]> git.hungrycats.org Git - linux/log
linux
22 years agoMerge intel.com:/home/lenb/bk/linux-2.6.4
Len Brown [Tue, 24 Feb 2004 12:14:48 +0000 (07:14 -0500)]
Merge intel.com:/home/lenb/bk/linux-2.6.4
into intel.com:/home/lenb/src/linux-acpi-test-2.6.4

22 years ago[PATCH] CONFIG_REGPARM breaks non-asmlinkage syscalls
Andreas Gruenbacher [Tue, 24 Feb 2004 04:29:27 +0000 (20:29 -0800)]
[PATCH] CONFIG_REGPARM breaks non-asmlinkage syscalls

With CONFIG_REGPARM=y, syscalls must be declared asmlinkage or else
calling them will fail.

This fix adds a few missing declarations for sys_fadvise64_64() and
sys_remap_file_pages().

22 years ago[PATCH] fix SCSI non-sector bio backed IO
Jens Axboe [Tue, 24 Feb 2004 04:26:03 +0000 (20:26 -0800)]
[PATCH] fix SCSI non-sector bio backed IO

This fixes the SCSI layer to handle non-sector-aligned requests from
SG_IO (and potentially anything else producing these requests) that
could stall the machine and cause all sorts of funnies depending on the
low level driver used.

22 years ago[PATCH] Undo SCSI 8-byte alignment relaxation
James Bottomley [Tue, 24 Feb 2004 04:25:53 +0000 (20:25 -0800)]
[PATCH] Undo SCSI 8-byte alignment relaxation

This makes the default alignment requirements be 512 bytes for SCSI,
the way it used to be.

Jens will fix the SCSI layer problems, but low-level drivers might have
other restrictions on alignment.

22 years ago[PATCH] fix IRQBALANCE Kconfig dependencies
James Bottomley [Tue, 24 Feb 2004 01:53:03 +0000 (17:53 -0800)]
[PATCH] fix IRQBALANCE Kconfig dependencies

CONFIG_IRQBALANCE only affects arch/i386/kernel/io_apic.c, so you should
only see it as an option if you actually have one of those.  This patch
makes IRQBALANCE depend on X86_IO_APIC.

22 years agoMerge bk://linux-scsi.bkbits.net/scsi-for-linus-2.6
Linus Torvalds [Tue, 24 Feb 2004 01:35:01 +0000 (17:35 -0800)]
Merge bk://linux-scsi.bkbits.net/scsi-for-linus-2.6
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years ago[PATCH] intermezzo ->permission() idiocy
Alexander Viro [Tue, 24 Feb 2004 00:09:51 +0000 (16:09 -0800)]
[PATCH] intermezzo ->permission() idiocy

... yes, Virginia, some perverts *do* tweak widely shared
method tables and do that in very sensitive spots, at that...

22 years ago[PATCH] ppc64 oops on /proc/cpuinfo
Benjamin Herrenschmidt [Tue, 24 Feb 2004 00:03:49 +0000 (16:03 -0800)]
[PATCH] ppc64 oops on /proc/cpuinfo

This fixes the ppc64 /proc/cpuinfo oops introduced with the lmb
cleanups.

We are trying to slowly phase out the LMB stuff (it mostly duplicates
functionality of bootmem...), this part got overlooked.

22 years agoMerge intel.com:/home/lenb/bk/linux-2.6.4
Len Brown [Mon, 23 Feb 2004 15:21:42 +0000 (10:21 -0500)]
Merge intel.com:/home/lenb/bk/linux-2.6.4
into intel.com:/home/lenb/src/linux-acpi-test-2.6.4

22 years agoMerge bk://bk.linux1394.org/ieee1394-2.6
Linus Torvalds [Mon, 23 Feb 2004 13:42:38 +0000 (05:42 -0800)]
Merge bk://bk.linux1394.org/ieee1394-2.6
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years agoIEEE1394(r1161): Major cleanup and addition of a ud-class, to make things even cleaner.
Ben Collins [Mon, 23 Feb 2004 18:54:17 +0000 (13:54 -0500)]
IEEE1394(r1161): Major cleanup and addition of a ud-class, to make things even cleaner.

22 years agoohci1394(r1160): Disable cross_bound() check, and add code to try to enable connected...
Ben Collins [Mon, 23 Feb 2004 14:41:21 +0000 (09:41 -0500)]
ohci1394(r1160): Disable cross_bound() check, and add code to try to enable connected ports.

22 years agoIEEE1394(r1159): Merge run_packet_complete into kernel thread.
Ben Collins [Mon, 23 Feb 2004 14:30:53 +0000 (09:30 -0500)]
IEEE1394(r1159): Merge run_packet_complete into kernel thread.

22 years agoSBP2(r1158): We don't need a remove_host callback in sbp2.
Ben Collins [Mon, 23 Feb 2004 14:10:20 +0000 (09:10 -0500)]
SBP2(r1158): We don't need a remove_host callback in sbp2.

22 years agoIEEE1394(r1157): Create a kernel thread to handle all run_packet_complete() calls.
Ben Collins [Mon, 23 Feb 2004 13:34:39 +0000 (08:34 -0500)]
IEEE1394(r1157): Create a kernel thread to handle all run_packet_complete() calls.

22 years ago[PATCH] ppc64: fix cmd_line bugs
Andrew Morton [Mon, 23 Feb 2004 12:50:28 +0000 (04:50 -0800)]
[PATCH] ppc64: fix cmd_line bugs

From: Anton Blanchard <anton@samba.org>

There were a number of bugs in our cmd_line handling:

- We were looking at cmd_line after it had been tokenised. Use saved_command
  line instead
- Use strlcpy instead of memcpy/explicit NULL set.
- Remove magic numbers in iseries cmd_line parsing

22 years agoeth1394(r1156): Fix oddities when MacOSX is IRM. Also MacOSX sets an invalid sspd...
Ben Collins [Mon, 23 Feb 2004 12:49:16 +0000 (07:49 -0500)]
eth1394(r1156): Fix oddities when MacOSX is IRM. Also MacOSX sets an invalid sspd in the ARP reply.

22 years ago[PATCH] ppc64: uniprocessor compile fixes
Andrew Morton [Mon, 23 Feb 2004 12:46:17 +0000 (04:46 -0800)]
[PATCH] ppc64: uniprocessor compile fixes

From: Anton Blanchard <anton@samba.org>

UP compile fixes.

22 years ago[PATCH] ppc64: restore cpu names
Andrew Morton [Mon, 23 Feb 2004 12:46:07 +0000 (04:46 -0800)]
[PATCH] ppc64: restore cpu names

From: Anton Blanchard <anton@samba.org>

Restore the old cpu names in /proc/cpuinfo, some applications parse this
field.

22 years ago[PATCH] trivial oops formatting cleanups
Andrew Morton [Mon, 23 Feb 2004 12:45:56 +0000 (04:45 -0800)]
[PATCH] trivial oops formatting cleanups

From: Anton Blanchard <anton@samba.org>

Some trivial oops cleanups.

22 years ago[PATCH] fix for NUMA kernel on non NUMA machine
Andrew Morton [Mon, 23 Feb 2004 12:45:46 +0000 (04:45 -0800)]
[PATCH] fix for NUMA kernel on non NUMA machine

From: Anton Blanchard <anton@samba.org>

Set node 0 online, otherwise the recent cpu/memory topology stuff will oops
on a non NUMA machine.

22 years ago[PATCH] ppc64: set err to -ENODEV when a new node doesn't have "interrupt" property.
Andrew Morton [Mon, 23 Feb 2004 12:45:35 +0000 (04:45 -0800)]
[PATCH] ppc64: set err to -ENODEV when a new node doesn't have "interrupt" property.

From: Linda Xie <lxiep@us.ibm.com>

set err to -ENODEV when a new node doesn't have "interrupt" property.

22 years ago[PATCH] ppc64: Fix __get_SP()
Andrew Morton [Mon, 23 Feb 2004 12:45:25 +0000 (04:45 -0800)]
[PATCH] ppc64: Fix __get_SP()

From: Anton Blanchard <anton@samba.org>

__get_SP used to be a function call which meant we allocated a stack
frame before calling it. This meant the SP it returned was one frame
below the current function. Lets call that bogusSP (and the real one
SP).

The new dump_stack was being tail call optimised so it remained one
frame above bogusSP. dump_stack would then store below SP (as the ABI
allows us to) and would stomp over the back link that bogusSP pointed
to (__get_SP had set the back link up so it worked sometimes, just not
all the time).

Fix this by just making __get_SP an inline that returns the current SP.

22 years ago[PATCH] ppc64: don't link some non iSeries stuff
Andrew Morton [Mon, 23 Feb 2004 12:45:14 +0000 (04:45 -0800)]
[PATCH] ppc64: don't link some non iSeries stuff

From: Stephen Rothwell <sfr@canb.auug.org.au>

I think Ben introduced this new file and iSeries doesn't need it.

22 years ago[PATCH] ppc64: PER_CPU irq optimisations
Andrew Morton [Mon, 23 Feb 2004 12:45:03 +0000 (04:45 -0800)]
[PATCH] ppc64: PER_CPU irq optimisations

From: Anton Blanchard <anton@samba.org>

Another decent optimisation found in the ia64 port, dont take the irq
descriptor lock or do the note_interrupt stuff on PER_CPU irqs (ie IPIs).

22 years ago[PATCH] ppc64: Clean up per cpu usage in segment table code.
Andrew Morton [Mon, 23 Feb 2004 12:44:53 +0000 (04:44 -0800)]
[PATCH] ppc64: Clean up per cpu usage in segment table code.

From: Anton Blanchard <anton@samba.org>

Clean up per cpu usage in segment table code.

22 years ago[PATCH] ppc64: Make a number of segment table functions static.
Andrew Morton [Mon, 23 Feb 2004 12:44:42 +0000 (04:44 -0800)]
[PATCH] ppc64: Make a number of segment table functions static.

From: Anton Blanchard <anton@samba.org>

Make a number of segment table functions static.
Clean up one duplicated slbia.

22 years ago[PATCH] Add cpus and NUMA memory nodes to sysfs. Also add cpu physical id.
Andrew Morton [Mon, 23 Feb 2004 12:44:31 +0000 (04:44 -0800)]
[PATCH] Add cpus and NUMA memory nodes to sysfs. Also add cpu physical id.

From: Anton Blanchard <anton@samba.org>

Add cpus and NUMA memory nodes to sysfs. Also add cpu physical id.

22 years ago[PATCH] ppc64: remove get_users in backtrace code
Andrew Morton [Mon, 23 Feb 2004 12:44:21 +0000 (04:44 -0800)]
[PATCH] ppc64: remove get_users in backtrace code

From: Anton Blanchard <anton@samba.org>

The might_sleep infrastructure doesnt like our get_users in the backtrace
code, we often end up with might_sleep warnings inside might_sleep warnings.

Instead just be careful about pointers before dereferencing them.

Also remove the hack where we only printed the bottom 32bits of the WCHAN
value.

22 years ago[PATCH] ppc64: Add stack overflow debugging
Andrew Morton [Mon, 23 Feb 2004 12:44:10 +0000 (04:44 -0800)]
[PATCH] ppc64: Add stack overflow debugging

From: Anton Blanchard <anton@samba.org>

This adds stack overflow debugging much like x86. Its pretty agressive
and checks if > 8kB out of 16kB of stack is used.

22 years ago[PATCH] ppc64: This cleans up the rtasd logic, and also makes it hotplug CPU safe.
Andrew Morton [Mon, 23 Feb 2004 12:44:00 +0000 (04:44 -0800)]
[PATCH] ppc64: This cleans up the rtasd logic, and also makes it hotplug CPU safe.

From: Rusty Russell <rusty@rustcorp.com.au>

This cleans up the rtasd logic, and also makes it hotplug CPU safe.

22 years ago[PATCH] ppc64: remove useless smp_message_pass args
Andrew Morton [Mon, 23 Feb 2004 12:43:49 +0000 (04:43 -0800)]
[PATCH] ppc64: remove useless smp_message_pass args

From: Anton Blanchard <anton@samba.org>

smp_message_pass has a few args that are never used. Remove em.
Also add memory missing memory barrier in smp_call_function_interrupt.

22 years ago[PATCH] ppc64 cpu spinup fixes
Andrew Morton [Mon, 23 Feb 2004 12:43:38 +0000 (04:43 -0800)]
[PATCH] ppc64 cpu spinup fixes

From: Rusty Russell <rusty@rustcorp.com.au>

1) Check for cpuids >= NR_CPUS in prom_hold_cpus: we previously
   overflowed arrays when this happened, and checking it here ensures
   that it doesn't happen elsewhere.  Still move processors to
   secondary_hold (out of OF), but we won't release them.

2) smp_iSeries_message_pass and smp_xics_message_pass don't need to
   iterate if given a specific CPU: cleanup and fix.

3) kick_cpu variants don't need to check cpu range (caller is screwed
   anyway if this happened): replaced with BUG_ON for now.

4) smp_openpic_probe and smp_xics_probe can use cpus_weight(): it's
   faster and simpler.

5) User for_each_cpu/for_each_online_cpu as appropriate.

22 years ago[PATCH] ppc64: fix warning and compile error without CONFIG_SMP
Andrew Morton [Mon, 23 Feb 2004 12:43:29 +0000 (04:43 -0800)]
[PATCH] ppc64: fix warning and compile error without CONFIG_SMP

From: Hollis Blanchard <hollisb@us.ibm.com>

fix warning and compile error without CONFIG_SMP

22 years ago[PATCH] fix naming collision with asm-ppc64/vio.h
Andrew Morton [Mon, 23 Feb 2004 12:43:19 +0000 (04:43 -0800)]
[PATCH] fix naming collision with asm-ppc64/vio.h

From: Dave Boutcher <sleddog@us.ibm.com>

fix naming collision with asm-ppc64/vio.h. The define would be set there and
this file would not get pulled in.

22 years ago[PATCH] ppc64: Fix for valid nvram rtas tokens.
Andrew Morton [Mon, 23 Feb 2004 12:43:07 +0000 (04:43 -0800)]
[PATCH] ppc64: Fix for valid nvram rtas tokens.

From: Olaf Hering <olh@suse.de>

Fix for valid nvram rtas tokens.

22 years ago[PATCH] Add 970FX entry into the cputable.
Andrew Morton [Mon, 23 Feb 2004 12:42:57 +0000 (04:42 -0800)]
[PATCH] Add 970FX entry into the cputable.

From: Jake Moilanen <moilanen@austin.ibm.com>

Add 970FX entry into the cputable.

22 years ago[PATCH] ppc64: Add -funit-at-a-time
Andrew Morton [Mon, 23 Feb 2004 12:42:46 +0000 (04:42 -0800)]
[PATCH] ppc64: Add -funit-at-a-time

From: Anton Blanchard <anton@samba.org>

Be like x86 and enable -funit-at-a-time.
Use the check_gcc hook when enabling -mcpu=power4/-mtune=power4.

22 years ago[PATCH] ppc64: Add DEBUG_STACK_USAGE
Andrew Morton [Mon, 23 Feb 2004 12:42:36 +0000 (04:42 -0800)]
[PATCH] ppc64: Add DEBUG_STACK_USAGE

From: Anton Blanchard <anton@samba.org>

Add DEBUG_STACK_USAGE to ppc64

22 years ago[PATCH] ppc64: print useful flags in oops, like x86
Andrew Morton [Mon, 23 Feb 2004 12:42:24 +0000 (04:42 -0800)]
[PATCH] ppc64: print useful flags in oops, like x86

From: Anton Blanchard <anton@samba.org

print useful flags in oops, like x86

22 years ago[PATCH] ppc64: Add a ppc64 archhelp.
Andrew Morton [Mon, 23 Feb 2004 12:42:14 +0000 (04:42 -0800)]
[PATCH] ppc64: Add a ppc64 archhelp.

From: Jeremy Kerr <jk@ozlabs.org>

Add a ppc64 archhelp.

22 years ago[PATCH] Allow CROSS32_COMPILE to be set via environment variable
Andrew Morton [Mon, 23 Feb 2004 12:42:01 +0000 (04:42 -0800)]
[PATCH] Allow CROSS32_COMPILE to be set via environment variable

From: Dave Hansen <haveblue@us.ibm.com>

Allow CROSS32_COMPILE to be set via environment variable

22 years ago[PATCH] ppc64: add log_rtas_error()
Andrew Morton [Mon, 23 Feb 2004 12:41:50 +0000 (04:41 -0800)]
[PATCH] ppc64: add log_rtas_error()

From: Nathan Fontenot <nfont@austin.ibm.com>

Add log_rtas_error(), to be called in the event that a rtas call
returns a hardware error (-1) so we can log additional information
about the hardware error.

22 years ago[PATCH] ppc64: add rtas slot-error-detail information
Andrew Morton [Mon, 23 Feb 2004 12:41:41 +0000 (04:41 -0800)]
[PATCH] ppc64: add rtas slot-error-detail information

From: Nathan Fontenot <nfont@austin.ibm.com>

Add call to rtas to retrieve slot-error-detail information and
log this data to nvram in an EEH event.

22 years ago[PATCH] ppc64: Fix for ppc64 SMT enablement bug provided by Jimi Xenidis and Michael Day
Andrew Morton [Mon, 23 Feb 2004 12:41:30 +0000 (04:41 -0800)]
[PATCH] ppc64: Fix for ppc64 SMT enablement bug provided by Jimi Xenidis and Michael Day

From: Anton Blanchard <anton@samba.org>

Fix for ppc64 SMT enablement bug provided by Jimi Xenidis and Michael Day

22 years ago[PATCH] ppc64: Remove duplicate pcibios_scan_all_fns definition
Andrew Morton [Mon, 23 Feb 2004 12:41:21 +0000 (04:41 -0800)]
[PATCH] ppc64: Remove duplicate pcibios_scan_all_fns definition

From: Anton Blanchard <anton@samba.org>

Remove duplicate pcibios_scan_all_fns definition

22 years ago[PATCH] ppc64: Potentially avoid an atomic operation in switch_mm.
Andrew Morton [Mon, 23 Feb 2004 12:41:10 +0000 (04:41 -0800)]
[PATCH] ppc64: Potentially avoid an atomic operation in switch_mm.

From: Anton Blanchard <anton@samba.org>

This appeared in the ia64 port and sounds like a good idea. Potentially
avoid an atomic operation in switch_mm.

22 years ago[PATCH] ppc64: cleanup lmb code
Andrew Morton [Mon, 23 Feb 2004 12:40:59 +0000 (04:40 -0800)]
[PATCH] ppc64: cleanup lmb code

From: Anton Blanchard <anton@samba.org>

- remove LMB_MEMORY_AREA, LMB_IO_AREA, we only allocate/reserve memory
  areas now
- remove lmb_property->type, lmb_region->iosize, lmb_region->lcd_size,
  no longer used
- bump number of regions to 128, we'll hit this limit sooner or later
  with our big boxes (if we have more than 64 PCI host bridges the
  reserved array will fill up for example)
- make all the lmb stuff __init
- no need to explicitly zero struct lmb lmb now we zero the BSS early
- we had two functions to dump the lmb array, kill one of them
- move the inline functions into lmb.c, they are only ever called from
  there

22 years ago[PATCH] ppc64: remove dump_regs
Andrew Morton [Mon, 23 Feb 2004 12:40:48 +0000 (04:40 -0800)]
[PATCH] ppc64: remove dump_regs

From: Anton Blanchard <anton@samba.org>

There is a prototype for dump_regs on ppc32 and ppc64 but its never
defined.  Remove it since it conflicts with something in
drivers/atm/horizon.c.

Also clean up ppc64 system.h a bit to match the ppc32 version.

22 years agoSCSI: 53c700: reduce default tag depth to 4
James Bottomley [Mon, 23 Feb 2004 09:07:54 +0000 (03:07 -0600)]
SCSI: 53c700: reduce default tag depth to 4

This is a nice safe depth for older devices (which
we've been having trouble with at larger depths). It
is still adjustable upwards using sysfs if anyone
really needs a larger queue.

22 years ago[PATCH] tty utf8 mode
Andries E. Brouwer [Mon, 23 Feb 2004 08:34:15 +0000 (00:34 -0800)]
[PATCH] tty utf8 mode

This sounds like a good moment to come with this patch.

This is essentially a five-year-old patch by Bruno Haible.  It
introduces utf8 mode, and the effect it has on erasing input characters.

(Side note - without stty support, this is not very useful, but now
the infrastructure is in place)

22 years ago[PATCH] removal of bogus CONFIG_BINFMT_ELF uses
Alexander Viro [Mon, 23 Feb 2004 08:16:36 +0000 (00:16 -0800)]
[PATCH] removal of bogus CONFIG_BINFMT_ELF uses

We never use CONFIG_BINFTM_ELF and CONFIG_BINFMT_ELF_MODULE
anywhere in the source (they are used in makefiles, obviously).
32bit-emulation binfmt_elf_... in arch/* still define those guys
even though it hadn't been needed for years.

Junk removed.

PS: BTW, CONFIG_BINFMT_ELF32 _is_ used in several places in arch/mips and it
looks like they really want
#if defined(CONFIG_BINFMT_ELF32) || defined(CONFIG_BINFMT_ELF32_MODULE)
in these ifdefs, but that's a separate story...

22 years ago[PATCH] blkdev_open/bd_claim vs BLKBSZSET
Nathan Scott [Mon, 23 Feb 2004 08:16:26 +0000 (00:16 -0800)]
[PATCH] blkdev_open/bd_claim vs BLKBSZSET

I was modifying mkfs.xfs to use O_EXCL for 2.6, and hit a snag.
It seems that once I've opened a block dev with O_EXCL I can no
longer issue the BLKBSZSET ioctl to it.

Making this change, the ioctl succeeds cos the original filp bdev owner
from open now matches with the owner in the ioctl call.

22 years agoMerge bk://kernel.bkbits.net/davem/sparc-2.6
Linus Torvalds [Mon, 23 Feb 2004 08:13:17 +0000 (00:13 -0800)]
Merge bk://kernel.bkbits.net/davem/sparc-2.6
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years agoMerge bk://kernel.bkbits.net/davem/tg3-2.6
Linus Torvalds [Mon, 23 Feb 2004 08:11:12 +0000 (00:11 -0800)]
Merge bk://kernel.bkbits.net/davem/tg3-2.6
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years agoMerge davem@nuts.davemloft.net:/disk1/BK/sparc-2.6
David S. Miller [Mon, 23 Feb 2004 07:31:32 +0000 (23:31 -0800)]
Merge davem@nuts.davemloft.net:/disk1/BK/sparc-2.6
into kernel.bkbits.net:/home/davem/sparc-2.6

22 years agoMerge davem@nuts.davemloft.net:/disk1/BK/tg3-2.6
David S. Miller [Mon, 23 Feb 2004 07:27:11 +0000 (23:27 -0800)]
Merge davem@nuts.davemloft.net:/disk1/BK/tg3-2.6
into kernel.bkbits.net:/home/davem/tg3-2.6

22 years agoMerge davem@nuts.davemloft.net:/disk1/BK/net-2.6
David S. Miller [Mon, 23 Feb 2004 07:23:57 +0000 (23:23 -0800)]
Merge davem@nuts.davemloft.net:/disk1/BK/net-2.6
into kernel.bkbits.net:/home/davem/net-2.6

22 years ago[PPPOE]: Kill unneeded ifdef/endif.
Luiz Capitulino [Mon, 23 Feb 2004 07:13:37 +0000 (23:13 -0800)]
[PPPOE]: Kill unneeded ifdef/endif.

22 years ago[PPPOE]: Handle disabled PROC_FS properly.
Luiz Capitulino [Mon, 23 Feb 2004 07:12:40 +0000 (23:12 -0800)]
[PPPOE]: Handle disabled PROC_FS properly.

22 years ago[IPV6]: Handle kmem_cache_create() failure in route.c
Walter Harms [Mon, 23 Feb 2004 07:11:34 +0000 (23:11 -0800)]
[IPV6]: Handle kmem_cache_create() failure in route.c

22 years ago[IPV6]: Handle kmem_cache_create() failure in ip6_fib.c
Walter Harms [Mon, 23 Feb 2004 07:10:59 +0000 (23:10 -0800)]
[IPV6]: Handle kmem_cache_create() failure in ip6_fib.c

22 years ago[IPV4]: Handle kmem_cache_create() failure in ipmr.c
Walter Harms [Mon, 23 Feb 2004 07:10:06 +0000 (23:10 -0800)]
[IPV4]: Handle kmem_cache_create() failure in ipmr.c

22 years ago[IPV4]: Handle kmem_cache_create() failure in inetpeer.c
Walter Harms [Mon, 23 Feb 2004 07:09:27 +0000 (23:09 -0800)]
[IPV4]: Handle kmem_cache_create() failure in inetpeer.c

22 years ago[NET]: Handle kmem_cache_create() failure in neighbour.c
Walter Harms [Mon, 23 Feb 2004 07:08:34 +0000 (23:08 -0800)]
[NET]: Handle kmem_cache_create() failure in neighbour.c

22 years ago[DECNET]: Zap MODULE ifdefs.
Alexander Viro [Mon, 23 Feb 2004 06:54:03 +0000 (22:54 -0800)]
[DECNET]: Zap MODULE ifdefs.

22 years agoIEEE1394(r1155): Get rid of another user of round_up_to_page().
Ben Collins [Mon, 23 Feb 2004 06:46:44 +0000 (01:46 -0500)]
IEEE1394(r1155): Get rid of another user of round_up_to_page().

22 years agoIEEE1394(r1154): Cleanup dma routines, and use vmalloc for sglist allocation.
Ben Collins [Mon, 23 Feb 2004 06:32:41 +0000 (01:32 -0500)]
IEEE1394(r1154): Cleanup dma routines, and use vmalloc for sglist allocation.

22 years ago[PATCH] Fix do_shmat() for CONFIG_SYSVIPC=n
Andrew Morton [Mon, 23 Feb 2004 06:16:13 +0000 (22:16 -0800)]
[PATCH] Fix do_shmat() for CONFIG_SYSVIPC=n

- The do_shmat() stub needs `static inline', not `inline'

- We need errno.h for ENOSYS.

22 years ago[NET]: Set default socket rmem/wmem values more sanely and consistently.
David S. Miller [Mon, 23 Feb 2004 06:09:29 +0000 (22:09 -0800)]
[NET]: Set default socket rmem/wmem values more sanely and consistently.

1) Take into account the fact that struct sk_buff, which counts towards
   socket buffer limits, changes across different platforms.
2) Give ipv4/ipv6 ICMP sockets more accurately the wmem limits they
   want.

22 years ago[PATCH] rename shmat to make it clear it isn't a system call entrypoint
Manfred Spraul [Mon, 23 Feb 2004 05:28:40 +0000 (21:28 -0800)]
[PATCH] rename shmat to make it clear it isn't a system call entrypoint

This renames sys_shmat to do_shmat.  Additionally, I've replaced the
cond_syscall with a conditional inline function.

It touches all archs - only i386 is tested.

22 years agoIEEE1394/Video1394(r1151): Fix typo in last commit
Ben Collins [Mon, 23 Feb 2004 04:49:42 +0000 (23:49 -0500)]
IEEE1394/Video1394(r1151): Fix typo in last commit

22 years agoMerge http://lia64.bkbits.net/to-linus-2.5
Linus Torvalds [Mon, 23 Feb 2004 04:02:46 +0000 (20:02 -0800)]
Merge http://lia64.bkbits.net/to-linus-2.5
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years agoIEEE1394/Video1394(r1150): Initialize d->link list so that failures cleanup properly.
Ben Collins [Mon, 23 Feb 2004 03:48:09 +0000 (22:48 -0500)]
IEEE1394/Video1394(r1150): Initialize d->link list so that failures cleanup properly.

22 years ago[TIGON3]: Bump version and reldate.
David S. Miller [Mon, 23 Feb 2004 03:35:32 +0000 (19:35 -0800)]
[TIGON3]: Bump version and reldate.

22 years ago[TIGON3]: Comment out card RAM validation in tg3_test_dma() for now.
David S. Miller [Mon, 23 Feb 2004 03:33:18 +0000 (19:33 -0800)]
[TIGON3]: Comment out card RAM validation in tg3_test_dma() for now.

22 years ago[TIGON3]: Add Apple tigon3 PCI device id.
ken@miriam.com [Mon, 23 Feb 2004 03:25:38 +0000 (19:25 -0800)]
[TIGON3]: Add Apple tigon3 PCI device id.

22 years agoMerge bk://kernel.bkbits.net/wesolows/sparc32-2.6
David S. Miller [Mon, 23 Feb 2004 02:05:36 +0000 (18:05 -0800)]
Merge bk://kernel.bkbits.net/wesolows/sparc32-2.6
into nuts.davemloft.net:/disk1/BK/sparc-2.6

22 years agoMerge intel.com:/home/lenb/bk/linux-2.6.4
Len Brown [Sun, 22 Feb 2004 22:16:33 +0000 (17:16 -0500)]
Merge intel.com:/home/lenb/bk/linux-2.6.4
into intel.com:/home/lenb/src/linux-acpi-test-2.6.4

22 years agoMerge bk://kernel.bkbits.net/davem/bt-2.6
Linus Torvalds [Sun, 22 Feb 2004 17:09:57 +0000 (09:09 -0800)]
Merge bk://kernel.bkbits.net/davem/bt-2.6
into ppc970.osdl.org:/home/torvalds/v2.5/linux

22 years ago[PATCH] Disable debugging verbosity in macio_asic.c
Benjamin Herrenschmidt [Sun, 22 Feb 2004 15:14:46 +0000 (07:14 -0800)]
[PATCH] Disable debugging verbosity in macio_asic.c

This undef's DEBUG by default in macio_asic, this is now working fairly
well, and the actual output isn't really useful anyway.

22 years ago[PATCH] runtime PM deadlock fix
Andrew Morton [Sun, 22 Feb 2004 15:08:06 +0000 (07:08 -0800)]
[PATCH] runtime PM deadlock fix

From: Alexander Atanasov <alex@ssi.bg>

dpm_runtime_suspend holds dpm_sem and calls dpm_runtime_resume which
deadlocks.  Directly call runtime_resume instead.

22 years ago[PATCH] Require GNU Make version 3.79.1 or later
Andrew Morton [Sun, 22 Feb 2004 15:07:55 +0000 (07:07 -0800)]
[PATCH] Require GNU Make version 3.79.1 or later

From: Adrian Bunk <bunk@fs.tum.de>

GNU make 3.78 reproducible segfaults when building kernel 2.6.

See http://marc.theaimsgroup.com/?l=linux-kernel&m=107333742902438&w=2

Therefore, the minimum requirement for GNU make has to be raised in
Documentation/Changes.

GNU make 3.79.1 works.

22 years ago[PATCH] powernow-k8 frequency handling fix
Andrew Morton [Sun, 22 Feb 2004 15:07:45 +0000 (07:07 -0800)]
[PATCH] powernow-k8 frequency handling fix

From: Pavel Machek <pavel@ucw.cz>

In 2.6.3, cleanups arrived, and they got test for invalid frequencies
wrong.

Only first frequency is allowed to be < FID_TABLE_BOTTOM, but due to
bug it cleanups it allowed no frequencies < FID_TABLE_BOTTOM. This
fixes it.

22 years ago[PATCH] Fix fs/partitions/efi.c printk warnings
Andrew Morton [Sun, 22 Feb 2004 15:07:34 +0000 (07:07 -0800)]
[PATCH] Fix fs/partitions/efi.c printk warnings

Lots of these:

fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 3)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:387: warning: long unsigned int format, u64 arg (arg 3)
fs/partitions/efi.c:395: warning: long unsigned int format, u64 arg (arg 2)
fs/partitions/efi.c:395: warning: long unsigned int format, u64 arg (arg 3)

because it is assuming that you can print a u64 with %ld on 64-bit machines.

22 years ago[PATCH] Fix ISDN v.110.
Andrew Morton [Sun, 22 Feb 2004 15:07:24 +0000 (07:07 -0800)]
[PATCH] Fix ISDN v.110.

From: David Woodhouse <dwmw2@infradead.org>

Now that we coalesce ACKs for outgoing packets, the v.110 code needs to
look at the size it's given rather than assuming it'll get precisely one
callback for each packet sent...

22 years ago[PATCH] telephony: use kernel min/max
Andrew Morton [Sun, 22 Feb 2004 15:07:14 +0000 (07:07 -0800)]
[PATCH] telephony: use kernel min/max

From: "Randy.Dunlap" <rddunlap@osdl.org>

From: Michael Veeck <michael.veeck@gmx.net>

Removes unnecessary min/max macros and use kernel.h macros instead.

22 years ago[PATCH] janitor: media: use kernel min/max
Andrew Morton [Sun, 22 Feb 2004 15:07:04 +0000 (07:07 -0800)]
[PATCH] janitor: media: use kernel min/max

From: "Randy.Dunlap" <rddunlap@osdl.org>

From: Michael Veeck <michael.veeck@gmx.net>

Remove unnecessary min/max macros and changes calls to use kernel.h macros
instead.

22 years ago[PATCH] From: "Randy.Dunlap" <rddunlap@osdl.org>
Andrew Morton [Sun, 22 Feb 2004 15:06:52 +0000 (07:06 -0800)]
[PATCH] From: "Randy.Dunlap" <rddunlap@osdl.org>

From: <WHarms@bfs.de>(Walter Harms)

Hi list,
while checking 2.6.1 for missing checks(). fs/jffs/inode-v23.c
got overlooked somehow.

This patch fixes:
2 unguarded kmem_cache_create
1 unguarded proc_mkdir

22 years ago[PATCH] security oops fix
Andrew Morton [Sun, 22 Feb 2004 15:06:42 +0000 (07:06 -0800)]
[PATCH] security oops fix

From: Stephen Smalley <sds@epoch.ncsc.mil>

Looking at d_splice_alias, it appears that the security_d_instantiate call
in the first case (where it ends up moving an entry in place of the given
dentry) should be passing 'new' rather than 'dentry' to ensure that the
passed dentry is associated with the inode.  This patch should fix the
problem.

(Addresses bugzilla #2153)

22 years ago[PATCH] ext3: fix scheduling-in-spinlock bug
Andrew Morton [Sun, 22 Feb 2004 15:06:31 +0000 (07:06 -0800)]
[PATCH] ext3: fix scheduling-in-spinlock bug

ext3_error() sleeps, so don't call it with the lock held.

22 years ago[PATCH] Another x86-64 fix for problems from the recent merge
Andrew Morton [Sun, 22 Feb 2004 15:06:20 +0000 (07:06 -0800)]
[PATCH] Another x86-64 fix for problems from the recent merge

From: Andi Kleen <ak@suse.de>,
      From Suresh.B.Siddha

Fix another bug caused by the recent IA32e merge.  The full L1 cache size
padding of the GDT didn't work with smaller cache line sizes.  Just revert
that.

I must admit I still don't understand why it didn't work, but it's better
to have a working system for now.

22 years ago[PATCH] Fixes to CodingStyle
Andrew Morton [Sun, 22 Feb 2004 15:06:10 +0000 (07:06 -0800)]
[PATCH] Fixes to CodingStyle

From: Xose Vazquez Perez <xose@wanadoo.es>

Even master's books bring errors :-).  URL has the errata and more
information about the books

22 years ago[PATCH] mtrr: init section usage
Andrew Morton [Sun, 22 Feb 2004 15:06:00 +0000 (07:06 -0800)]
[PATCH] mtrr: init section usage

From: "Randy.Dunlap" <rddunlap@osdl.org>

This patch tries to clean up init section usage in
arch/i386/kernel/cpu/mtrr/*.c.

22 years ago[PATCH] IPMI warning fixes
Andrew Morton [Sun, 22 Feb 2004 15:05:49 +0000 (07:05 -0800)]
[PATCH] IPMI warning fixes

From: Geert Uytterhoeven <geert@linux-m68k.org>

When compiling the IPMI drivers on m68k, I needed a few more includes:

  - <asm/irq.h> (for disable_irq_nosync() and enable_irq())

  - <linux/types.h> (for size_t)

  - <asm/system.h> (for printk())

22 years ago[PATCH] cleanup condsyscall for sysv ipc
Andrew Morton [Sun, 22 Feb 2004 15:05:38 +0000 (07:05 -0800)]
[PATCH] cleanup condsyscall for sysv ipc

From: Manfred Spraul <manfred@colorfullife.com>

Attached is a patch that replaces the #ifndef CONFIG_SYSV syscall stubs
with cond_syscall stubs.

22 years ago[PATCH] nbd: remove PARANOIA and other cleanups
Andrew Morton [Sun, 22 Feb 2004 15:05:27 +0000 (07:05 -0800)]
[PATCH] nbd: remove PARANOIA and other cleanups

From: Paul Clements <Paul.Clements@SteelEye.com>

This patch removes the PARANOIA define (it was always defined anyway).  It
also removes the requests_in/out counters (which weren't always accurate,
and are superfluous anyway, since the block layer keeps its own in_flight
counter).  Also some minor cleanup of comments.

22 years ago[PATCH] nbd: fix set_capacity call
Andrew Morton [Sun, 22 Feb 2004 15:05:17 +0000 (07:05 -0800)]
[PATCH] nbd: fix set_capacity call

From: Paul Clements <Paul.Clements@SteelEye.com>

This patch fixes the initial set_capacity call so that it matches nbd's
internal device size (nbd_device->bytesize).

22 years ago[PATCH] Kconfig help: dm-crypto && cryptoloop
Andrew Morton [Sun, 22 Feb 2004 15:05:05 +0000 (07:05 -0800)]
[PATCH] Kconfig help: dm-crypto && cryptoloop

From: bert hubert <ahu@ds9a.nl>

Current bitkeeper features a cryptoloop which is not safe for journaled
file systems and also a Device Mapper target which is.  The patch below
updates Kconfig to that effect.

22 years ago[PATCH] add range checking to sys_sysctl()
Andrew Morton [Sun, 22 Feb 2004 15:04:55 +0000 (07:04 -0800)]
[PATCH] add range checking to sys_sysctl()

If you pass a huge nlen value into sys_sysctl() it will loop for hours
printing billions of question marks, so range-check the arguments.

Also, remove the code which informs the user that the sysctl is obsolete: it
allows unprivileged users to spam the logs.