]> git.hungrycats.org Git - linux/log
linux
23 years ago[PATCH] dev_t handling cleanups (1/12)
Alexander Viro [Sun, 31 Aug 2003 05:50:17 +0000 (22:50 -0700)]
[PATCH] dev_t handling cleanups (1/12)

removed unused kdev_t stuff, fixed a typo left from the
console->device() conversion.

23 years ago[PATCH] use irqreturn_t in m68knommu/5272 config.c
Greg Ungerer [Sun, 31 Aug 2003 05:44:54 +0000 (22:44 -0700)]
[PATCH] use irqreturn_t in m68knommu/5272 config.c

Fix up interrupt handler type to be irqreturn_t.

23 years ago[PATCH] use irqreturn_t in m68knommu/5249 config.c
Greg Ungerer [Sun, 31 Aug 2003 05:44:46 +0000 (22:44 -0700)]
[PATCH] use irqreturn_t in m68knommu/5249 config.c

Fix up interrupt handler type to be irqreturn_t.

23 years ago[PATCH] use irqreturn_t in m68knommu/5206e config.c
Greg Ungerer [Sun, 31 Aug 2003 05:44:38 +0000 (22:44 -0700)]
[PATCH] use irqreturn_t in m68knommu/5206e config.c

Fix up interrupt handler type to be irqreturn_t.

23 years ago[PATCH] use irqreturn_t in m68knommu/5206 config.c
Greg Ungerer [Sun, 31 Aug 2003 05:44:30 +0000 (22:44 -0700)]
[PATCH] use irqreturn_t in m68knommu/5206 config.c

23 years ago[PATCH] fix in NFSv4 server for bad sequence id errors
Neil Brown [Sun, 31 Aug 2003 05:44:22 +0000 (22:44 -0700)]
[PATCH] fix in NFSv4 server for bad sequence id errors

From: "William A.(Andy) Adamson" <andros@citi.umich.edu>

this patch fixes the share state sequenceid bookeeping.

- increment the sequence id on an open that is confirmed
- increment the sequence id on close

23 years ago[PATCH] Track nfsv4 open files by "struct inode" rather than dev/ino/generation
Neil Brown [Sun, 31 Aug 2003 05:44:14 +0000 (22:44 -0700)]
[PATCH] Track nfsv4 open files by "struct inode" rather than dev/ino/generation

23 years ago[PATCH] Fix md superblock incompatabilities with 2.4 kernels.
Neil Brown [Sun, 31 Aug 2003 05:44:06 +0000 (22:44 -0700)]
[PATCH] Fix md superblock incompatabilities with 2.4 kernels.

2.4 kernels are very fussy about some values in the superblock, and
2.6 got them wrong.  This fixes it.

23 years ago[PATCH] Set max_sectors for raid0 only, not for all raid levels.
Neil Brown [Sun, 31 Aug 2003 05:43:58 +0000 (22:43 -0700)]
[PATCH] Set max_sectors for raid0 only, not for all raid levels.

raid1 and multipath have not concept of a chunksize, so basing
max_sectors on it is obviously wrong.

Similary 'linear' has a very different concept of chunksize and
max_sectors doesn't apply.

raid5 does have relevant chunk_size concept, but it has code to
effectively handle any chunksize.

So we only need to set max_sectors based on chunk_size in raid0.

23 years ago[PATCH] Honour the read-ahead for for reads in raid5.
Neil Brown [Sun, 31 Aug 2003 05:43:49 +0000 (22:43 -0700)]
[PATCH] Honour the read-ahead for for reads in raid5.

If we get a failure trying to allocate a stripe_head for a read-ahead
request (the only time we can get a failure), we skip the rest of the
request and fail the whole bio.

23 years ago[PATCH] Fix module ref counting for md.
Neil Brown [Sun, 31 Aug 2003 05:43:41 +0000 (22:43 -0700)]
[PATCH] Fix module ref counting for md.

We don't need to explicitly count references as:
 - refcounting already happens for opens of /dev/md?
 -  when an array is active, a daughter module is loaded which
      locks "md" in.
We just need to make sure we clean up properly on unload. (export_array)

Also, xor needs a null module_exit so that it can be unloaded.

23 years ago[PATCH] bio.c: reduce verbosity at boot
Matthew Wilcox [Sun, 31 Aug 2003 05:40:10 +0000 (22:40 -0700)]
[PATCH] bio.c: reduce verbosity at boot

The queue init is really far too verbose at boot time.  I don't think
these messages add anything to either the end user experience or debug
ability.

Acked by Jens

23 years ago[PATCH] cciss init problem
Jens Axboe [Sun, 31 Aug 2003 05:40:01 +0000 (22:40 -0700)]
[PATCH] cciss init problem

This assigns the queue properly.

23 years ago[PATCH] x86-64 update
Andi Kleen [Sun, 31 Aug 2003 05:33:55 +0000 (22:33 -0700)]
[PATCH] x86-64 update

Make everything compile and boot again.

 - Update defconfig
 - Some minor cleanup
 - Introduce physid_t for APIC masks (fixes UP kernels)
 - Add bandaid for CardBus bridges and broken BIOS (Vojtech)
 - Add bandaid for unsynchronized TSCs  (Vojtech)
 - Fix ffs(0) return value (fixes XFS)
 - Fix compilation with software suspend

23 years ago[PATCH] sis190 driver fix
François Romieu [Sun, 31 Aug 2003 05:32:04 +0000 (22:32 -0700)]
[PATCH] sis190 driver fix

synchronize_irq() requires an argument when built with CONFIG_SMP.

23 years ago[PATCH] Do 32bit addresses in /proc/self/maps if possible
Andi Kleen [Sun, 31 Aug 2003 05:31:55 +0000 (22:31 -0700)]
[PATCH] Do 32bit addresses in /proc/self/maps if possible

As discussed earlier. This implements Linus' idea of printing
the addresses in /proc/<pid>/maps as 32bit if possible.

This works around some broken 32bit programs that cannot parse
64bit addresses as generated by x86-64 kernels.

Also simplifies the code slightly.

23 years ago[PATCH] Make ACPI_SLEEP select SOFTWARE_SUSPEND
Andi Kleen [Sun, 31 Aug 2003 05:31:47 +0000 (22:31 -0700)]
[PATCH] Make ACPI_SLEEP select SOFTWARE_SUSPEND

CONFIG_ACPI_SLEEP doesn't compile without SOFTWARE_SUSPEND.
Make it select it automagically.

This fixes some bugzilla bug whose number I forgot.

23 years ago[PATCH] kill ide_register()
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:50 +0000 (22:17 -0700)]
[PATCH] kill ide_register()

23 years ago[PATCH] kill ide_module_t
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:43 +0000 (22:17 -0700)]
[PATCH] kill ide_module_t

23 years ago[PATCH] remove unused ide_chipsets and IDE_CHIPSET_MODULE
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:35 +0000 (22:17 -0700)]
[PATCH] remove unused ide_chipsets and IDE_CHIPSET_MODULE

23 years ago[PATCH] remove unused exports from ide-probe.c
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:27 +0000 (22:17 -0700)]
[PATCH] remove unused exports from ide-probe.c

export_ide_init_queue() and export_probe_for_drive()

23 years ago[PATCH] kill ide_init_drive() in ide-probe.c
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:18 +0000 (22:17 -0700)]
[PATCH] kill ide_init_drive() in ide-probe.c

also fix comment in init_irq()

23 years ago[PATCH] do not set drive->dn twice in probe_hwif()
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:10 +0000 (22:17 -0700)]
[PATCH] do not set drive->dn twice in probe_hwif()

23 years ago[PATCH] kill ide_modes.h
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:17:02 +0000 (22:17 -0700)]
[PATCH] kill ide_modes.h

23 years ago[PATCH] allow drivers (ie. mediabay) to set hwif->gendev.parent
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:16:54 +0000 (22:16 -0700)]
[PATCH] allow drivers (ie. mediabay) to set hwif->gendev.parent

From Benjamin Herrenschmidt <benh@kernel.crashing.org>.

23 years ago[PATCH] fix ide-lib.c warning when compiling IDE without DMA support
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:16:46 +0000 (22:16 -0700)]
[PATCH] fix ide-lib.c warning when compiling IDE without DMA support

From Mikael Pettersson <mikpe@csd.uu.se>.

23 years ago[PATCH] fix ide.c warning when compiling IDE for non-PCI systems
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:16:39 +0000 (22:16 -0700)]
[PATCH] fix ide.c warning when compiling IDE for non-PCI systems

From Stephane Ouellette <ouellettes@videotron.ca>.

23 years ago[PATCH] fix PowerMac driver breakage caused by recent dynamic queue change
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:16:29 +0000 (22:16 -0700)]
[PATCH] fix PowerMac driver breakage caused by recent dynamic queue change

From Mikael Pettersson <mikpe@csd.uu.se>.

23 years ago[PATCH] cable detection fixes for HPT37x controllers
Bartlomiej Zolnierkiewicz [Sun, 31 Aug 2003 05:16:21 +0000 (22:16 -0700)]
[PATCH] cable detection fixes for HPT37x controllers

From Duncan Laurie <duncan@sun.com>

This same patch made its way into 2.4 via the -ac tree but hasn't
been put in 2.6 yet.

It fixes some cable detect issues that stem from the fact that the
cable detect pins are also used as address/data lines, so they need
to first be configured as inputs to read valid cable detect state.

23 years ago[PATCH] More ->pid to ->tgid changes
Ulrich Drepper [Sun, 31 Aug 2003 04:32:59 +0000 (21:32 -0700)]
[PATCH] More ->pid to ->tgid changes

One more overlooked area where the proper process ID has to be used:
SysV IPC "pid" values should use the thread group ID, not the per-thread
one.

23 years ago[PATCH] vx_entry.c: remove release timer
Vinay K. Nallamothu [Fri, 22 Aug 2003 22:23:23 +0000 (15:23 -0700)]
[PATCH] vx_entry.c: remove release timer

sound/pcmcia/vx/vx_entry.c:

This patch removes the PCMCIA timer release functionality which is no
longer required. Without this the module does not compile.

23 years agoInput: typo in device matching.
Linus Torvalds [Fri, 22 Aug 2003 22:16:32 +0000 (15:16 -0700)]
Input: typo in device matching.

Too much cut-and-paste, noticed by Dmitry Torokhov <dtor_core@ameritech.net>

23 years agoPPC32: Update some of the example configs
Paul Mackerras [Sun, 24 Aug 2003 05:33:49 +0000 (15:33 +1000)]
PPC32: Update some of the example configs

23 years agoMerge samba.org:/home/paulus/kernel/linux-2.5
Paul Mackerras [Sun, 24 Aug 2003 03:59:25 +0000 (13:59 +1000)]
Merge samba.org:/home/paulus/kernel/linux-2.5
into samba.org:/home/paulus/kernel/for-linus-ppc

23 years agoPPC32: Declare cpu_online_map and cpu_possible_map as cpumask_t.
Paul Mackerras [Sat, 23 Aug 2003 22:58:19 +0000 (08:58 +1000)]
PPC32: Declare cpu_online_map and cpu_possible_map as cpumask_t.

23 years agoPPC32: Add the fadvise64_64 system call.
Paul Mackerras [Sat, 23 Aug 2003 22:15:37 +0000 (08:15 +1000)]
PPC32: Add the fadvise64_64 system call.

On PPC32 we reorder the arguments so they fit into 6 registers.  Glibc will
need a two-line stub to change them from the standard order to the ordering
used by the system call: (fd, advice, offset, len).

23 years agoMerge samba.org:/home/paulus/kernel/linux-2.5
Paul Mackerras [Sat, 23 Aug 2003 20:09:40 +0000 (06:09 +1000)]
Merge samba.org:/home/paulus/kernel/linux-2.5
into samba.org:/home/paulus/kernel/for-linus-ppc

23 years agoPPC32: Define MCA_bus__is_a_macro. From Christoph Hellwig.
Paul Mackerras [Sat, 23 Aug 2003 08:26:30 +0000 (18:26 +1000)]
PPC32: Define MCA_bus__is_a_macro.  From Christoph Hellwig.

23 years agoMerge samba.org:/stuff/paulus/kernel/linux-2.5
Paul Mackerras [Sat, 23 Aug 2003 04:36:05 +0000 (14:36 +1000)]
Merge samba.org:/stuff/paulus/kernel/linux-2.5
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years agoMerge bk://stop.crashing.org/linux-2.6-misc
Paul Mackerras [Fri, 22 Aug 2003 19:54:45 +0000 (05:54 +1000)]
Merge bk://stop.crashing.org/linux-2.6-misc
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years agoMerge bk://stop.crashing.org/linux-2.6-8260
Paul Mackerras [Fri, 22 Aug 2003 19:42:10 +0000 (05:42 +1000)]
Merge bk://stop.crashing.org/linux-2.6-8260
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years agoMerge samba.org:/stuff/paulus/kernel/linux-2.5
Paul Mackerras [Fri, 22 Aug 2003 18:33:59 +0000 (04:33 +1000)]
Merge samba.org:/stuff/paulus/kernel/linux-2.5
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years ago[PATCH] Fix cdrom error handling in 2.6
Erik Andersen [Fri, 22 Aug 2003 11:29:33 +0000 (04:29 -0700)]
[PATCH] Fix cdrom error handling in 2.6

In both 2.4 and in 2.6, error handling for bad cdrom media is
wrong.  And it is my fault I'm afraid, since I botched an earlier
fix for the problem by putting the fix in the wrong spot.

My kids have a "Jumpstart Toddlers" cd they have long since
completely killed, which makes a great test disc.  Without this
fix, the best time projection I can get for completing a dd type
sector copy is about 2 years...  Most of that is spent thrashing
about in kernel space trying to re-read sectors we already know
are not correctable....  After the fix, I was able to rip a copy
the CD (or rather muddle through it getting lots of EIO errors)
in about 15 minutes.

Attached is the fix for 2.6.x,

23 years agoLinux 2.6.0-test4 v2.6.0-test4
Linus Torvalds [Fri, 22 Aug 2003 09:38:56 +0000 (02:38 -0700)]
Linux 2.6.0-test4

23 years agoMerge samba.org:/stuff/paulus/kernel/linux-2.5
Paul Mackerras [Fri, 22 Aug 2003 06:44:16 +0000 (16:44 +1000)]
Merge samba.org:/stuff/paulus/kernel/linux-2.5
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years agoMake USB storage select SCSI support automatically, instead
Linus Torvalds [Fri, 22 Aug 2003 05:01:57 +0000 (22:01 -0700)]
Make USB storage select SCSI support automatically, instead
of requiring users to select SCSI in order to see it.

23 years agoMerge http://nfsclient.bkbits.net/linux-2.5
Trond Myklebust [Fri, 22 Aug 2003 04:05:41 +0000 (21:05 -0700)]
Merge http://nfsclient.bkbits.net/linux-2.5
into fys.uio.no:/home/linux/bitkeeper/nfsclient-2.5

23 years agoVT requires INPUT support: make it be automatically included.
Linus Torvalds [Fri, 22 Aug 2003 02:52:24 +0000 (19:52 -0700)]
VT requires INPUT support: make it be automatically included.

From Christoph Hellwig

23 years ago[PATCH] /proc/kallsyms problem
Rusty Russell [Fri, 22 Aug 2003 02:46:47 +0000 (19:46 -0700)]
[PATCH] /proc/kallsyms problem

From David Mosberger:

> We definitely are seeing some strange crashes in get_ksymbol_core().
> I think the reason is the lack of iterator initialization.

Yes, looks good: reset the iterator at allocation.

23 years ago[PATCH] sedlbauer_cs.c: remove release timer
Christoph Hellwig [Fri, 22 Aug 2003 01:16:24 +0000 (18:16 -0700)]
[PATCH] sedlbauer_cs.c: remove release timer

Remove the PCMCIA release timer that got missed in earlier patch.  This
is required for successful compilation of the driver.

23 years agoMerge http://nfsclient.bkbits.net/linux-2.5
Trond Myklebust [Thu, 21 Aug 2003 17:09:04 +0000 (10:09 -0700)]
Merge http://nfsclient.bkbits.net/linux-2.5
into fys.uio.no:/home/linux/bitkeeper/nfsclient-2.5

23 years agoFix compiler warning about using a wrong type as the argument for
Trond Myklebust [Thu, 21 Aug 2003 17:05:15 +0000 (10:05 -0700)]
Fix compiler warning about using a wrong type as the argument for
nfsroot_mount().

23 years agoFix problem with open(O_EXCL) not creating hashed dentries.
Trond Myklebust [Thu, 21 Aug 2003 17:03:36 +0000 (10:03 -0700)]
Fix problem with open(O_EXCL) not creating hashed dentries.
Problem analysis and fix provided by Andries Brouwer.
Cleanup of nfs_lookup() by me...

23 years ago[PATCH] h8300 support fix (2/2)
Yoshinori Sato [Thu, 21 Aug 2003 15:56:22 +0000 (08:56 -0700)]
[PATCH] h8300 support fix (2/2)

o inline assembler funcs cleanup
o machine depend header cleanup
o interrupt management fix

23 years ago[PATCH] h8300 support fix (1/2)
Yoshinori Sato [Thu, 21 Aug 2003 15:56:04 +0000 (08:56 -0700)]
[PATCH] h8300 support fix (1/2)

o build script update
o interrupt management fix
o mtd support fix
o signal handling fix

23 years agoMerge http://klibc.bkbits.net/early_userspace
Linus Torvalds [Thu, 21 Aug 2003 15:54:21 +0000 (08:54 -0700)]
Merge http://klibc.bkbits.net/early_userspace
into home.osdl.org:/home/torvalds/v2.5/linux

23 years agoMerge bk://cifs.bkbits.net/linux-2.5cifs
Linus Torvalds [Thu, 21 Aug 2003 15:51:07 +0000 (08:51 -0700)]
Merge bk://cifs.bkbits.net/linux-2.5cifs
into home.osdl.org:/home/torvalds/v2.5/linux

23 years agoThe Intel Instruction set manual is wrong on how to test for
Linus Torvalds [Thu, 21 Aug 2003 15:50:34 +0000 (08:50 -0700)]
The Intel Instruction set manual is wrong on how to test for
a valid SEP bit. The errata has it right: you have to have at
least model 3, stepping 3. Not "model >= 3 or stepping >= 3".

Fix the test accordingly.

23 years agoMerge camp4.serpentine.com:/export/bos/bk/pristine/linux-2.5
Bryan O'Sullivan [Thu, 21 Aug 2003 15:08:10 +0000 (08:08 -0700)]
Merge camp4.serpentine.com:/export/bos/bk/pristine/linux-2.5
into camp4.serpentine.com:/export/bos/bk/linux-early

23 years agoDocumentation for initramfs, klibc, and early userspace.
Bryan O'Sullivan [Thu, 21 Aug 2003 14:56:56 +0000 (07:56 -0700)]
Documentation for initramfs, klibc, and early userspace.

23 years agoAdd missing CIFS VFS entry to maintainers list
Steve French [Thu, 21 Aug 2003 14:00:57 +0000 (07:00 -0700)]
Add missing CIFS VFS entry to maintainers list

23 years agoMerge http://nfsclient.bkbits.net/linux-2.5
Trond Myklebust [Thu, 21 Aug 2003 11:45:26 +0000 (04:45 -0700)]
Merge http://nfsclient.bkbits.net/linux-2.5
into fys.uio.no:/home/linux/bitkeeper/nfsclient-2.5

23 years agoFix a problem whereby READDIRPLUS was causing lookups to result in
Trond Myklebust [Thu, 21 Aug 2003 11:43:45 +0000 (04:43 -0700)]
Fix a problem whereby READDIRPLUS was causing lookups to result in
ESTALE errors.

23 years agoA request cannot be used as part of the RTO estimation if it gets
Trond Myklebust [Thu, 21 Aug 2003 11:42:05 +0000 (04:42 -0700)]
A request cannot be used as part of the RTO estimation if it gets
resent since you don't know whether the server is replying to the
first or the second transmission. However we're currently setting
the cutoff point to be the timeout of the first transmission.

This patch moves the cutoff point to the actual start of the
retransmission.

23 years agoIncrease the minimum RTO timer value to 1/10 second. This is more
Trond Myklebust [Thu, 21 Aug 2003 11:33:03 +0000 (04:33 -0700)]
Increase the minimum RTO timer value to 1/10 second. This is more
in line with what is done for TCP.

Signedness corrections when updating RTT.

Be conservative when calculating RTO. Round up the residues.

23 years agoBack out some congestion control changes that were causing trouble,
Trond Myklebust [Thu, 21 Aug 2003 11:29:14 +0000 (04:29 -0700)]
Back out some congestion control changes that were causing trouble,
among other things, for the "soft" mount option.

23 years agoIf an RPC request has to be resent due to a timeout, it turns out
Trond Myklebust [Thu, 21 Aug 2003 11:27:54 +0000 (04:27 -0700)]
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.

Solve by adding rq_private_buf, which is updated atomically from
rq_rcv_buf.

23 years agoVarious RPC client fixes:
Trond Myklebust [Thu, 21 Aug 2003 11:23:34 +0000 (04:23 -0700)]
Various RPC client fixes:

   - Ensure that we reset req->rq_received before resending if the
     server has sent us a garbage reply.

   - Whenever we grab the xprt_lock_write "semaphore" ensure that we
     reset req->rq_bytes_sent.

   - When resending a TCP request, do not interrupt in the
     middle of transmission even if we do get a reply from the server.

   - Protect the call to rpc_calc_rto() against modification while
     we are calculating

23 years agoMerge http://lia64.bkbits.net/to-linus-2.5
Linus Torvalds [Thu, 21 Aug 2003 10:40:41 +0000 (03:40 -0700)]
Merge http://lia64.bkbits.net/to-linus-2.5
into home.osdl.org:/home/torvalds/v2.5/linux

23 years ago[PATCH] Fix APIC ID handling
William Lee Irwin III [Thu, 21 Aug 2003 10:06:22 +0000 (03:06 -0700)]
[PATCH] Fix APIC ID handling

Fix APIC ID lookup.  In the bios_cpu_apicid[] case, it would walk off
the end of bios_cpu_apicid[] and attempt to send APIC INIT messages to
garbage without this patch, and in the NUMA-Q case, it would attempt to
send NMI wakeups to destinations in the broadcast cluster (which is
harmless, but very poor form) without this patch.

23 years ago[PATCH] OProfile: don't assume MSRs stay the same across CPU models
John Levon [Thu, 21 Aug 2003 09:57:28 +0000 (02:57 -0700)]
[PATCH] OProfile: don't assume MSRs stay the same across CPU models

The MSRs might change between CPU models, error out if we haven't met
the CPU before. From Venkatesh Pallipadi

23 years ago[PATCH] OProfile: add a useful statistic
John Levon [Thu, 21 Aug 2003 09:57:19 +0000 (02:57 -0700)]
[PATCH] OProfile: add a useful statistic

Add a stat counting the (relatively common) case where a PC value is logged
but there is no (longer) any executable mapping covering that address.

23 years ago[PATCH] OProfile: export kernel pointer size in oprofilefs
John Levon [Thu, 21 Aug 2003 09:57:11 +0000 (02:57 -0700)]
[PATCH] OProfile: export kernel pointer size in oprofilefs

Tell user-space how big kernel pointers are, as preferable to sniffing /proc/kcore.
Improve the oprofilefs_ulong_to_user() prototype.

23 years ago[PATCH] OProfile: reduce allocations of MSR structs
John Levon [Thu, 21 Aug 2003 09:57:03 +0000 (02:57 -0700)]
[PATCH] OProfile: reduce allocations of MSR structs

Andi Kleen pointed out the MSRs array was a massive bloat source. Reduce
it somewhat by only allocating the amount actually needed for the CPU type.

Untested on Pentium IV - I don't have a machine.

23 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Thu, 21 Aug 2003 09:39:23 +0000 (02:39 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.osdl.org:/home/torvalds/v2.5/linux

23 years agoia64: perfmon update.
David Mosberger [Thu, 21 Aug 2003 08:46:04 +0000 (01:46 -0700)]
ia64: perfmon update.

Here is a ChangeLog for the patch:

        - The perfmon core will invoke the sampling module handler
          routine once for each overflowed PMD. When multiple PMDs
          overflow at the same time (with the same PMU interrupt),
          then up to 64 distinct calls can happen. A common timestamp
          parameter allows the module to identify this kind of
          entries.

        - Changed the module ovfl_ctrl arguments to simplify the reset
          field. Now it is a simple boolean.

        - Updated perfmon.h to convert the "set" field to ushort from
          uint.  Other structure updates to get better layout.

        - Update perfmon_default_smpl.h to reflect the change in
          overflow processing mentioned above.

        - Cleanup some state checking code to use switch-case instead
          of if-then with macros. Make the code more readable and
          easier to optmize for gcc. Thanks to David for the
          suggestion.

        - Added extra safety checks on pfm_context_load() to verify
          that the task actually exists.

        - The default sampling format module now supports the
          fmt_restart_active callbacks. Patch from David.

23 years agoPPC32: Add Magic SysRq support to the MPC8260 platforms.
Tom Rini [Thu, 21 Aug 2003 08:26:49 +0000 (01:26 -0700)]
PPC32: Add Magic SysRq support to the MPC8260 platforms.
From Wade Farnsworth <wfarnsworth@mvista.com>.

23 years agoMerge bk://kernel.bkbits.net//home/mochel/linux-2.5-power
Linus Torvalds [Thu, 21 Aug 2003 08:14:26 +0000 (01:14 -0700)]
Merge bk://kernel.bkbits.net//home/mochel/linux-2.5-power
into home.osdl.org:/home/torvalds/v2.5/linux

23 years agoMerge bk://cifs.bkbits.net/linux-2.5cifs
Linus Torvalds [Thu, 21 Aug 2003 08:13:11 +0000 (01:13 -0700)]
Merge bk://cifs.bkbits.net/linux-2.5cifs
into home.osdl.org:/home/torvalds/v2.5/linux

23 years agoremove spurious logging of message on "create if file does not exist" case (without...
Steve French [Thu, 21 Aug 2003 09:44:51 +0000 (02:44 -0700)]
remove spurious logging of message on "create if file does not exist" case (without O_EXCL) when file exists

23 years agoPPC32: A number of minor KGDB fixes and tweaks.
Tom Rini [Thu, 21 Aug 2003 08:10:44 +0000 (01:10 -0700)]
PPC32: A number of minor KGDB fixes and tweaks.

23 years ago[NET]: Make sure interval member of struct tc_estimator is signed.
Michel Dänzer [Thu, 21 Aug 2003 08:09:09 +0000 (01:09 -0700)]
[NET]: Make sure interval member of struct tc_estimator is signed.

23 years agoretry socket write on EAGAIN. Fix oops in write when tcp session dead.
Steve French [Thu, 21 Aug 2003 08:08:44 +0000 (01:08 -0700)]
retry socket write on EAGAIN.  Fix oops in write when tcp session dead.

23 years agoPPC32: Fix a warning in the 'mktree' boot util.
Tom Rini [Thu, 21 Aug 2003 08:00:49 +0000 (01:00 -0700)]
PPC32: Fix a warning in the 'mktree' boot util.

23 years ago[IPV6]: Fix dangling multicast device references.
David S. Miller [Thu, 21 Aug 2003 07:50:53 +0000 (00:50 -0700)]
[IPV6]: Fix dangling multicast device references.

When addrconf_ifdown() calls ipv6_mc_destroy_dev(), it has NULL'd
out dev->ip6_ptr, which means all in6_dev_get() calls will fail.
So pass an explicit idev into ipv6_dev_mc_dec() in this case so
that we don't leak the all-nodes multicast address reference to the
idev.

23 years ago[cpufreq] Update resume method.
Patrick Mochel [Thu, 21 Aug 2003 07:19:31 +0000 (00:19 -0700)]
[cpufreq] Update resume method.

From Dominik Brodowski.

- Can simplify, since the policy is already known, and we only need to kick
  the CPU.

- Make sure we set the right method in the sysdev driver.

23 years ago[apm] Fix calls to device_{suspend,resume}
Patrick Mochel [Thu, 21 Aug 2003 07:15:20 +0000 (00:15 -0700)]
[apm] Fix calls to device_{suspend,resume}

23 years ago[power] Fix typo.
Patrick Mochel [Thu, 21 Aug 2003 07:12:23 +0000 (00:12 -0700)]
[power] Fix typo.

23 years agoMerge samba.org:/stuff/paulus/kernel/linux-2.5
Paul Mackerras [Thu, 21 Aug 2003 06:53:35 +0000 (16:53 +1000)]
Merge samba.org:/stuff/paulus/kernel/linux-2.5
into samba.org:/stuff/paulus/kernel/for-linus-ppc

23 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Thu, 21 Aug 2003 05:56:09 +0000 (22:56 -0700)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.osdl.org:/home/torvalds/v2.5/linux

23 years ago[power] Update documentation.
Patrick Mochel [Thu, 21 Aug 2003 05:49:17 +0000 (22:49 -0700)]
[power] Update documentation.

23 years ago[power] Fix locking in device_{suspend,resume}
Patrick Mochel [Thu, 21 Aug 2003 05:48:57 +0000 (22:48 -0700)]
[power] Fix locking in device_{suspend,resume}

We don't need to protect against multiple suspend calls, since subsequent
ones will happen on an empty (or at worst - short) list.

23 years agoMerge osdl.org:/home/mochel/src/kernel/linux-2.5-virgin
Patrick Mochel [Thu, 21 Aug 2003 05:27:06 +0000 (22:27 -0700)]
Merge osdl.org:/home/mochel/src/kernel/linux-2.5-virgin
into osdl.org:/home/mochel/src/kernel/linux-2.5-power

23 years agoMerge tiger.hpl.hp.com:/data1/bk/vanilla/linux-2.5
David Mosberger [Thu, 21 Aug 2003 05:24:49 +0000 (22:24 -0700)]
Merge tiger.hpl.hp.com:/data1/bk/vanilla/linux-2.5
into tiger.hpl.hp.com:/data1/bk/lia64/to-linus-2.5

23 years ago[acpi] Make a dummy mp_congig_ioapic_for_sci() function.
Patrick Mochel [Thu, 21 Aug 2003 05:02:33 +0000 (22:02 -0700)]
[acpi] Make a dummy mp_congig_ioapic_for_sci() function.

23 years agoMerge http://linux-sound.bkbits.net/linux-sound
Linus Torvalds [Thu, 21 Aug 2003 04:58:51 +0000 (21:58 -0700)]
Merge http://linux-sound.bkbits.net/linux-sound
into home.osdl.org:/home/torvalds/v2.5/linux

23 years ago[power] Make swsusp-only mm functions available when CONFIG_PM=y
Patrick Mochel [Thu, 21 Aug 2003 04:57:54 +0000 (21:57 -0700)]
[power] Make swsusp-only mm functions available when CONFIG_PM=y

Calls were moved to the PM core, so they must be compiled in to use them.

23 years ago[dmi] Ugh, fixup broken merge once and for all.
Patrick Mochel [Thu, 21 Aug 2003 04:56:44 +0000 (21:56 -0700)]
[dmi] Ugh, fixup broken merge once and for all.

- Didn't realize that mount_root_failed_msg() had been removed, so it was
  accidentally merged back in. It's gone again now, as well as the entire
  (unused) dmi_dump_system() function.

23 years ago[PATCH] remove cramfs maintainership
Jörn Engel [Thu, 21 Aug 2003 04:51:48 +0000 (21:51 -0700)]
[PATCH] remove cramfs maintainership

Dan hasn't been maintaining cramfs for some time.

Mark it orphan until somebody else takes it up.

23 years ago[power] Make sure MTRR uses right methods in sysdev_driver.
Patrick Mochel [Thu, 21 Aug 2003 04:48:21 +0000 (21:48 -0700)]
[power] Make sure MTRR uses right methods in sysdev_driver.