ChangeSet@1.1536, 2004-08-31 07:56:06-03:00, lcapitulino@conectiva.com.br [PATCH] Fix missing `return NULL' missing in ext3_get_journal() Some time ago I fixed this bug in 2.6. There is a `return NULL' missing in ext3_get_journal() if the call to journal_init_inode() fail. Note that if the error happens, `journal' will be NULL and used. (agains't 2.4.28-pre2). Signed-off-by: Luiz Capitulino ChangeSet@1.1535, 2004-08-31 07:55:24-03:00, lcapitulino@conectiva.com.br [PATCH] - drivers/pci/pci.c NULL pointer fix. Marcelo, This is another bug fix: the function pci/pci.c::pci_add_new_bus() does not check the return of pci_alloc_bus(). If this function returns NULL, `child' can be utilised. In the patch bellow I added the check, if the call that pci_scan_bridge() is doing to pci_alloc_bus() return NULL, pci_scan_bridge() returns `max' (Greg accepted this fix for 2.6, a while ago). (against 2.4.28-pre2) Signed-off-by: Luiz Capitulino drivers/pci/pci.c | 9 +++++++++ 1 files changed, 9 insertions(+) ChangeSet@1.1531, 2004-08-31 07:38:21-03:00, davem@davemloft.net [PATCH] LVM ioctl fix - Trying to vfree() nonexistent vm area On Tue, 31 Aug 2004 01:15:40 +0100 Richard Mortimer wrote: > I'm seeing problems with lvm on sparc64. I have a reproducible test case > using snapshots where I can reliably reproduce an error similar to > > Trying to vfree() nonexistent vm area (0000000140072000) For once it's not sparc64's fault, it's a bug in the generic LVM ioctl handling :-) It saves both pointers, clobbers the userspace copy, then only restores one of the two pointers correctly. Easy to fix, see below. Marcelo, please apply, thanks. # This is a BitKeeper generated diff -Nru style patch. # # ChangeSet # 2004/08/30 17:24:05-07:00 davem@nuts.davemloft.net # [LVM]: Do not forget to restore both user pointers. # # This in particular can make compatability layers # crash, and it is a bug for regular applications # too. # # Signed-off-by: David S. Miller # # drivers/md/lvm.c # 2004/08/30 17:23:48-07:00 davem@nuts.davemloft.net +12 -0 # [LVM]: Do not forget to restore both user pointers. # # This in particular can make compatability layers # crash, and it is a bug for regular applications # too. # # Signed-off-by: David S. Miller # ChangeSet@1.1525.4.2, 2004-08-31 03:21:17-04:00, achew@nvidia.com [PATCH] i810_audio.c and pci_ids.h: add support for nforce MCP2S, ChangeSet@1.1523.1.5, 2004-08-31 03:01:34-04:00, jgarzik@pobox.com [libata] resync with 2.6 (very minor, mostly cosmetic) ChangeSet@1.1493.3.13, 2004-08-30 15:36:32-07:00, davem@nuts.davemloft.net [TG3]: Update driver version and reldate. ChangeSet@1.1493.3.12, 2004-08-30 15:35:31-07:00, davem@nuts.davemloft.net [TG3]: Fix clock control programming on 5705/5750. Need to clear one bit at a time, so if we are clearing both 625_CORE_CLOCK and ALTCLOCK we first clear the latter then the final write will clear the former. Signed-off-by: David S. Miller ChangeSet@1.1493.3.11, 2004-08-30 15:25:46-07:00, davem@nuts.davemloft.net [TG3]: Add capacitive coupling support. Signed-off-by: David S. Miller ChangeSet@1.1493.3.10, 2004-08-30 15:12:10-07:00, mchan@broadcom.com [TG3]: Check MAC_STATUS_SIGNAL_DET in serdes polling. Otherwise we do not handle properly the case where the switch/hub does not support auto- negotiation. This is what was breaking 5704 hw fiber autoneg. Signed-off-by: David S. Miller ChangeSet@1.1523.1.4, 2004-08-30 14:43:55-04:00, bzolnier@elka.pw.edu.pl [PATCH] libata: ata_piix.c PIO fix [patch] libata: ata_piix.c PIO fix "[libata] transfer mode cleanup" introduced bug in ata_piix.c: previously PIO number (not mode) was passed to piix_set_piomode(). Fortunately this function is only used for (disabled) PATA support. I bet that this is the reason why MWDMA didn't work for PATA. Signed-off-by: Bartlomiej Zolnierkiewicz ChangeSet@1.1525.5.1, 2004-08-30 16:00:55+10:00, nathans@sgi.com Merge sgi.com:/source1/linux-2.4 into sgi.com:/source1/xfs-linux-2.4 ChangeSet@1.1525.1.28, 2004-08-29 21:37:35-03:00, mikpe@csd.uu.se [PATCH] drivers/ide/pci/sc1200.c cast-as-lvalue fix Fix for cast-as-lvalue in drivers/ide/pci/sc1200.c. Backport from 2.6. /Mikael ChangeSet@1.1525.4.1, 2004-08-28 19:40:16-04:00, ileong@nvidia.com [ac97_codec] add new codec ChangeSet@1.1525.3.2, 2004-08-28 19:34:39-04:00, ha505@hszk.bme.hu [netdrvr fealnx] fix spin_unlock_irqrestore() usage ChangeSet@1.1523.1.3, 2004-08-28 19:29:58-04:00, achew@nvidia.com [PATCH] sata_nv: fix CK804 support This patch fixes a problem introduced when CK804 support was added. mmio_base can only be set in the CK804 case, else libata will attempt to iounmap mmio_base, which isn't iomapped for the non-CK804 case. Still need the bar 5 address, so steal from host_set->ports[0]->ioaddr.scr_addr. Jeff, let me know if this is a bad thing to do. ChangeSet@1.1523.1.2, 2004-08-28 19:04:47-04:00, jeremy@sgi.com [PATCH] Fix DMA boundary overflow bug We seem to have found an overflow problem in libata-core.c. We were trying to DMA to the address range 0xffff8000-0xffffbfff. In the original version of the code, given that address and count (0xffff8000 and 0x4000), the variable "boundary" would be set to 0, causing len to be set to 0x8000 (which is greater than sg_len). Then at the bottom of the loop, sg_len would be set to 0xffffc000 (0x4000 - 0x8000), which would then cause the loop never to terminate (until much of memory was scribbled over or the kernel died). The code below should be functionally identical, but not be subject to the same overflow problem (boundary needs to be a u33). Signed-off-by: jeremy@sgi.com ===== drivers/scsi/libata-core.c 1.94 vs edited ===== ChangeSet@1.1525.3.1, 2004-08-28 18:57:31-04:00, tharbaugh@lnxi.com [netdrvr e1000] disable DITR, which apparently hurts performance ChangeSet@1.1528, 2004-08-28 15:15:47-05:00, shaggy@austin.ibm.com JFS: fix memory leak in __invalidate_metapages Signed-off-by: Dave Kleikamp ChangeSet@1.1525.1.25, 2004-08-28 02:19:26-07:00, davem@kernel.bkbits.net [CREDITS]: Update my email and home address. ChangeSet@1.1525.1.24, 2004-08-28 02:19:16-07:00, davem@kernel.bkbits.net [MAINTAINERS]: Update my email address. Also fix a mis-merge that clobbered the crypto maintainers entry by putting the MARVELL YUKON syskonnect entry smack in the middle of it :-) ChangeSet@1.1525.1.23, 2004-08-27 13:09:39-03:00, mikpe@user.it.uu.se [PATCH] more gcc34 lvalue fixes Marcelo, Here are some more fixes for broken lvalues in 2.4.28-pre2. - drivers/net/ne2k-pci.c: cast-as-lvalue, fixes from 2.6 - drivers/scsi/53c7,8xx.c: cast-as-lvalue, new 2.4 fix since 2.6 doesn't seem to have this code any more - drivers/scsi/advansys.c: cast-as-lvalue, fixes from 2.6 - drivers/scsi/dpt_i2o.c: cast-as-lvalue, fixes from 2.6 - drivers/scsi/seagate.c: cast-as-lvalue, fix from 2.6 - fs/affs/super.c: conditional-as-lvalue, new 2.4 fix since the 2.6 code is completely different ChangeSet@1.1525.1.22, 2004-08-27 13:02:39-03:00, jbaron@redhat.com [PATCH] ppos cleanups 1) vc_screen.c This just adds back a removed -EINVAL 2) hysdyn_procconf.c retores the FMODE_READ check, which was dropped, and makes the code a bit easier to read, at least to me. 3) file_hdr.c off_t -> loff_t ChangeSet@1.1525.1.21, 2004-08-27 12:27:11-03:00, Jack_Hammer@adaptec.com [PATCH] ServeRAID driver (ips) Version 7.10.18 The following patch updates the ServeRAID driver ( ips ). One important bug was fixed ( the sorting of 7K adapters in function ips_order_controllers). This also includes Arjan's recommendation for 2.6 to remove an unnecessary #ifdef around MODULE_DESCRIPTION. The rest of the patch is primarily concerned with IBM version numbers. The patch to bring the 2.6 kernel driver up to the same 7.10.18 level has already been issued. ChangeSet@1.1525.1.20, 2004-08-27 10:12:03-03:00, bunk@fs.tum.de [PATCH] dscc4.c: fix gcc 3.4 compilation I got compile errors starting with the following when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-modular/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -DMODULE -DMODVERSIONS -include /home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-modular/include/linux/modversions.h -nostdinc -iwithprefix include -DKBUILD_BASENAME=dscc4 -c -o dscc4.o dscc4.c dscc4.c: In function `dscc4_found1': dscc4.c:369: sorry, unimplemented: inlining failed in call to 'dscc4_set_quartz': function body not available dscc4.c:921: sorry, unimplemented: called from here make[3]: *** [dscc4.o] Error 1 make[3]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-modular/drivers/net/wan' <-- snip --> The patch below fixes this issue (similar to how it was done in 2.6). diffstat output: drivers/net/wan/dscc4.c | 33 ++++++++++++++++----------------- 1 files changed, 16 insertions(+), 17 deletions(-) Signed-off-by: Adrian Bunk ChangeSet@1.1525.1.19, 2004-08-27 10:10:29-03:00, bunk@fs.tum.de [PATCH] asm-i386/smpboot.h: fix gcc 3.4 compilation I got the following compile error when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -nostdinc -iwithprefix include -DKBUILD_BASENAME=process -c -o process.o process.c In file included from process.c:47: /home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include/asm/smpboot.h: In function `target_cpus': /home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include/asm/smpboot.h:133: error: label at end of compound statement make[1]: *** [process.o] Error 1 make[1]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/arch/i386/kernel' <-- snip --> The patch below fixes this issue. Signed-off-by: Adrian Bunk ChangeSet@1.1525.1.18, 2004-08-27 10:03:06-03:00, bunk@fs.tum.de [PATCH] irlmp.c: fix gcc 3.4 compilation I got the following compile error when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -nostdinc -iwithprefix include -DKBUILD_BASENAME=irlmp -c -o irlmp.o irlmp.c irlmp.c: In function `irlmp_flow_indication': irlmp.c:1244: error: parse error before "__FUNCTION__" irlmp.c:1258: error: parse error before "__FUNCTION__" irlmp.c:1277: error: parse error before "__FUNCTION__" irlmp.c:1284: error: parse error before "__FUNCTION__" make[3]: *** [irlmp.o] Error 1 make[3]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/net/irda' <-- snip --> The patch below fixes this issue (similar to how it was done in 2.6). Signed-off-by: Adrian Bunk ChangeSet@1.1525.1.17, 2004-08-27 10:01:47-03:00, bunk@fs.tum.de [PATCH] ircomm_param.c: fix __FUNCTION__ paste error I got the following compile error when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -nostdinc -iwithprefix include -DKBUILD_BASENAME=ircomm_param -c -o ircomm_param.o ircomm_param.c ircomm_param.c: In function `ircomm_param_service_type': ircomm_param.c:201: error: parse error before "__FUNCTION__" make[4]: *** [ircomm_param.o] Error 1 make[4]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/net/irda/ircomm' <-- snip --> The patch below fixes this issue by removing the superfluous __FUNCTION__ (similar to how it is in 2.6). Signed-off-by: Adrian Bunk ChangeSet@1.1525.1.16, 2004-08-27 10:01:18-03:00, bunk@fs.tum.de [PATCH] lmc_media.c: fix gcc 3.4 compilation I got the following compile error when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -I. -nostdinc -iwithprefix include -DKBUILD_BASENAME=lmc_media -c -o lmc_media.o lmc_media.c lmc_media.c: In function `lmc_t1_get_link_status': lmc_debug.h:50: sorry, unimplemented: inlining failed in call to 'lmc_trace': function body not available lmc_media.c:1073: sorry, unimplemented: called from here lmc_debug.h:50: sorry, unimplemented: inlining failed in call to 'lmc_trace': function body not available lmc_media.c:1168: sorry, unimplemented: called from here make[5]: *** [lmc_media.o] Error 1 make[5]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/drivers/net/wan/lmc' <-- snip --> The patch below fixes this issue by uninlining lmc_trace. Signed-off-by: Adrian Bunk ChangeSet@1.1525.2.3, 2004-08-27 09:13:01-03:00, marcelo@logos.cnet Fix mm.h typo introduced by s390 changes ChangeSet@1.1525.2.2, 2004-08-27 09:02:40-03:00, bunk@fs.tum.de [PATCH] ibmphp_res.c: fix gcc 3.4 compilation I got the following compile error when trying to build 2.4.28-pre2 using gcc 3.4: <-- snip --> ... gcc-3.4 -D__KERNEL__ -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -fno-unit-at-a-time -D_LINUX -I/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/drivers/acpi -nostdinc -iwithprefix include -DKBUILD_BASENAME=ibmphp_res -c -o ibmphp_res.o ibmphp_res.c ibmphp_res.c: In function `ibmphp_rsrc_init': ibmphp_res.c:45: sorry, unimplemented: inlining failed in call to 'find_bus_wprev': function body not available ibmphp_res.c:237: sorry, unimplemented: called from here ibmphp_res.c:45: sorry, unimplemented: inlining failed in call to 'find_bus_wprev': function body not available ibmphp_res.c:261: sorry, unimplemented: called from here ibmphp_res.c:45: sorry, unimplemented: inlining failed in call to 'find_bus_wprev': function body not available ibmphp_res.c:284: sorry, unimplemented: called from here make[3]: *** [ibmphp_res.o] Error 1 make[3]: Leaving directory `/home/bunk/linux/kernel-2.4/linux-2.4.28-pre2-full/drivers/hotplug' <-- snip --> The patch below fixes this issue by uninlining find_bus_wprev (as done in 2.6). Signed-off-by: Adrian Bunk ChangeSet@1.1525.2.1, 2004-08-27 08:52:10-03:00, paulkf@microgate.com [PATCH] synclinkmp transmit eom fix Bug Fixes: * Fix transmit end of message (EOM) processing to work correctly with hardware auto CTS feature * Fix oops in error path if hardware diags fail during device initialization Cosmetic change: * Use existing macros for address space size instead of hardcoded values Signed-off-by: Paul Fulghum ChangeSet@1.1493.5.33, 2004-08-27 16:28:05+10:00, nathans@sgi.com [XFS] Ensure maxagi not updated early during growfs, conflicts with concurrent inode allocations. Fix from ASANO Masahiro. SGI Modid: xfs-linux:xfs-kern:177699a Signed-off-by: Nathan Scott ChangeSet@1.1493.5.32, 2004-08-27 16:24:52+10:00, hch@sgi.com [XFS] Fix warnings in xfs_bmap.c SGI Modid: xfs-linux:xfs-kern:177663a Signed-off-by: Nathan Scott ChangeSet@1.1493.5.31, 2004-08-27 16:21:42+10:00, nathans@sgi.com [XFS] Support for default quota limits via the zero dquot (ala grace times). SGI Modid: xfs-linux:xfs-kern:177627a Signed-off-by: Nathan Scott ChangeSet@1.1493.5.30, 2004-08-27 16:09:42+10:00, roehrich@sgi.com [XFS] Change DMAPI dm_punch_hole to punch holes, rather than just truncate files. SGI Modid: xfs-linux:xfs-kern:177585a Signed-off-by: Nathan Scott ChangeSet@1.1525.1.14, 2004-08-26 14:58:43-07:00, davem@nuts.davemloft.net [SPARC64]: Fix arg passing to copy_in_user(). ChangeSet@1.1525.1.13, 2004-08-26 14:57:52-07:00, davem@nuts.davemloft.net [SPARC64]: Remove memcpy/bzero symbol usage in sparc64_do_profile. Signed-off-by: David S. Miller ChangeSet@1.1525.1.12, 2004-08-26 14:42:46-07:00, davem@nuts.davemloft.net [SPARC64]: Update defconfig. ChangeSet@1.1525.1.11, 2004-08-26 14:26:22-07:00, davem@nuts.davemloft.net [SPARC64]: Revamped memcpy infrastructure. - Make it easier to maintain the Ultra-I vs. Ultra-III memcpy implementations. Before you had to maintain 3 different entire copies of the routines. - Kill %asi register writing Ultra-I single memcpy loop for both user and kernel. Was not worth it. - Simplify exception detection and handling enormously. Backported from 2.6.x Signed-off-by: David S. Miller ChangeSet@1.1525.1.10, 2004-08-26 14:16:05-07:00, davem@nuts.davemloft.net [SPARC64]: Use saner local label names in Ultra3 copies. This makes the kernel profiles look much more meaningful. Signed-off-by: David S. Miller ChangeSet@1.1525.1.9, 2004-08-26 14:15:23-07:00, davem@nuts.davemloft.net [SPARC64]: Remove memcpy Ultra3 PCACHE patching trick. It could never be enabled safely, so just toss the code. Signed-off-by: David S. Miller ChangeSet@1.1525.1.7, 2004-08-25 08:10:33-03:00, marcelo@logos.cnet Changed EXTRAVERSION to -pre2 TAG: v2.4.28-pre2