to fix this is attached. --Mark -- Mark Fasheh Software Developer, Oracle Corp mark.fasheh@oracle.com ChangeSet@1.1019.1.4, 2003-07-21 11:28:51-03:00, bhavesh@avaya.com [PATCH] Fix aha152x hangs on pcmcia card eject > Attached is a patch that takes into account comments I have > received in response to the original patch I posted. Please > peruse it, and if it looks okay, please recommend that Marcelo > pick it up for 2.4.22. > > FYI, I have tested this patch on a PIII/440BX with a TI1225 based > PCMCIA card reader and Adaptec SlimSCSI 1460D SCSI adapter card > connected to a Fujitsu SCSI MO drive. ChangeSet@1.1019.1.2, 2003-07-21 09:30:01-03:00, neilb@cse.unsw.edu.au [PATCH] knfsd: Only set ->reuse for tcp sockets, not udp It doesn't really make sense for udp. ----------- Diffstat output ------------ ./net/sunrpc/svcsock.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletion(-) ChangeSet@1.1019.2.1, 2003-07-21 05:02:37-07:00, chas@cmf.nrl.navy.mil [ATM]: Get config/build dependencies correct. ChangeSet@1.1019.1.1, 2003-07-20 13:15:00-04:00, lethal@unusual.internal.linux-sh.org sh64: sh-sci support for SH-5 101/103 Re-add support for the SH-5 SCIF that was previously lost due to some merge breakage with the SH7300. ChangeSet@1.1032, 2003-07-19 22:56:28-04:00, shmulik.hen@intel.com [bonding] fix ifenslave ABI bug ChangeSet@1.1031, 2003-07-19 22:40:52-04:00, achirica@telefonica.net [wireless airo] Update to wireless extensions 16 (new spy API). ChangeSet@1.1030, 2003-07-19 22:39:33-04:00, achirica@telefonica.net [wireless airo] Update to wireless extensions 15 (add monitor mode). ChangeSet@1.1029, 2003-07-19 22:38:00-04:00, achirica@telefonica.net [wireless airo] Return channel in infrastructure mode. ChangeSet@1.1028, 2003-07-19 22:36:36-04:00, achirica@telefonica.net [wireless airo] Checks for small packets before transmitting them. ChangeSet@1.1027, 2003-07-19 22:35:05-04:00, achirica@telefonica.net [wireless airo] Returns proper status in case of transmission error. ChangeSet@1.1026, 2003-07-19 22:33:51-04:00, achirica@telefonica.net [wireless airo] Fix small endianness bug. ChangeSet@1.1025, 2003-07-19 22:32:22-04:00, achirica@telefonica.net [wireless airo] Don't call MIC functions if the card doesn't support them. ChangeSet@1.1024, 2003-07-19 22:26:18-04:00, achirica@telefonica.net [wireless airo] Don't sleep when the stats are requested. ChangeSet@1.1023, 2003-07-19 22:22:44-04:00, achirica@telefonica.net [wireless airo] Make locking "per thread" so it's fully preemptive. ChangeSet@1.1022, 2003-07-19 22:20:27-04:00, achirica@telefonica.net [wireless airo] Update structs with the new fields in latest firmwares. ChangeSet@1.1021, 2003-07-19 22:18:31-04:00, achirica@telefonica.net [wireless airo] Simplify dynamic buffer code in Cisco extensions. ChangeSet@1.1020, 2003-07-19 22:16:06-04:00, achirica@telefonica.net [wireless airo] sync with 2.6 Trivialities: spelling, stack usage, checking return vals, etc. ChangeSet@1.1016.1.20, 2003-07-19 18:00:20-03:00, bcollins@debian.org [PATCH] Update IEEE1394 (r1010) I know I'm a little late in the 2.4.22 release, but these are real fixes, and have been tested for several weeks. I'd hate to see .22 release without them. Thanks! ALL : Consolidate and make consistent, the NODE_BUS_{FMT,ARGS} macros. AMDTP : Fix unchecked copy_from_user. IEEE1394 : - Replace panic() with BUG and return value in dma routine. - Reset reset_retries on success or loop recover. - Add RECV_FLUSH to flush buffers on > 1 irq interval or in buffer-fill mode. - Probe logic-unit-subdirectories as children of the unit-dir. SBP2 : - Fix logical-unit-directory LUN's. - Fix usage of initdata after init. HOST Con : - Change irq output to use __irq_itoa() for sparc. - Do more to notify bus when module is unloaded, to clear IRM, etc. OHCI1394 : Fix endianess of async stream packet headers. ETH1394 : Fix incorrect etherdev flags. ChangeSet@1.1016.1.19, 2003-07-19 12:47:34-04:00, rl@hellgate.ch [PATCH] via-rhine 1.19: One more Rhine-I fix This patch fixes another way the Rhine-I found to break down under load. It should bring Rhine-I behavior on par with the Rhine-II. ChangeSet@1.1016.1.18, 2003-07-19 12:09:16-04:00, scott.feldman@intel.com [e1000] misc cleanup * whitespace cleanup * removal of unused members of netdev priv struct * extendable arrangement of h/w reset logic ChangeSet@1.1016.1.17, 2003-07-19 12:09:03-04:00, scott.feldman@intel.com [e1000] s/int/unsigned int/ for descriptor ring indexes * Perf cleanup: s/int/unsigned int/ for descriptor ring indexes [suggestion by Jeff Garzik]. * Perf cleanup: cache references to ring elements using local pointer ChangeSet@1.1016.1.16, 2003-07-19 12:08:51-04:00, scott.feldman@intel.com [e1000] h/w workaround for mis-fused parts * h/w workaround: several 10's of thousands of 82547 controllers where mis-fused during manufacturing, resulting in PHY Tx amplitude to be too high and out of spec. This workaround detects those parts, and compensates the Tx amplitude by subtracting ~80mV. ChangeSet@1.1016.1.15, 2003-07-19 12:08:39-04:00, scott.feldman@intel.com [e1000] ethtool diag cleanup * Cleanup: ethtool diags: only reset if not if_running. ChangeSet@1.1016.1.14, 2003-07-19 12:08:26-04:00, scott.feldman@intel.com [e1000] alloc_etherdev failure didn't cleanup regions * Bug fix: alloc_etherdev failure didn't cleanup regions in probe. ChangeSet@1.1016.1.13, 2003-07-19 12:08:14-04:00, scott.feldman@intel.com [e1000] missing Tx cleanup opportunities during intr handling * Bug fix: missing Tx cleanup opportunities during interrupt handling. ChangeSet@1.1016.1.12, 2003-07-19 12:08:03-04:00, scott.feldman@intel.com [e1000] fix VLAN support on PPC64 * Bug fix: fix VLAN support on PPC64 [Mark Rakes (mrakes@vivato.net)] ChangeSet@1.1016.1.11, 2003-07-19 12:07:51-04:00, scott.feldman@intel.com [e1000] request_irq() failure resulted in freeing twice * Bug fix: request_irq() failure resulted in freeing resources twice! [Don Fry (brazilnut@us.ibm.com)] ChangeSet@1.1016.1.10, 2003-07-19 12:05:23-04:00, scott.feldman@intel.com [netdrvr ethtool] add ethtool TSO get/set * Add TSO get/set command to ethtool interface. Applies to both 2.4/2.5. Ethtool application patch sent under separate cover. ChangeSet@1.1016.1.9, 2003-07-19 12:05:11-04:00, taowenhwa@intel.com [e100] misc * Allow changing Wake On LAN when EEPROM disabled * Change Log updated * Version changed ChangeSet@1.1016.1.8, 2003-07-19 12:04:59-04:00, taowenhwa@intel.com [e100] cu_start: timeout waiting for cu * Bug fix: 82557 (with National PHY) timeout during init [Adam Kropelin] akropel1@rochester.rr.com ChangeSet@1.1016.1.7, 2003-07-19 12:04:48-04:00, taowenhwa@intel.com [e100] read skb->len after freeing skb * Bug fix: read skb->len after freeing skb [Andrew Morton] akpm@zip.com.au ChangeSet@1.1016.1.6, 2003-07-19 12:04:37-04:00, daniel.ritz@gmx.ch [PATCH] fix ne2k-pci memleak ne2k-pci leaks memory on unload. dev->priv is allocated in ethdev_init(), but never freed. against 2.4-bk, but also applies to 2.5-bk with offset. ChangeSet@1.1016.1.5, 2003-07-19 11:57:58-04:00, gorgo@thunderchild.debian.net [netdrvr wan] note comx maintainer change, by request ChangeSet@1.1016.1.4, 2003-07-18 16:02:04-03:00, marcelo@freak.distro.conectiva buffer.c: remove unused out_putf label TAG: v2.4.22-pre7 ChangeSet@1.1016.1.3, 2003-07-18 15:56:15-03:00, hch@lst.de [PATCH] vmap() backport Hi Marcelo, this patch backports vmap() from 2.5. vmap() is needed for - XFS - proper AGP support on ia64 and ppc (- latest ALSA seems to have some use for in on 2.5) the 2.4 implementation is not as sound as the 2.5 one but far less intrusive. It has been tested for more than a year in the XFS tree and -aa and in -ac for a while. ChangeSet@1.1016.1.2, 2003-07-18 11:23:58-03:00, bcollins@debian.org [PATCH] Fix ALi15x3 DMA on sparc64 (maybe others) Attached is a patch that is known to fix atleast the DMA corruption occuring on sparc64 in 2.4.x kernels for ALi15x3 IDE chipsets. I think Alan hinted that someone also reported it fixed a similar problem on alpha. I've been using this patch for some time now, with great results, and Debian unstable sparc64 users have unknowingly been testing the patch aswell. No reports of problems. Please apply for 2.4.22. Original patch was from Ivan Kokshaysky. ChangeSet@1.1016.1.1, 2003-07-18 09:35:27-03:00, ink@jurassic.park.msu.ru [PATCH] typecast bug in sched.c bites reschedule_idle Posted on l-k by Andy Isaacson , found by Bruce Keller on SMP Alpha. On Thu, Jul 17, 2003 at 04:51:39PM -0500, Andy Isaacson wrote: > Since asm-alpha/timex.h defines cycles_t as unsigned int, this > comparison is always false. Changing it to (cycles_t)-1 fixes the > problem. Other archs which have cycles_t != unsigned long long must be affected as well. Ivan. ChangeSet@1.1017, 2003-07-17 23:30:35-07:00, bcollins@debian.org [SPARC64]: Clear all IRQs at probe time in PCI sabre driver. Fixes operation under newer OBP versions. ChangeSet@1.1016, 2003-07-17 17:19:54-03:00, kraxel@bytesex.org [PATCH] Update tv card i2c helper modules Hi, This patch updates various tv card i2c helper modules. Changes: * misc chip-specific updates / fixes. * various cleanups (drop private MIN/MAX macros, ...). * catch kernel_thread() failures. * i2c changes (2.5.x adaptions). Also uses the new i2c-compat.h header and thus depends on the bttv update sent earlier today. Please apply, Gerd ChangeSet@1.1015, 2003-07-17 17:19:29-03:00, kraxel@bytesex.org [PATCH] bttv documentation update Hi, This patch updates the bttv documentation. Please apply, Gerd ChangeSet@1.1014, 2003-07-17 17:18:24-03:00, kraxel@bytesex.org [PATCH] tuner driver update Hi, This patch updates the tuner module. Changes: * Support for two more tuner chips. * i2c changes (2.5.x adaptions). Also uses the new i2c-compat.h header and thus depends on the bttv update sent earlier today. Please apply, Gerd ChangeSet@1.1013, 2003-07-17 16:43:51-03:00, kraxel@bytesex.org [PATCH] bttv driver update Hi, This patch updates the bttv driver. Changes: * lots of card specific changes and new TV cards. * minor cleanups / fixes here and there (use ARRAY_SIZE macro, ...). * i2c changes (2.5.x adaptions) + new i2c-compat.h header which backports some 2.5.x i2c changes to 2.4.x Please apply, Gerd ChangeSet@1.1003.41.1, 2003-07-17 10:28:43-07:00, trini@kernel.crashing.org Merge ChangeSet@1.1003.1.91, 2003-07-17 14:20:22-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: typo bits ChangeSet@1.1003.1.90, 2003-07-17 08:58:46-03:00, marcelo@freak.distro.conectiva Merge http://linux-sh.bkbits.net/linux-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 ChangeSet@1.1003.38.6, 2003-07-17 08:55:30-03:00, zippel@linux-m68k.org [PATCH] hfs+: Makefile update Hi, On Wed, 16 Jul 2003, David S. Miller wrote: > The file fs/hfsplus/btiter.c got removed but > the reference to btiter.o still remains in > fs/hfsplus/Makefile > > Probably the correct fix is to just delete the > reference from the Makefile. Indeed. Sorry about this, I have a different Makefile so I missed this change. Marcelo, please apply. bye, Roman -- Remove last reference to deleted btiter.c ChangeSet@1.1003.38.4, 2003-07-17 00:21:33-07:00, davem@nuts.ninka.net [SPARC64]: Do not break out of PCI controller probing loop too early. ChangeSet@1.1003.38.3, 2003-07-16 23:31:20-07:00, bcollins@debian.org [SPARC64]: Fix OBP 4.6+ PCI probing, use pcic_present() consistently. ChangeSet@1.1003.39.10, 2003-07-16 23:07:03-07:00, davem@nuts.ninka.net [Bluetooth]: Fix buggy CONFIG_ISDN test in cmtp Config.in ChangeSet@1.1003.39.8, 2003-07-16 21:04:25-07:00, marcel@holtmann.org [Bluetooth] Update the maintainer entries for the Bluetooth subsystem This patch puts a little bit more detailed information about the maintainers of the Bluetooth subsystem into the MAINTAINERS file. ChangeSet@1.1003.39.7, 2003-07-16 21:03:22-07:00, marcel@holtmann.org [Bluetooth] Declare the function l2cap_load() The function l2cap_load() must be declared in the l2cap.h header file. ChangeSet@1.1003.39.6, 2003-07-16 21:02:23-07:00, marcel@holtmann.org [Bluetooth] Handle command complete event for inquiry cancel The command complete event of the inquiry cancel command must clear the HCI_INQUIRY flag and finish the HCI request. ChangeSet@1.1003.39.5, 2003-07-16 21:01:19-07:00, maxk@qualcomm.com [Bluetooth] CMTP protocol depends on ISDN and ISDN CAPI. ChangeSet@1.1003.39.4, 2003-07-16 21:00:04-07:00, marcel@holtmann.org [Bluetooth] Add l2cap_load() function This adds the dummy function l2cap_load() to trigger automatic L2CAP module loading by other modules that use L2CAP sockets but do not use any other symbols from it. ChangeSet@1.1003.39.3, 2003-07-16 20:58:41-07:00, marcel@holtmann.org [Bluetooth] Support for AVM BlueFRITZ! USB This adds a driver for the AVM BlueFRITZ! USB devices. The devices from AVM do not follow the Bluetooth specification part H:2 and thus they need their own driver. ChangeSet@1.1003.39.2, 2003-07-16 20:57:08-07:00, marcel@holtmann.org [Bluetooth] Support for inquiry with unlimited responses This patch handles the special case if the number of responses is set to zero, which means unlimited responses. ChangeSet@1.1003.39.1, 2003-07-16 20:50:42-07:00, marcel@holtmann.org [Bluetooth] Make READ_TRANSMIT_POWER_LEVEL available for normal users This makes the HCI command READ_TRANSMIT_POWER_LEVEL available for normal users. ChangeSet@1.1003.38.2, 2003-07-16 19:57:36-07:00, davem@nuts.ninka.net [SPARC64]: Update defconfig. ChangeSet@1.1003.36.2, 2003-07-16 18:36:36-07:00, davem@nuts.ninka.net [SPARC{,64}]: Add barrier() to cpu_relax() for consistency with 2.5.x ChangeSet@1.1003.1.89, 2003-07-16 16:57:05-04:00, lethal@unusual.internal.linux-sh.org SH update Add support for the SH Mobile Solution Engine and subsequently add support for the SH7300 processor, patches from Takashi YOSHII. ChangeSet@1.1003.1.88, 2003-07-16 15:34:05-03:00, benh@kernel.crashing.org [PATCH] radeonfb: fix artifacts during boot This patch applies on top of what is currently in bk (0.1.8 + my stuffs) and fixes the reported artifact problem due to the fact that I was setting up the initial mode too early. BTW. I'm not sure it's worth keeping all that history of fixes around what do you think Ani ? Ben. ChangeSet@1.1003.1.87, 2003-07-16 15:32:34-03:00, axboe@suse.de [PATCH] more iosched work Hi Marcelo, The low latency elevator in 2.4 is a good thing, however it removes some logic that I always liked and that in the past few days have been shown to be really beneficial to io performance. - With the removal of the two request free lists, reads no longer have reserved slots. This means that a writer can steal all the request slots. We have always given reads a higher priority there, due to the nature of them. It's trivial to fill the queue with writes, not so with reads (doesn't trivially happen in normal use). Patch allows reads to go a bit beyond what the oversized logic protects. - Use BH_Sync to indicate a sync request. A sync request will unplug the queue immediately. We might have to tweak this a bit more (maybe even kill it), the logic is appealing though. There are some debug printk()s in there that should not trigger, when it has seen some testing I will remove them. ===== drivers/block/ll_rw_blk.c 1.47 vs edited ===== ChangeSet@1.1003.37.11, 2003-07-16 14:20:59-03:00, zippel@linux-m68k.org [PATCH] hfs+: export mark_page_accessed hfsplus requires mark_page_accessed (the same export is also in 2.5/6) ChangeSet@1.1003.37.10, 2003-07-16 14:20:25-03:00, zippel@linux-m68k.org [PATCH] hfs+: link hfsplus before hfs hfsplus can be wrapped within a hfs volume, so we have to link hfsplus before hfs. ChangeSet@1.1003.37.9, 2003-07-16 14:19:20-03:00, zippel@linux-m68k.org [PATCH] hfs+: various cleanups various cleanups: - removing unused function - fix up some variable names ChangeSet@1.1003.37.8, 2003-07-16 14:18:33-03:00, zippel@linux-m68k.org [PATCH] hfs+: check size of inode and sb info check that size of inode and sb info doesn't exceed available space ChangeSet@1.1003.37.7, 2003-07-16 14:17:53-03:00, zippel@linux-m68k.org [PATCH] hfs+: fix rename of links use correct cnid to rename links ChangeSet@1.1003.37.6, 2003-07-16 14:16:50-03:00, zippel@linux-m68k.org [PATCH] hfs+: volume/permission fixes Patch provided by Ethan Benson : - fix file type/creator - fix permission handling - improve volume handling ChangeSet@1.1003.37.5, 2003-07-16 14:16:01-03:00, zippel@linux-m68k.org [PATCH] hfs+: remove some smaller files Merge some smaller files into other files ChangeSet@1.1003.37.4, 2003-07-16 14:14:18-03:00, zippel@linux-m68k.org [PATCH] hfs+: update copyright update copyright ChangeSet@1.1003.37.2, 2003-07-16 14:01:05-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre7 ChangeSet@1.1003.37.1, 2003-07-16 11:28:17-04:00, lethal@unusual.internal.linux-sh.org SH Merge Several new things here. Most notably, support for several more boards: - HS7729PCI support (David Woodhouse) - SecureEdge 5410 support (David McCullough) - KeyWest support (David McCullough) other changes include: - kGDB support (Jeremy Siegel and others) - ST40 merge (Alex Bennee) In addition to this, there are random other SH updates all around. ChangeSet@1.1003.1.84, 2003-07-16 09:11:52-03:00, lethal@linux-sh.org [PATCH] sh64: Cayman IRQ handler updates Marcelo, This updates the cayman irq handler. Mostly compile fixes and things of that nature. Please apply. arch/sh64/mach-cayman/irq.c | 53 +++++++++++++++++++++++--------------------- 1 files changed, 28 insertions(+), 25 deletions(-) ChangeSet@1.1003.1.83, 2003-07-16 09:10:55-03:00, lethal@linux-sh.org [PATCH] sh64: defconfig update Marcelo, Nothing exciting here.. this updates the defconfig for sh64 for recent changes. arch/sh64/defconfig | 24 ++++++++++++++++++++++++ 1 files changed, 24 insertions(+) ChangeSet@1.1003.1.82, 2003-07-16 09:05:52-03:00, lethal@linux-sh.org [PATCH] sh64: Fix ATM module build Marcelo, This patch updates arch/sh64/config.in in accordance with the ATM updates on the rest of the architectures. Thus, allowing us to build ATM as a module. arch/sh64/config.in | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.36.1, 2003-07-16 01:13:39-07:00, davem@nuts.ninka.net [SPARC64]: Fix assumptions about data section ordering and objects ending up in .data vs .bss ChangeSet@1.1003.35.1, 2003-07-15 21:49:46-07:00, ja@ssi.bg [IPV4/IPV6]: Fix use-after-free bugs in tunneling drivers. ChangeSet@1.1003.34.1, 2003-07-15 18:09:44+04:00, green@angband.namesys.com Fix link/unlink race. By Chris Mason concurrent link/unlinks can create savelinks for files that still exist, the fix here is to be somewhat smarter about when we change the link count ChangeSet@1.1003.29.44, 2003-07-15 10:17:29-03:00, mikpe@csd.uu.se [PATCH] clean crc temp files in lib/ A 2.4.22-pre kernel build leaves two temp files in lib/, crc32table.h and gen_crc32table, that mrproper doesn't remove. This is ugly. Fixed in the patch below. /Mikael ChangeSet@1.1003.1.78, 2003-07-15 15:05:10+10:00, paulus@samba.org PPC32: Fix IRQ sense and polarity setting on 405 and 440 cpus. ChangeSet@1.1003.29.42, 2003-07-14 19:53:05-03:00, jack@ucw.cz [PATCH] Fix misc quota bugs Hi Marcelo, I went through -pre5 vs my latest patches diff and I found one bug which isn't fixed in 2.4.22-pre5 (in 2.5 it's not needed because of a bit different locking). It's possible deadlock on chgrp(1). The patch is attached. I also got one report about possible appearance of bogus values in old format quota file when new user is being added so the fix of this problem is also attached. TAG: v2.4.22-pre6 ChangeSet@1.1003.29.41, 2003-07-14 19:15:55-03:00, marcelo@freak.distro.conectiva add radeonfb.h ChangeSet@1.1003.33.1, 2003-07-14 13:52:30-07:00, stern@rowland.harvard.edu [PATCH] USB: usb-storage US_FL_FIX_CAPACITY fix Pat LaVarre uncovered a bug where you could throw a well-formed (but arguably meaningless) INQUIRY command at a device via the sg interface and OOPS the usb-storage driver. This prevents that from happening. This has been in 2.5 for a while now. A 2.4 backport is probably a good thing. ChangeSet@1.1003.29.39, 2003-07-14 17:28:30-03:00, marcelo@freak.distro.conectiva Cset exclude: alan@lxorguk.ukuu.org.uk|ChangeSet|20030714133559|12582 ChangeSet@1.1003.29.38, 2003-07-14 14:24:47-03:00, lethal@linux-sh.org [PATCH] sh64: 2-3 pgtable level updates Marcelo, This patch makes page table levels on the sh64 configurable (choice between 2 and 3 levels), and updates the rest of the relevant code for this accordingly. There's also a couple of other minor fixes grouped in with this, as it's not worth breaking these out into other patches. Patches from Richard Curnow, please apply. arch/sh64/config.in | 6 + arch/sh64/mm/fault.c | 34 ------ arch/sh64/mm/tlbmiss.c | 10 - include/asm-sh64/pgalloc-3level.h | 78 --------------- include/asm-sh64/pgalloc.h | 32 ++++-- include/asm-sh64/pgtable-3level.h | 152 ------------------------------ include/asm-sh64/pgtable.h | 192 +++++++++++++++++++++++++++++++++++++- include/asm-sh64/processor.h | 12 +- 8 files changed, 230 insertions(+), 286 deletions(-) ChangeSet@1.1003.32.1, 2003-07-14 14:07:59-03:00, hch@lst.de [PATCH] Fix ext3 quota deadlock I've now extracted the ext3 quota changes from -ac for you. Note that their only relation of them to the new quota code is that the patch is ontop of it and thus takes it into account, the problem is an old one. Here's the patch, probably the last 2.4 patch from me if we really want to play the hot potatoe game. ChangeSet@1.1003.29.36, 2003-07-14 09:50:56-07:00, greg@kroah.com Merge kroah.com:/home/greg/linux/BK/bleed-2.4 into kroah.com:/home/greg/linux/BK/gregkh-2.4 ChangeSet@1.1003.31.3, 2003-07-14 09:48:18-07:00, stern@rowland.harvard.edu [PATCH] USB: Implement US_FL_FIX_CAPACITY for 2.4 On Wed, 9 Jul 2003, Greg KH wrote: > After applying all three patches here (as49, as50, as51) I get the > following build error: > > In file included from usb.c:190: > unusual_devs.h:562: `US_FL_FIX_CAPACITY' undeclared here (not in a function) > unusual_devs.h:562: initializer element is not constant > unusual_devs.h:562: (near initialization for `us_unusual_dev_list[72].flags') > > Care to send me a patch to fix this up? This should fix everything. It defines the new flag and adds the corresponding implementation. ChangeSet@1.1003.29.35, 2003-07-14 13:46:49-03:00, marcelo@freak.distro.conectiva Cset exclude: alan@lxorguk.ukuu.org.uk|ChangeSet|20030714130500|09040 ChangeSet@1.1003.29.34, 2003-07-14 13:42:32-03:00, jbourne@hardrock.org [PATCH] Add missing break in Apollo P4X400 AGP code On Mon, 14 Jul 2003, Alan Cox wrote: > diff -u --new-file --recursive --exclude-from /usr/src/exclude linux.22-pre5/include/linux/agp_backend.h linux.22-pre5-ac1/include/linux/agp_backend.h > --- linux.22-pre5/include/linux/agp_backend.h 2003-07-14 12:27:43.000000000 +0100 > +++ linux.22-pre5-ac1/include/linux/agp_backend.h 2003-07-14 13:05:58.000000000 +0100 > @@ -66,6 +66,7 @@ > VIA_APOLLO_KM266, > VIA_APOLLO_KT400, > VIA_APOLLO_P4M266, > + VIA_APOLLO_P4X400, > SIS_GENERIC, > AMD_GENERIC, > AMD_IRONGATE, Hi, you'll also note a missing break; at the end of the case statement in agpsupport.c as follows ChangeSet@1.1003.29.33, 2003-07-14 11:21:51-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: use the right function in reiserfs (resend #3) #ra1 ChangeSet@1.1003.29.32, 2003-07-14 11:11:17-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre6 ChangeSet@1.1003.29.31, 2003-07-14 10:50:06-03:00, marcelo@freak.distro.conectiva drm_agpsupport.h: Remove ugly comments which used to fix compilation ChangeSet@1.1003.29.30, 2003-07-14 10:46:25-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix agpgart list ChangeSet@1.1003.29.29, 2003-07-14 10:45:50-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: qdio headers for S/390 and S/390x ChangeSet@1.1003.29.28, 2003-07-14 10:44:57-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make rep-nop a barrier as in 2.5 ChangeSet@1.1003.29.27, 2003-07-14 10:35:59-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add quota autoload ChangeSet@1.1003.29.26, 2003-07-14 10:33:20-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update intelfb #ra1 ChangeSet@1.1003.29.25, 2003-07-14 10:29:26-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix vicam with old gcc ChangeSet@1.1003.29.24, 2003-07-14 10:29:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add intellinet to the usb idents ChangeSet@1.1003.29.23, 2003-07-14 10:28:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: re-enable POST on via audio ChangeSet@1.1003.29.22, 2003-07-14 10:28:02-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix i810 and cs46xx crashes ChangeSet@1.1003.29.21, 2003-07-14 10:27:10-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: example ac97 plugin codec ChangeSet@1.1003.29.20, 2003-07-14 10:26:49-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix a race in the plugin api for ac97 ChangeSet@1.1003.29.19, 2003-07-14 10:11:21-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: more warning fixes ChangeSet@1.1003.29.18, 2003-07-14 10:10:56-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: qeth/qdio driver layer ChangeSet@1.1003.29.17, 2003-07-14 10:06:43-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix yenta hang on some laptops ChangeSet@1.1003.29.16, 2003-07-14 10:05:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix sbni driver ChangeSet@1.1003.29.15, 2003-07-14 10:05:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make pcmcica devices report pcmcia bus info in gdrvinfo ChangeSet@1.1003.29.14, 2003-07-14 09:56:28-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: warning fixes ChangeSet@1.1003.29.13, 2003-07-14 09:56:07-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix hpt ide crash, floppy noise ChangeSet@1.1003.29.12, 2003-07-14 09:55:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: run late loaded ide modules ChangeSet@1.1003.29.11, 2003-07-14 09:34:01-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: allow legacy free hw with no smi cmd port ChangeSet@1.1003.29.10, 2003-07-14 09:33:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add qdio options ChangeSet@1.1003.29.9, 2003-07-14 09:33:12-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: clear mp bus array properly ChangeSet@1.1003.29.8, 2003-07-14 09:32:55-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix compile warning ChangeSet@1.1003.29.7, 2003-07-14 09:32:11-03:00, geert@linux-m68k.org [PATCH] kmap_types.h for m68k On Mon, 14 Jul 2003, William Lee Irwin III wrote: > On Sat, 5 Jul 2003, Marcelo Tosatti wrote: > >> o [CRYPTO-2.4]: Add dummy kmap_types.h header for sparc64 > > On Mon, Jul 14, 2003 at 11:55:40AM +0200, Geert Uytterhoeven wrote: > > What are the actual purpose and semantics of the KM_* types? I need to add them > > for m68k to make crypto compile. > > Gr{oetje,eeting}s, > > Geert > > They're per-cpu virtualspace reservations for predetermined purposes. > It's for a variant of kmap() usable under spinlocks and in interrupt > handlers. If you don't have highmem, you just don't care. OK, in that case the one below (copied from PPC/SPARC64) should work fine. Marcelo, please apply. Thx! ChangeSet@1.1003.29.6, 2003-07-14 08:56:59-03:00, bunk@fs.tum.de [PATCH] Configure.help updates from -ac Hi Marcelo, the patch below adds some Configure.help entries that are in pre3-ac1 but are missing in -pre5. Please apply Adrian ChangeSet@1.1003.31.2, 2003-07-13 23:16:40-07:00, greg@kroah.com Merge kroah.com:/home/linux/BK/bleed-2.4 into kroah.com:/home/linux/BK/gregkh-2.4 ChangeSet@1.1003.21.19, 2003-07-13 22:51:44-07:00, ganesh@vxindia.veritas.com [PATCH] USB: more ids for ipaq 3 more device ids added to ipaq.c and ipaq.h Added ids for the Rover P5, Toshiba E310 and E335. Thanks to Matthijs van der Molen, Pavel Stoliarov, Tod B. Schmidt and Matt Hartley. ChangeSet@1.1003.30.8, 2003-07-12 22:47:30-07:00, jmorris@intercode.com.au [NETLINK]: Just drop packets for kernel netlink socket with no data_ready handler. ChangeSet@1.1003.31.1, 2003-07-12 12:04:19-03:00, trond.myklebust@fys.uio.no [PATCH] 2.4.22 NFS O_DIRECT a la mode ->direct_IO2() The following patch reimplements NFS O_DIRECT using a new address_space_operation. Cheers, Trond ChangeSet@1.1003.29.4, 2003-07-12 00:25:45-07:00, davem@nuts.ninka.net [SPARC64]: Fix sys32_rt_sigtimedwait, noticed by Roland McGrath and Jakub Jelinek. ChangeSet@1.1003.30.7, 2003-07-12 00:00:14-07:00, davem@nuts.ninka.net [ATM]: Fix lec.c warning with bridging disabled. ChangeSet@1.1003.30.6, 2003-07-11 23:58:10-07:00, davem@nuts.ninka.net [ATM]: Fix build, missing lec_priv member. ChangeSet@1.1003.30.5, 2003-07-11 23:09:20-07:00, chas@cmf.nrl.navy.mil [ATM]: Eliminate cli, make function names sane in net/atm/lec.c ChangeSet@1.1003.30.4, 2003-07-11 23:06:59-07:00, chas@cmf.nrl.navy.mil [ATM]: Make ATM buildable as a module. ChangeSet@1.1003.29.3, 2003-07-11 22:40:27-07:00, davem@nuts.ninka.net [SPARC64]: Update defconfig. ChangeSet@1.1003.30.3, 2003-07-11 22:39:10-07:00, davem@nuts.ninka.net [NETFILTER]: Fix build warnings in ipv6 modules, thanks Geert. ChangeSet@1.1003.30.2, 2003-07-11 22:14:14-07:00, davem@nuts.ninka.net [SUNRPC]: Fix compiler warning in svcsock.c. ChangeSet@1.1003.29.2, 2003-07-11 21:17:15-07:00, davem@nuts.ninka.net [SPARC64]: Fix warning in drivers/sbus/sbus.c build. ChangeSet@1.1003.1.76, 2003-07-11 16:11:39-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/jgarzik/net-drivers-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: v2.4.22-pre5 ChangeSet@1.1003.28.12, 2003-07-11 15:00:36-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre5 ChangeSet@1.1003.28.11, 2003-07-11 14:15:09-03:00, ak@muc.de [PATCH] Fix compiling on x86-64 Minimal fix to compile x86-64 again on 2.4.22pre4. -Andi ChangeSet@1.1003.28.10, 2003-07-11 12:08:20-03:00, mason@suse.com [PATCH] Fix deadlocks in IO scheduler changes On Thu, 2003-07-10 at 14:01, Marcelo Tosatti wrote: > On Thu, 10 Jul 2003, Stephan von Krawczynski wrote: > > > On Thu, 10 Jul 2003 20:20:02 +0400 > > "Peter Lojkin" wrote: > > > > > Hello, > > > > > > here is exact patch i've used. i made it by cutting pre2-pre3 diff, > > > so apply it o top of 2.4.22-pre3 with -R option to patch... > > > > Hello Peter > > Hello Marcelo > > > > I can confirm that pre3 works when reversing the attached patch. Thanks very > > much, Peter. > > Fine Stephan. Now can youplease get us the task backtraces from sysrq when > the hang happens? > > Andrea, Chris, any idea of why this is happening? My first guess is that blk_oversized_queue is false but there aren't any requests left. That will pretty much spin in __get_request_wait with irqs off, which sounds similar to what he's hitting. I think we need this hunk even if it doesn't fix his problem. Stephan, if this patch doesn't help, could you please boot with nmi_watchdog=1? An earlier email said sysrq wasn't working, so we'll probably need the nmi_watchdog to get a backtrace. ===== drivers/block/ll_rw_blk.c 1.46 vs edited ===== ChangeSet@1.1003.28.9, 2003-07-11 11:47:49-03:00, johnstul@us.ibm.com [PATCH] Cleanup x440 acpi fix On Wed, 2003-07-09 at 20:22, Marcelo Tosatti wrote: > I just applied it John, it will be in bk soon. > > But cant that be done in a cleaner way? Think I've sorted a slightly cleaner way, and bonus points for making it compile with CONFIG_ACPI & !CONFIG_X86_CLUSTERED_APIC (gah! sorry about that). This patch moves the summit detection being done in acpi_madt_oem_check() to detect_clustered_apic(). Also fixes the compile failure when compiling with CONFIG_ACPI and without CONFIG_X86_CLUSTERED_APIC. Applies on top of 2.4.22-bkcurrent. Thanks -john ChangeSet@1.1003.28.8, 2003-07-11 11:44:01-03:00, mike.miller@hp.com [PATCH] cciss: Second part of PCI changes/driver cleanup This is the 6th of 6 updates for cciss. Apply this patch AFTER applying the 5th of 6. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Note the caveat below. Patch name: p2_cciss_2447_pci_read_fixes_for_lx2422p3.patch Changes: 1. Reduces reading directly from PCI config space. Most of the info we require is already in the pci_dev struct (except the command register). Request from Redhat. 2. Miscellaneous code cleanup, remove some debug printk()'s, etc. Caveats: This patch and p1_cciss_2448_pci_enable_fix_for_lx2422p3.patch both touch code very close to each other. Apply p1_cciss_2448_pci_enable_fix_for_lx2422p3.patch first, then apply this patch. Otherwise, they may not patch cleanly. ChangeSet@1.1003.28.7, 2003-07-11 11:36:46-03:00, mike.miller@hp.com [PATCH] cciss: First part of PCI changes/driver cleanup This is the 5th of 6 updates. Apply this patch before applying the 6th patch. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Note the caveat below. Patch name: p1_cciss_2447_pci_enable_fix_for_lx2422p3.patch Changes: 1. Checks the command register BEFORE trying to enable the controller. If the controller was disabled via the ROM Based Setup Utility or the System Configuration Utility the driver may Oops while loading. Bug fix. Caveats: This patch and p2_cciss_2448_pci_read_fixes_for_lx2422p3.patch both touch code very close to each other. Apply this patch first, then apply p2_cciss_2448_pci_read_fixes_lx2422p3.patch. Otherwise, they may not patch cleanly. ChangeSet@1.1003.28.6, 2003-07-11 11:32:01-03:00, mike.miller@hp.com [PATCH] cciss: update version These patches can be installed in any order EXCEPT the final 2 of the 6. They are named p1* & p2* respectively. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Patch name: cciss_2447_version_change_for_lx2422p3.patch Changes: 1. Bumps the version number from 2.4.44 to 2.4.47. 2.4.47 is the preferred number version to use in kernel 2.4.22. ChangeSet@1.1003.28.5, 2003-07-11 11:30:22-03:00, mike.miller@hp.com [PATCH] cciss: Fix potential overrun. These patches can be installed in any order EXCEPT the final 2 of the 6. They are named p1* & p2* respectively. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Patch name: cciss_2447_scsi_overrun_fix_for_lx2422p3.patch Changes: 1. Fixes a potential overrun if a medium changer is attached. Bug fix. ChangeSet@1.1003.28.4, 2003-07-11 11:27:55-03:00, mike.miller@hp.com [PATCH] cciss: PCI BAR fix These patches can be installed in any order EXCEPT the final 2 of the 6. They are name p1* & p2* respectively. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Patch name: cciss_2447_PCI_BAR_fix.patch Changes: 1. Changes the PCI_BASE_ADDRESS_MEM_PREFETCH case in the find_PCI_BAR_index() function to a default case. This should never happen in PCI version 2.2. Bug fix. ChangeSet@1.1003.28.3, 2003-07-11 11:19:00-03:00, mike.miller@hp.com [PATCH] cciss: change names and correct subsystem device ID for U320. These patches can be installed in any order EXCEPT the final 2 of the 6. They are name p1* & p2* respectively. This patch was built & tested using kernel 2.4.21 with the 2.4.22pre3 patch applied. It is intended for inclusion in the 2.4.22 kernel. Patch name: cciss_2447_6400_name_fix_for_lx2422p3.patch Changes: 1. Changes the marketing names (again) for the 6400 Smart Array base controller and the U320 expansion module. Required by marketing. 2. Corrects the subsystem device ID for the U320 Expansion Module. Bug fix, thanks to Janos Farkas for pointing this out. ChangeSet@1.1003.28.2, 2003-07-11 10:21:12-03:00, geert@linux-m68k.org [PATCH] Fix adbhid m68k screwup On Sat, 5 Jul 2003, Marcelo Tosatti wrote: > Summary of changes from v2.4.22-pre2 to v2.4.22-pre3 > ============================================ > Benjamin Herrenschmidt : > o ppc32: Update adbhid driver This change breaks the build for Mac/m68k (cfr. 2.5.x). The patch below cures that, cfr. the similar so-far-unapplied patch for 2.5.x (it's CONFIG_ALL_PPC in 2.4.x and CONFIG_PPC_PMAC in 2.5.x, right)? --snip-- ADB HID: Exclude PowerMac-specific things on classic Macs ChangeSet@1.1003.28.1, 2003-07-11 09:32:51-03:00, benh@kernel.crashing.org [PATCH] radeonfb 0.1.8 + my stuffs Ok, merge was a bit painful (mostly my fault as I did shuffle things around to avoid some of the prototypes and I broke set_var/write_mode in pieces). Hopefully, I haven't screwed it up too much, it still works on my laptop, 2.4.22-pre should be around long enough to fix merge errors if any... Ben. # This is a BitKeeper generated patch for the following project: # Project Name: Linux kernel tree # This patch format is intended for GNU patch command version 2.5 or higher. # This patch includes the following deltas: # ChangeSet 1.1070 -> 1.1072 # include/linux/pci_ids.h 1.65 -> 1.67 # MAINTAINERS 1.99 -> 1.100 # drivers/video/radeonfb.c 1.16 -> 1.18 # drivers/video/radeon.h 1.5 -> 1.7 # # The following is the BitKeeper ChangeSet Log # -------------------------------------------- # 03/07/11 benh@kernel.crashing.org 1.1071 # Ani's 0.1.8 version of radeonfb # -------------------------------------------- # 03/07/11 benh@kernel.crashing.org 1.1072 # Merge back my changes # -------------------------------------------- # ChangeSet@1.1003.26.3, 2003-07-11 01:10:46-07:00, davem@nuts.ninka.net [SPARC64]: Delete bogus icmpv6 filter translation code. A long time ago in a land far far away the icmpv6 filter code bogusly used the "unsigned long" based bitops and thus we had all kinds of endianness problems wrt. 32-bit apps running on 64-bit machines. These days the icmpv6 code explicitly uses u32 fixed types and thus no translations are necessary. ChangeSet@1.1003.27.2, 2003-07-10 19:26:44-07:00, dlstevens@us.ibm.com [IPV4]: Do not sent IGMP leave messages unless IFF_UP. ChangeSet@1.1003.26.2, 2003-07-10 18:10:15-07:00, ebrower@usa.net [SPARC]: Missing part of 2.5.x interrupt decoders fix backport. ChangeSet@1.1003.22.5, 2003-07-10 17:44:18-07:00, davem@nuts.ninka.net [NET]: Ok, sunhme is VLAN challenged after all. ChangeSet@1.1003.22.4, 2003-07-10 17:39:45-07:00, tgraf@suug.ch [NET]: Return EDESTADDRREQ as appropriate in sendmsg implementations. Another 1003.1 fix: [EDESTADDRREQ] The socket is not connection-mode and does not have its peer address set, and no destination address was specified. fixes sendmsg in ipv{4,6}/{raw,udp} ChangeSet@1.1003.1.75, 2003-07-10 19:07:25-04:00, jcchen@icplus.com.tw [netdrvr sundance] increase eeprom read timeout ChangeSet@1.1003.21.18, 2003-07-10 14:34:28-07:00, dhollis@davehollis.com [PATCH] USB: ax8817x.c - add Intellinet USB 2.0 Ethernet device ids Thanks to Zeph for submitting the ids and testing. ChangeSet@1.1003.21.17, 2003-07-10 14:34:23-07:00, david@csse.uwa.edu.au [PATCH] USB: Adding DSS-20 SyncStation to ftdi_sio Patch for 2.4.22 ftdi_sio to add DSS-20 SyncStation to ftdi_sio - the SyncStation is the usb cradle for the Sony Ericsson P800 phone. ChangeSet@1.1003.26.1, 2003-07-10 14:16:47-07:00, davem@nuts.ninka.net [SPARC64]: Port over IPC msg{snd,rcv} compat32 fixes from ia64. ChangeSet@1.1003.25.1, 2003-07-10 12:50:19-07:00, trini@kernel.crashing.org An ethernet driver for the IBM PPC 4xx series of machines. Originally written by Armin Kuster , and then massively re-done and cleaned up by Ben Herrenschmidt and David Gibson . ChangeSet@1.1003.24.1, 2003-07-10 09:09:49-03:00, vandrove@vc.cvut.cz [PATCH] Fix matroxfb on PPC64 Hi Marcelo, can you please apply patch below? It fixes matroxfb's idea of accessing PCI devices on ppc64. Thanks, Petr Vandrovec ChangeSet@1.1003.23.1, 2003-07-10 22:03:17+10:00, paulus@samba.org PPC32: Fix the debug check in kunmap_atomic. ChangeSet@1.1003.22.3, 2003-07-09 22:59:05-07:00, tgraf@suug.ch [NET]: Make {send,recv}msg return EMSGSIZE when msg_iovelen is too big, as per 1003.1 ChangeSet@1.1003.21.16, 2003-07-09 22:18:06-07:00, greg@kroah.com [PATCH] USB: fix up previous sl811 patch. ChangeSet@1.1003.21.15, 2003-07-09 22:17:36-07:00, yinah@couragetech.com.cn [PATCH] USB: patch for sl811 usb host controller driver Ok, I have get rid of the typedefs and make it build on 2.4.21. I'm not try to merge this with the existing sl811 driver, I modifed most of it for support isochronous transfer mode, especially the structrue and the urb schedule part, so it's almost a new one. Some one need this feature can use this one, others can still use the existing. So may be it makes sense to have two different drivers. Note: I wrote it on 2.4.18, at our board which use SA1110, and I test it with usb mouse, usb moving disk and a web camera which use ov511 chipset. ChangeSet@1.1003.21.14, 2003-07-09 21:41:24-07:00, stern@rowland.harvard.edu [PATCH] USB: Updates for unusual_devs.h These changes to unusual_devs.h can be applied to both 2.5 and 2.4. They reflect reports from users in response to the new kernel log messages about unneeded entries in the file. ChangeSet@1.1003.21.13, 2003-07-09 21:41:09-07:00, stern@rowland.harvard.edu [PATCH] USB: Final reconciliation for unusual_devs.h in 2.4 This final patch for 2.4 will make unusual_devs.h identical with the version in 2.5, so from now on new entries can be made to both files in parallel. This simply adds in Andries Brouwer's initialization function and defines it in the source as a dummy do-nothing routine. ChangeSet@1.1003.21.12, 2003-07-09 21:40:56-07:00, stern@rowland.harvard.edu [PATCH] USB: Reconcile unusual_devs.h for 2.4 and 2.5 This patch for 2.4 removes the now-unneeded code for handling START-STOP commands and fixes up the unusual_devs.h file. Instances of the START_STOP flag are removed as are unnecessary entries. The entries are sorted into numerical order and the US_SC_DEVICE and US_PR_DEVICE flags have been added in. Also a few changes in whitespace were made. When this patch and as48 are applied, unusual_devs.h should be identical between 2.4 and 2.5, except for one line. An entry created by Andries Brouwer refers in 2.5 to an initialization function that is #ifdef'ed out in 2.4. I'm asking him how best to fix this up. ChangeSet@1.1003.21.11, 2003-07-09 21:34:21-07:00, david-b@pacbell.net [PATCH] USB: usbnet updates This patch pulls some more 2.5 changes into usbnet: - Adds a "read the endpoint descriptors" mode. * Finally resolves the ARM-Linux "usb-eth" mess, where sa1100 and pxa2xx drivers are using different endpoints though the drivers re-use product/vendor codes. * Means that high speed packet sizes (512 bytes/packet) can work automatically, on dual speed devices (full/high). - Now recognizes the BLOB boot loader. - Resolves that DROP-vs-SUCCESS issue noted by Martin Diehl (patch just sent in for 2.5). ChangeSet@1.1003.21.10, 2003-07-09 21:34:08-07:00, david-b@pacbell.net [PATCH] USB: usb_string(), don't use bogus ids This is the 2.4 version of that patch I just sent for 2.5, making sure usbcore doesn't use un-initialized memory for its language IDs. ChangeSet@1.1003.21.9, 2003-07-09 21:33:50-07:00, greg@kroah.com [PATCH] USB: fix up my USB Bluetooth entry to help prevent confusion in the future. ChangeSet@1.1003.22.2, 2003-07-09 21:02:13-07:00, davem@nuts.ninka.net [SUNHME]: Set RXMAX/TXMAX large enough to handle VLAN frames. ChangeSet@1.1003.22.1, 2003-07-09 20:28:56-07:00, chas@cmf.nrl.navy.mil [ATM]: Add reference counting to atm_dev. ChangeSet@1.1003.21.8, 2003-07-10 00:17:47-03:00, johnstul@us.ibm.com [PATCH] Fix boot crash of x440's in full acpi mode Marcelo, Andrew, All, Due to the new ACPI code, when booting in full ACPI mode, we do not go through the mps tables, thus we do not execute the summit detection code required for booting an x440. This patch insures that when booting in full ACPI mode we check to see if we're running on a summit based system and enable clustered apic mode. Without this patch the x440s hang while booting in full ACPI mode. Thanks to James Cleverdon for the original version of this patch. Please apply, thanks -john ChangeSet@1.1003.21.7, 2003-07-10 00:13:56-03:00, jamagallon@able.es [PATCH] hfsplus: group Apple FS's and help text > > Agreed with maintainer some time ago... > I should learn to write with hands instead of feet... ChangeSet@1.1003.21.6, 2003-07-09 17:40:31-03:00, marcelo@freak.distro.conectiva Comment out VIA_APOLLO_P4X400 handling in drm_agpsupport.h: Alan will fix that up later. TAG: v2.4.22-pre4 ChangeSet@1.1003.21.5, 2003-07-09 15:51:49-03:00, marcelo@freak.distro.conectiva Add missing fs/quota_v2.c file ChangeSet@1.1003.21.4, 2003-07-09 15:50:09-03:00, rddunlap@osdl.org [PATCH] make profile= doc. clearer Please apply to 2.4.22-pre. patch_name: kernproftext_24.patch patch_version: 2003-07-09.11:14:56 author: Randy.Dunlap description: make profile= clearer product: Linux product_versions: 2.4.22-pre diffstat: = Documentation/kernel-parameters.txt | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.21.3, 2003-07-09 15:42:40-03:00, marcelo@freak.distro.conectiva Cset exclude: remove NFS direct IO patches Cset exclude: trond.myklebust@fys.uio.no|ChangeSet|20030706143259|16957 ChangeSet@1.1003.21.2, 2003-07-09 15:42:33-03:00, marcelo@freak.distro.conectiva Cset exclude: Remove NFS direct IO patches Cset exclude: trond.myklebust@fys.uio.no|ChangeSet|20030708095239|55752 ChangeSet@1.1003.21.1, 2003-07-09 15:05:39-03:00, marcelo@freak.distro.conectiva CRIS architecture update ChangeSet@1.1003.1.72, 2003-07-09 17:28:22+01:00, dwmw2@infradead.org Backport vsprintf/scanf fixes from 2.5.74. From akpm@digeo.com: Support %z size_t qualifier From louis.zhuang@intel.com: missed field_width reset in vsscanf() From pazke@orbita1.ru: Make char arrays static not on-stack From rddunlap@osdl.org: Fix sscanf("-1", "%d", &i); ChangeSet@1.1003.1.71, 2003-07-09 10:20:54-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/davem/sparc-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1003.20.3, 2003-07-09 22:14:08+10:00, paulus@samba.org PPC32: Make __kernel_ino_t be unsigned long like on other architectures. ChangeSet@1.1003.20.2, 2003-07-09 22:12:18+10:00, paulus@samba.org PPC32: Define screen_info if CONFIG_FB is set for the sake of vesafb. ChangeSet@1.1003.20.1, 2003-07-09 22:09:23+10:00, paulus@samba.org PPC32: Minor boot wrapper cleanups ChangeSet@1.1003.19.3, 2003-07-08 23:58:38-07:00, davem@nuts.ninka.net [SPARC]: SEMTIMEDOP for both Sparc ports. ChangeSet@1.1003.19.2, 2003-07-08 22:45:28-07:00, davem@nuts.ninka.net [FS]: Provide unshare_files() declaration and export to modules. ChangeSet@1.1003.19.1, 2003-07-08 22:03:06-07:00, davem@nuts.ninka.net [SPARC64]: sys_sparc32.c needs linux/quotacompat.h ChangeSet@1.1003.18.3, 2003-07-08 18:58:10-07:00, lode_leroy@hotmail.com [IPV4] display bootserver in /proc/net/pnp. With modification from niv@us.ibm.com to only print bootserver if there was one. ChangeSet@1.1003.18.2, 2003-07-08 18:53:23-07:00, dtor_core@ameritech.net [NET] Attach inner qdiscs to TBF. ChangeSet@1.1003.1.68, 2003-07-08 18:33:06-03:00, m.c.p@wolk-project.de [PATCH] [RESEND 5th] Fix oom killer braindamage [ The following text is in the "iso-8859-15" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] Hi Marcelo, sorry, I've sent this patch 4 times now and no one but Alan (yesterday) told me, it does not apply to mainline. This is due to stuff I had in WOLK and haven't removed it in the previous attached patches :-( This one now works/applies for real. attached patch fixes the oom killer braindamage where it tries to kill processes again and again and again w/o any ending or successfull killing of the selected processes in an OOM case. The attached, very simple but effective, patch fixes it. All the kudos go to Rik van Riel. Patch tested and works, and also for a long time in my tree (and maybe also others, and for sure also in 2.5 for some time.) ChangeSet@1.1003.1.67, 2003-07-08 18:19:46-03:00, marcelo@freak.distro.conectiva Fixes ext3 quota/truncate oops TAG: MailDone ChangeSet@1.1003.1.66, 2003-07-08 18:05:16-03:00, hch@lst.de [PATCH] fix Q_SYNC for dev == 0 Need to sync all quota-enabled filesystems in that case. Backported from Jan's 2.5 work. ChangeSet@1.1003.1.65, 2003-07-08 17:54:24-03:00, lethal@linux-sh.org [PATCH] sh64: IDE support This patch implements rudimentary IDE support for the sh64. Patch by Richard Curnow, please apply. arch/sh64/config.in | 20 ++++++------ arch/sh64/kernel/sh_ksyms.c | 3 + include/asm-sh64/hdreg.h | 13 ++++++++ include/asm-sh64/ide.h | 65 +++++++++++++++++++++++++++++++++++++++++ include/asm-sh64/mc146818rtc.h | 7 ++++ 5 files changed, 98 insertions(+), 10 deletions(-) ChangeSet@1.1003.1.64, 2003-07-08 17:54:03-03:00, lethal@linux-sh.org [PATCH] sh64: Fix privileged insn handling This patch fixes up a case where an attempt is made to run a privileged instruction from user-mode. Patch from Richard Curnow. Please apply. arch/sh64/kernel/traps.c | 95 +++++++++++++++++++++++++++++++++-------------- 1 files changed, 68 insertions(+), 27 deletions(-) ChangeSet@1.1003.1.63, 2003-07-08 17:50:09-03:00, lethal@linux-sh.org [PATCH] sh64: SH-5 PCI updates This updates some of the SH-5 PCI code. Notably, this fixes up some of the IRQ routing semantics, which allow cards to now properly use INTB/INTC/INTD instead of just INTA. In addition, this also fixes up the PCI arbiter to use round-robin arbitration as opposed to fixed-priority, which fixes numerous PCI ARB interrupts that were causing deadlock at seemingly random intervals before. Both of these patches are from Richard Curnow, please apply. arch/sh64/kernel/pci_sh5.c | 41 +++++++++++++++++++++++++++-------------- 1 files changed, 27 insertions(+), 14 deletions(-) ChangeSet@1.1003.1.62, 2003-07-08 17:49:34-03:00, lethal@linux-sh.org [PATCH] sh64: Fixup Cayman IRQ reporting The Cayman IRQ code currently drops the first interrupt from PCI bus 2 in the /proc/interrupts output. This patch from Richard Curnow fixes that. Please apply. arch/sh64/mach-cayman/irq.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.1.61, 2003-07-08 17:49:15-03:00, lethal@linux-sh.org [PATCH] sh64: Add FIOQSIZE definition Marcelo, Can you try this to see if it works? I suspect that the missing changeset you needed was the 1.1060 merge point, as 1.1059 is already common between our trees. If this still produces the same error, please just apply this as a normal patch. You can import this changeset into BK by piping this whole message to: '| bk receive [path to repository]' or apply the patch as usual. =================================================================== ChangeSet@1.1061, 2003-07-08 00:40:12-04:00, lethal@unusual.internal.linux-sh.org include/asm-sh64/ioctls.h: Add FIOQSIZE definition, so new quota code compiles. ChangeSet@1.1060, 2003-07-08 00:10:52-04:00, lethal@unusual.internal.linux-sh.org Merge unusual.internal.linux-sh.org:/home/lethal/devel/bk/vanilla-2.4 into unusual.internal.linux-sh.org:/home/lethal/devel/bk/shmedia-2.4 ioctls.h | 1 + 1 files changed, 1 insertion(+) ChangeSet@1.1003.1.60, 2003-07-08 17:07:20-03:00, hch@infradead.org [PATCH] quota patch breaks kernel build On Tue, Jul 08, 2003 at 06:16:08AM -0300, Marcelo Tosatti wrote: > > Mind sending a patch Christoph? Oh, I meant to attach it, sorry. ===== fs/dquot.c 1.18 vs edited ===== ChangeSet@1.1003.17.1, 2003-07-08 16:32:05-03:00, marcelo@freak.distro.conectiva Cset exclude: hannal@us.ibm.com|ChangeSet|20030707180059|08076 TAG: MailDone ChangeSet@1.1003.1.58, 2003-07-08 11:59:02+02:00, trond.myklebust@fys.uio.no Back out some congestion control changes that were causing trouble among other things for the "soft" mount option. ChangeSet@1.1003.1.57, 2003-07-08 11:52:39+02:00, trond.myklebust@fys.uio.no Add missing NFS O_DIRECT file fs/nfs/direct.c ChangeSet@1.1003.1.56, 2003-07-07 17:01:15-03:00, hch@lst.de [PATCH] new quota code Okay, here's the quota patch. Basically all changes are from Jan Kara and I backport them from 2.5. The 32bit quota code has been shipped by the commercial vendors ever since they used Linux 2.4 and this particular codebase with backwards compatiblity support is around in the 2.5, the XFS tree, -ac and -aa for a long time. The only change over that version is that support for the old 16bit quota format and the old quotactl ABI is enabled unconditionally, i.e. there's no way to render your system unusable by wrong make config choices [1]. [1] This also mean completely dropping support for the interim ABI used in the early 32bit quota patches as it's mutally incompatible to the old ABI. But we never ever shipped that in any mainline kernels so there's no problem. TAG: MailDone ChangeSet@1.1003.1.55, 2003-07-07 16:49:40-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre4 TAG: MailDone ChangeSet@1.1003.1.54, 2003-07-07 16:47:08-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix up z85230 queue wake logic ChangeSet@1.1003.1.53, 2003-07-07 16:46:38-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: S/390 CLAW bits ChangeSet@1.1003.1.52, 2003-07-07 16:45:59-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: kernel/fork helper for exec security fix ChangeSet@1.1003.1.51, 2003-07-07 16:43:57-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add the new agp modes to the headers ChangeSet@1.1003.1.50, 2003-07-07 16:43:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix definition of boot_DT ChangeSet@1.1003.1.49, 2003-07-07 16:43:17-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix inverted dnotify ChangeSet@1.1003.1.48, 2003-07-07 16:41:17-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: exec part of security fix ChangeSet@1.1003.1.47, 2003-07-07 16:40:57-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: re-fix printk level for buffer cachehash ChangeSet@1.1003.1.46, 2003-07-07 16:40:33-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: resend - fix security bits in binfmt_exec/som ChangeSet@1.1003.1.45, 2003-07-07 16:37:54-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add code for missing c7000 driver ChangeSet@1.1003.1.44, 2003-07-07 16:37:19-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: move sdla to mod_timer ChangeSet@1.1003.1.43, 2003-07-07 16:37:04-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix the eexpress ChangeSet@1.1003.1.42, 2003-07-07 16:36:48-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update mpt fusion driver ChangeSet@1.1003.1.41, 2003-07-07 16:36:25-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: ebda check in ibm hotplug is insufficient ChangeSet@1.1003.1.40, 2003-07-07 16:35:31-03:00, marcelo@freak.distro.conectiva Remove bogus diff from drivers/char/Config.in ChangeSet@1.1003.1.39, 2003-07-07 16:33:03-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: AGP update - new intel, add nvidia ChangeSet@1.1003.1.38, 2003-07-07 16:29:20-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: parisc sync up (resend of resend of resend ... 8)) Hopefully conectiva's mail is working again ChangeSet@1.1003.1.37, 2003-07-07 16:22:24-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: config for new Nvidia AGP ChangeSet@1.1003.1.36, 2003-07-07 16:22:04-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: Make ACPI work on lots more boxes Older AML compilers got some range rules wrong as well as generating load localn, localn (not with 0). The Intel acpi blows up on these (technically invalid) constructs. This patch allows systems with the ranges wrong to run fine - notably most toshiba's ChangeSet@1.1003.1.35, 2003-07-07 16:04:31-03:00, marcelo@freak.distro.conectiva Cset exclude: rusty@rustcorp.com.au|ChangeSet|20030707182325|08049 ChangeSet@1.1003.1.34, 2003-07-07 16:03:07-03:00, stelian@popies.net [PATCH] Export 'acpi_disabled' symbol to modules... Hi, 'acpi_disabled' is not exported by the current 2.4-pre kernel, but is used by at least the sonypi module. I have submitted a similar patch for 2.5 some weeks ago (and it got applied) but for some reason the 2.4 ACPI branch don't have it. Andy, Marcelo, please apply. (Alan, a similar change is needed for -ac, but the current patch will not apply on top of the ac tree. Do you want me to send you a patch correcting this or you'll do the change by hand ?) Thanks, Stelian. ===== arch/i386/kernel/setup.c 1.68 vs edited ===== ChangeSet@1.1003.1.33, 2003-07-07 16:00:07-03:00, rusty@rustcorp.com.au [PATCH] esssolo1.c doesn't free resources correctly (OK from maintainer Thomas Sailer ) From: Marcus Alanen personally I don't really care that much if these get added or not, but i happened to recognize it wasn't in -- Marcus Alanen * Embedded Systems Laboratory * http://www.eslab.cs.abo.fi/ marcus.alanen@abo.fi ---------- Forwarded message ---------- Date: Mon, 1 Jul 2002 20:47:56 +0300 (EEST) From: Marcus Alanen To: Alan Cox , t.sailer@alumni.ethz.ch Cc: linux-kernel@vger.kernel.org Subject: [patch, 2.4] esssolo1.c doesn't free resources correctly In drivers/sound/esssolo1.c, a failed solo1_probe() frees resources in the wrong order (Look at the allocation order). Also, the wrong unregister_sound_* routines are used. I don't know if register_sound_special should use unit 10 ("dmfm") instead of 15 ("unknown"). Compiles, not tested. Marcus ===== drivers/sound/esssolo1.c 1.11 vs edited ===== ChangeSet@1.1003.1.32, 2003-07-07 15:29:39-03:00, rusty@rustcorp.com.au [PATCH] 2.4 patch for more debug safety (Included in 2.5) From: Rusty Russell I don't think it's misused anywhere, but it's better to be safe. Pointed out by Joern Engel. Rusty. ChangeSet@1.1003.1.31, 2003-07-07 15:29:03-03:00, rusty@rustcorp.com.au [PATCH] [2.5 patch] two small MTD fixes (Included in 2.5) (OK from maintainer David Woodhouse ) From: Adrian Bunk The patch below contains the following fixes for MTD in 2.5.73: - postfix a constat in mptbase.c with ULL - add an extern declaration for a function that is later called in gen_probe.c Please apply Adrian ChangeSet@1.1003.1.30, 2003-07-07 15:27:57-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_filesystems_befs.txt (Included in 2.5) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/filesystems/befs.txt more readible by fixing the linewrap. ChangeSet@1.1003.1.29, 2003-07-07 15:25:47-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_arm_SA1100_CERF (Included in 2.5) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/arm/SA1100/CERF more readible by fixing the linewrap. ChangeSet@1.1003.1.28, 2003-07-07 15:23:25-03:00, rusty@rustcorp.com.au [PATCH] fix undefined reference for sis drm. [ CONFIG_DRM_SIS needs CONFIG_FB_SIS to compile, of course. --RR ] (OK from maintainer Rik Faith ) From: Geoffrey Lee Yep. Take one was pretty silly. Here's a better one. -- G. -- char p[] = "xebx1fx5ex89x76x08x31xc0x88x46x07x89x46x0cxb0x0b" "x89xf3x8dx4ex08x8dx56x0cxcdx80x31xdbx89xd8x40xcd" "x80xe8xdcxffxffxff/bin/sh"; ChangeSet@1.1003.1.27, 2003-07-07 15:22:49-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_ia64_efirtc.txt (Included in 2.5) (OK from maintainer David Mosberger-Tang) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/ia64/efirtc.txt more readible by fixing the linewrap. ChangeSet@1.1003.1.26, 2003-07-07 15:21:08-03:00, rusty@rustcorp.com.au [PATCH] Re: setrlimit incorrectly allows hard limits to exceed [ This adds a sanity check for setting rlimits: you can't set rlim_cur past rlim_max. --RR ] (Included in 2.5) From: Kingsley Cheung Oops, should be after the copy :-( ChangeSet@1.1003.1.25, 2003-07-07 15:19:42-03:00, rusty@rustcorp.com.au [PATCH] Decision PCCOM4_PCCOM8 serial support for 2.4.19 (OK from maintainer Luca Montecchiani ) From: Luca Montecchiani This patch add support for the Decision PCCOM 4 and 8 ports pci cards. The patch was made and tested on a 2.4.19 kernel with the standard serial driver. thanks, lucadiff -ur linux/drivers/char/serial.c.orig linux/drivers/char/serial.c ChangeSet@1.1003.1.24, 2003-07-07 15:01:43-03:00, marcelo@freak.distro.conectiva Merge bk://ppc.bkbits.net/for-marcelo-ppc into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1003.16.11, 2003-07-07 15:00:59-03:00, hannal@us.ibm.com [PATCH] Fastwalk: reduce cacheline bouncing of d_count Marcelo, Here is the dcache fastwalk patch to reduce cacheline bouncing of the d_count reference counter while traversing paths that are already in the dcache. The testing I have done this year shows a small increase in throughput on a 2-way system. I expect the benefit to be greatest on slower SMP systems which could suffer from cacheline problems. Please consider it for inclusion. Thanks. Hanna ChangeSet@1.1003.16.10, 2003-07-07 14:23:26-03:00, rusty@rustcorp.com.au [PATCH] fs_bfs_dir.c unused variables [ Unused (but not uninitialized) variables. Also unused in 2.4. --RR ] (Included in 2.5) (OK from maintainer Tigran Aivazian ) From: Rusty Russell Hi Will, Want to take these? Rusty. TAG: MailDone ChangeSet@1.1003.16.9, 2003-07-07 14:20:37-03:00, rusty@rustcorp.com.au [PATCH] fix sound doc typos [ Just typo/spelling fixes. --RR ] From: Christoph Hellwig This patch is from debian's kernel tree. I still don't get why no maintainer of vendor trees is able to properly submit such fixes upstream. ChangeSet@1.1003.16.8, 2003-07-07 14:19:59-03:00, rusty@rustcorp.com.au [PATCH] trivial patch From: T.Lukaszewski@elka.pw.edu.pl I don't know if it is a good address for that patch, but i hope that it won't be lost. This patch adds KERN_INFO in two printk() calls (in i2c_core.c and i2c_algo_bit.c). Tomasz Lukaszewski T.Lukaszewski@elka.pw.edu.pl ChangeSet@1.1003.16.7, 2003-07-07 14:19:37-03:00, rusty@rustcorp.com.au [PATCH] 2.4 drivers_char_random.c fix sample shellscripts (Included in 2.5) From: dave@qix.net [tyop in first sumbissoin, fixed here] This fixes the sample shellscripts given in the comments of drivers/char/random.c. The scripts save and restore random seeds for /dev/random across reboots. ChangeSet@1.1003.16.6, 2003-07-07 14:19:00-03:00, rusty@rustcorp.com.au [PATCH] unreachable code in drivers_media_video_cpia_pp.c (Included in 2.5) From: Scott Russell - Rearranged unreachable code reported via kbugs.org ChangeSet@1.1003.16.5, 2003-07-07 14:18:13-03:00, rusty@rustcorp.com.au [PATCH] Configure.help Polish translation location update [ From the man himself... ] From: Dominik Mierzejewski Will do, thanks. Attached patch formatted according to http://www.kernel.org/pub/linux/kernel/people/rusty/trivial/ -- "The third principle of sentient life is its capacity for self-sacrifice. For a cause, a loved one, for a friend." -- Delenn to Draal in Babylon 5:"A Voice in the Wilderness #2" ChangeSet@1.1003.16.4, 2003-07-07 11:41:39-03:00, benh@kernel.crashing.org [PATCH] ppc32: Fix races in low level adb drivers Hi Marcelo ! You seem to have missed this patch in the bunch I sent you earlier, it fixes some subtle SMP races in the low level ADB drivers and cleans up via-cuda a bit. Please apply. Ben. ChangeSet@1.1003.16.3, 2003-07-06 22:02:46-03:00, mikpe@csd.uu.se [PATCH] i386 cpufeature.h cleanup + comment The 2.4.22-pre3 changelog includes: >Alan Cox : ... > o add the extra cpu bit test flags This added a second cpu_has() macro to cpufeature.h, which the patch below cleans up. Note that FEATURE_EST currently mustn't be used: it resides in the fifth word ([4]) but only four words are allocated (NCAPINTS). If you bump NCAPINTS, you must also adjust the hard-coded struct offsets in arch/i386/kernel/head.S or Very Bad Things happen. /Mikael TAG: MailDone ChangeSet@1.1003.16.2, 2003-07-06 11:32:59-03:00, trond.myklebust@fys.uio.no [PATCH] Experimental O_DIRECT support for NFS Hi Marcelo, The following patch from Chuck Lever adds experimental support for O_DIRECT file access under NFS. It is mainly meant for use by database programs such as Oracle, that need to manage their own caches rather than relying on the page cache. Cheers, Trond TAG: MailDone ChangeSet@1.1003.16.1, 2003-07-06 10:53:53-03:00, willy@debian.org [PATCH] pci_name() Hi Marcelo. This patch introduces the pci_name() compatibility function to ease backporting drivers from 2.5. ChangeSet@1.1003.15.1, 2003-07-06 13:00:58+10:00, paulus@samba.org PPC32: Minor updates to comments and processor register definitions. ChangeSet@1.1003.1.21, 2003-07-05 20:59:23-03:00, mason@suse.com [PATCH] Fix potential IO hangs and increase interactiveness during heavy IO io-stalls-10: ===== drivers/block/ll_rw_blk.c 1.45 vs edited ===== TAG: MailDone TAG: v2.4.22-pre3 ChangeSet@1.1003.9.71, 2003-07-05 19:58:37-03:00, marcelo@freak.distro.conectiva Changed EXTRAVERSION to -pre3 TAG: MailDone ChangeSet@1.1003.9.70, 2003-07-05 19:57:01-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make i810 audio compile This one still oopses when you unload it but thats a known bug with a known fix I need to merge later ChangeSet@1.1003.1.18, 2003-07-05 16:55:15-04:00, shmulik.hen@intel.com [netdrvr bonding] Fix change active for ALB/TLB The following patch fixes bonding's change active interface operation for ALB/TLB modes. It used to incorrectly set the old active interface's state to BACKUP (which is required only for active-backup mode) and would cause that slave not to take part in load sharing. ChangeSet@1.1003.1.17, 2003-07-05 16:55:01-04:00, rl@hellgate.ch [netdrvr via-rhine] via-rhine 1.18-rc1: Fix Rhine-I regression This patch addresses a minor regression reported by Rhine-I users (leading to occasional Tx timeouts). I also merged some cosmetic changes (including spelling fix from 2.5), but with the recent interrupt handling changes 2.4 and 2.5 versions of via-rhine seem to have diverged for good. ChangeSet@1.1003.1.16, 2003-07-05 15:10:42-04:00, jgarzik@redhat.com [netdrvr 8139too] fix debug printk printk args were accidentally reversed. ChangeSet@1.1003.9.69, 2003-07-05 13:43:21-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: (resend) collected semaphore fixes and semtimedop TAG: MailDone ChangeSet@1.1003.9.68, 2003-07-05 13:43:05-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: (new) Turn on the IDE modular stuff in the Makefile This isnt perfect but it is a start ChangeSet@1.1003.9.67, 2003-07-05 13:42:35-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: CMD640 update This patch seems to have vanished in the post first time around. It cleans up the CMD640 logic a little and sorts out the registration bits Its a requirement to making modular IDE work again ChangeSet@1.1003.9.66, 2003-07-05 13:03:58-03:00, marcelo@freak.distro.conectiva Added missing BROKEN_PNP_BIOS declaration ChangeSet@1.1003.9.65, 2003-07-05 12:56:10-03:00, hugh@veritas.com [PATCH] shmem loopable tmpfs [again] Hi Marcelo, Many thanks for putting in my tmpfs patches - but one I sent (originally 10/16) has gone missing. I think the current -bk version is safe, but peculiar (11/16 went in, but only fixes a leftover from 10/16; 12/16 went in and includes documentation of changes in 10/16). I imagine BK won't allow you to retrospectively insert the missing patch? So here is a replacement patch against -bk to go on top to set it right. Thanks, Hugh shmem_readpage, shmem_prepare_write, shmem_commit_write added so that tmpfs files can be used as backing for loop devices, and for sendfile. Most change is in shmem_getpage: which normally works just as before, but now has to distinguish between the filepage passed in (NULL unless readpage or prepare_write) and the swappage which might already exist. Exceptionally needs to copy swappage to filepage: copy_highpage macro needs copy_page macro, already defined for all arches except sparc64. ChangeSet@1.1003.9.64, 2003-07-04 20:50:15-03:00, hugh@veritas.com [PATCH] mremap VM_LOCKED move_vma mremap's move_vma VM_LOCKED case was wrong. Although that do_munmap will never free the new_vma structure, it may split the area into two extents, and [new_vma->vm_start, new_vma->vm_end) afterwards may not be the new range to make present. We need [new_addr, new_addr+new_len) to be locked down; but move_page_tables already transferred the locked pages [new_addr, new_addr+old_len), and they're either held in a VM_LOCKED vma throughout, or temporarily in no vma: in neither case can be swapped out, so no need to run over that range again. And shorten a line to 80 columns (akpm). TAG: MailDone ChangeSet@1.1003.9.63, 2003-07-04 20:47:59-03:00, hugh@veritas.com [PATCH] shmem_file_write precheck_file_write 2.4.22-pre1 introduced precheck_file_write: just what shmem_file_write has been wanting to avoid repeating all those tedious checks. Except we'd prefer to avoid mark_inode_dirty_sync cluttering the sync list with tmpfs, so separate out remove_suid etc. like 2.5's generic_write_checks. ChangeSet@1.1003.9.62, 2003-07-04 20:46:44-03:00, hugh@veritas.com [PATCH] madvise_willneed check readpage Andrea Arcangeli's fix to Andrew Morton's realization of my prediction: madvise_willneed must not page_cache_read if there's no a_ops->readpage. Though this won't affect the 2.4.22 tmpfs now it's been given a readpage. ChangeSet@1.1003.9.61, 2003-07-04 20:45:58-03:00, hugh@veritas.com [PATCH] loop file use highmem When loop restricts underlying file's allocation mask to avoid deadlock, it unintentionally masks out its highmem capability, making failures at the low level much more likely. Though this is more likely to affect loop on other filesystems than tmpfs - which uses read_swap_cache_async (ignoring mapping->gfp_mask) once initial blank page has been allocated. ChangeSet@1.1003.9.60, 2003-07-04 20:44:25-03:00, hugh@veritas.com [PATCH] shmem_removepage replace recalc_inode 2.4.21-pre3 added the removepage callback to address_space_operations, but without any uses. Now add shmem_removepage to do the accounting when vmscan frees an undirtied hole page, and remove shmem_recalc_inode (which only noticed when the file in question was touched again). This has long been in -ac, though I'm reluctant to push removepage into 2.5. ChangeSet@1.1003.9.59, 2003-07-04 20:43:44-03:00, hugh@veritas.com [PATCH] shmem mount percentile size >From CaT : What this patch does is allow you to specify the max amount of memory tmpfs can use as a percentage of available real ram. This (in my eyes) is useful so that you do not have to remember to change the setting if you want something other then 50% and some of your ram goes. Hugh redid the arithmetic to not overflow at 4GB; the particular order of lines helps RH2.96-110 not to get confused in the do_div. He also updated the tmpfs Documentation for this and for loopable tmpfs. There's an argument that the percentage should be of ram+swap, that's what Christoph originally intended. But we set the default at 50% of ram only, and 2.5 is already following that precedent with this option. ChangeSet@1.1003.9.58, 2003-07-04 20:42:16-03:00, hugh@veritas.com [PATCH] swapoff loopable tmpfs The preceding loopable tmpfs patch needs a little more to handle swapoff correctly: shmem_unuse_inode may rarely be temporarily unable to swizzle the page from swap to file because concurrent loopio or sendfile has just inserted a distinct, not yet uptodate, filepage into the cache. Back off if so (with inelegant but effective interface to try_to_unuse). ChangeSet@1.1003.9.57, 2003-07-04 20:39:31-03:00, hugh@veritas.com [PATCH] shmem misc minor mods mark_page_accessed and SetPageReferenced like generic filemap does. shmem_mmap will never meet a null inode->i_sb (from James Morris). Remove unused shmem_nrpages, even -ac does nothing with its count. Replace info->locked by info->flags, though only VM_LOCKED used here. Correct VM_ACCT and shmem_nopage for PAGE_CACHE_SIZE != PAGE_SIZE. Use strsep instead of strtok, use BUG_ON instead of BUG. Make shmem_get_inode and shmem_sync_file static. Update copyrights. ChangeSet@1.1003.9.56, 2003-07-04 20:34:16-03:00, hugh@veritas.com [PATCH] shmem whitespace only Previous patches have been surreptitiously chipping away at the erratic whitespace in shmem.c, now clean up the rest. Builds exactly as before. ChangeSet@1.1003.9.55, 2003-07-04 20:33:41-03:00, hugh@veritas.com [PATCH] init_tmpfs shm_mnt error Renamed init_shmem_fs to init_tmpfs, adjusted its coding style; in the unlikely case of error leave shm_mnt as error code for shmem_file_setup. Removed exit_shmem_fs, this is never a loadable module (mmap uses it). ChangeSet@1.1003.9.54, 2003-07-04 20:32:42-03:00, hugh@veritas.com [PATCH] shmem_file_write and _read Update shmem_file_write: the -ac tree has long contained MAX_NON_LFS and s_maxbytes (SHMEM_MAX_BYTES) checking, bring those here at last; and the 2.5 tree has recently corrected its partial -EFAULT handling. Update shmem_file_read, less indented, and -EINVAL on negative count. ChangeSet@1.1003.9.53, 2003-07-04 20:32:04-03:00, hugh@veritas.com [PATCH] shmem account metablocks akpm and wli each discovered unfortunate behaviour of dbench on tmpfs: after tmpfs has reached its data memory limit, dbench continues to lseek and write, and tmpfs carries on allocating unlimited metadata blocks to accommodate the data it then refuses. That particular behaviour could be simply fixed by checking earlier; but I think tmpfs metablocks should be subject to the memory limit, and included in df and du accounting. Manipulate inode->i_blocks with sbinfo->free_blocks under stat_lock. Strengthen resemblance between shmem_free_swp and shmem_find_swp. Avoid shmem_free_swp search when (common case) nothing is swapped. shmem_truncate_direct adjust alloced count by metadata so shmem_recalc_ inode works out right. shmem_truncate_indirect use % instead of mask when calculating baseidx, to allow testing with small ENTRIES_PER_PAGE. ChangeSet@1.1003.9.52, 2003-07-04 20:31:01-03:00, hugh@veritas.com [PATCH] shmem truncation swizzled Fix two races between truncation and bringing tmpfs page back from swap. The first race, in which a partial page wasn't cleared because it was assigned to swap during truncate_inode_pages, already had an effective but inadequate fix in shmem_truncate. Now shmem_getpage(,,,SGP_READ), called from shmem_notify_change before truncation, gives a complete fix: holds page in memory across truncation, without having to allocate hole. The second race is that shmem_unuse_inode can bring a tmpfs page from swap to page cache, in the window between truncate_inode_pages (which frees page cache) and shmem_truncate (which frees swap). Check i_size and deletion's I_FREEING before add_to_page_cache there. (Yes, in rearranging that code, a SetPageDirty vanishes: try_to_unuse does it.) ChangeSet@1.1003.9.51, 2003-07-04 20:30:27-03:00, hugh@veritas.com [PATCH] shmem_getpage read,cache,write The first of these patches left shmem_getpage with an inadequate next_index test to guard against races with truncation. Now remove that check and settle the issue with checks against i_size within shmem_swp_alloc, which needs to know whether reading or writing. And let shmem_getpage use the empty_zero_page instead of allocating when a hole is read; but that cannot be done when it's being mapped, since nowadays the nopage doesn't know if page will be copied or not. ChangeSet@1.1003.9.50, 2003-07-04 20:29:06-03:00, hugh@veritas.com [PATCH] shmem_getpage absorb _locked The distinction between shmem_getpage and shmem_getpage_locked is not helpful, particularly now info->sem is gone; and shmem_getpage was confusingly tailored to shmem_nopage's expectations. Put the code of shmem_getpage_locked into the frame of shmem_getpage. shmem_getpage does need to lock its page (to secure it against shmem_writepage), but return with it unlocked. The only caller who appeared to be using the page lock was shmem_file_write, but it wasn't actually protecting against anything - i_sem prevents concurrent writes and truncates, and do_shmem_file_read dropped the lock before copying. ChangeSet@1.1003.9.49, 2003-07-04 20:28:01-03:00, hugh@veritas.com [PATCH] remove shmem info->sem Between inode->i_sem and info->lock comes info->sem; but it doesn't guard thoroughly against the difficult races (truncate during read), and serializes reads from tmpfs unlike other filesystems. I'd prefer to work with just i_sem and info->lock, backtracking when necessary (when another task allocates block or metablock at the same time). Oh, and remember to fix up free_blocks when page_cache_alloc fails. (I am not satisfied with the locked setting of next_index at the start of shmem_getpage_locked: it's one lock hold too many, and it doesn't really fix races against truncate better than before: a later patch will resolve that.) ChangeSet@1.1003.9.48, 2003-07-04 20:16:31-03:00, jgarzik@pobox.com [PATCH] fix Via pci irq routing Marcelo, Via irq routing has a funky PIRQD location. I checked my datasheets and, yep, this is correct all the way back to via686a. Bug spotted and original fix by Phoenix Tech. Please apply. ===== arch/i386/kernel/pci-irq.c 1.20 vs edited ===== ChangeSet@1.1003.9.47, 2003-07-04 19:42:55-03:00, benh@kernel.crashing.org [PATCH] ppc32: Fix a problem with both gmac and sungem Hi Marcelo ! This patch fixes a kernel crash when both gmac and sungem are in the kernel. TAG: MailDone ChangeSet@1.1003.9.46, 2003-07-04 19:42:17-03:00, benh@kernel.crashing.org [PATCH] ppc32: Minimal ethtool for bmac and mace Hi Marcelo ! bmac and mace are 2 powermac-only drivers. This patch adds minimal ethtool support for them (mostly for driver identification purposes from userland, a bit of port control for mace). please apply, Ben. ChangeSet@1.1003.9.45, 2003-07-04 19:41:01-03:00, benh@kernel.crashing.org [PATCH] ppc32: Update battery calculation code & via-pmu Hi Marcelo ! This patch updates the battery calculation code both in the via-pmu driver and the APM emulation driver to properly work with multiple batteries, among other fixes. The via-pmu driver also gets a couple of low level i2c access routines that are meant to be user later on, but that I'd like to get in now. Please apply, Ben. ChangeSet@1.1003.9.44, 2003-07-04 19:34:22-03:00, benh@kernel.crashing.org [PATCH] ppc32: Update adbhid driver Hi Marcelo ! This patch updates the adbhid driver for ADB keyboards and mice. It fixes some issues with reporting of backlight keys and the Power key on some PowerBooks. Please apply, Ben. ChangeSet@1.1003.9.43, 2003-07-04 19:32:33-03:00, benh@kernel.crashing.org [PATCH] ppc32: Add a "query" function to core ADB Hi Marcelo ! ADB (Apple Desktop Bus) is the bus used on Macs to plug mice and keyboards, along as a few other low speed devices. This patch adds a simple way for userland to retreive a list of devices along with their original address (before relocation). Without this, it's a lot more difficult for userland to probe the type of devices on the ADB bus as the original address is no longer available, and it indicates the "family" of device. This patch is already present in various distribution kernels to allow their installer to probe for ADB keyboards & mice, please apply. Ben. ChangeSet@1.1003.9.42, 2003-07-04 19:30:16-03:00, benh@kernel.crashing.org [PATCH] ppc32: Add TotalImpact briQ panel driver Hi Marcelo ! This driver was missing from 2.4.21, it drives the front-panel of Total Impact briQ machines, written by Karsten Jeppesen. Please apply, Ben. ChangeSet@1.1003.9.41, 2003-07-04 18:02:32-03:00, benh@kernel.crashing.org [PATCH] ppc32: Update swim3 floppy driver Hi Marcelo ! This patch contains various fixes to the PowerMac swim3 floppy driver. Those fixes are from Paul Mackerras. They improve significantly reliability of this driver. Please apply, Ben. ChangeSet@1.1003.9.40, 2003-07-04 18:01:57-03:00, benh@kernel.crashing.org [PATCH] ppc32: Update PowerMac motherboard support Hi Marcelo ! This patch update the low level PowerMac motherboard support to properly identify and deal with newer PowerBooks and XServe rev 2. Please apply, Ben. ChangeSet@1.1003.9.39, 2003-07-04 18:01:11-03:00, benh@kernel.crashing.org [PATCH] ppc32: Enable use of USB2 on recent G4s Hi Marcelo ! This patch enables the firmware-disabled EHCI part that Apple hide in some recent G4 models, thus allowing the PCI EHCI driver to pick it up. Please apply to your tree, Ben. ChangeSet@1.1003.9.38, 2003-07-04 17:42:41-03:00, benh@kernel.crashing.org [PATCH] ppc32: support for 750FX rev2 CPU Hi Marcelo ! This patch adds support for the IBM 750FX rev2. CPU so that the HID2 register is properly saved/restored accross machine sleep on ibooks using this CPU. Please apply, Ben. ChangeSet@1.1003.9.37, 2003-07-04 17:22:10-03:00, lethal@linux-sh.org [PATCH] sh64: Fixes for Cayman LEDs This patch adds support for adjusting brightness on the alphanumeric display on the cayman, and fixes up some misc issues with the led code. arch/sh64/kernel/led.c | 2 +- arch/sh64/mach-cayman/led.c | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-) ChangeSet@1.1003.9.36, 2003-07-04 17:21:40-03:00, lethal@linux-sh.org [PATCH] sh64: export more needed symbols This patch just exports some more symbols that were needed for various modules. Please apply. arch/sh64/kernel/sh_ksyms.c | 12 ++++++++++++ 1 files changed, 12 insertions(+) ChangeSet@1.1003.9.35, 2003-07-04 17:21:20-03:00, lethal@linux-sh.org [PATCH] sh64: Make memcpy safe on SH5-101 cut2 This patch does some cleanup of copy_user_memcpy() and memcpy_toio() to make it safe on cut2 of the SH5-101 processor. Please apply. arch/sh64/lib/copy_user_memcpy.S | 14 ++++++++++---- arch/sh64/lib/io.c | 9 +++++---- 2 files changed, 15 insertions(+), 8 deletions(-) ChangeSet@1.1003.9.34, 2003-07-04 17:20:39-03:00, lethal@linux-sh.org [PATCH] sh64: Fix PTRACE_POKEUSR to ignore changes of privileged This patch changes some of the PTRACE_POKEUSR semantics to drop changes of privileged bits of the SR register .. otherwise a gaping security hole. Please apply. arch/sh64/kernel/ptrace.c | 13 ++++++------- 1 files changed, 6 insertions(+), 7 deletions(-) ChangeSet@1.1003.9.33, 2003-07-04 17:20:02-03:00, lethal@linux-sh.org [PATCH] sh64: Cleanup sleep usage This patch does some cleanup of usage of the sleep instruction. Usage of sleep in the idle loop is now configurable, and we also use it to enter deep standby mode for machine_power_off(). Please apply. arch/sh64/kernel/process.c | 62 +++++++++++++++++++++++++++++++++++++++++++++++++------------- arch/sh64/kernel/time.c | 33 +++++++++++++++++++++++++++++++++ 2 files changed, 82 insertions(+), 13 deletions(-) ChangeSet@1.1003.9.32, 2003-07-04 17:19:42-03:00, lethal@linux-sh.org [PATCH] sh64: Add workarounds for cache aliasing issues This patch adds a HAVE_ARCH_UNMAPPED_AREA / arch_get_unmapped_area() that essentially follows the SH implementation to get around the same issues. This also fixes up some deadlocks that were occuring due to problems with flush_page_to_ram() by wrapping it back to flush_dcache_page(). arch/sh64/kernel/sys_sh64.c | 41 +++++++++++++++++++++++++++++++++++++++++ arch/sh64/mm/cache.c | 29 +++++++++++++++++++++++++++-- include/asm-sh64/pgtable.h | 4 ++++ 3 files changed, 72 insertions(+), 2 deletions(-) ChangeSet@1.1003.9.31, 2003-07-04 17:17:55-03:00, lethal@linux-sh.org [PATCH] sh64: Don't startup the irq in make_intc_irq() This patch reverts previous behavior that started up the IRQ automatically in make_intc_irq(). Instead, we want this going through disable_intc_irq() after assigning the handler type so that it can be started up by the normal handler->startup() route. Please apply. arch/sh64/kernel/irq_intc.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.9.30, 2003-07-04 17:16:20-03:00, lethal@linux-sh.org [PATCH] sh64: tlbmiss handler updates This patch includes updates for the fast-path tlbmiss handler by Richard Curnow. Please apply. arch/sh64/kernel/entry.S | 64 +++++++++++++++++++++++++++++++++------------------------ arch/sh64/mm/Makefile | 2 - arch/sh64/mm/fault.c | 45 +++++++++++++++++++++++++--------------- arch/sh64/mm/tlbmiss.c | 11 ++++++++- 4 files changed, 77 insertions(+), 45 deletions(-) ChangeSet@1.1003.9.29, 2003-07-04 17:15:45-03:00, lethal@linux-sh.org [PATCH] sh64: Add an onchip_unmap() to clean up after sh64 needs to remap onchip peripherals in order to get to them, this patch implements a method for these mappings to be unmapped. The DMAC patch also depends on this, please apply. arch/sh64/mm/ioremap.c | 110 ++++++++++++++++++++++++++++++++++++++++++++++++++++++-------- 1 files changed, 96 insertions(+), 14 deletions(-) ChangeSet@1.1003.9.28, 2003-07-04 16:28:46-03:00, lethal@linux-sh.org [PATCH] sh64: Fix SHMBLA compile error We got rid of CACHE_OC_NUM_SETS awhile ago, this updates the SHMBLA definition to use the new value. Primarily a compilation fix, please apply. include/asm-sh64/shmparam.h | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.9.27, 2003-07-04 16:28:02-03:00, lethal@linux-sh.org [PATCH] sh64 PCI DMA coherency fixups This patch implements cache coherency fixups for PCI DMA on the sh64, and also does proper cleanup for pci_free_consistent(). Please apply. arch/sh64/kernel/pci-dma.c | 12 +++++--- include/asm-sh64/pci.h | 56 ++++++++++++++++++++++++++++++++++++----- include/asm-sh64/scatterlist.h | 13 +++++---- 3 files changed, 65 insertions(+), 16 deletions(-) ChangeSet@1.1003.9.26, 2003-07-04 16:25:13-03:00, lethal@linux-sh.org [PATCH] SH-5 DMAC Support This patch adds support for the SH-5 DMAC. Please apply. arch/sh64/config.in | 6 arch/sh64/kernel/Makefile | 2 arch/sh64/kernel/dma.c | 293 ++++++++++++++++++++++++++++++++++++++++++++ include/asm-sh64/dma.h | 7 - include/asm-sh64/hardware.h | 26 +++ include/asm-sh64/io.h | 80 ++++++------ 6 files changed, 372 insertions(+), 42 deletions(-) ChangeSet@1.1003.14.2, 2003-07-03 09:56:38-07:00, trini@kernel.crashing.org PPC32: Remove trailing whitespace in numerous files. ChangeSet@1.1003.14.1, 2003-07-03 09:15:28-07:00, trini@kernel.crashing.org PPC32: Add support for the Motorola PowerPlus family of boards. ChangeSet@1.1003.9.25, 2003-07-02 21:22:49-03:00, marcelo@freak.distro.conectiva Merge http://nfsclient.bkbits.net/linux-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1003.10.121, 2003-07-02 19:35:10-03:00, olh@suse.de [PATCH] missing asm-ppc64/kmap_types.h [ The following text is in the "utf-8" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] copied from 2.5: TAG: MailDone ChangeSet@1.1003.10.120, 2003-07-02 19:34:21-03:00, lethal@linux-sh.org [PATCH] Update MAINTAINERS for sh/sh64 This patch adds a MAINTAINERS entry for sh64, and also includes an update for my email address and sh. MAINTAINERS | 26 +++++++++++++++++++++----- 1 files changed, 21 insertions(+), 5 deletions(-) ChangeSet@1.1003.10.119, 2003-07-02 19:29:52-03:00, lethal@linux-sh.org [PATCH] Add SH-5 support to tulip_core This patch adds SH-5 support to the tulip driver. The patch itself is harmless, and doesn't touch any common tulip code. This is needed for sh64, please apply. tulip_core.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) ChangeSet@1.1003.10.118, 2003-07-02 19:23:06-03:00, lethal@linux-sh.org [PATCH] Add SH-5 support to SH-SCI This patch adds support for the SH-5 101 and SH-5 103 to the SH-SCI serial driver. This is needed for serial on sh64, please apply. sh-sci.c | 18 ++++++++++++++++-- sh-sci.h | 29 ++++++++++++++++++++++++++++- 2 files changed, 44 insertions(+), 3 deletions(-) ChangeSet@1.1003.10.117, 2003-07-02 19:22:30-03:00, ink@jurassic.park.msu.ru [PATCH] alpha: finalize Sable/Lynx support (Jay Estabrook) Make X servers finally work on Lynx and Sable platforms. Ivan. ChangeSet@1.1003.10.116, 2003-07-02 19:21:58-03:00, ink@jurassic.park.msu.ru [PATCH] alpha: fix generic kernel build (Jay Estabrook) This fixes build problem introduced by recent Lynx patch. Ivan. ChangeSet@1.1003.10.114, 2003-07-02 19:10:49-03:00, rmk@arm.linux.org.uk [PATCH] ARM merge part 3 - drivers/acorn ARM merge - drivers/acorn, mainly: - fix previous cleanups to fd1772.c so we clean up properly - fix ether driver ETH_ZLEN leaks - make acorn serial cards use iomem accesses - Update fas216 and associated scsi drivers with performance improvements ChangeSet@1.1003.10.113, 2003-07-02 19:09:58-03:00, rmk@arm.linux.org.uk [PATCH] ARM merge part 2 - include/asm-arm ARM merge - include/asm-arm - add new machine types, epxa10db becomes epxa, and updates for 2.4.20 and .21 changes. ChangeSet@1.1003.10.112, 2003-07-02 19:06:54-03:00, rmk@arm.linux.org.uk [PATCH] ARM merge part 1 - arch/arm ARM Merge - merge arch/arm (new machine types, new CPU types, nwfpe performance improvements.) ChangeSet@1.1003.10.111, 2003-07-02 18:48:08-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add vram to vesafb docs TAG: MailDone ChangeSet@1.1003.10.110, 2003-07-02 18:47:55-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update cciss docs to match new driver ChangeSet@1.1003.10.109, 2003-07-02 18:47:33-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add scripts ready to merge kconfig ChangeSet@1.1003.10.108, 2003-07-02 18:47:14-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: declare semtimedop function ChangeSet@1.1003.10.107, 2003-07-02 18:46:56-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update ac97 codec headers ChangeSet@1.1003.10.106, 2003-07-02 18:46:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add semtimedop to x86 headers ChangeSet@1.1003.10.105, 2003-07-02 18:44:19-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update sis fb drivers ChangeSet@1.1003.10.104, 2003-07-02 18:43:52-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update vesafb memory handling for big cards ChangeSet@1.1003.10.103, 2003-07-02 18:43:30-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: core fbcon fixes ChangeSet@1.1003.10.102, 2003-07-02 18:43:06-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update to new ac97_codec ChangeSet@1.1003.10.101, 2003-07-02 18:42:45-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: Update via audio - fix problems esd, mpg321 ChangeSet@1.1003.10.100, 2003-07-02 18:42:15-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update trident, fix printks, new ac97 ChangeSet@1.1003.10.99, 2003-07-02 18:41:42-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix long standing doc typo ChangeSet@1.1003.10.98, 2003-07-02 18:41:13-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: ac97 updates ChangeSet@1.1003.10.97, 2003-07-02 18:40:51-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: switch i810 to generalised digital out, new ac97 ChangeSet@1.1003.10.96, 2003-07-02 18:40:32-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: switch cards to new ac97_audio ChangeSet@1.1003.10.95, 2003-07-02 18:39:30-03:00, bunk@fs.tum.de [PATCH] postfix a constant in efi.h with ULL On Mon, Jun 23, 2003 at 04:27:07PM -0500, Matt_Domsch@Dell.com wrote: > > The patch below postfixes a constant in efi.h with ULL, on 32 bit archs > > this constant is too big for an int. > > -#define GPT_HEADER_SIGNATURE 0x5452415020494645L > > +#define GPT_HEADER_SIGNATURE 0x5452415020494645ULL > > Sounds good. Please submit for 2.4.x also. Marcelo, this trivial patch is below. Please apply. > Thanks, > Matt cu Adrian ChangeSet@1.1003.10.94, 2003-07-02 18:39:08-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update AC97 codec core This has a couple of items to follow ChangeSet@1.1003.10.93, 2003-07-02 18:38:52-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix copy from user bug in cmpci ChangeSet@1.1003.10.92, 2003-07-02 18:38:16-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: remove noise ChangeSet@1.1003.10.91, 2003-07-02 18:36:52-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update scsi tape driver ChangeSet@1.1003.10.90, 2003-07-02 18:36:33-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: new segate bios string ChangeSet@1.1003.10.89, 2003-07-02 18:36:14-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: resync scsi blacklist ChangeSet@1.1003.10.88, 2003-07-02 18:35:54-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: let the ide layer fail commands This gets us out of fixing the problems with ide reset with ide-scsi by going back to old behaviour. Does mean Bart has a mess on his hands fixing 2.5 properly of course 8) ChangeSet@1.1003.10.87, 2003-07-02 18:35:27-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: send_diag wants long timeout default ChangeSet@1.1003.10.86, 2003-07-02 18:27:23-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: megaraid broke config tools ChangeSet@1.1003.10.85, 2003-07-02 18:26:56-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update scsi tape docs ChangeSet@1.1003.10.84, 2003-07-02 18:26:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: gdth register failure path ChangeSet@1.1003.10.83, 2003-07-02 18:08:51-03:00, andersen@codepoet.org [PATCH] fix 2.4.22-pre broken x86 math-emu As of today's "fix up gcc 3.3 bits" patch [1], x86 math emulation is now even more broken, since this latest patch has added some mismatched quotes while still failing to address the actual problems preventing this code from compiling with gcc 3.3. This patch, first sent to you on Jun 21st, fixes the missing semicolons and missing quotes in the x86 math-emu code, allowing it to compile with gcc 3.3. I have updated things to also fix the mismatched quotes that were added today. Unlike the patch you applied earlier today, my patch is actually tested... Please apply, -Erik [1] http://www.kernel.org/diff/diffview.cgi?file=/pub/linux/kernel/v2.4/testing/cset/cset-alan@lxorguk.ukuu.org.uk|ChangeSet|20030701183359|14011.txt -- Erik B. Andersen http://codepoet-consulting.com/ --This message was written using 73% post-consumer electrons-- ChangeSet@1.1003.10.82, 2003-07-02 18:03:49-03:00, marcelo@freak.distro.conectiva Merge http://lia64.bkbits.net/to-marcelo-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4-cloned ChangeSet@1.1003.10.81, 2003-07-02 18:03:05-03:00, marcelo@freak.distro.conectiva Cset exclude: rusty@rustcorp.com.au|ChangeSet|20030625201246|52045 ChangeSet@1.1003.10.80, 2003-07-02 17:42:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: aic7xxx allow db4 TAG: MailDone ChangeSet@1.1003.10.79, 2003-07-02 17:28:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update aacraid ChangeSet@1.1003.10.78, 2003-07-02 17:27:46-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: pci routing for ti cardbus ChangeSet@1.1003.10.77, 2003-07-02 17:27:01-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix rsrc manager ChangeSet@1.1003.10.76, 2003-07-02 17:26:48-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add cirrus support to i82092 ChangeSet@1.1003.10.75, 2003-07-02 17:24:54-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update pci.ids ChangeSet@1.1003.10.74, 2003-07-02 17:24:35-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update orinoco drivers ChangeSet@1.1003.10.73, 2003-07-02 17:24:06-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update sonic ChangeSet@1.1003.10.72, 2003-07-02 17:21:38-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix plip hang on ifdown/ifup ChangeSet@1.1003.10.71, 2003-07-02 17:21:14-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix leak in aironet4500_cs ChangeSet@1.1003.10.70, 2003-07-02 17:20:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: minor m68k fixes ChangeSet@1.1003.10.69, 2003-07-02 17:20:14-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix ma600 gcc 3.3 ChangeSet@1.1003.10.68, 2003-07-02 17:19:34-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: typo fix in atari_pamsnet driver ChangeSet@1.1003.10.67, 2003-07-02 17:19:18-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: new 3c59x. plus handle power bits (ACPI needs the power bits) ChangeSet@1.1003.10.66, 2003-07-02 17:18:51-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix leak in octagon ChangeSet@1.1003.10.65, 2003-07-02 17:18:32-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update motion eye drivers ChangeSet@1.1003.10.64, 2003-07-02 16:18:27-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: d_path can return an error code, must handle it ChangeSet@1.1003.10.63, 2003-07-02 16:04:46-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: small setup-pci cleanups ChangeSet@1.1003.10.62, 2003-07-02 16:01:34-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: hptraid updates ChangeSet@1.1003.10.61, 2003-07-02 15:56:49-03:00, ak@muc.de [PATCH] non executable stack support for x86-64 AMD64 optionally supports non executable stacks now. This requires some minor changes in generic code: - Allow the architecture to overwrite VM_STACK_FLAGS - Look up the protection of the stack mapping in exec instead of hardcoding PAGE_COPY find_vma is very cheap here because the VMA list only contains one element at this point. ChangeSet@1.1003.13.1, 2003-07-03 01:55:31+10:00, lunz@falooley.org [NET] Fix refcounting of dev->promiscuity for af_packet. ChangeSet@1.1003.10.60, 2003-07-02 11:59:08-03:00, rusty@rustcorp.com.au [PATCH] 2.5.43 export _end From: Keith Owens Some programs such as ps, lkcd and others need to validate that System.map matches the kernel. Comparing all symbol names from ksyms against map breaks badly when faced with function descriptors (ia64 has hundreds of mismatches because of function descriptors). lkcd attempts to solve this problem by adding kernel_magic which contains the value of _end, but that requires /dev/kmem access to read kernel_magic. Trivial fix - export _end. Every arch *lds* file defines _end. No special access is required to match ksyms _end against System.map _end. Index: 43.1/kernel/ksyms.c ChangeSet@1.1003.10.59, 2003-07-02 11:56:38-03:00, ak@muc.de [PATCH] Support exception-trace sysctl for x86-64 Adds a sysctl to control the exception_trace variable in the x86-64 architecture specific code. ChangeSet@1.1003.10.58, 2003-07-02 11:41:51-03:00, ak@muc.de [PATCH] x86-64 merge x86-64 merge for 2.4.22pre1. Only changes architecture specific code. This needs a few other patches I'm sending separately. - document all boot options - Set -Wno-sign-compare for gcc 3.3 prereleases - Support new ACPI code (including suspend code from 2.5, but it's not usable because the generic ACPI code only does S1 and S5 currently) - Update defconfig - Support 3GB personality to make buggy JVMs work - Support for non executable stacks - Inline assembly fixes for latest gcc 3.3-hammerbranch - various bugfixes for the IA32 emulation - New extended reboot code; support warm reboot, CPU reboot, better SMP reboot - Better flush strategy for PCI IOMMU - Unmap aperture in PCI IOMMU - More support for NX bit. - Only print backtrace once at oops - Add more oops locking - Print exception trace only for unhandled signals - Support 16 IO-APICs - Fix minor bug in x87 FPU handling. could access uninitialized memory. - Tell generic code we have an NMI watchdog - Test properly for local apic support - Fix siginfo to match glibc/POSIX ChangeSet@1.1003.10.57, 2003-07-02 09:42:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: SiS IDE updates TAG: MailDone ChangeSet@1.1003.10.56, 2003-07-02 09:42:20-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: siimage updates, add aar-1210sa ChangeSet@1.1003.10.55, 2003-07-02 09:04:37-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: clean up older pdc ChangeSet@1.1003.10.54, 2003-07-02 08:38:43-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix promise sx6000 newer board problems ChangeSet@1.1003.10.53, 2003-07-02 08:24:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix hpt speed bits ChangeSet@1.1003.10.52, 2003-07-02 08:22:40-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add generic support for toshiba piccolo ChangeSet@1.1003.10.51, 2003-07-01 16:29:12-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make pdc4030 module happy TAG: MailDone ChangeSet@1.1003.10.50, 2003-07-01 16:28:29-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: Herbert's fix for ide proc oops ChangeSet@1.1003.10.49, 2003-07-01 16:26:46-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make pnpide module happy ChangeSet@1.1003.10.48, 2003-07-01 16:26:24-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix ide dma timeout bugs ChangeSet@1.1003.10.47, 2003-07-01 16:25:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: make IDE modularisable ChangeSet@1.1003.10.46, 2003-07-01 16:25:04-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix config.in bits for IDE ChangeSet@1.1003.10.45, 2003-07-01 16:24:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: parisc gsc driver sync ChangeSet@1.1003.10.44, 2003-07-01 16:23:46-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update sonypi driver ChangeSet@1.1003.10.43, 2003-07-01 16:23:05-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: & v && fixes in sysrq.c ChangeSet@1.1003.10.42, 2003-07-01 16:22:24-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add comtrol note in case we need to know in the future ChangeSet@1.1003.10.41, 2003-07-01 15:50:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: Jens floppy locking fixes TAG: MailDone ChangeSet@1.1003.10.40, 2003-07-01 15:48:18-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: bump cciss to new vendor driver ChangeSet@1.1003.10.39, 2003-07-01 15:47:42-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix iphase leak ChangeSet@1.1003.10.38, 2003-07-01 15:47:19-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: typo fix ChangeSet@1.1003.10.37, 2003-07-01 15:35:29-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: collected m68k core diffs ChangeSet@1.1003.10.36, 2003-07-01 15:34:43-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: copy the right data in mips emulation ChangeSet@1.1003.10.35, 2003-07-01 15:33:59-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix up gcc 3.3 bits ChangeSet@1.1003.10.34, 2003-07-01 15:30:53-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add semtimedop to ia64 emu too ChangeSet@1.1003.10.33, 2003-07-01 15:30:06-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix error in vm86 fixups ChangeSet@1.1003.10.32, 2003-07-01 15:29:24-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix up semops and return, allow timedop ChangeSet@1.1003.10.31, 2003-07-01 15:28:20-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: merge long standing reboot fix form -ac ChangeSet@1.1003.10.30, 2003-07-01 15:23:08-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: large scale DMI table updates ChangeSet@1.1003.10.29, 2003-07-01 15:20:32-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: dont idle if forbid_idle set ChangeSet@1.1003.10.28, 2003-07-01 15:19:01-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix make rpm ChangeSet@1.1003.10.27, 2003-07-01 15:13:06-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: dont corrupt utsname on failed copy ChangeSet@1.1003.9.21, 2003-07-01 12:33:12+02:00, Richard.Curnow@superh.com Ensure that the 'unlink' XDR structures are correctly aligned on 64-bit architectures. ChangeSet@1.1003.10.25, 2003-07-01 01:21:58-07:00, davem@nuts.ninka.net [SPARC64]: Update defconfig. ChangeSet@1.1003.11.4, 2003-06-30 23:10:13-07:00, shemminger@osdl.org [BRIDGE]: Ethernet bridge fixes. 1. STP protocol has no security, so malcontents can fuck with the bridge's topology. The fixes are to ship with STP turned off to protect the ignorant, and run STP packets through ebtables netfilter for the smart. Got this one via a russian hacker "Oleg K. Artemjev" before he published the paper. Bridge netfilter still needs work to give a nice face on this but this patch gives the hooks to filter. 2. STP input processing was lax in it's length checking so I bet you could make up a bomb packet. My inspection while doing #1. 3. Forwarding table could be abused by sending forged packets with bogus source address same as the local host. This came via Lennart from Jerry Kreuscher who ran into it by mistake. ChangeSet@1.1003.11.3, 2003-06-30 21:15:59-07:00, davem@nuts.ninka.net Merge nuts.ninka.net:/home/davem/src/BK/network-2.4 into nuts.ninka.net:/home/davem/src/BK/net-2.4 ChangeSet@1.930.1.250, 2003-06-30 16:52:41-06:00, bjorn_helgaas@hp.com ia64: Ignore empty address ranges from _CRS to workaround buggy Big Sur firmware. See linux-ia64 archives, May 9, 2003: "Re: [Linux-ia64] [PATCH] 1/4 multi-ioport space support for 2.5" ChangeSet@1.930.1.249, 2003-06-30 15:02:35-06:00, jh@sgi.com ia64: SN2 update 030630 Here's an update of SN2 specific files relative to to-marcelo-2.4. It includes: Clean-up/remove early prototype code. hwgraph changes Fix some error return values. pci bridge RRB tweaks Generic kernel tweaks for SN. Fix for IO port space on SN (fixes slow console) Support for changing memory protections and registering nofault ranges. Other miscellaneous bug fixes. ChangeSet@1.1003.12.25, 2003-06-30 17:21:25-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: assorted module race fixe ChangeSet@1.1003.12.24, 2003-06-30 17:18:55-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add timedop stub for IPC=n ChangeSet@1.1003.12.23, 2003-06-30 17:16:39-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: support cramfs initrd ChangeSet@1.1003.12.22, 2003-06-30 15:09:29-03:00, ak@muc.de [PATCH] Personality fixes for x86-64 Remove useless printk in sys_personality that made the system call quite annoying. Add a 3GB personality for x86-64. This is needed to get some buggy JVMs to work. ChangeSet@1.930.1.248, 2003-06-30 09:50:40-06:00, jsm@udlkern.fc.hp.com [PATCH] ia64: don't let PTRACE_POKEDATA write the NaT bits of syscall args ChangeSet@1.930.1.247, 2003-06-30 09:50:20-06:00, arun.sharma@intel.com [PATCH] ia64: IA-32 support patch: mmap should return ENOMEM Single Unix Specification says: [ENOMEM] MAP_FIXED was specified, and the range [addr, addr + len) exceeds that allowed for the address space of a process. Therefore if addr + len > IA32_PAGE_OFFSET, we should return ENOMEM. ChangeSet@1.930.1.246, 2003-06-30 09:50:09-06:00, arun.sharma@intel.com [PATCH] ia64: IA-32 support patch: munmap should return EINVAL if size == 0 Native IA-32 returns EINVAL on a unmap of size 0. Our behavior is currently not compatible. ChangeSet@1.930.1.245, 2003-06-30 09:49:59-06:00, arun.sharma@intel.com [PATCH] ia64: IA-32 support patch: msgsnd/msgrcv return value off by 4 ChangeSet@1.1003.12.21, 2003-06-30 12:14:08-03:00, marcelo@freak.distro.conectiva Merge bk://namesys.com/bk/reiser3-linux-2.4-relocated into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1003.12.20, 2003-06-30 11:20:09-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add new entry to sisfb types TAG: MailDone ChangeSet@1.1003.12.19, 2003-06-30 11:19:30-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: bring PCI_IDS back into sync ChangeSet@1.1003.12.18, 2003-06-30 11:19:00-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: update fat docs - we now use the field ChangeSet@1.1003.12.17, 2003-06-30 11:18:32-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add EDP2 protocol id ChangeSet@1.1003.12.16, 2003-06-30 11:17:54-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add the ide_register_driver defines ChangeSet@1.1003.12.15, 2003-06-30 11:16:50-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add a flag so we can forbid APM idling ChangeSet@1.1003.12.14, 2003-06-30 10:06:17-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: collated copy of Geerts patches for m68k headers TAG: MailDone ChangeSet@1.1003.12.13, 2003-06-30 09:30:26-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: we moved these so this copy can go ChangeSet@1.1003.12.12, 2003-06-30 09:29:41-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix false sharing of mm info ChangeSet@1.1003.12.11, 2003-06-30 09:28:55-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add the MSR's for IA32 perf ctl ChangeSet@1.1003.12.10, 2003-06-30 09:28:30-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: remove io_apic_modify - this doesnt work on some APICs ChangeSet@1.1003.12.9, 2003-06-29 20:09:29-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add the extra cpu bit test flags TAG: MailDone ChangeSet@1.1003.12.8, 2003-06-29 20:08:53-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: add xapic checking define ChangeSet@1.1003.12.7, 2003-06-29 20:08:29-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix decnet gcc 3.3 build ChangeSet@1.1003.12.6, 2003-06-29 20:06:36-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: fix leak in llc 802 ChangeSet@1.1003.12.5, 2003-06-29 20:06:10-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: makefile for HFSPLUS ChangeSet@1.1003.12.4, 2003-06-29 20:05:44-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: config.in for HFSPLUS ChangeSet@1.1003.12.3, 2003-06-29 20:05:20-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: NLS config.in for hfsplus ChangeSet@1.1003.12.2, 2003-06-29 20:04:11-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATH: add hfsplus file system (stands alone) ChangeSet@1.1003.12.1, 2003-06-29 20:03:44-03:00, alan@lxorguk.ukuu.org.uk [PATCH] PATCH: Optimise FAT handling using the prev_free info as DOS does ChangeSet@1.1003.10.23, 2003-06-29 15:00:18-07:00, alan@lxorguk.ukuu.org.uk [SPARC]: d_path() can return an error code, must handle it. ChangeSet@1.1003.11.2, 2003-06-29 14:41:34-07:00, alan@lxorguk.ukuu.org.uk [NET]: Add EDP2 ethernet protocol ID. ChangeSet@1.1003.11.1, 2003-06-29 14:36:26-07:00, alan@lxorguk.ukuu.org.uk [NETFILTER]: Fix nat_helper warnings with gcc 3.3 ChangeSet@1.1003.10.22, 2003-06-28 15:53:29-07:00, davem@nuts.ninka.net [SPARC64]: Fix build error from OBP parsing patch. ChangeSet@1.1003.10.21, 2003-06-28 15:32:28-07:00, tcallawa@redhat.com [SPARC64]: Fix OBP version parsing on newer systems. ChangeSet@1.1003.10.20, 2003-06-27 20:21:16-07:00, davem@nuts.ninka.net [NET]: Scale DST/ipv6 intervals like we did for ipv4. TAG: MailDone ChangeSet@1.1003.10.19, 2003-06-27 20:07:21-07:00, davem@nuts.ninka.net [NET]: net/bluetooth/cmtp/core.c needs linux/init.h ChangeSet@1.1003.10.18, 2003-06-27 18:24:40-07:00, hch@lst.de [CRYPTO-2.4]: Missing ULL postfixes and statics. ChangeSet@1.1003.10.17, 2003-06-27 18:21:33-07:00, davem@nuts.ninka.net Merge nuts.ninka.net:/home/davem/src/BK/network-2.4 into nuts.ninka.net:/home/davem/src/BK/net-2.4 ChangeSet@1.1003.1.14, 2003-06-27 21:17:27-04:00, shmulik.hen@intel.com [PATCH] Fix 802.3ad long fail over with high UDP Tx stress Hi, This patch fixes a problem detected by our QA group. On very high bi-directional stress traffic, removing the last slave of the active aggregator results in long failover time to another aggregator (upto 90 sec). The fix is to send LACPDU packets with the highest priority (TC_PRIO_CONTROL), to overcome the possibility of packets being dropped from the adapter's queue. This further fixes the original long failover problem reported by Jay Vosburgh on April 3rd and fixed by us on May 20th. We verified it fixes the problem for 1000Mbps adapters, but it may still not entirely fix it for 10/100 adapters since they simply can't handle the load. In the latter case, the failover may have to wait the entire timeout. Since we still don't use bitkeeper, this patch should be applied on top of Marcelo's 2.4.22-pre1 patch plus Jeff Garzik's 2.4 net driver updates from from June 20Th 2.4.22-pre1-netdrvr1. There is also a small fix for a non-printable character that somehow snuck into bond_3ad.h. ChangeSet@1.1003.1.13, 2003-06-27 21:17:13-04:00, shmulik.hen@intel.com [PATCH] Fix load balance problem with high UDP Tx stress Hi, This patch fixes a problem detected by our QA group. On very high UDP Tx stress traffic on 10/100 adapters, load sharing would collapse to only one slave after very short time. The bug is due to unsigned to signed conversions that caused calculation errors (outgoing traffic "exceeds" adapter's actual capability). Since we still don't use bitkeeper, this patch should be applied on top of Marcelo's 2.4.22-pre1 patch plus Jeff Garzik's 2.4 net driver updates from from June 20Th (2.4.22-pre1-netdrvr1). ChangeSet@1.1003.1.12, 2003-06-27 21:17:01-04:00, scott.feldman@intel.com [PATCH] Remove CAP_NET_ADMIN check for SIOCETHTOOL's dev_ioctl already checks capable(CAP_NET_ADMIN), so no need to do so in drivers. ChangeSet@1.1003.10.16, 2003-06-27 16:59:55-07:00, judd@jpilot.org [PATCH] USB: visor.h[c] USB device IDs Add ability to specify USB vendor and product ids as module options. TAG: MailDone ChangeSet@1.1003.2.13, 2003-06-27 16:01:19-07:00, greg@kroah.com [PATCH] USB: compiler fixes for previous vicam patches. ChangeSet@1.1003.2.12, 2003-06-27 16:01:07-07:00, oliver@neukum.org [PATCH] USB: fix to previous vicam patch OK, I'll think next time. - fix my own stupid oversight regarding disconnect() ChangeSet@1.1003.2.11, 2003-06-27 16:01:00-07:00, oliver@neukum.org [PATCH] USB: disconnect of v4l devices in 2.4 in 2.4 video_unregister_device() has lost its magic properties breaking most USB v4l drivers. IMHO they should be converted to delayed freeing resources just like ordinary character devices. Here's the change for vicam.c. ChangeSet@1.1003.2.10, 2003-06-27 15:23:21-07:00, david@csse.uwa.edu.au [PATCH] USB: usb-ohci handling of one-shot interrupt transfers A long standing problem has existed with usb-ohci handling of one-shot interrupt transfers (they never worked). Attached is a fix which was originally proposed by P.C. Chan and subsequently modified and re-presented by Frode Isaksen. The Lego USB driver does not work with ohci without this fix and so I would really appreciate it being applied. ChangeSet@1.1003.2.9, 2003-06-27 15:23:13-07:00, david@csse.uwa.edu.au [PATCH] USB: usb-uhci fix for one-shot interrupt problem A change introduced into usb-uhci.c in 2.4.21 causes the kernel to freeze when usb-uhci is used with any driver using one-shot interrupt transfers. The attached fix was originally proposed by Frode Isaksen and improved by Pete Zaitcev. Pete Zaitcev has applied this patch as an errata fix for the RedHat 9.0 kernel. Other than the serious problem that this causes with the Lego USB driver (and yes, this is used pretty heavily in Universities for teaching and some research), there are other drivers (e.g. Visor Treo 90) that this causes problems for. ChangeSet@1.1003.2.8, 2003-06-27 12:02:02-07:00, abbotti@mev.co.uk [PATCH] USB: several ftdi_sio driver patches I have attached several patches for the ftdi_sio (USB serial device) driver that I have been accumulating over the last month or so as the official maintainer (Bill Ryder) has been rather quiet of late. He hasn't responded to any patches or other messages on ftdi-usb-sio-devel since the end of March. The last patch Bill sent to linux-usb-devel was for ftdi_sio version 1.3.3, which is the latest available for download from the sourceforge project page. Greg had some criticisms about whitespace, braces, etc. which was not replied to by Bill. In this sequence of patches, I have tidied some things up, accepted patches and vid/pids for extra device support and fixed a spinlock bug. The patches apply cleanly in the sequence presented here. I have split the patches by function, but have attempted to preserve the chronology where possible - there is a certain amount of time-warping going on as can be seen from the file header comments changed by the patches! The patches are as follows: 2.4.21-ftdi_sio-p01-xonxoff.patch - John Wilkins Xon/Xoff patch (included in ftdi_sio-1.3.3) 2.4.21-ftdi_sio-p02-homechoice.patch - John Wilkins vid/pid for Homechoice (included in ftdi_sio-1.3.3) 2.4.21-ftdi_sio-p03-readspeed.patch - Richard Shooter's read speed-up code (included in ftdi_sio-1.3.3), but I've tidied up the source and moved some stuff around. I've bumped the version to 1.3.3a to distinguish it from the 1.3.3 that Bill previously sent. 2.4.21-ftdi_sio-p04-spinlockbug.patch - my patch to avoid copying user data with a spinlock held (and interrupts disabled). 2.4.21-ftdi_sio-p05-sealink.patch - Adds Sealevel vid/pids - based on a patch by Tuan Hoang but with less bloat. 2.4.21-ftdi_sio-p06-usbuirt.patch - David Norwood's patch for USB-UIRT device using a preset custom divisor. 2.4.21-ftdi_sio-p07-writepooltidy.patch - my patch to take account of write urb pool table entries that failed allocation, and to free the write urb and transfer buffer allocated by the usbserial driver. 2.4.21-ftdi_sio-p08-relais.patch - support for USB Relais pid, backported from 2.5.x. 2.4.21-ftdi_sio-p09-tira1.patch - half of Erik Nygren's patch to support Home Electronics' Tira-1 IR tranceiver using a preset custom divisor. 2.4.21-ftdi_sio-p10-forcebaud.patch - the other half of Erik Nygren's patch forces the baud rate setting to B38400 for USB-UIRT and Tira-1 devices and also forces RTS/CTS on for Tira-1. 2.4.21-ftdi_sio-p11-paranoid.patch - my patch to make sure pointers that fail paranoid checks are not dereferenced. 2.4.21-ftdi_sio-p12-versionbump.patch - my patch to bump the version. This is stepping on Bill's toes a little, but I think whatever ends up in the 2.4.22 kernel should be labelled version 1.3.4. I have a 2.5.x driver version as a work in progress containing most of the above changes. I just need to finish it off a little and maybe replace the write urb pool stuff with something resembling the changes in the Visor driver. ChangeSet@1.1003.2.7, 2003-06-27 11:59:24-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: use common CRC library Remove the speedtouch CRC library. With this change, the speedtch module is no longer a multi-part object, so fix that up too. ChangeSet@1.1003.2.6, 2003-06-27 11:58:51-07:00, grigouze@noos.fr [PATCH] USB: zaurus SL-C700 This is a patch for usbnet for working with Zaurus SL-C700. The productid is different from other Zaurus, so i add an entry for it :) ChangeSet@1.1003.10.14, 2003-06-27 15:56:12-03:00, lethal@linux-sh.org [PATCH] SH64 Merge Here's the patch for sh64! This adds arch/sh64 and include/asm-sh64. All changes are localized to these directories. Here's the diffstat output, nothing too terribly exciting: arch/sh64/Makefile | 91 + arch/sh64/boot/Makefile | 34 arch/sh64/boot/compressed/Makefile | 57 arch/sh64/boot/compressed/cache.c | 39 arch/sh64/boot/compressed/head.S | 164 ++ arch/sh64/boot/compressed/install.sh | 56 arch/sh64/boot/compressed/misc.c | 251 +++ arch/sh64/boot/compressed/vmlinux.lds.S | 65 arch/sh64/config.in | 298 ++++ arch/sh64/defconfig | 467 +++++++ arch/sh64/kernel/Makefile | 38 arch/sh64/kernel/entry.S | 2099 ++++++++++++++++++++++++++++++++ arch/sh64/kernel/fpu.c | 171 ++ arch/sh64/kernel/head.S | 347 +++++ arch/sh64/kernel/init_task.c | 36 arch/sh64/kernel/irq.c | 706 ++++++++++ arch/sh64/kernel/irq_intc.c | 269 ++++ arch/sh64/kernel/led.c | 69 + arch/sh64/kernel/pci-dma.c | 46 arch/sh64/kernel/pci_sh5.c | 607 +++++++++ arch/sh64/kernel/pci_sh5.h | 107 + arch/sh64/kernel/pcibios.c | 129 + arch/sh64/kernel/process.c | 903 +++++++++++++ arch/sh64/kernel/ptrace.c | 375 +++++ arch/sh64/kernel/semaphore.c | 137 ++ arch/sh64/kernel/setup.c | 362 +++++ arch/sh64/kernel/sh_ksyms.c | 78 + arch/sh64/kernel/signal.c | 821 ++++++++++++ arch/sh64/kernel/sys_sh64.c | 268 ++++ arch/sh64/kernel/time.c | 552 ++++++++ arch/sh64/kernel/traps.c | 263 ++++ arch/sh64/lib/Makefile | 26 arch/sh64/lib/c-checksum.c | 330 +++++ arch/sh64/lib/checksum.S | 656 ++++++++++ arch/sh64/lib/copy_user_memcpy.S | 205 +++ arch/sh64/lib/dbg.c | 319 ++++ arch/sh64/lib/io.c | 200 +++ arch/sh64/lib/memcpy.c | 82 + arch/sh64/lib/old-checksum.c | 17 arch/sh64/lib/page_clear.S | 46 arch/sh64/lib/page_copy.S | 77 + arch/sh64/lib/panic.c | 60 arch/sh64/lib/syscalltab.h | 311 ++++ arch/sh64/lib/udelay.c | 53 arch/sh64/mach-cayman/Makefile | 15 arch/sh64/mach-cayman/irq.c | 188 ++ arch/sh64/mach-cayman/led.c | 47 arch/sh64/mach-cayman/setup.c | 209 +++ arch/sh64/mach-harp/Makefile | 14 arch/sh64/mach-harp/setup.c | 139 ++ arch/sh64/mach-sim/Makefile | 14 arch/sh64/mach-sim/setup.c | 164 ++ arch/sh64/mm/Makefile | 43 arch/sh64/mm/cache.c | 1062 ++++++++++++++++ arch/sh64/mm/extable.c | 95 + arch/sh64/mm/fault.c | 716 ++++++++++ arch/sh64/mm/init.c | 203 +++ arch/sh64/mm/ioremap.c | 358 +++++ arch/sh64/mm/tlb.c | 166 ++ arch/sh64/mm/tlbmiss.c | 276 ++++ arch/sh64/vmlinux.lds.S | 158 ++ include/asm-sh64/a.out.h | 37 include/asm-sh64/atomic.h | 102 + include/asm-sh64/bitops.h | 364 +++++ include/asm-sh64/bugs.h | 38 include/asm-sh64/byteorder.h | 49 include/asm-sh64/cache.h | 139 ++ include/asm-sh64/cayman.h | 20 include/asm-sh64/checksum.h | 321 ++++ include/asm-sh64/current.h | 31 include/asm-sh64/delay.h | 11 include/asm-sh64/div64.h | 21 include/asm-sh64/dma.h | 38 include/asm-sh64/elf.h | 101 + include/asm-sh64/errno.h | 143 ++ include/asm-sh64/fcntl.h | 87 + include/asm-sh64/hardirq.h | 42 include/asm-sh64/hardware.h | 19 include/asm-sh64/hw_irq.h | 16 include/asm-sh64/init.h | 17 include/asm-sh64/io.h | 215 +++ include/asm-sh64/ioctl.h | 83 + include/asm-sh64/ioctls.h | 110 + include/asm-sh64/ipc.h | 42 include/asm-sh64/ipcbuf.h | 40 include/asm-sh64/irq.h | 142 ++ include/asm-sh64/keyboard.h | 74 + include/asm-sh64/linux_logo.h | 47 include/asm-sh64/mman.h | 49 include/asm-sh64/mmu.h | 7 include/asm-sh64/mmu_context.h | 209 +++ include/asm-sh64/module.h | 12 include/asm-sh64/msgbuf.h | 42 include/asm-sh64/namei.h | 24 include/asm-sh64/page.h | 129 + include/asm-sh64/param.h | 43 include/asm-sh64/pci.h | 230 +++ include/asm-sh64/pgalloc-3level.h | 78 + include/asm-sh64/pgalloc.h | 173 ++ include/asm-sh64/pgtable-3level.h | 152 ++ include/asm-sh64/pgtable.h | 336 +++++ include/asm-sh64/platform.h | 69 + include/asm-sh64/poll.h | 36 include/asm-sh64/posix_types.h | 128 + include/asm-sh64/processor.h | 273 ++++ include/asm-sh64/ptrace.h | 38 include/asm-sh64/registers.h | 199 +++ include/asm-sh64/resource.h | 47 include/asm-sh64/scatterlist.h | 36 include/asm-sh64/segment.h | 6 include/asm-sh64/semaphore-helper.h | 100 + include/asm-sh64/semaphore.h | 139 ++ include/asm-sh64/sembuf.h | 36 include/asm-sh64/serial.h | 33 include/asm-sh64/shmbuf.h | 53 include/asm-sh64/shmparam.h | 20 include/asm-sh64/sigcontext.h | 30 include/asm-sh64/siginfo.h | 233 +++ include/asm-sh64/signal.h | 184 ++ include/asm-sh64/smp.h | 15 include/asm-sh64/smplock.h | 77 + include/asm-sh64/socket.h | 64 include/asm-sh64/sockios.h | 24 include/asm-sh64/softirq.h | 30 include/asm-sh64/spinlock.h | 17 include/asm-sh64/stat.h | 88 + include/asm-sh64/statfs.h | 36 include/asm-sh64/string.h | 21 include/asm-sh64/system.h | 405 ++++++ include/asm-sh64/termbits.h | 183 ++ include/asm-sh64/termios.h | 118 + include/asm-sh64/timex.h | 36 include/asm-sh64/tlb.h | 95 + include/asm-sh64/types.h | 66 + include/asm-sh64/uaccess.h | 288 ++++ include/asm-sh64/ucontext.h | 23 include/asm-sh64/unaligned.h | 30 include/asm-sh64/unistd.h | 416 ++++++ include/asm-sh64/user.h | 71 + 139 files changed, 23750 insertions(+) and here's the patch.. TAG: MailDone ChangeSet@1.1003.2.5, 2003-06-27 11:54:20-07:00, greg@kroah.com Cset exclude: cweidema@indiana.edu|ChangeSet|20030620002017|05386 ChangeSet@1.1003.10.13, 2003-06-27 15:41:53-03:00, schwidefsky@de.ibm.com [PATCH] console semaphore fix. Avoid BUG if panic is called from an interrupt context. This patch has been accepted to linux-2.5. diffstat: kernel/printk.c | 9 ++++++++- 1 files changed, 8 insertions(+), 1 deletion(-) TAG: MailDone ChangeSet@1.1003.10.12, 2003-06-27 15:39:44-03:00, schwidefsky@de.ibm.com [PATCH] s390 defconfigs update New default configurations. diffstat: arch/s390/defconfig | 218 ++++++++++++++++++++++++++++++++++++++++++--------- arch/s390x/defconfig | 166 ++++++++++++++++++++++++++++++-------- 2 files changed, 312 insertions(+), 72 deletions(-) ChangeSet@1.1003.10.11, 2003-06-27 15:38:36-03:00, schwidefsky@de.ibm.com [PATCH] s390 iucv network driver. Changes for the iucv network driver: - Added mising call of release_param(). - Allow '$' in username. - Workaround for VM bug. - Don't rely on IUCV ipmsgtags. - Additional debug code. - Fix deadlock when starting more than 160 devices. diffstat: drivers/s390/net/iucv.c | 125 +++++++++++++++++++++++++++++++++++++-------- drivers/s390/net/iucv.h | 4 + drivers/s390/net/netiucv.c | 83 +++++++++++++++++++++++------ 3 files changed, 172 insertions(+), 40 deletions(-) ChangeSet@1.1003.10.10, 2003-06-27 15:34:42-03:00, schwidefsky@de.ibm.com [PATCH] s390 ctc network driver update Changes for the ctc network driver: - Fixed vary on/off issue. - Implemented restart after -EIO. - Changed severity of some warnings to debug. - Fixed physical link lost problems. diffstat: drivers/s390/net/ctcmain.c | 969 ++++++++++++++++++++++++++++++--------------- drivers/s390/net/ctctty.c | 106 +++- drivers/s390/net/fsm.h | 2 3 files changed, 735 insertions(+), 342 deletions(-) ChangeSet@1.1003.10.9, 2003-06-27 15:23:28-03:00, schwidefsky@de.ibm.com [PATCH] s390 3215 driver update Changes for the 3215 driver and new control character helper functions. diffstat: drivers/s390/char/con3215.c | 39 +++++++++++------- drivers/s390/char/ctrlchar.c | 92 +++++++++++++++++++------------------------ drivers/s390/char/ctrlchar.h | 12 ++++- 3 files changed, 75 insertions(+), 68 deletions(-) ChangeSet@1.1003.10.8, 2003-06-27 15:11:06-03:00, schwidefsky@de.ibm.com [PATCH] Add Configure.help entries for s390 options Update of s/390 specific documentation and help texts. diffstat: Documentation/Configure.help | 14 ++++++++++++++ Documentation/devices.txt | 12 ++++++------ 2 files changed, 20 insertions(+), 6 deletions(-) ChangeSet@1.1003.10.7, 2003-06-27 15:06:03-03:00, schwidefsky@de.ibm.com [PATCH] s390 documentation update S/390 Documentation changes: - Updated section about /proc/subchannels. - Added description for /proc/chpids. - Added descriptions for get_irq_first() / get_irq_next(). - Added description for read_conf_data(). - Added description for s390_request_irq_special(). - Typos and minor improvements. diffstat: Documentation/s390/CommonIO | 48 +++ Documentation/s390/Debugging390.txt | 8 Documentation/s390/cds.txt | 442 ++++++++++++++++++++++++++++-------- 3 files changed, 397 insertions(+), 101 deletions(-) ChangeSet@1.1003.10.6, 2003-06-27 15:03:29-03:00, schwidefsky@de.ibm.com [PATCH] s390 31 bit compat. 31 bit emulation changes: - Support for PER_LINUX32 personality added. - Added [un]register_ioctl32_conversion. - Don't do float/integer conversion in save/restore_sigregs32. - Add system call emulation for sys_readahead, sys_gettid, sys_tkill, sys_sysctl and sys_stime. - Add ioctl emulation for BLKBSZGET, BLKELVGET, BLKELVSET, BLKFLSBUF BLKFRAGET, BLKFRASET, BLKGETSIZE, BLKGETSIZE64, BLKPG, BLKRASET BLKROGET, BLKROSET, BLKSECTGET, BLKSECTSET, BLKSSZGET, LOOP_CLR_FD LOOP_GET_STATUS, LOOP_SET_FD, LOOP_SET_STATUS, RAW_GETBIND RAW_SETBIND and SIOCATMARK. - Signal backchain fix for 31 bit emulation signal handler. - Added missing check for SIGURG in emulation signal handler. - sys_msgsnd and sys_msgrcv emulation fixes. - Fix emulation for sys_getrlimit. - Add check for (ssize_t32) count < 0 in read and write system call emulation. - Check offset in pwrite system call emulation. diffstat: arch/s390x/kernel/Makefile | 2 arch/s390x/kernel/binfmt_elf32.c | 8 arch/s390x/kernel/entry.S | 12 - arch/s390x/kernel/exec_domain32.c | 30 +++ arch/s390x/kernel/ioctl32.c | 181 ++++++++++++++++++ arch/s390x/kernel/linux32.c | 364 +++++++++++++++++++++++++++++++++++--- arch/s390x/kernel/s390_ksyms.c | 14 + arch/s390x/kernel/signal32.c | 14 + arch/s390x/kernel/wrapper32.S | 24 ++ 9 files changed, 608 insertions(+), 41 deletions(-) ChangeSet@1.1003.10.5, 2003-06-27 14:55:53-03:00, schwidefsky@de.ibm.com [PATCH] s390 dasd driver update Big patch for the dasd driver. New features: - Add 'set on/off' to /proc/dasd/statistics. - Added BIODASDPSRD ioctl (Read Subsystem Performance Statistics). - Added BIODASDSATTR ioctl (Set Cache Attributes). - Support for XRC timestamping. - Add support for breaking the reservation of a dasd (boxed dasd access). - Implemented hotplug support for dasd. - Support for ESS dasd devices. Bug-Fixes: - Use internal timer instead of DOIO_TIMEOUT option (cio). - Use 'get_clock' instead of 'STCK....' - Switch off autodetect/probeonly as default behaviour. - Rework of dasd messages. - Reseve IOCTL-NR 240-255 for EMC - Fix statistics counting. - Clear request queue in dasd_disable_blkdev. - Fix for a race between dasd_format and sleep_on_req. - Fix for a race between reserve timeout and successful completion. - Set maximun end-cylinder to geometry cylinder -1 (in Define Extent). - Private implementation of BLKROSET. - Allow sharing of external interrupt 0x2603 between pfault and dasd diag. - Plug devices during bringup. - Get diag discipline to work again. - Check diag discipline forst on dynamic attach of a device. - EXPORT dasd_device_from_devno (needed by EMC) - Fix problem with ext3 doing modifications to the first request on the request queue. - Prevent scheduling in timer_bh. - Check for empty queue after state change pending interrupt. - Register a dasd only if the blocksize and the number of blocks are valid. - Check for spurious interrupts while waiting for basic sense data. - Check for unformatted dasd in BIODASDINFO ioctl. - Check for unformatted dasd in dasd_disable_blkdev. - Fix oops during boot if an invalid dasd= parameter has been specified. - Prevent dasd driver from creating /proc/partition names for scsi devices. - Return error for non-dasd-devices in dasd ioctls. - Add missing MODULE_LICENSE("GPL"). - Don't accept invalid device numbers in /proc/dasd/devices interface. - Add check in dasd_discipline_del if discipline to be removed has been added. - Remove static initializer from dasd_major_info and use proper list_for_each operations for dasd_major_info list. - Disable dasd diag for 64 bit. - Fix race condition on timer variable in dasd_device_t. - Fix timeout processing for reserve/release requests. - Fix race condition between setup and first use of request_queue. - make dasd_eckd compile with gcc 3.3. - Post requests with invalid blocksize with i/o-error. - Various fixes for ESS dasd device support. - Fixed path revalidation. - Retry i/o after path failure. - Retry i/o after interface control check. - Fix major&minor number for dynamically attached dasd devices. - Fix low memory handling. - Free spinlock in case of an error in dasd_device_from_devno. - Head queue diag discipline to give it a chance to grab is device before eckd. - Remove some warnings. - Fix reserve/release for 64 bit. - Add module licence to fba discipline. diffstat: drivers/s390/block/dasd.c | 3568 ++++++++++++++++++++++++------------- drivers/s390/block/dasd_3990_erp.c | 1347 ++++++------- drivers/s390/block/dasd_diag.c | 273 +- drivers/s390/block/dasd_diag.h | 20 drivers/s390/block/dasd_eckd.c | 729 +++++-- drivers/s390/block/dasd_eckd.h | 152 - drivers/s390/block/dasd_fba.c | 116 - drivers/s390/block/dasd_fba.h | 12 drivers/s390/block/dasd_int.h | 353 ++- include/asm-s390/ccwcache.h | 38 include/asm-s390/dasd.h | 187 + include/asm-s390/vtoc.h | 1 include/asm-s390x/ccwcache.h | 38 include/asm-s390x/dasd.h | 187 + include/asm-s390x/vtoc.h | 1 15 files changed, 4532 insertions(+), 2490 deletions(-) ChangeSet@1.1003.10.4, 2003-06-27 14:47:57-03:00, schwidefsky@de.ibm.com [PATCH] s390 common i/o layer fixes Common i/o fixes: - Don't confuse device drivers with zero sense data. - Check return code for the start of the basic sense ccw. - Fix deadlock in enable_cpu_sync_isc. - Fix check for path not operational condition. - Use unsigned long for flags variable in read_dev_chars. - Only try sense path group id on available paths. - Retry sense path group id on another path after deferred cc=3. - Fix deadlock in link incident handler. - Fix output in /proc/chpids. - Adapt to latest path group algorithm. - Fix handling of condition code 1 (status pending) on i/o operations. - Only process status pending conditions when doing sync. i/o. - Path revalidation fixes: * Add a handler for machine checks with chpid sources. * Distinguish between device gone and device not accessible in the device not operational handler. * Don't accept i/o from the device drivers while doing path revalidation. * Use a bottom half for doing path verification from interrupt context. * Don't do path verification if sync. isc is alread in use. Reschedule bh. * Kill pending i/o before doing path verification. * Always start with a logical path mask of 0xff because the information store by stsch() can be outdated. - Always use a tpi loop for basic sense. - Lowered level of 'not operational' messages. - Check after store event information if there are more crws pending. - Make diag210 a non-inline function to avoid problems with modules loaded above 2G. - Show reserved devices as "boxed" instead of "unkown". - Fix for path no operational condition in cio_start. - Fix /proc/cio_ignore string parsing. - Fix parsing of verbs in chandev. diffstat: drivers/s390/misc/chandev.c | 2 drivers/s390/s390io.c | 2422 +++++++++++++++++++++++++------------------- drivers/s390/s390mach.c | 14 include/asm-s390/irq.h | 43 include/asm-s390/s390io.h | 14 include/asm-s390/setup.h | 1 include/asm-s390x/irq.h | 44 include/asm-s390x/s390io.h | 16 include/asm-s390x/setup.h | 1 9 files changed, 1494 insertions(+), 1063 deletions(-) ChangeSet@1.1003.10.3, 2003-06-27 14:33:40-03:00, schwidefsky@de.ibm.com [PATCH] s390 base update s/390 base fixes: - docu: Correct description of 3270 device nodes - arch: Do reate_proc_entry for debug feature outside spin locked code. - arch: Set CR5 to get program checks for space switching instructions. - arch: Use sig_exit in 64 bit signal handler. - arch: Avoid warning in idals.h. - arch: Do pfix early in the boot process. No pfix for 64 bit. - arch: Fix linker output format for 64 bit kernels. - arch: Fix race condition in dirty bit clearing. - arch: Fix deadlock in pgd_populate. diffstat: arch/s390/kernel/debug.c | 20 ++++-- arch/s390/kernel/head.S | 2 arch/s390/kernel/s390_ksyms.c | 5 + arch/s390/kernel/setup.c | 133 ++++++++++++++++++++++++++++++++++++++++- arch/s390/mm/init.c | 5 - arch/s390x/kernel/debug.c | 20 ++++-- arch/s390x/kernel/head.S | 17 ----- arch/s390x/kernel/s390_ksyms.c | 2 arch/s390x/kernel/setup.c | 6 - arch/s390x/kernel/signal.c | 5 - arch/s390x/mm/init.c | 74 ++++++++++++++++------ arch/s390x/vmlinux-shared.lds | 2 arch/s390x/vmlinux.lds | 2 include/asm-s390/idals.h | 4 - include/asm-s390/page.h | 61 +----------------- include/asm-s390/pgtable.h | 36 ++--------- include/asm-s390x/idals.h | 4 - include/asm-s390x/page.h | 61 +----------------- include/asm-s390x/pgtable.h | 36 ++--------- include/asm-s390x/setup.h | 2 include/linux/mm.h | 10 ++- 21 files changed, 261 insertions(+), 246 deletions(-) ChangeSet@1.1003.10.2, 2003-06-27 11:53:41-03:00, bcollins@debian.org [PATCH] Update IEEE1394 (r972) IEEE1394 : Add OUI database. DV1394 : Fix endian conversion brokeness. ETH1394 : Updates for async streams, EUI based ARP and packet fragmentation. IEEE1394 : Host key lookup improvements. SBP2 : Fix > S400 max_payload setting. IEEE1394 : Move hotplug declerations around to more generic place. IEEE1394 : Fix possible memory leak in ISO code. IEEE1394 : Fix proc output for > page size. OHCI1394 : Async stream packets. OHCI1394 : Trivial CONFIG_PM support. SBP2 : Only allocate scsi_host for ieee1394_hosts that have sbp2 devices attached to it (on demand scsi_host allocation). SBP2 : Code cleanups to bring closer to 2.5 code. SBP2 : Handle Logical_Unit_Number entries. VIDEO1394: Handle user pointer correctly. IEEE1394 : Macro namespace cleanups. ALL : Cleanups of some C constructs. ETH1394 : Limited multicast support. TAG: MailDone ChangeSet@1.1003.10.1, 2003-06-27 11:48:29-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/davem/crypto-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1003.9.13, 2003-06-27 12:29:08+02:00, trond.myklebust@fys.uio.no Replace buggy version of xdr_shift_buf() with the version from 2.5.x. This has the added bonus that we also get rid of the need for doing kmap() of multiple pages at the same time. ChangeSet@1.1003.9.12, 2003-06-27 08:23:39+02:00, trond.myklebust@fys.uio.no Ensure that the lockd clients always use one of the reserved ports. ChangeSet@1.1003.9.11, 2003-06-27 08:23:05+02:00, trond.myklebust@fys.uio.no Fix a TCP client corruption problem affecting resent requests. ChangeSet@1.1003.9.10, 2003-06-27 08:22:41+02:00, trond.myklebust@fys.uio.no Ensure that if we need to reconnect the socket, we also resend the entire message. Assorted TCP reconnection fixes. Temporarily raise the necessary CAP_NET_BIND_SERVICE capability if we need to bind the socket to a reserved port during a TCP reconnection. Check for CAP_NET_BIND_SERVICE at mount time. ChangeSet@1.1003.9.9, 2003-06-27 08:20:43+02:00, trond.myklebust@fys.uio.no Don't use an RPC child process when reconnecting to a TCP server. This is more efficient, and also fixes an existing deadlock situation in which the child could be waiting for an xprt_write_lock that was being held by the parent. ChangeSet@1.1003.9.8, 2003-06-27 08:18:15+02:00, trond.myklebust@fys.uio.no Fix a race: Ensure that requests retry if the remote server disconnects us while we're inside xprt_transmit(). ChangeSet@1.1003.9.7, 2003-06-27 08:17:37+02:00, trond.myklebust@fys.uio.no Add standard spinlocks to protect the socket from being released by one CPU while the other is in a soft interrupt. ChangeSet@1.1003.9.6, 2003-06-27 08:16:47+02:00, trond.myklebust@fys.uio.no A patch to ensures that blocks which are not going to time out are placed last on the nlm_block list (problem reported by Olaf Kirch). ChangeSet@1.1003.9.5, 2003-06-27 08:16:05+02:00, trond.myklebust@fys.uio.no A patch by Patrice Dumas to add a check in order to ensure that we really were requesting a blocking lock when we get a reply from the server asking us to block. ChangeSet@1.1003.9.4, 2003-06-27 08:14:43+02:00, trond.myklebust@fys.uio.no A patch by Patrice Dumas to implement nlmsvc_proc_granted_res. When a server receives that callback it deallocates the corresponding blocked lock, using the nlmsvc_grant_reply function. ChangeSet@1.1003.9.3, 2003-06-27 08:11:52+02:00, trond.myklebust@fys.uio.no Another patch by Chuck Lever that ensures that the PG_uptodate bit gets set when the entire page gets written by nfs_writepage_sync() ChangeSet@1.1003.9.2, 2003-06-27 08:10:58+02:00, trond.myklebust@fys.uio.no A patch by Chuck Lever with further cleanups of the RPC socket slot allocation code. ChangeSet@1.1003.9.1, 2003-06-27 08:09:02+02:00, trond.myklebust@fys.uio.no A patch by Chuck Lever that cleans up the RPC socket slot allocation code. ChangeSet@1.1003.5.22, 2003-06-26 23:03:09-07:00, chas@cmf.nrl.navy.mil [ATM]: make clip buildable as a module ChangeSet@1.1003.5.21, 2003-06-26 23:02:22-07:00, chas@cmf.nrl.navy.mil [ATM]: eliminate ATM_PDU_OVHD, ops->free_rx_skb and ops->alloc_tx ChangeSet@1.1003.5.20, 2003-06-26 23:01:22-07:00, chas@cmf.nrl.navy.mil [ATM]: make sub skb->cb is clear before upcall to network ChangeSet@1.1003.5.19, 2003-06-26 23:00:26-07:00, chas@cmf.nrl.navy.mil [ATM]: lock neighbor entry during update in clip.c ChangeSet@1.1003.5.18, 2003-06-26 22:59:39-07:00, chas@cmd.nrl.navy.mil [ATM]: remove iovcnt member in struct atm_skb ChangeSet@1.1003.5.17, 2003-06-26 22:58:51-07:00, chas@cmf.nrl.navy.mil [ATM]: ixmicro puts esi in different location ChangeSet@1.1003.5.16, 2003-06-26 22:48:20-07:00, bdschuym@pandora.be [NETFILTER]: Add arptables mangle module. ChangeSet@1.1003.1.10, 2003-06-27 01:25:19-04:00, dave@thedillows.org Merge ssh://typhoon@typhoon.bkbits.net/typhoon-2.4 into thedillows.org:/home/il1/projects/typhoon/bk/typhoon-2.4 ChangeSet@1.1003.7.1, 2003-06-27 00:00:57-04:00, dave@thedillows.org Fix misreporting of card type and spurious "already scheduled" messages. ChangeSet@1.1003.1.9, 2003-06-26 20:32:08-07:00, typhoon.adm@hostme.bitkeeper.com Merge bk://linux.bkbits.net/linux-2.4 into hostme.bitkeeper.com:/repos/t/typhoon/typhoon-2.4 ChangeSet@1.1003.1.8, 2003-06-26 18:25:43-03:00, trini@kernel.crashing.org [PATCH] Add /proc/sys/kernel/l3cr Hello. The following patch is from Mark Greer. This adds read-only support for the L3 cache register found in the MPC745x line of CPUs. ===== arch/ppc/kernel/ppc_htab.c 1.7 vs edited ===== TAG: MailDone TAG: v2.4.22-pre2 ChangeSet@1.1003.1.7, 2003-06-26 18:24:02-03:00, rddunlap@osdl.org [PATCH] unexpected IO-APIC code update Hi, Recently there has been a rash of Unexpected IO APIC reports on the linux-smp mailing list. Most of the most recent ones are due to some newer Intel chipsets (865, 875). I have an patch that addresses these chipsets. It has been tested by a few people with good results and has been blessed by Maciej Rozycki. Other than conditionally decoding IO APIC registers 2 and 3, we could alternately ignore them since Linux doesn't use the values for anything other than printing them. This patch ignores IO APIC register 2 if it's the same value as IO APIC register 1. It also reads IO APIC register 3 if the IO APIC version is >= 0x20, but some chipsets don't support this register, so it is also ignored if it's value if the same as IO APIC register 1 or 2. The IO APIC Version register doesn't indicate the differences in these IO APICs. Patch for 2.4.22-pre1 is below. Please apply. -- ~Randy ~ http://developer.osdl.org/rddunlap/ ~ http://www.xenotime.net/linux/ ~ patch_name: ioapic_update_2422.patch patch_version: 2003-06-24.15:25:38 author: Randy.Dunlap description: support newer Intel chipset IO APICs; product: Linux product_versions: linux-2422-pre1 maintainer: Maciej W. Rozycki diffstat: = arch/i386/kernel/io_apic.c | 23 ++++++++++++++++++++++- include/asm-i386/io_apic.h | 5 +++++ 2 files changed, 27 insertions(+), 1 deletion(-) ChangeSet@1.1003.6.33, 2003-06-26 18:18:59-03:00, marcelo@freak.distro.conectiva Merge bk://linux-bt.bkbits.net/bt-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 ChangeSet@1.1003.6.32, 2003-06-26 18:16:39-03:00, green@linuxhacker.ru [PATCH] current bk ipmi build fix Hello! Not that I really have the hardware, but it breaks my "allyesconfig" build. So here is this compile fix for ipmi driver in current 2.4 bk tree. (I see that Alan have some similarly named fix in his tree and actually there is whole new version of the driver on the net somewhere, but it is unclear when it is planned to be pushed to 2.4 tree, so I'd better post this now ;) ). Bye, Oleg ===== drivers/char/ipmi/ipmi_kcs_intf.c 1.3 vs edited ===== ChangeSet@1.1003.6.31, 2003-06-26 16:29:49-03:00, rusty@rustcorp.com.au [PATCH] reorganize for unreachable code (Included in 2.5) From: Scott Russell - moved return to eliminate unreachable code reported by kbugs.org ChangeSet@1.1003.6.30, 2003-06-26 16:24:53-03:00, marcelo@freak.distro.conectiva Merge bk://linux-scsi.bkbits.net/scsi-misc-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 ChangeSet@1.1003.6.29, 2003-06-26 15:47:43-03:00, marcelo@freak.distro.conectiva Merge TAG: MailDone ChangeSet@1.1003.6.28, 2003-06-26 15:46:13-03:00, marcelo@freak.distro.conectiva Cset exclude: jamagallon@able.es|ChangeSet|20030620200318|50799 ChangeSet@1.1003.6.27, 2003-06-26 14:31:24-03:00, hugh@veritas.com [PATCH] remove unsafe BUG() in __remove_inode_page() PageDirty BUG in __remove_inode_page is, and always has been, unsafe for SMP: truncation may be racing against unmapping's set_page_dirty in __free_pte (amongst a few other possibilities). ChangeSet@1.1003.6.26, 2003-06-26 12:48:23-03:00, olh@suse.de [PATCH] RAID_AUTORUN is a compatible ioctl [ The following text is in the "utf-8" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] The RAID_AUTORUN ioctl can be handled as compatible. ChangeSet@1.1003.6.25, 2003-06-26 12:48:03-03:00, olh@suse.de [PATCH] remove TIOCGDEV from asm/ioctls.h [ The following text is in the "utf-8" character set. ] [ Your display is set for the "ISO-8859-1" character set. ] [ Some characters may be displayed incorrectly. ] The TIOCGDEV ioctl is SuSE specific, the bootscripts can do better logging when they know what the real console device is. No other distribution uses it, so it must not be in asm-ppc64/ioctls.h Please remove it. ChangeSet@1.1003.6.24, 2003-06-26 10:57:28-03:00, ink@jurassic.park.msu.ru [PATCH] alpha: nautilus poweroff This makes soft power-off work on UP1100 and UP1500. Sync up with 2.5. Ivan. ChangeSet@1.1003.6.23, 2003-06-26 10:56:52-03:00, ink@jurassic.park.msu.ru [PATCH] alpha: initrd fix (Wiedemeier, Jeff) This's clear bugfix... Sync up with 2.5. Ivan. On Mon, Apr 14, 2003 at 07:46:18AM -0400, Wiedemeier, Jeff wrote: > While testing our upcoming kernel update for 7.2 alpha, I've encountered > a problem with move_initrd. It allocates a page-aligned chunk to move > the initrd into, but it doesn't allocate the entire last > page. Subsequent bootmem allocations can then be filled from the last > page used be the initrd. This then becomes a problem when the initrd > memory is released. > > /jeff ChangeSet@1.1003.6.22, 2003-06-26 10:51:29-03:00, ink@jurassic.park.msu.ru [PATCH] alpha: Lynx platform support (Jay Estabrook) This adds support for AlphaServer 2100A-based systems. Sync up with 2.5. Ivan. On Sun, Apr 06, 2003 at 10:34:48PM -0400, Jay Estabrook wrote: > Here's one I *finally* got around to, after interminable axp-list and > debian-alpha grumbling... ;-} > > Also, apparently a not insignificant number of these machines have > finally made it to "legacy" status, ie been retired from active > NT/VMS/OSF use... :-) > > --Jay++ ChangeSet@1.1003.5.15, 2003-06-25 23:49:03-07:00, chas@cmf.nrl.navy.mil [ATM]: Backport HE driver fixes from 2.5.x ChangeSet@1.1003.5.14, 2003-06-25 23:20:45-07:00, bdschuym@pandora.be [NETFILTER]: Missing return in arp_packet_match(). ChangeSet@1.1003.5.13, 2003-06-25 21:58:33-07:00, romieu@fr.zoreil.com [NETFILTER]: Fix leaks in error paths of ip_recent_ctrl. ChangeSet@1.930.144.11, 2003-06-25 18:34:16-05:00, jejb@raven.il.steeleye.com Backport from 2.5: scsi allow devices to restrict start on add From: Eddie.Williams@steeleye.com When a SCSI disk is added and it returns a NOT READY the SD driver is automatically sending a START_UNIT command to spin the device up. While this may be the desired behavior for many if not most devices not all devices either want or need this. The attached patch provides a mechanism via the device_list that allows a device to be defined to disable the automatic start being issued on an add. The patch also modifies the device_list for several devices that would prefer to not have the start command issued. ChangeSet@1.930.144.10, 2003-06-25 17:55:39-05:00, jejb@raven.il.steeleye.com sd.c: Backport wild spin loop mitigation from 2.5 This problem was reported against 2.4 by Eddie.Williams@SteelEye.com There's a problem in the sd spinup code in that if the unit returns NOT READY, we begin to spin it up, but thereafter if it returns anything other than NOT READY or success, the while loop in the spinup code will be executed *without* the 1s delay that's in the NOT READY case. The problem was seen with a real device: Compaq multi-path storage arrays return NOT READY to probes down inactive paths, but when the start unit is sent to activate the path, they can then respond back with error conditions. The fix is to terminate the while loop for any unexpected return. ChangeSet@1.1003.6.21, 2003-06-25 18:19:39-03:00, ak@muc.de [PATCH] Don't enable I2O for AMD64 I2O isn't 64bit clean and doesn't work. Disable it in the configuration to prevent user mistakes. TAG: MailDone ChangeSet@1.1003.6.20, 2003-06-25 18:19:30-03:00, ak@muc.de [PATCH] ACPI compile fixes for 2.4.22pre1 Without this ACPI doesn't compile on AMD64 on 2.4.22-pre1 It fixes a mismatched prototype. Andrew Grover stated that this is the way he wants it to be fixed. Also add an missing linux/init.h include in ACPI. ChangeSet@1.1003.6.19, 2003-06-25 17:51:58-03:00, rusty@rustcorp.com.au [PATCH] misc register fix on ds1286 (Included in 2.5) From: Michael Still This patch corrects an error found whilst auditing the use of misc_register in 2.5.44. misc_register returns 0 on success, and a negative number on failure... TAG: MailDone ChangeSet@1.1003.6.18, 2003-06-25 17:49:31-03:00, rusty@rustcorp.com.au [PATCH] misc_register audit fix of wdt_pci From: Michael Still This patch corrects an error found whilst auditing the use of misc_register in 2.5.44. misc_register returns 0 on success, and a negative number on failure... ChangeSet@1.1003.6.17, 2003-06-25 17:45:03-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_pci.txt (Included in 2.5) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/pci.txt more readible by fixing the linewrap. ChangeSet@1.1003.6.16, 2003-06-25 17:44:33-03:00, rusty@rustcorp.com.au [PATCH] set b_page to null in fake buffer_head for O_DIRECT From: Christoph Hellwig currently it contains garbage, but knowing what kind of get_block() this is is needed at least for some versions of reiserfs O_DIRECT. ChangeSet@1.1003.6.15, 2003-06-25 17:43:07-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_filesystems_sysv-fs.txt (Included in 2.5) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/filesystems/sysv-fs.txt more readible by fixing the linewrap. ChangeSet@1.1003.6.14, 2003-06-25 17:38:30-03:00, rusty@rustcorp.com.au [PATCH] Squash warning in ppc64 addnote tool (Included in 2.5) From: David Gibson Anton, please apply. addnote in arch/ppc64/boot (a userspace tool, not kernel code) uses exit() without including stdlib.h. ChangeSet@1.930.144.9, 2003-06-25 15:36:23-05:00, tonyb@cybernetics.com [PATCH] make sym53c8xx_2 not reject autosense IWR This patch against recent 2.4.x makes sym53c8xx_2 silently ignore the Ignore Wide Residue message on autosense commands rather than rejecting it. This makes the SCSI communications cleaner for targets that return an odd number of sense bytes. ChangeSet@1.1003.6.13, 2003-06-25 17:24:40-03:00, rusty@rustcorp.com.au [PATCH] remove superflous if in wait_kio From: Christoph Hellwig wait_on_buffer is declared as: static inline void wait_on_buffer(struct buffer_head * bh) { if (test_bit(BH_Lock, &bh->b_state)) __wait_on_buffer(bh); } so the buffer_locked() that is nothing but test_bit(BH_Lock, &bh->b_state)) is superflous. ChangeSet@1.1003.6.12, 2003-06-25 17:24:03-03:00, rusty@rustcorp.com.au [PATCH] add some missing init.h inclusions From: Christoph Hellwig unfortunately i386 gets it implicitly through some asm/* headers, but many other ports don't. ChangeSet@1.1003.6.11, 2003-06-25 17:16:30-03:00, rusty@rustcorp.com.au [PATCH] redundant declarations (#1_15) (Included in 2.5) From: dent@cosy.sbg.ac.at (Thomas Mirlacher) hi rusty, this patch fixes redundant declarations in 2.5.24 (same as sent yesterday, but this time automacially splitted into several mails) ------------------------- BEGIN PATCH ------------------------- ChangeSet@1.1003.6.10, 2003-06-25 17:15:26-03:00, rusty@rustcorp.com.au [PATCH] [TRIVIAL PATCH] include_asm-i386_dma.h: wrong lowest DMA (Included in 2.5) From: Uros Bizjak ChangeSet@1.1003.6.9, 2003-06-25 17:14:04-03:00, rusty@rustcorp.com.au [PATCH] Typos in drivers_s390_net_iucv.h (Included in 2.5) From: James Mayer ChangeSet@1.1003.6.8, 2003-06-25 17:12:46-03:00, rusty@rustcorp.com.au [PATCH] include_asm-ia64_sal.h, typo: the the (Included in 2.5) From: James Mayer ChangeSet@1.1003.6.7, 2003-06-25 17:12:01-03:00, rusty@rustcorp.com.au [PATCH] fix linewrap in Documentation_power_pci.txt (Included in 2.5) From: ookhoi@humilis.net Hi, With this patch I tried to make Documentation/arm/SA1100/CERF more readible by fixing the linewrap. ChangeSet@1.1003.6.6, 2003-06-25 17:11:54-03:00, rusty@rustcorp.com.au [PATCH] missing headers in i82092.c From: Christoph Hellwig again i386 seems to get them implicitly from somewhere, but at least alpha doesn't. ChangeSet@1.1003.6.5, 2003-06-25 17:06:40-03:00, rusty@rustcorp.com.au [PATCH] fix documentation in include_asm-i386_bitops.h [ Documentation changes only. ] (Included in 2.5) From: "Vitezslav Samel" Whean I was searching for prototype for set_bit() I found IMHO wrong doc entries in include/asm-i386/bitops.h. Please consider applying. Cheers, Vita ChangeSet@1.1003.6.4, 2003-06-25 17:03:49-03:00, rusty@rustcorp.com.au [PATCH] [TRIVIAL PATCH 2.4] update README file to current From: Petri Koistinen Hi! I forgot to CC: patch below to you. By the way, page is extremely hard to find for some reason. Could you add link to that page from too? E-mail address would be nice too. =) Best regards, Petri Koistinen ---------- Forwarded message ---------- Date: Wed, 4 Jun 2003 01:20:39 +0300 (EEST) From: Petri Koistinen To: Marcelo Tosatti Cc: linux-kernel@vger.kernel.org Subject: [TRIVIAL PATCH 2.4] update README file to current realities Hello! Similar little clarification patch got accepted in 2.5.63 and I think this could be helpful in 2.4.x kernel tree too. Best regards, Petri Koistinen ChangeSet@1.1003.6.3, 2003-06-25 16:52:48-03:00, rusty@rustcorp.com.au [PATCH] fs_autofs4_root.c unused variable (Included in 2.5) (OK from maintainer trivial/3673) From: Rusty Russell Hi HPA! Trivial unused var... TAG: MailDone ChangeSet@1.1003.6.2, 2003-06-25 16:52:16-03:00, rusty@rustcorp.com.au [PATCH] 2.4.20 wait.h doc typo From: Martin Pool ChangeSet@1.1003.6.1, 2003-06-25 16:50:54-03:00, rusty@rustcorp.com.au [PATCH] improve signal-to-noise ratio in atm code (Included in 2.5) From: Paul P Komkoff Jr ChangeSet@1.930.1.244, 2003-06-25 13:20:41-06:00, garyhade@us.ibm.com ia64: fix sysinfo(2) memory value truncation for 32-bit apps ChangeSet@1.930.1.243, 2003-06-25 09:32:58-06:00, bjorn_helgaas@hp.com Cset exclude: rohit.seth@intel.com[helgaas]|ChangeSet|20030623203306|58862 ChangeSet@1.1003.5.12, 2003-06-24 15:28:28-07:00, laforge@netfilter.org [NETFILTER]: Fix conntrack master_ct refcounting. ChangeSet@1.1003.5.11, 2003-06-24 15:26:47-07:00, sfrost@snowman.net [NETFILTER]: Add iptables "recent" module. ChangeSet@1.1003.5.10, 2003-06-24 15:21:50-07:00, laforge@netfilter.org [NETFILTER]: ip{,6}tables enhancement, add new /proc/net files. ChangeSet@1.1003.5.9, 2003-06-24 15:20:06-07:00, laforge@netfilter.org [NETFILTER]: Cosmetic changes. - The 04_newnat-udp-helper patch submitted to 2.4.20 introduced an unused variable. This patch gets rid of it. - Removes dead code in case CONFIG_SYSCTL is not defined - Change order of NFC_ defines in netfilter.h to be more logical. Add comment that <= 0x2000 are already used - Removes some old outdated comments - Removes an obslete check (master_ct() checks for ct->master, no need for the caller to do it before). - Remove some unneeded initializiations / memsets - Clean up some ip_conntrack_core code (use some common macro instead of reimplementing list iteration again). ChangeSet@1.1003.5.8, 2003-06-24 15:17:43-07:00, kaber@trash.net [NETFILTER]: Dont call helpers expectfn() for unconfirmed connections. Do not delete timer / call expect function for connections matching expectations from unconfirmed connections. ChangeSet@1.1003.5.7, 2003-06-24 15:16:24-07:00, qboosh@pld.org.pl [NETFILTER]: Fix endianness bugs in ipt_nat. This patch fixes bugs related to machine endianess (harmless on little-endians, but causing that some of conntrack modules dont work on big-endian machines) in netfilter/ip_nat code. Bugs are similar to those fixed in 42_conntrack-bigendian-fixes.patch (which fixed netfilter/ip_conntrack code). ChangeSet@1.1003.5.6, 2003-06-24 15:15:05-07:00, qboosh@pld.org.pl [NETFILTER]: Fix endianness bugs in conntrack. This patch fixes bugs related to machine endianess (harmless on little-endians, but causing that some of conntrack modules dont work on big-endian machines). ChangeSet@1.1003.5.5, 2003-06-24 15:13:43-07:00, qboosh@pld.org.pl [NETFILTER]: Fix ip6tables alignment (64bit archs). The calculated padding used for alignment in struct ip6t_table_info type definition is wrong. char[SMP_ALIGN((NF_IP6_NUMHOOKS*2+2)*sizeof(unsigned int))] has _its size_ (not end address) aligned to cache line size; also note that there are 3 ints, not 2 before arrays. And if whole ip6t_table_info started on address aligned to 8, entries were placed on address not aligned to 8. On some machines (at least ppc - IBM RS/6000) __alignof__(struct ip6t_entry) is 8 - so loading any ip6table_* module failed with "Bad offset". Probably padding was meant to be something like char padding[SMP_ALIGN((NF_IP6_NUMHOOKS*2+3)*sizeof(unsigned int)) - ((NF_IP6_NUMHOOKS*2+3)*sizeof(unsigned int))] but its simpler to use gccs aligned attribute through ____cacheline_aligned macro (just like in ipt_table_info struct in ip_tables module for ipv4/netfilter). ChangeSet@1.1003.5.4, 2003-06-24 15:11:37-07:00, mort@wildopensource.com [NETFILTER]: Fix processor shifts in lockhelp.h There are a bunch of (1< UNINITIALISED o [CORRECT] Accept data from TTY before link initialisation This seems necessary to avoid chat (via pppd) dropping chars o [CRITICA] Remember allocated skb size to avoid to over-write it o [FEATURE] Remove LM-IAS object once connected o [CORRECT] Avoid declaring link ready when it's not true ChangeSet@1.1002.1.6, 2003-06-24 02:41:29-03:00, jt@bougret.hpl.hp.com [PATCH] irda-usb driver fixes Hi Marcelo, Various fixes for the irda-usb driver. o [FEATURE] Update various comments to current state o [CORRECT] Handle properly failure of URB with new speed o [CORRECT] Don't test for (self != NULL) after using it (doh !) o [FEATURE] Other minor cleanups o [CORRECT] Add ID for new USB device (thanks to Sami Kyostila) o [CORRECT] Fix for big endian platforms (thanks to Jacek Jakubowski) ChangeSet@1.1002.1.5, 2003-06-24 02:39:57-03:00, jt@bougret.hpl.hp.com [PATCH] irda: Mask C/R bit from connection Hi Marcelo, Some devices add bits where they should not. Let's not get ourselves confused by it. Please apply ;-) Jean ir241_caddr_mask.diff : ChangeSet@1.1002.1.4, 2003-06-24 02:37:50-03:00, jt@bougret.hpl.hp.com [PATCH] irda: Export CRC routine to drivers Hi Marcelo, This export avoid users to duplicate this code. Driver fixes will come later on. Please apply ;-) Jean ir241_export_crc-3.diff : ChangeSet@1.1002.1.3, 2003-06-24 02:35:43-03:00, jt@bougret.hpl.hp.com [PATCH] irda: static init fixes Hi Marcelo, This make the static initialisation of some IrDA driver a bit less broken. Please apply ;-) Jean ir241_static_init.diff : ChangeSet@1.1002.1.2, 2003-06-24 01:55:56-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/davem/net-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1002.5.8, 2003-06-23 19:30:16-04:00, rusty@rustcorp.com.au [PATCH] namespace pollution in skfddi driver (Included in 2.5) From: Arnd Bergmann The skfp driver has a global function named 'set_int'. This makes it static to avoid namespace pollution. ===== drivers/net/skfp/fplustm.c 1.2 vs edited ===== ChangeSet@1.1002.5.7, 2003-06-23 19:30:05-04:00, rusty@rustcorp.com.au [PATCH] Remove naked GFP_DMA from drivers_net_macmace.c (Included in 2.5) From: Matthew Wilcox Can use GFP_KERNEL since this is a netdevice ->open routine. ChangeSet@1.1002.5.6, 2003-06-23 19:29:56-04:00, rusty@rustcorp.com.au [PATCH] [patch, 2.5] fix errorpath in apne.c (Included in 2.5) From: Marcus Alanen ===== drivers/net/apne.c 1.4 vs edited ===== ChangeSet@1.1002.5.5, 2003-06-23 19:29:46-04:00, rusty@rustcorp.com.au [PATCH] Clear up GFP confusion in rcpci45.c (Included in 2.5) From: Matthew Wilcox - Move PCI ID definitions to pci_ids.h - The GFP_DMA in rcpci45_init_one should be GFP_KERNEL because it's a pci_driver ->probe method, so it can sleep. - The GFP_DMA in RC_allocate_and_post_buffers should be GFP_ATOMIC because it's called from a timer function, so it must not sleep. ChangeSet@1.1002.5.4, 2003-06-23 19:29:36-04:00, rusty@rustcorp.com.au [PATCH] [2.4 patch] fix wavelan_cs compile warning [ Trivial, removes compile warning --RR ] From: Adrian Bunk I saw the following compile warning in 2.4.21-rc3: <-- snip --> ... make[3]: Entering directory `/home/bunk/linux/kernel-2.4/linux-2.4.21-rc3-modular/drivers/net/pcmcia' ... gcc -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.21-rc3-modular/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march=k6 -DMODULE -DMODVERSIONS -include /home/bunk/linux/kernel-2.4/linux-2.4.21-rc3-modular/include/linux/modversions.h -nostdinc -iwithprefix include -DKBUILD_BASENAME=wavelan_cs -c -o wavelan_cs.o wavelan_cs.c In file included from wavelan_cs.c:67: wavelan_cs.h:492:33: warning: extra tokens at end of #undef directive ... <-- snip --> The fix is trivial: ChangeSet@1.1002.5.3, 2003-06-23 19:29:24-04:00, rusty@rustcorp.com.au [PATCH] namespace pollution in cosa driver (Included in 2.5) From: Arnd Bergmann Variables named 'io' and 'dma' should not be global ===== drivers/net/wan/cosa.c 1.15 vs edited ===== ChangeSet@1.1002.5.2, 2003-06-23 19:29:14-04:00, rusty@rustcorp.com.au [PATCH] [patch, 2.5] dgrs doesn't free on error path (Included in 2.5) From: Marcus Alanen No Status Update. ChangeSet@1.1002.5.1, 2003-06-23 19:21:35-04:00, jgarzik@redhat.com Merge redhat.com:/garz/repo/marcelo-2.4 into redhat.com:/garz/repo/net-drivers-2.4 ChangeSet@1.930.1.242, 2003-06-23 16:52:06-06:00, davidm@tiger.hpl.hp.com ia64: Fixups for GCC v3.3. ChangeSet@1.1002.3.3, 2003-06-23 15:26:22-07:00, dlstevens@us.ibm.com [IPV{4,6}]: Fix "slow multicast on 2.5.69" bug. ChangeSet@1.1004, 2003-06-23 16:45:51-05:00, shaggy@shaggy.austin.ibm.com JFS: Possible trap/data loss when fixing directory index table ChangeSet@1.1002.4.68, 2003-06-23 14:01:01-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: Make sure crypto config is before lib config on ia64. ChangeSet@1.1002.4.67, 2003-06-23 14:00:52-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: Emulate module_name semantics correctly to avoid OOPS. ChangeSet@1.1002.4.66, 2003-06-23 14:00:43-07:00, mk@linux-ipv6.org [CRYPTO]: Update deflate dependencies. ChangeSet@1.1002.4.65, 2003-06-23 14:00:34-07:00, akpm@digeo.com [CRYPTO]: Fix memcpy/memset args. ChangeSet@1.1002.4.64, 2003-06-23 14:00:25-07:00, jmorris@intercode.com.au [CRYPTO]: Fix config dependencies. ChangeSet@1.1002.4.63, 2003-06-23 14:00:16-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: Fix condition typos in crypto/Config.in ChangeSet@1.1002.4.62, 2003-06-23 14:00:07-07:00, davem@nuts.ninka.net [CRYPTO]: deflate.c needs slab.h ChangeSet@1.1002.4.61, 2003-06-23 13:59:58-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: const static --> static const. ChangeSet@1.1002.4.60, 2003-06-23 13:59:49-07:00, jmorris@intercode.com.au [CRYPTO]: deflate module: workaround zlib bug. ChangeSet@1.1002.4.59, 2003-06-23 13:59:39-07:00, jmorris@intercode.com.au [CRYPTO]: Add Deflate algorithm to crypto API. ChangeSet@1.1002.4.58, 2003-06-23 13:59:29-07:00, jmorris@intercode.com.au [CRYPTO]: Make use of crypto_exit_ops() during crypto_free_tfm(). ChangeSet@1.1002.4.57, 2003-06-23 13:59:20-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: module_name does not exist in 2.4.x ChangeSet@1.1002.4.56, 2003-06-23 13:59:11-07:00, davem@nuts.ninka.net [CRYPTO]: Include linux/errno.h as appropriate. ChangeSet@1.1002.4.55, 2003-06-23 13:59:03-07:00, davem@nuts.ninka.net [CRYPTO-2.4]: Add dummy kmap_types.h header for sparc64. ChangeSet@1.1002.4.54, 2003-06-23 13:58:54-07:00, jmorris@intercode.com.au [CRYPTO]: Documentation updates. ChangeSet@1.1002.4.53, 2003-06-23 13:58:45-07:00, jmorris@intercode.com.au [CRYPTO]: Eliminate crypto_tfm.crt_ctx, from Adam Richter. ChangeSet@1.1002.4.52, 2003-06-23 13:58:36-07:00, jmorris@intercode.com.au [CRYPTO]: Add encrypt_iv() and decrypt_iv() methods. ChangeSet@1.1002.4.51, 2003-06-23 13:58:27-07:00, jmorris@intercode.com.au [CRYPTO]: Move km_types out of header. ChangeSet@1.1002.4.50, 2003-06-23 13:58:18-07:00, jmorris@intercode.com.au [CRYPTO]: in/out scatterlist support for ciphers. - Merge scatterwalk patch from Adam J. Richter API change: cipher methods now take in/out scatterlists and nbytes params. - Merge gss_krb5_crypto update from Adam J. Richter - Add KM_SOFTIRQn (instead of KM_CRYPTO_IN etc). - Add asm/kmap_types.h to crypto/internal.h - Update cipher.c credits. - Update cipher.c documentation. ChangeSet@1.1002.4.49, 2003-06-23 13:58:09-07:00, jmorris@intercode.com.au [CRYPTO] Add AES and MD4 to tcrypto crypto_alg_available() test. ChangeSet@1.1002.4.48, 2003-06-23 13:58:00-07:00, jmorris@intercode.com.au [CRYPTO] remove superfluous goto from des module init exception path ChangeSet@1.1002.4.47, 2003-06-23 13:57:50-07:00, jmorris@intercode.com.au [CRYPTO]: Add support for SHA-386 and SHA-512 - Merged SHA-384 and SHA-512 code from Kyle McMartin - Added test vectors. - Documentation and credits updates. ChangeSet@1.1002.4.46, 2003-06-23 13:57:40-07:00, jmorris@intercode.com.au [CRYPTO]: More credits for AES. ChangeSet@1.1002.4.45, 2003-06-23 13:57:31-07:00, davem@nuts.ninka.net [CRYPTO]: Use appropriate defaults if AH/ESP is enabled. ChangeSet@1.1002.4.44, 2003-06-23 13:57:22-07:00, jmorris@intercode.com.au [CRYPTO]: Add AES algorithm. - Merged AES code from Adam J. Richter - Add kconfig help and test vector code from Martin Clausen - Minor cleanups: removed EXPORT_NO_SYMBOLS (not needed for 2.5), removed debugging code etc. - Documentation updates. ChangeSet@1.1002.4.43, 2003-06-23 13:57:12-07:00, davem@nuts.ninka.net [CRYPTO]: internal.h needs init.h ChangeSet@1.1002.4.42, 2003-06-23 13:57:03-07:00, adam@yggdrasil.com [CRYPTO]: Simplify crypto memory allocation. ChangeSet@1.1002.4.41, 2003-06-23 13:56:54-07:00, jmorris@intercode.com.au [CRYPTO]: Dont compile procfs stuff if procfs is not enabled. ChangeSet@1.1002.4.40, 2003-06-23 13:56:45-07:00, jmorris@intercode.com.au [CRYPTO]: Documentation update. ChangeSet@1.1002.4.39, 2003-06-23 13:56:35-07:00, jmorris@intercode.com.au [CRYPTO]: Add serpent algorithm. ChangeSet@1.1002.4.38, 2003-06-23 13:56:25-07:00, jmorris@intercode.com.au [CRYPTO]: Add twofish algorithm. ChangeSet@1.1002.4.37, 2003-06-23 13:56:14-07:00, jmorris@intercode.com.au [CRYPTO]: Kill stray CRYPTO_ALG_TYPE_COMP. ChangeSet@1.1002.4.36, 2003-06-23 13:56:05-07:00, jmorris@intercode.com.au [CRYPTO]: Add null algorithms and minor cleanups. ChangeSet@1.1002.4.35, 2003-06-23 13:55:55-07:00, jgarzik@redhat.com [CRYPTO]: Kill accidental double memset. ChangeSet@1.1002.4.34, 2003-06-23 13:55:46-07:00, jmorris@intercode.com.au [CRYPTO] Minor doc update. ChangeSet@1.1002.4.33, 2003-06-23 13:55:37-07:00, jmorris@intercode.com.au [CRYPTO] Add maintainers entry. ChangeSet@1.1002.4.32, 2003-06-23 13:55:28-07:00, jmorris@intercode.com.au [CRYPTO] kstack cleanup (v0.28) ChangeSet@1.1002.4.31, 2003-06-23 13:55:18-07:00, jmorris@intercode.com.au [CRYPTO]: minor updates - Fixed min keysize bug for Blowfish (it is 32, not 64). - Documentation updates. ChangeSet@1.1002.4.30, 2003-06-23 13:55:09-07:00, davem@nuts.ninka.net [CRYPTO]: Make sha256.c more palatable to GCCs optimizers. ChangeSet@1.1002.4.29, 2003-06-23 13:54:59-07:00, jmorris@intercode.com.au [CRYPTO]: Add blowfish algorithm. ChangeSet@1.1002.4.28, 2003-06-23 13:54:49-07:00, jmorris@intercode.com.au [CRYPTO]: Add SHA256 plus bug fixes. - Bugfix in sha1 copyright - Add support for SHA256, test vectors and HMAC test vectors - Remove obsolete atomic messages. ChangeSet@1.1002.4.27, 2003-06-23 13:54:39-07:00, torvalds@transmeta.com The crypto auto-load should be enabled if crypto is enabled. ChangeSet@1.1002.4.26, 2003-06-23 13:54:30-07:00, davem@nuts.ninka.net [CRYPTO]: Allocate work buffers instead of using kstack. ChangeSet@1.1002.4.25, 2003-06-23 13:54:20-07:00, davem@nuts.ninka.net [CRYPTO]: Include kernel.h in crypto.h ChangeSet@1.1002.4.24, 2003-06-23 13:54:11-07:00, jmorris@intercode.com.au [CRYPTO]: Rework HMAC interface. ChangeSet@1.1002.4.23, 2003-06-23 13:54:01-07:00, jmorris@intercode.com.au [CRYPTO]: Add crypto_alg_available interface. ChangeSet@1.1002.4.22, 2003-06-23 13:53:52-07:00, jmorris@intercode.com.au [CRYPTO]: Cleanups based upon feedback from jgarzik. - make crypto_cipher_flags() return u32 (this means it will return the actual flags reliably, instead of being just a boolean op). - simplify error path in crypto_init_flags(). ChangeSet@1.1002.4.21, 2003-06-23 13:53:43-07:00, jmorris@intercode.com.au [CRYPTO]: Uninline some functions to save some bloat. ChangeSet@1.1002.4.20, 2003-06-23 13:53:34-07:00, jmorris@intercode.com.au [CRYPTO]: Cleanups based upon suggestions by Jeff Garzik. - Changed unsigned to unsigned int in algos. - Consistent use of u32 for flags throughout api. - Use of unsigned int rather than int for counting things which must be positive, also replaced size_ts to keep code simpler and lessen bloat on some archs. - got rid of some unneeded returns. - const correctness. ChangeSet@1.1002.4.19, 2003-06-23 13:53:25-07:00, jmorris@intercode.com.au [CRYPTO]: Fix some credits. ChangeSet@1.1002.4.18, 2003-06-23 13:53:15-07:00, davem@nuts.ninka.net [CRYPTO]: Clean up header file usage. ChangeSet@1.1002.4.17, 2003-06-23 13:53:06-07:00, jmorris@intercode.com.au [CRYPTO]: Add some documentation. ChangeSet@1.1002.4.16, 2003-06-23 13:52:57-07:00, davem@nuts.ninka.net [CRYPTO]: Build/warning fixups. ChangeSet@1.1002.4.15, 2003-06-23 13:52:48-07:00, davem@nuts.ninka.net [CRYPTO]: kunmap does not return a value. ChangeSet@1.1002.4.14, 2003-06-23 13:52:39-07:00, jmorris@intercode.com.au [CRYPTO]: Update to IV get/set interface. ChangeSet@1.1002.4.13, 2003-06-23 13:52:30-07:00, jmorris@intercode.com.au [CRYPTO]: Cleanups and more consistency checks. - Removed local_bh_disable() from kmap wrapper, not needed now with two atomic kmaps. - Nuked atomic flag, use in_softirq() instead. - Converted crypto_kmap() and crypto_yield() to check in_softirq(). - Check CRYPTO_MAX_CIPHER_BLOCK_SIZE during alg init. - Try to initialize as much at compile time as possible (feedback from Christoph Hellwig). - Clean up list handling a bit (feedback from Christoph Hellwig). ChangeSet@1.1002.4.12, 2003-06-23 13:52:21-07:00, jmorris@intercode.com.au [CRYPTO]: Assert that interfaces are called on correct cipher type. ChangeSet@1.1002.4.11, 2003-06-23 13:52:12-07:00, jmorris@intercode.com.au [CRYPTO]: Run tcrypt through lindent, plus doc update. ChangeSet@1.1002.4.10, 2003-06-23 13:52:03-07:00, jmorris@intercode.com.au [CRYPTO]: Algorithm lookup API change plus bug fixes. - API change: implemented simplest version of algorithm lookup by name (feedback from Rusty Russell and Herbert Valerio Riedel). - Now need to add the following line to to /etc/modules.conf for dynamic module loading: alias des3_ede des ChangeSet@1.1002.4.9, 2003-06-23 13:51:54-07:00, jmorris@intercode.com.au [CRYPTO]: Add MD4. ChangeSet@1.1002.4.8, 2003-06-23 13:51:44-07:00, jmorris@intercode.com.au [CRYPTO]: More bug fixes and cleanups. - added back USAGI copyright for HMAC (lost earlier during some refactoring). - bugfix: make sure tfm pointer is set to NULL during post allocation failure path in crypto_alloc_tfm() ChangeSet@1.1002.4.7, 2003-06-23 13:51:33-07:00, jmorris@intercode.com.au [CRYPTO]: Bug fixes and cleanups. - try_inc_mod_count() already does what crypto_alg_get() was trying to do. (feedback from Andrew Morton.) - Moved the BUG_ON() in crypto_unregister_alg() further up, no need to bother iterating over the list. - Always use kmap_atomic (feedback from Andrew Morton). Implemented two atomic kmaps, KM_USER for user context and KM_SOFTIRQ for softirq context. - Fixup KM_CRYPTO_ placement so Dave does not go crazy. ChangeSet@1.1002.4.6, 2003-06-23 13:51:25-07:00, jmorris@intercode.com.au [CRYPTO]: Use kmod to try to autoload modules. ChangeSet@1.1002.4.5, 2003-06-23 13:51:15-07:00, jmorris@intercode.com.au [CRYPTO]: Use try_inc_mod_count and semaphore for alg list. ChangeSet@1.1002.4.4, 2003-06-23 13:51:07-07:00, jmorris@intercode.com.au [CRYPTO]: Cleanups based upon feedback from Rusty and jgarzik - s/__u/u/ - s/char/u8/ - Fixed bug in cipher.c, page remapped was off by one block ChangeSet@1.1002.4.3, 2003-06-23 13:50:57-07:00, jmorris@intercode.com.au [CRYPTO]: Cleanups based upon feedback from Rusty and jgarzik - s/__u/u/ - s/char/u8/ - Fixed bug in cipher.c, page remapped was off by one block ChangeSet@1.1002.4.2, 2003-06-23 13:50:47-07:00, jmorris@intercode.com.au [CRYPTO]: Add initial crypto api subsystem. ChangeSet@1.1002.4.1, 2003-06-23 13:50:32-07:00, davem@nuts.ninka.net [BK]: Add *~ to ignore regexps. ChangeSet@1.930.1.241, 2003-06-23 14:33:06-06:00, rohit.seth@intel.com ia64: Use "hint @pause" in cpu_relax() and locking routines (if supported by gas). ChangeSet@1.1002.3.2, 2003-06-23 13:16:21-07:00, solt@dns.toxicfilms.tv [IPV4]: Be more verbose about invalid ICMPs sent to broadcast. ChangeSet@1.1002.3.1, 2003-06-23 12:56:02-07:00, hadi@shell.cyberus.ca [NET]: Fix OOPSes with RSVP. ChangeSet@1.997.2.2, 2003-06-22 23:41:28-07:00, agrover@groveronline.com ACPI: make it so acpismp=force works (reported by Andrew Morton) ChangeSet@1.997.2.1, 2003-06-22 23:38:39-07:00, agrover@groveronline.com ACPI: Fix config.in (Jeff Garzik) ChangeSet@1.1002.1.1, 2003-06-22 19:13:03-03:00, bunk@fs.tum.de [PATCH] add three ACPI Configure.help entries The patch below adds Configure.help entries for three ACPI options added in 2.4.22-pre1 (help texts stolen from 2.5). Please apply Adrian TAG: MailDone ChangeSet@1.1002, 2003-06-20 19:16:49-03:00, marcelo@freak.distro.conectiva Merge http://linux-isdn.bkbits.net/linux-2.4.isdn into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.1001, 2003-06-20 16:50:14-05:00, kai@tp1.ruhr-uni-bochum.de ISDN: Fix bug in ST5481 D-Channel state machine Thanks to Nicholas Robinson for debugging logs. ChangeSet@1.1000, 2003-06-20 16:47:43-05:00, kai@tp1.ruhr-uni-bochum.de ISDN: Fix Fritz!PCI v2 xmit irq underrun recovery Thanks to Holger Metschulat for finding / debugging the problem. ChangeSet@1.997.1.1, 2003-06-20 18:41:35-03:00, marcelo@freak.distro.conectiva Added missing "-" to EXTRAVERSION TAG: v2.4.22-pre1 ChangeSet@1.999, 2003-06-20 16:40:09-05:00, kaber@trash.net ISDN: Add CONFIG_IPPP_FILTER This patch adds ippp filters to isdn similar to ppp filters (active-filter & pass-filter). ChangeSet@1.930.164.3, 2003-06-20 14:31:04-07:00, greg@kroah.com [PATCH] USB: 2.4 fix UHCI debug kmalloc() usage Here's a patch from Tony Luck that fixes a problem with the UHCI debugging code on architectures with big page sizes. We end up allocating more than kmalloc allows. ChangeSet@1.930.164.2, 2003-06-20 14:22:23-07:00, greg@kroah.com [PATCH] USB: remove stupid conversions and use of floating point from aiptek.c ChangeSet@1.997, 2003-06-20 17:50:13-03:00, marcelo@freak.distro.conectiva Merge http://linux-acpi.bkbits.net/linux-2.4-acpi into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.930.165.1, 2003-06-20 17:03:18-03:00, jamagallon@able.es [PATCH] Allow aicasm to be built with db4-devel Hi. This enables the build of aicasm with db4-devel. ChangeSet@1.930.163.4, 2003-06-20 16:28:49-03:00, marcelo@freak.distro.conectiva Delete autogenerated lib/crc32table.h TAG: MailDone ChangeSet@1.930.164.1, 2003-06-20 11:50:29-07:00, will@sowerbutts.com [PATCH] USB: Update for the powermate driver to work with newer devices this patch updates the powermate driver to work with a minor revision of the powermate firmware. ChangeSet@1.930.163.3, 2003-06-20 14:46:40-03:00, marcelo@freak.distro.conectiva Merge master.kernel.org:/home/dwmw2/BK/crc32-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 TAG: MailDone ChangeSet@1.930.163.1, 2003-06-20 14:17:44-03:00, marcelo@freak.distro.conectiva Merge bk://kernel.bkbits.net/davem/net-2.4 into freak.distro.conectiva:/home/marcelo/bk/linux-2.4 ChangeSet@1.930.162.1, 2003-06-20 14:00:05-03:00, sct@redhat.com [PATCH] Fix O_DIRECT races in 2.4 Hi, I've found a few races in O_DIRECT in 2.4. There are multiple places where races can occur, mostly affecting sparse files or truncate: O_DIRECT reads against buffered writes: Read with O_DIRECT to a sparse area, then submit a buffered write to the same area. The file flush that the O_DIRECT read does initially can happen before the write, so you end up with newly written data in the area which has not yet been flushed to disk by the time the direct read is serviced. Stale data from the disk can be returned. O_DIRECT writes against buffered reads: Similar to the above, submit an O_DIRECT write into a sparse region of a file then read from that region while the write is still in progress. The write doesn't lock pages in the page cache so there's no synchronisation against the read: stale data can be returned. O_DIRECT IOs against truncate: Submit direct IO against a file then truncate it while the IO is in progress. Writes are OK because direct writes currently hold i_sem, but reads don't --- the data blocks can be deallocated, reallocated to somebody else, and we potentially get to read that other data. The i_sem on writes is actually a problem --- it prevents multiple threads from submitting parallel direct IOs at once, as the semaphore effectively serialises these IOs synchronously. The patch below fixes this by: * Prevent direct IO into sparse regions of a file. For reads, zeros are filled in anyway; for writes, fall back to buffered IO followed by fdatasync(). * Lock against truncate. Add a new, rwsem (i_alloc_sem) to guard against deallocation of data blocks while a direct IO is in progress: held in shared mode for the duration of all direct IOs, taken exclusively for truncate. * Guard all direct IO getblk()s with i_sem The direct IO read path takes an extra i_sem, which means we can't look up data blocks which are still in the process of being filled in by a buffered write. But also * Drop i_sem for the actual direct IO, once we've done the getblk() lookups. Allows multiple direct IOs to be in progress against a file at once. Holding i_sem just for the getblk() ensures that direct writes beyond EOF are still consistent even with O_APPEND, while still allowing parallelism in the IOs once we've mapped the file blocks. The bulk of the race prevention is in preventing IO to sparse regions and dealing with the truncate locking. --Stephen ChangeSet@1.930.149.43, 2003-06-19 22:55:29-04:00, bernie@develer.com [PATCH] PATCH: fix bug in drivers/net/cs89x0.c:set_mac_address() Hello Andrew, Jeff and Alan, the following patch fixes a bug in the CS89xx net device which would set new MAC address through SIOCSIFHWADDR _only_ when net_debug is set, which is obviously not what it was meant to do. The original code bogusly interpreted the addr argument as a buffer containing the MAC address instead of a struct sockaddr. Applies as-is to 2.4.20 and with offset to 2.5.69. Please forward it to Linus and Marcelo. This bug has been found and fixed by Stefano Fedrigo . ChangeSet@1.930.149.42, 2003-06-19 22:54:44-04:00, jgarzik@redhat.com [netdrvr sis900] minor fixes from 2.5 spelling, C99 initializers, jiffy wrap, set_bit ChangeSet@1.930.149.41, 2003-06-19 22:42:43-04:00, sam@mars.ravnborg.org [netdrvr sis900] make function headers readable by kernel-doc tool ChangeSet@1.930.149.40, 2003-06-19 22:28:19-04:00, zwane@linuxpower.ca [PATCH] Remove warning due to comparison in drivers/net/pcnet32.c drivers/net/pcnet32.c: In function `pcnet32_init_ring': drivers/net/pcnet32.c:1006: warning: comparison between pointer and integer ChangeSet@1.930.149.39, 2003-06-19 22:05:20-04:00, linux-kernel@vger.kernel.org [PATCH] new eepro100 PDI ID [PATCH] new eepro100 PDI ID From: Tom Alsberg Add support for a new eepro100 PCI ID. ChangeSet@1.930.149.38, 2003-06-19 22:04:48-04:00, ak@muc.de [PATCH] Remove copied inet_aton code in bond_main.c According to a report the my_inet_aton code in bond_main.c is copied from 4.4BSD, but it doesn't carry a BSD copyright license. In addition it is somewhat redundant with the standard in_aton. Convert it to use the linux function. Error handling is a bit worse than before, but not much. Patch for 2.5 bonding. The 2.4 version has the same problem, but afaik it is scheduled to be replaced by the 2.5 codebase anyways. -Andi ChangeSet@1.930.149.37, 2003-06-19 22:04:23-04:00, akpm@digeo.com [PATCH] Additional 3c980 device support From: "J.A. Magallon" Adds support for a couple of 3c980 variants which are in pci.ids, but not in the driver. ChangeSet@1.995, 2003-06-19 17:50:03-07:00, agrover@groveronline.com ACPI: Interpreter update to 20030619 - Fix To/FromBCD, eliminating the need for an arch-specific #define - Do not acquire a semaphore in the S5 shutdown path - Fix ex_digits_needed for 0 (Takayoshi Kochi) - Fix sleep/stall code reversal (Andi Kleen) - Revert a change having to do with control method calling semantics ChangeSet@1.930.158.10, 2003-06-19 17:20:17-07:00, cweidema@indiana.edu [PATCH] USB: pentax optio S ChangeSet@1.930.158.9, 2003-06-19 17:16:37-07:00, david-b@pacbell.net [PATCH] USB: EHCI update for 2.4 Here's another one -- the EHCI driver fixes that are now in Linus' tree, and are mostly in 2.4.21-ac1 (but that's missing an important one-liner). It goes on top of the patches Greg just sent. The "meat" of this patch is fixing a handful of logic bugs in the qh_completions() code which usually did a good job of covering for each other. ChangeSet@1.930.97.35, 2003-06-19 14:23:11-07:00, davem@nuts.ninka.net [NET]: Size hh_cache->hh_data more appropriately. ChangeSet@1.930.97.34, 2003-06-19 12:36:19-07:00, devik@cdi.cz [NET]: Fix jiffies races in net/sched/sch_htb.c ChangeSet@1.930.114.59, 2003-06-19 21:04:09+02:00, marcel@holtmann.org [Bluetooth] Quirk for devices with no ISOC endpoints This patch checks the existence of ISOC endpoints before it starts the URB for them. ChangeSet@1.930.1.240, 2003-06-19 10:30:08-06:00, bjorn_helgaas@hp.com ia64: Rename EFI systab tags (no spaces, etc, for easier parsing). ChangeSet@1.930.151.27, 2003-06-19 07:51:16-07:00, trini@kernel.crashing.org PPC32: Describe when we want to do a CPM reset on MPC8xx. From Dan Malek . ChangeSet@1.930.151.25, 2003-06-19 22:29:58+10:00, paulus@samba.org PPC32: Allow for PCI host bridges that need explicit type 1 cycle indication. ChangeSet@1.930.151.24, 2003-06-19 22:23:16+10:00, paulus@samba.org PPC32: Add support for PCI and time-of-day clock on 405GP-based systems. ChangeSet@1.930.151.23, 2003-06-19 22:03:18+10:00, paulus@samba.org PPC32: Add definitions for the UIC interrupt controller on the 405GP processor. ChangeSet@1.930.151.22, 2003-06-19 21:56:33+10:00, paulus@samba.org PPC32: Don't run `checks' program on make zImage. The program itself is a bit problematical since it relies on including both kernel and system headers, and it hasn't found any problems for ages. ChangeSet@1.930.151.21, 2003-06-19 21:53:35+10:00, paulus@samba.org PPC32: Cleanups for PPC 405GP-based systems; add file of OCP ids. ChangeSet@1.930.151.20, 2003-06-19 21:44:08+10:00, paulus@samba.org PPC32: remove ppc4xx_serial.h, it is no longer used. ChangeSet@1.930.151.19, 2003-06-19 21:36:34+10:00, paulus@samba.org PPC32: Move PC-style serial port definitions out to asm/pc_serial.h. ChangeSet@1.930.151.18, 2003-06-19 21:32:46+10:00, paulus@samba.org PPC32: Extra register and other definitions for the PPC 405GP processor. ChangeSet@1.930.151.17, 2003-06-19 21:26:27+10:00, paulus@samba.org PPC32: Add support for PPC 405GP interrupt controller. ChangeSet@1.930.151.16, 2003-06-19 21:21:18+10:00, paulus@samba.org PPC32: Make debug exceptions usable on 4xx-class processors, and improve trap handling. ChangeSet@1.930.151.15, 2003-06-19 19:47:46+10:00, paulus@samba.org PPC32: Update the support for the "Walnut" 405GP platform. ChangeSet@1.930.151.14, 2003-06-19 19:20:20+10:00, paulus@samba.org PPC32: Add generic IBM PPC405GP support and use it on the walnut platform. ChangeSet@1.930.151.13, 2003-06-19 18:27:05+10:00, paulus@samba.org PPC32: Introduce a new config symbol, CONFIG_40x, used for PPC 40x cpus. This allows us to distinguish code which is appropriate for all 4xx cpus from code that applies to the 403 and 405 but not to the 440. ChangeSet@1.930.151.12, 2003-06-19 18:08:42+10:00, paulus@samba.org PPC32: Add a new framework for on-chip peripherals for the IBM 4xx embedded processors. ChangeSet@1.930.151.11, 2003-06-19 15:41:59+10:00, paulus@samba.org PPC32: Update for PPC 4xx TLB and exception handling. ChangeSet@1.930.97.33, 2003-06-18 20:29:40-07:00, davem@nuts.ninka.net [NET]: Fix build failure from recent sunrpc changes. ChangeSet@1.930.97.31, 2003-06-18 18:04:31-07:00, dlstevens@us.ibm.com [IPV4/IPV6]: Fix IGMP device refcount leaks, with help from yoshfuji@linux-ipv6.org. ChangeSet@1.930.159.64, 2003-06-18 17:27:07-07:00, stern@rowland.harvard.edu [PATCH] USB: US_SC_DEVICE and US_PR_DEVICE for 2.4 This patch implements the new subclass and protocol unusual_devs.h codes for 2.4.21. Now new entries can remain consistent between 2.4 and 2.5. ChangeSet@1.930.159.63, 2003-06-18 17:24:52-07:00, dhollis@davehollis.com [PATCH] USB: AX8817X Driver for 2.4 Kernels ChangeSet@1.930.159.62, 2003-06-18 17:21:31-07:00, henning@meier-geinitz.de [PATCH] USB: New vendor/product ids for scanner driver This patch adds some new vendor/product ids for the USB scanner driver. It's the same that already went into 2.5. ChangeSet@1.930.159.61, 2003-06-18 17:21:26-07:00, vojtech@suse.cz [PATCH] USB: Fix HID logical min/max for 2.4 Some UPSes have logical minimum and maximum value in HID descriptors the same. The HID driver rejects such devices. This patch fixes this. Please apply for 2.4, I'll take care of a 2.5 patch. ChangeSet@1.930.159.60, 2003-06-18 17:21:20-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: parametrize the module ChangeSet@1.930.158.7, 2003-06-18 21:09:14-03:00, pdelaney@lsil.com [PATCH] Critical bug fix for fusion driver Hello Marcelo, Alan - In our IA64 testing we have discovered a bug that causes a critical failure with the fusion driver. To date we have been unable to reproduce the failure on 32 bit systems, but given the fix, the failure could occur on any architecture.  Please accept the attached patch for 2.4.21-rc9. Changes: o critical bug fix: 2 line patch to zero unused CDB bytes in driver qcmd function o use old eh code instead of new o change mailto labels to lstephen. TAG: MailDone ChangeSet@1.930.159.59, 2003-06-18 17:08:12-07:00, greg@kroah.com USB: clean up extra whitespace in visor.c driver. ChangeSet@1.930.159.58, 2003-06-18 17:07:21-07:00, greg@kroah.com [PATCH] USB: fixup aiptek driver for older compilers ChangeSet@1.930.159.57, 2003-06-18 17:07:07-07:00, bwheadley@earthlink.net [PATCH] USB: Aiptek kernel driver 1.0 for Kernel 2.4 ChangeSet@1.930.159.56, 2003-06-18 17:03:51-07:00, greg@kroah.com [PATCH] USB: add error reporting functionality to the pl2303 driver. ChangeSet@1.930.159.55, 2003-06-18 16:55:52-07:00, hanno@gmx.de [PATCH] USB: Patch for Vivicam 355 ChangeSet@1.930.159.54, 2003-06-18 16:55:46-07:00, vsu@altlinux.ru [PATCH] USB: HIDDEV / UPS patches ChangeSet@1.930.159.53, 2003-06-18 16:55:41-07:00, bcollins@debian.org [PATCH] USB: Actually Fix 2.4 HID input Slight typo on my part for the 2.4 fix (no keyboard LEDs). ChangeSet@1.930.159.52, 2003-06-18 16:55:36-07:00, olh@suse.de [PATCH] USB: incorrect ethtool -i driver name This converts 2.4 to use driver_name ChangeSet@1.930.159.51, 2003-06-18 16:55:30-07:00, johannes@erdfelt.com [PATCH] USB: fix 2.4 usbdevfs race Here's a patch to fix a race condition in usbdevfs. The fix is in hub.c but the race is related to usbdevfs. The race goes like this: Process 1 (khubd) Process 2 (mount) usb_hub_port_connect_change() hub->children[port] = dev usb_new_device() usbdevfs_read_super() recurse_new_dev_inode() new_dev_inode() list_add_tail(..., &dev->inodes) usbdevfs_add_device() new_dev_inode() list_add_tail(..., &dev->inodes) The problem is that the inode gets added twice, corrupting dev->inodes. This will cause a problems at disconnect where the same inode will be freed twice, causing a neverending loop, or an oops. I think it will also cause problems at unmount. The fix is to just move setting hub->children to later in the enumeration process. This way usbdevfs_read_super won't see the device before it has been through the usbdevfs_add_device path. I didn't see this on x86, but apparentely others have looking at the RedHat 9 kernel sources. (RedHat bugzilla #81091) Pete, could you give this patch a shot for the problem you found in that bug? I'm pretty sure they are the same problem. ChangeSet@1.930.159.50, 2003-06-18 16:55:25-07:00, greg@kroah.com [PATCH] USB: pegasus ethtool fixup. ChangeSet@1.930.159.49, 2003-06-18 16:55:20-07:00, olh@suse.de [PATCH] USB: incorrect ethtool -i driver name ethtool -i ethX should return the driver name instead of a 'verbose' string. Other tools rely on the output. 2.5 might need a similar fix. smirnow:~ # ethtool -i eth0 driver: 3c59x version: LK1.1.16 firmware-version: bus-info: 00:0a.0 smirnow:~ # ethtool -i eth1 driver: Pegasus/Pegasus II USB Ethernet v0.4.30 (2003/04/01) version: v0.4.30 (2003/04/01) firmware-version: bus-info: usb1:8 ChangeSet@1.930.159.48, 2003-06-18 16:55:14-07:00, stewart@inverse.wetlogic.net [PATCH] USB: HIDDev uref backport for 2.4? ChangeSet@1.930.159.47, 2003-06-18 16:55:08-07:00, bcollins@debian.org [PATCH] USB: fix keyboard leds ChangeSet@1.930.159.46, 2003-06-18 16:55:03-07:00, petkan@users.sourceforge.net [PATCH] USB: pegasus patch better EPIPE handling, ethtool improvements, and another attempt to fix HOME_PNA support was made. HP added as a vendor and device ID. ChangeSet@1.930.159.45, 2003-06-18 16:54:57-07:00, hwahl@hwahl.de [PATCH] USB: Patch for Samsung Digimax 410 *** a/drivers/usb/storage/unusual_devs.h 2003-05-22 20:54:26.000000000 +0200 ChangeSet@1.930.159.44, 2003-06-18 16:54:52-07:00, bcollins@debian.org [PATCH] USB Multi-input quirk ChangeSet@1.930.159.43, 2003-06-18 16:54:47-07:00, wahrenbruch@kobil.de [PATCH] USB: kobil_sct.c added support for KAAN SIM Reader here is a patch for the kobil_sct.c usb-serial driver which adds support for the KAAN SIM. ChangeSet@1.930.159.42, 2003-06-18 16:54:42-07:00, per.winkvist@telia.com [PATCH] Re: unusual_devs.h patch that was in 2.5.68 On Friday 02 May 2003 09:03, Greg KH wrote: > On Thu, May 01, 2003 at 11:15:32AM +0200, Per Winkvist wrote: > > I've modified James Courtier-Dutton description from Optio 430 => Optio > > 2/3/400 since it applies to Optio 330 and possible 230 too. > > > > Attached are the patch that was tried with Pentax Optio S and Optio 330 > > RS. I've talked to Pete Zaitcev and he said I should be using CB instead > > since more devices liked that better... so I leave that to you! > > > > Thanks a lot and I hope this will end up in 2.4 sooon :=) > > I've applied these to my 2.4 and 2.5 trees. The 2.4 ones will have to > wait until after 2.4.21 comes out. The 2.5 one will go out with my next > round of 2.5 usb patches, probably in a few days. Hmm... The unusual patch for Pentax Optio S (and description fix for Pentax 2/3/400) I sent to you was against 2.4.20, not against your bk usb repository. This caused Pentax entries to get duplicated for 2.4 and 2.5 (bk current too). Then another Pentax Optio S patch have been applied as well... Right now you have three 0x0004 and two 0x0006 entries ;=) ChangeSet@1.930.159.41, 2003-06-18 16:54:36-07:00, vojtech@suse.cz [PATCH] USB: Make Olympus cameras work with usb-storage According to Jan Derfinak, Olympus USB cameras mistakenly report an 'USBU' signature instead of 'USBS', while using the normal USB Storage protocol. ChangeSet@1.930.159.40, 2003-06-18 16:54:31-07:00, Walter.Harms@Informatik.Uni-Oldenburg.DE [PATCH] USB: fixes kernel_thread ChangeSet@1.930.159.39, 2003-06-18 16:54:26-07:00, Walter.Harms@Informatik.Uni-Oldenburg.DE [PATCH] USB: fixes kernel_thread ChangeSet@1.930.159.38, 2003-06-18 16:54:21-07:00, Richard.Curnow@superh.com [PATCH] USB: ehci-hcd.c needs to include When I try to configure in EHCI support without this patch, I get generic_ffs undefined at link time. (This is with 2.4.21-rc2 on our sh64 (SH-5) port). Perhaps there are other ways to achieve this, but this worked for me. ChangeSet@1.930.159.37, 2003-06-18 16:54:16-07:00, bcollins@debian.org [PATCH] USB: Happ UGCI added as BADPAD for workaround Greg, I sent these patches to Vojtech, and haven't heard anything back, but they are straight forward. One simply backports the BADPAD handling to 2.4, and both patches add Happ UGCI joysticks under the BADPAD workaround. ChangeSet@1.930.159.36, 2003-06-18 16:54:10-07:00, greg@kroah.com [PATCH] USB: fix break control for pl2303 driver Thanks to Martin Evans for pointing this out. ChangeSet@1.930.159.35, 2003-06-18 16:54:05-07:00, smb@smbnet.de [PATCH] USB: another usb storage addition ChangeSet@1.930.159.34, 2003-06-18 16:54:00-07:00, david-b@pacbell.net [PATCH] USB: SMP ehci-q.c 1010 BUG() Stefano Barbato wrote: > Dual PIII > kernel 2.4.21-rc2 (w/ SMP) (2.5.69 below) > ... > > I put a few printk before the BUG() and I found that the offending if() is > this: > if(qh->qh_state != QH_STATE_LINKED > && qh->qh_state != QH_STATE_UNLINK_WAIT) > > because qh_state were QH_STATE_COMPLETING. I got a similar SMP report recently, but without info about which clause was failing -- which is a key clue, thanks!! The COMPLETING state is used only while a QH is being scanned for completed TDs. (Think CPU-0 irq handler.) Looking at the handful of places that call the routine reporting the BUG(), a couple seem like they could make trouble with multiple CPUs in the driver. ChangeSet@1.930.159.33, 2003-06-18 16:53:54-07:00, nicolas@dupeux.net [PATCH] USB: UNUSUAL_DEV for aiptek pocketcam Here is the unusual_dev entry i'm using to get my digital camera. diff -cr linux/drivers/usb/storage/unusual_devs.h linux_vaxvms/drivers/usb/storage/unusual_devs.h *** linux/drivers/usb/storage/unusual_devs.h 2003-03-14 21:32:46.000000000 +0100 ChangeSet@1.930.159.32, 2003-06-18 16:49:03-07:00, philipp@void.at [PATCH] USB: unusual_devs.h patch ChangeSet@1.930.159.31, 2003-06-18 16:48:57-07:00, geert@linux-m68k.org [PATCH] USB: Big endian RTL8150 The RTL8150 USB Ethernet driver doesn't work on big endian machines. Here are patches (for both 2.4.x and 2.5.x) to fix that. The fix was tested on the 2.4.20 and 2.4.21-rc1 version of the driver on big endian MIPS. Changes: - Fix endianness of rx_creg (from Dimitri Torfs ) - Kill unused last parameter of async_set_registers() ChangeSet@1.930.159.30, 2003-06-18 16:48:52-07:00, david-b@pacbell.net [PATCH] USB: ehci i/o watchdog This patch adds a new "I/O watchdog" role to the existing timer code, and cleans it up a bit. If you want to run EHCI without IRQs, it's now simple: disable them, and tweak the timer appropriately. The patch should help with these reported problems. (a) Bulk I/O sometimes seems to stop progressing. Not trouble in itself, but usb-storage and scsi could wedge deeply because of bugs in their fault recovery; and then the problems could break khubd and rmmod... (b) Some periodic transfers need to be "jumpstarted". Usually seen with a high speed hub. ChangeSet@1.930.159.29, 2003-06-18 16:48:47-07:00, per.winkvist@telia.com [PATCH] USB: more unusual_devs.h changes I've modified James Courtier-Dutton description from Optio 430 => Optio 2/3/400 since it applies to Optio 330 and possible 230 too. Attached are the patch that was tried with Pentax Optio S and Optio 330 RS. I've talked to Pete Zaitcev and he said I should be using CB instead since more devices liked that better... so I leave that to you! ChangeSet@1.930.159.28, 2003-06-18 16:48:42-07:00, ccheney@cheney.cx [PATCH] USB: vicam.c copyright patches Here are two vicam.c patches to clean up and restore copyright notices for 2.4.21-rc1 and 2.5.68-bk9. At some point in time Pavel and my copyrights were remove inadvertenly from the code. ChangeSet@1.930.159.27, 2003-06-18 16:48:36-07:00, James@superbug.demon.co.uk [PATCH] USB: Add support for Pentax Still Camera to linux kernel ChangeSet@1.930.159.26, 2003-06-18 16:48:31-07:00, greg@kroah.com [PATCH] USB: add comment to storage/unusual_devs.h that specifies how to add new entries. ChangeSet@1.930.159.25, 2003-06-18 16:48:11-07:00, greg@kroah.com [PATCH] USB: attempt to track down pl2303 oopses on close... This doesn't solve it, but it's the proper thing to do right now. ChangeSet@1.930.159.24, 2003-06-18 16:47:26-07:00, greg@kroah.com [PATCH] USB: added support for Sony DSC-P8 Thanks to David Kimdon for the information. ChangeSet@1.930.159.23, 2003-06-18 16:47:20-07:00, linux-usb@gemeinhardt.info [PATCH] USB: add support for Mello MP3 Player ChangeSet@1.930.158.6, 2003-06-18 20:37:02-03:00, neilb@cse.unsw.edu.au [PATCH] kNFSd: Make sure an early close on a nfs/tcp connection is handled properly. From: Hirokazu Takahashi In svc_tcp_listen_data_ready we should be waiting for TCP_LISTEN, not TCP_ESTABLISHED. The later only worked by accident. Also, if a socket is closed as soon as we accept it, we must shut it down straight away as we will never get a 'close' event. TAG: MailDone ChangeSet@1.930.159.22, 2003-06-18 16:35:23-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: set owner fields I think I'd better work on my plausible deniability... ChangeSet@1.930.159.21, 2003-06-18 16:35:17-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: remove MOD_XXX_USE_COUNT The ATM layer takes a reference with fops_get. ChangeSet@1.930.159.20, 2003-06-18 16:35:11-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: receive code rewrite Main points: - receive buffers are decoupled from urbs, so an urb can be resubmitted with a new buffer before the old buffer is processed. - the packet reconstruction code is much simpler. - locking is simplified by the fact that only the tasklet launches receive urbs ChangeSet@1.930.159.19, 2003-06-18 16:35:06-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: receive path micro optimization Make the most discriminating comparison first. ChangeSet@1.930.158.5, 2003-06-18 20:35:01-03:00, neilb@cse.unsw.edu.au [PATCH] kNFSd: TCP nfsd connection hangs when partial record header is received From: Olof Johansson Below patch resolves a hang where a TCP nfsd connection will hang even though new data is received on the socket. We've seen this a few times in our lab, but it usually happened every few weeks. If a short record header is received, the SK_BUSY flag is never cleared, and even though new data arrives, it will not be handled. This in turn leads to hangs of particular clients (while others will continue to work without problem). I also changed the return code for that condition to be the same as for a (regular) short read. ChangeSet@1.930.159.18, 2003-06-18 16:35:01-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: remove useless NULL pointer checks The stats field is never NULL. ChangeSet@1.930.159.17, 2003-06-18 16:34:55-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: kfree_skb -> dev_kfree_skb Always use dev_kfree_skb. ChangeSet@1.930.159.16, 2003-06-18 16:34:50-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: send path micro optimizations ChangeSet@1.930.159.15, 2003-06-18 16:34:45-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: use optimally sized reconstruction buffers Calculate the maximum size needed for the SAR reconstruction buffer from the supplied qos parameters. ChangeSet@1.930.159.14, 2003-06-18 16:34:39-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: verbose debugging Add a vdbg macro for verbose debugging, and convert some noisy debugging statements to use it. ChangeSet@1.930.159.13, 2003-06-18 16:34:34-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: spin_lock_irqsave -> spin_lock_irq in tasklets Replace spin_lock_irqsave/spin_unlock_irqrestore with spin_lock_irq/spin_unlock_irq in tasklet actions, since these are always called with local irqs enabled. ChangeSet@1.930.159.12, 2003-06-18 16:34:29-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: spin_lock_irqsave -> spin_lock_irq in process context Replace spin_lock_irqsave/spin_unlock_irqrestore with spin_lock_irq/spin_unlock_irq in routines that are only called in process context. ChangeSet@1.930.159.11, 2003-06-18 16:34:23-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: add defensive memory barriers Defend against future maintainers. ChangeSet@1.930.159.10, 2003-06-18 16:34:18-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: replace yield() Use set_current_state (TASK_RUNNING); schedule(); instead. ChangeSet@1.930.159.9, 2003-06-18 16:34:13-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: add missing #include Backport from 2.5. ChangeSet@1.930.159.8, 2003-06-18 16:34:08-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: trivial whitespace and name changes No code changes. ChangeSet@1.930.159.7, 2003-06-18 16:34:03-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: remove trailing semicolon ChangeSet@1.930.159.6, 2003-06-18 16:32:47-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: compile fix ChangeSet@1.930.159.5, 2003-06-18 16:32:42-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: crc optimization ChangeSet@1.930.159.4, 2003-06-18 16:32:37-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: bump the version number ChangeSet@1.930.158.4, 2003-06-18 20:32:36-03:00, neilb@cse.unsw.edu.au [PATCH] kNFSd: SVC sockets don't disable Nagle I noticed that the Nagle algorithm was disabled just recently on the client side, while it still seems to be enabled on the server side. The previous patch came from Chuck Lever. Below patch disables it on the server side as well. For latency reasons, this should be the desired behaviour NFS at both client and server. ChangeSet@1.930.159.3, 2003-06-18 16:30:13-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: discard packets for non-existant vcc's I broke part of the udsl_decode_rawcell logic in a previous patch, leading to possible hangs on startup/shutdown. Thanks to Subodh Srivastava and Ted Phelps for their bug reports. ChangeSet@1.930.159.2, 2003-06-18 16:30:08-07:00, baldrick@wanadoo.fr [PATCH] USB speedtouch: move MOD_INC_USE_COUNT ChangeSet@1.930.159.1, 2003-06-18 16:30:02-07:00, baldrick@wanadoo.fr [PATCH] USB: Backport of USB speedtouch driver to 2.4 Since the 2.5 crc library hasn't been backported to the 2.4 tree yet, I included a crc routine in the speedcrc files. ChangeSet@1.930.158.3, 2003-06-18 20:29:20-03:00, neilb@cse.unsw.edu.au [PATCH] Fix the check for execute permissions of parent directories in NFSd [4 patches for nfsd in 2.4.22-pre] The previous check fails on directories with uid=0, gid=0, and permissions in ACLs. A similar check to this one is also found in the 2.5 kernel in fs/nfsd/nfsfh.c:nfsd_acceptable(). ChangeSet@1.930.158.2, 2003-06-18 20:19:05-03:00, neilb@cse.unsw.edu.au [PATCH] Fix bug in /proc/mdstat [These 3 patches for 2.4.22-pre - thanks] If /proc/mdstat is large, or reads are for a small size, then the last line of /proc/mdstat is repeated nearly 65536 times. This patch will fix it. ChangeSet@1.930.158.1, 2003-06-18 20:16:55-03:00, neilb@cse.unsw.edu.au [PATCH] Handle concurrent failure of two drives in raid5 ### Comments for ChangeSet If two drives both fail during a write request, raid5 doesn't cope properly and will eventually oops. With this patch, blocks that have already been 'written' are failed when double drive failure is noticed, as well as blocks that are about to be written. ----------- Diffstat output ------------ ./drivers/md/raid5.c | 10 +++++++++- 1 files changed, 9 insertions(+), 1 deletion(-) ChangeSet@1.930.97.30, 2003-06-17 18:58:46-07:00, davem@nuts.ninka.net [IPV6]: Fix igmp6_timer_handler forward declaration. ChangeSet@1.994, 2003-06-17 17:37:58-07:00, agrover@groveronline.com ACPI: acpiphp update (Takayoshi Kochi) ChangeSet@1.930.97.29, 2003-06-17 16:17:46-07:00, davem@nuts.ninka.net Merge nuts.ninka.net:/home/davem/src/BK/net-2.4.22 into nuts.ninka.net:/home/davem/src/BK/net-2.4 ChangeSet@1.930.157.1, 2003-06-17 16:15:51-07:00, davem@nuts.ninka.net Merge. ChangeSet@1.930.151.8, 2003-06-17 20:05:31-03:00, marcelo@freak.distro.conectiva Changed VERSION to .22 TAG: MailDone ChangeSet@1.930.1.238, 2003-06-17 15:45:41-06:00, chad_smith@hp.com ia64: expose pointers from EFI system table in /proc. ChangeSet@1.930.151.7, 2003-06-17 18:39:01-03:00, marcelo@freak.distro.conectiva Removed EXTRAVERSION TAG: MailDone TAG: v2.4.21