]> git.hungrycats.org Git - linux/log
linux
23 years agoMerge http://linux-scsi.bkbits.net/scsi-for-linus-2.5
Linus Torvalds [Mon, 31 Mar 2003 10:38:28 +0000 (02:38 -0800)]
Merge http://linux-scsi.bkbits.net/scsi-for-linus-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge raven.il.steeleye.com:/home/jejb/BK/scsi-misc-2.5
James Bottomley [Mon, 31 Mar 2003 10:30:45 +0000 (04:30 -0600)]
Merge raven.il.steeleye.com:/home/jejb/BK/scsi-misc-2.5
into raven.il.steeleye.com:/home/jejb/BK/scsi-for-linus-2.5

23 years agoscsi_debug version 1.69 for lk 2.5.66
Douglas Gilbert [Mon, 31 Mar 2003 08:52:21 +0000 (02:52 -0600)]
scsi_debug version 1.69 for lk 2.5.66

As people are using scsi_debug to simulate a large
number of disks to play with the dev_t expansion,
some problems have come to light.

Changelog:
   - change num_devs to num_tgts
       so now the user can independently modify the
       number of hosts (add_host), the number of targets
       per host (num_tgts) and the number of luns per
       target (max_luns)
   - num_tgts sets scsi_host::max_id after allowance is
     made for the initiator's target id
   - max_luns default changed to 1 and it is sysfs
     writeable
   - REPORT LUN response cleaned up. Could now possibly
     generate a response indicatimg there were up to
     16,384 logical units available
   - clean up sense buffer generation

Updated http://www.torque.net/sg/sdebug25.html

23 years agoFix SMP/preemption race condition in vm86 entry mode.
Linus Torvalds [Mon, 31 Mar 2003 06:30:11 +0000 (22:30 -0800)]
Fix SMP/preemption race condition in vm86 entry mode.

23 years agoDid you know that C integer constant promotions are different
Linus Torvalds [Mon, 31 Mar 2003 06:29:20 +0000 (22:29 -0800)]
Did you know that C integer constant promotions are different
depending on whether the constant is a hexadecimal one as
opposed to a decimal one?

Let's make it all explicit. There are probably more lurking
around, these were found during development of my C checker tool.

23 years ago[PATCH] scsi queueing weirdness
Jens Axboe [Mon, 31 Mar 2003 01:52:42 +0000 (17:52 -0800)]
[PATCH] scsi queueing weirdness

The queueing logic in scsi_lib looks really odd right now.

- Defers calling elv_next_request() until we actually think we can queue
  something. Always want to do that, request will have been marked
  REQ_STARTED after this, so block layer cannot touch it or merge to it.

- Kill the queue empty check. If elv_next_request() returned a req,
  there's one to queue.

23 years ago[PATCH] Check for disabled local interrupts in "might_sleep()"
David S. Miller [Mon, 31 Mar 2003 01:50:40 +0000 (17:50 -0800)]
[PATCH] Check for disabled local interrupts in "might_sleep()"

Self explanatory, as per the discussion last week.

23 years agoMerge bk://kernel.bkbits.net/davem/net-2.5
Linus Torvalds [Mon, 31 Mar 2003 01:49:51 +0000 (17:49 -0800)]
Merge bk://kernel.bkbits.net/davem/net-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[NETFILTER]: ip_conntrack bugfix for LOCAL_NAT and PPTP.
Paul Mielke [Sun, 30 Mar 2003 23:04:22 +0000 (15:04 -0800)]
[NETFILTER]: ip_conntrack bugfix for LOCAL_NAT and PPTP.

23 years ago[NETFILTER]: ipt_REJECT bugfix for TCP RST packets + asymm. routing.
Harald Welte [Sun, 30 Mar 2003 23:01:47 +0000 (15:01 -0800)]
[NETFILTER]: ipt_REJECT bugfix for TCP RST packets + asymm. routing.

23 years ago[NETFILTER]: iptables iptable_mangle LOCAL_IN bugfix.
Harald Welte [Sun, 30 Mar 2003 22:57:27 +0000 (14:57 -0800)]
[NETFILTER]: iptables iptable_mangle LOCAL_IN bugfix.

23 years ago[NET]: Use might_sleep in alloc_skb.
David S. Miller [Sun, 30 Mar 2003 22:41:24 +0000 (14:41 -0800)]
[NET]: Use might_sleep in alloc_skb.

23 years ago[SPARC64]: Get ALI trident sound working again.
David S. Miller [Sun, 30 Mar 2003 21:52:24 +0000 (13:52 -0800)]
[SPARC64]: Get ALI trident sound working again.

23 years agoMerge nuts.ninka.net:/home/davem/src/BK/sparcwork-2.5
David S. Miller [Sun, 30 Mar 2003 21:31:18 +0000 (13:31 -0800)]
Merge nuts.ninka.net:/home/davem/src/BK/sparcwork-2.5
into nuts.ninka.net:/home/davem/src/BK/sparc-2.5

23 years ago[IPV6]: Fix warning with modular ipv6.
Christoph Hellwig [Sun, 30 Mar 2003 21:22:10 +0000 (13:22 -0800)]
[IPV6]: Fix warning with modular ipv6.

23 years ago[COMPAT]: Fix MSG_CMSG_COMPAT flag passing, kill cmsg_compat_recvmsg_fixup.
Randolph Chung [Sun, 30 Mar 2003 21:19:33 +0000 (13:19 -0800)]
[COMPAT]: Fix MSG_CMSG_COMPAT flag passing, kill cmsg_compat_recvmsg_fixup.

23 years ago[COMPAT]: Fix sock_fprog handling.
Randolph Chung [Sun, 30 Mar 2003 21:18:03 +0000 (13:18 -0800)]
[COMPAT]: Fix sock_fprog handling.

23 years agoMerge nuts.ninka.net:/home/davem/src/BK/network-2.5
David S. Miller [Sun, 30 Mar 2003 21:13:05 +0000 (13:13 -0800)]
Merge nuts.ninka.net:/home/davem/src/BK/network-2.5
into nuts.ninka.net:/home/davem/src/BK/net-2.5

23 years agoMerge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
Linus Torvalds [Sun, 30 Mar 2003 01:17:00 +0000 (17:17 -0800)]
Merge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PCMCIA] Reorganise SA11xx PCMCIA support.
Russell King [Sun, 30 Mar 2003 10:14:49 +0000 (11:14 +0100)]
[PCMCIA] Reorganise SA11xx PCMCIA support.

The SA1100 PCMCIA structure didn't lend itself well to the device
model.  With this reorganisation, we end up with a reasonable
structure which fits better with the driver model.  It is now
obvious that SA11x0-based socket drivers are separate from
SA1111-based socket drivers, and are treated as two separate drivers
by the driver model.

23 years ago[PATCH] readlink in /proc w/ overlong path
Andries E. Brouwer [Sun, 30 Mar 2003 01:14:38 +0000 (17:14 -0800)]
[PATCH] readlink in /proc w/ overlong path

Trying to read the overlong target of a /proc/*/fd/N file descriptor
leads to a SIGSEGV inside the kernel, because the code doesn't check
for an error pointer return from d_path.

23 years ago[PCMCIA] Fix "Removing wireless card triggers might_sleep warnings."
Dominik Brodowski [Sun, 30 Mar 2003 00:13:33 +0000 (00:13 +0000)]
[PCMCIA] Fix "Removing wireless card triggers might_sleep warnings."

Bug 516.

Use schedule_delayed_work instead of a timer should fix this. Thanks
to Andrew Morton and Russell King.

(Added flush_scheduled_work() to ensure our delayed work completes
before we free the pcmcia_bus_socket structure. --rmk)

23 years ago[NET]: Nuke CONFIG_FILTER.
David S. Miller [Sat, 29 Mar 2003 17:45:55 +0000 (09:45 -0800)]
[NET]: Nuke CONFIG_FILTER.

23 years agoUpdate direct-rendering to current DRI CVS tree.
Linus Torvalds [Sat, 29 Mar 2003 14:34:37 +0000 (06:34 -0800)]
Update direct-rendering to current DRI CVS tree.

This adds support for i830 interrupt handling, and new improved
lock context keying. See per-file comments for more detail, as this
commit sadly mixes up a few different things (that's what you get
for not tracking the changes at a fine enough granularity).

23 years agoMerge bk://ldm.bkbits.net/linux-2.5-core
Linus Torvalds [Sat, 29 Mar 2003 02:35:25 +0000 (18:35 -0800)]
Merge bk://ldm.bkbits.net/linux-2.5-core
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge bk://kernel.bkbits.net/jgarzik/misc-2.5
Linus Torvalds [Sat, 29 Mar 2003 02:33:25 +0000 (18:33 -0800)]
Merge bk://kernel.bkbits.net/jgarzik/misc-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[hw_random] add AMD pci id
Jeff Garzik [Sat, 29 Mar 2003 06:27:36 +0000 (01:27 -0500)]
[hw_random] add AMD pci id

Contributed by Andi Kleen

23 years agoMerge bk://kernel.bkbits.net/jgarzik/irda-2.5
Linus Torvalds [Sat, 29 Mar 2003 02:32:51 +0000 (18:32 -0800)]
Merge bk://kernel.bkbits.net/jgarzik/irda-2.5
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PATCH] : IrNET module fix
Jean Tourrilhes [Sat, 29 Mar 2003 04:46:47 +0000 (23:46 -0500)]
[PATCH] : IrNET module fix

o [CORRECT] Fix module refcounting (MOD_INC/DEC => .owner)
o [FEATURE] Add hints to discovery (control channel)

23 years ago[PATCH] : IrDA timer fix
Jean Tourrilhes [Sat, 29 Mar 2003 04:45:08 +0000 (23:45 -0500)]
[PATCH] : IrDA timer fix

o [FEATURE] Make IrDA timers use mod_timer instead of
add+del_timer

23 years ago[PATCH] : irtty-sir ZeroCopy Rx
Jean Tourrilhes [Sat, 29 Mar 2003 04:43:47 +0000 (23:43 -0500)]
[PATCH] : irtty-sir ZeroCopy Rx

o [FEATURE] Enable ZeroCopy Rx in irtty-sir/sir-dev
(provided by the new SIR wrapper in 2.5.61).

23 years ago[PATCH] : irda-usb Rx path cleanup + no clear_halt
Jean Tourrilhes [Sat, 29 Mar 2003 04:42:38 +0000 (23:42 -0500)]
[PATCH] : irda-usb Rx path cleanup + no clear_halt

o [CORRECT] Don't do usb_clear_halt() on USB control pipe
o [FEATURE] Cleanup and simplify the USB Rx path

23 years ago[PATCH] : IrLAP dynamic window code fix
Jean Tourrilhes [Sat, 29 Mar 2003 04:41:51 +0000 (23:41 -0500)]
[PATCH] : IrLAP dynamic window code fix

o [FEATURE] Fix the dynamic window code to properly send the pf bit.
Increase perf by 40% for large packets at SIR.

23 years ago[PATCH] : Discovery locking fixes
Jean Tourrilhes [Sat, 29 Mar 2003 04:41:15 +0000 (23:41 -0500)]
[PATCH] : Discovery locking fixes

o [CRITICA] Fix remaining locking problem with discovery log
o [CRITICA] Don't call expiry callback under spinlock
o [FEATURE] Simplify/cleanup/optimise discovery/expiry code

23 years ago[SERIAL] switch over 8250_cs to pcmcia_register_driver
Christoph Hellwig [Sat, 29 Mar 2003 00:01:38 +0000 (00:01 +0000)]
[SERIAL] switch over 8250_cs to pcmcia_register_driver

23 years ago[PCMCIA] don't inform "driver services" of cardbus-related events
Dominik Brodowski [Fri, 28 Mar 2003 23:55:14 +0000 (23:55 +0000)]
[PCMCIA] don't inform "driver services" of cardbus-related events

23 years ago[PCMCIA] generic suspend/resume capability
Dominik Brodowski [Fri, 28 Mar 2003 23:51:58 +0000 (23:51 +0000)]
[PCMCIA] generic suspend/resume capability

The socket drivers already offer suspend and resume
capability. Integrate this with the driver model, based on a
suggestion by Russell King.

Also, remove two never-used functions from the socket drivers (to_ns).

 drivers/pcmcia/cs.c             |   70 ++++++++++++++++++++--------------------
 drivers/pcmcia/cs_internal.h    |    1
 drivers/pcmcia/hd64465_ss.c     |    2 +
 drivers/pcmcia/i82092.c         |   17 ++++++---
 drivers/pcmcia/i82365.c         |    2 +
 drivers/pcmcia/pci_socket.c     |   15 +-------
 drivers/pcmcia/sa1100_generic.c |    2 +
 drivers/pcmcia/sa1111_generic.c |   14 +-------
 drivers/pcmcia/tcic.c           |    7 +---
 include/pcmcia/ss.h             |    5 ++
 10 files changed, 64 insertions(+), 71 deletions(-)

23 years ago[IPSEC]: IPV6 AH/ESP fixes.
Tom Lendacky [Fri, 28 Mar 2003 20:12:50 +0000 (12:12 -0800)]
[IPSEC]: IPV6 AH/ESP fixes.

23 years ago[IPSEC]: Missing ipv6 policy checks.
Tom Lendacky [Fri, 28 Mar 2003 20:08:41 +0000 (12:08 -0800)]
[IPSEC]: Missing ipv6 policy checks.

23 years ago[PATCH] remove warning for 3c509.c
Martin J. Bligh [Fri, 28 Mar 2003 19:43:27 +0000 (14:43 -0500)]
[PATCH] remove warning for 3c509.c

Get this compile warning:
drivers/net/3c509.c:207: warning: `el3_device_remove' declared `static' but never defined
because the function definition is under
"#if defined(CONFIG_EISA) || defined(CONFIG_MCA)".

This patch puts the declaration under the same conditions.
I'd be shocked if it wasn't correct ;-)

M.

23 years ago[PATCH] MACE ethernet driver update
Paul Mackerras [Fri, 28 Mar 2003 19:43:08 +0000 (14:43 -0500)]
[PATCH] MACE ethernet driver update

This patch updates the MACE ethernet driver, used on older powermacs,
to remove the uses of save_flags/restore_flags/cli/sti and use a
spinlock instead.

Jeff, please send this on to Linus.

Paul.

23 years ago[PATCH] finish init_etherdev conversion for gt96100eth
Dave Jones [Fri, 28 Mar 2003 19:42:48 +0000 (14:42 -0500)]
[PATCH] finish init_etherdev conversion for gt96100eth

- No need to alloc dev->priv (due to init_etherdev usage)
- No need to kfree dev->priv (kfree'd with (dev) already)

23 years ago[PATCH] fix .text.exit error in drivers/net/r8169.c
Adrian Bunk [Fri, 28 Mar 2003 19:42:29 +0000 (14:42 -0500)]
[PATCH] fix .text.exit error in drivers/net/r8169.c

In drivers/net/r8169.c the function rtl8169_remove_one is __devexit but
the pointer to it didn't use __devexit_p resulting in a.text.exit
compile error when !CONFIG_HOTPLUG.

The fix is simple:

23 years ago[SPARC64]: Implement pcibios_bus_to_resource.
David S. Miller [Fri, 28 Mar 2003 19:40:48 +0000 (11:40 -0800)]
[SPARC64]: Implement pcibios_bus_to_resource.

23 years ago[CRYPTO]: Add Deflate algorithm to crypto API.
James Morris [Fri, 28 Mar 2003 19:17:21 +0000 (11:17 -0800)]
[CRYPTO]: Add Deflate algorithm to crypto API.

23 years ago[bonding] bug fixes, and a few minor feature additions
Jay Vosburgh [Fri, 28 Mar 2003 19:14:20 +0000 (14:14 -0500)]
[bonding] bug fixes, and a few minor feature additions

Mainly sync w/ 2.4.x version.

23 years ago[tulip dmfe] add pci id
Dave Jones [Fri, 28 Mar 2003 19:04:22 +0000 (14:04 -0500)]
[tulip dmfe] add pci id

23 years ago[tulip] remove unnecessary linux/version.h includes
Burton N. Windle [Fri, 28 Mar 2003 18:43:48 +0000 (13:43 -0500)]
[tulip] remove unnecessary linux/version.h includes

23 years ago[SPARC64]: Update defconfig.
David S. Miller [Fri, 28 Mar 2003 18:05:07 +0000 (10:05 -0800)]
[SPARC64]: Update defconfig.

23 years agoMerge nuts.ninka.net:/home/davem/src/BK/sparcwork-2.5
David S. Miller [Fri, 28 Mar 2003 18:01:03 +0000 (10:01 -0800)]
Merge nuts.ninka.net:/home/davem/src/BK/sparcwork-2.5
into nuts.ninka.net:/home/davem/src/BK/sparc-2.5

23 years ago[SPARC64]: Fix interrupt enabling on trap return.
David S. Miller [Fri, 28 Mar 2003 14:15:50 +0000 (06:15 -0800)]
[SPARC64]: Fix interrupt enabling on trap return.

23 years ago[PATCH] Remove unused variable in nfs_readpage_result()
Andrew Morton [Fri, 28 Mar 2003 11:21:14 +0000 (03:21 -0800)]
[PATCH] Remove unused variable in nfs_readpage_result()

Remove unused local `fattr' in nfs_readpage_result().

23 years ago[PATCH] sync blockdevs on the final close only
Andrew Morton [Fri, 28 Mar 2003 11:21:07 +0000 (03:21 -0800)]
[PATCH] sync blockdevs on the final close only

From: davej@codemonkey.org.uk

We currently run sync_blockdev() on every close of a blockdev.  This patch
arranges for the sync to happen only on the final close.

23 years ago[PATCH] slab: cache sizes cleanup
Andrew Morton [Fri, 28 Mar 2003 11:21:00 +0000 (03:21 -0800)]
[PATCH] slab: cache sizes cleanup

From: Brian Gerst <bgerst@didntduck.org>

- Reduce code duplication by putting the kmalloc cache sizes into a header
  file.

- Tidy up kmem_cache_sizes_init().

23 years ago[PATCH] register_chrdev_region() leak and race fix
Andrew Morton [Fri, 28 Mar 2003 11:20:52 +0000 (03:20 -0800)]
[PATCH] register_chrdev_region() leak and race fix

- If two CPUs run register_chrdev_region(major == 0) at the same time they
  can get the same major.

  Fix that by extending the lock coverage.

- local variable `cd' was leaky on an error path.

- Add some API commentary.

23 years ago[PATCH] make cdevname() callable from interrupts
Andrew Morton [Fri, 28 Mar 2003 11:20:46 +0000 (03:20 -0800)]
[PATCH] make cdevname() callable from interrupts

We just made bdevname() irq-safe.  This patch makes cdevname() callable from
interrupts as well.

23 years ago[PATCH] Fix 64bit warnings in mm/page_alloc.c
Andrew Morton [Fri, 28 Mar 2003 11:20:39 +0000 (03:20 -0800)]
[PATCH] Fix 64bit warnings in mm/page_alloc.c

From: Andi Kleen <ak@muc.de>

Fix a warning on x86_64 (and presumably ia64) which arises from casting an
int to a pointer.

23 years ago[PATCH] honour VM_DONTEXPAND in vma merging
Andrew Morton [Fri, 28 Mar 2003 11:20:32 +0000 (03:20 -0800)]
[PATCH] honour VM_DONTEXPAND in vma merging

From: Gerd Knorr <kraxel@bytesex.org>

2.5.x kernels don't look at the VM_DONTEXPAND flag when merging multiple vmas
into one.

23 years ago[PATCH] init_inode_once() wants sizeof(struct hlist_head)
Andrew Morton [Fri, 28 Mar 2003 11:20:26 +0000 (03:20 -0800)]
[PATCH] init_inode_once() wants sizeof(struct hlist_head)

From: William Lee Irwin III <wli@holomorphy.com>

inode_init() wants to deal with things in in units of the size of
struct hlist_head, not struct list_head.

23 years ago[PATCH] use page_to_pfn() in __blk_queue_bounce()
Andrew Morton [Fri, 28 Mar 2003 11:20:19 +0000 (03:20 -0800)]
[PATCH] use page_to_pfn() in __blk_queue_bounce()

From: William Lee Irwin III <wli@holomorphy.com>

__blk_queue_bounce() contains an open-coded page_to_pfn() for the
discontig, non-MAP_NR_DENSE() case (wherever MAP_NR_DENSE() went).
This converts it to use the standard page_to_pfn() abstraction.

23 years ago[PATCH] ext3: remove dead code and variables
Andrew Morton [Fri, 28 Mar 2003 11:20:12 +0000 (03:20 -0800)]
[PATCH] ext3: remove dead code and variables

Spotted by Oleg Drokin: remove a couple of local variables which aren't being
used for anything.

23 years ago[PATCH] add a might_sleep() check to kmap()
Andrew Morton [Fri, 28 Mar 2003 11:20:05 +0000 (03:20 -0800)]
[PATCH] add a might_sleep() check to kmap()

kmap() can sleep, but rarely does.  Add a check for kmap() being called from
inappropriate contexts.

23 years ago[PATCH] speed up ext3_sync_file()
Andrew Morton [Fri, 28 Mar 2003 11:19:58 +0000 (03:19 -0800)]
[PATCH] speed up ext3_sync_file()

There is never a need to write out b_assoc_buffers() in ext3_sync_file().

23 years ago[PATCH] tty_io cleanup
Andrew Morton [Fri, 28 Mar 2003 11:19:51 +0000 (03:19 -0800)]
[PATCH] tty_io cleanup

From: Andries.Brouwer@cwi.nl

Adding the unregister_chrdev_region call that is the counterpart to
register_chrdev_region, we get a nice cleanup of tty_io.c.

23 years ago[PATCH] fix to support discontigmem for 16way x440
Andrew Morton [Fri, 28 Mar 2003 11:19:44 +0000 (03:19 -0800)]
[PATCH] fix to support discontigmem for 16way x440

From: Patricia Gaughen <gone@us.ibm.com>

The boot-time ioermap code needs to invalidate the tlb entries after setting
up the new pte.

23 years ago[PATCH] x86 clock override boot option
Andrew Morton [Fri, 28 Mar 2003 11:19:37 +0000 (03:19 -0800)]
[PATCH] x86 clock override boot option

From: john stultz <johnstul@us.ibm.com>

This patch allows one to manually specify the i386 gettimeofday time-source
by passing clock=[pit|tsc|cyclone|...] as a boot argument.  The argument will
override the default probled selection, and in case the selected time-source
not be avalible the code defaults to using the PIT (printing a warning saying
so).

23 years ago[PATCH] wait_on_buffer refcounting checks
Andrew Morton [Fri, 28 Mar 2003 11:19:30 +0000 (03:19 -0800)]
[PATCH] wait_on_buffer refcounting checks

It is generally illegal to wait on an unpinned buffer - another CPU could
free it up even before __wait_on_buffer() has taken a ref against the buffer.

Maybe external locking rules will prevent this in specific cases, but that is
really subtle and fragile as locking rules are evolved.

The patch detects people calling wait_on_buffer() against an unpinned buffer
and issues a diagnostic.

Also remove the get_bh() from __wait_on_buffer().  It is too late.

23 years ago[PATCH] ext3: fix max file size
Andrew Morton [Fri, 28 Mar 2003 11:19:22 +0000 (03:19 -0800)]
[PATCH] ext3: fix max file size

ext3 is only permitting files of up to 32G.  It should be allowing files to
be up to 2TB.

This is because it is running ext3_max_size(1024) before the filesystem's
blocksize has been determined.

23 years ago[PATCH] remove vm_enough_memory double counting
Andrew Morton [Fri, 28 Mar 2003 11:19:16 +0000 (03:19 -0800)]
[PATCH] remove vm_enough_memory double counting

From: Hugh Dickins <hugh@veritas.com>

Stop vm_enough_memory double counting total_swapcache_pages: it dates
from the days when we didn't free swap when freeing swapcache page.

23 years ago[PATCH] handle oom in tmpfs
Andrew Morton [Fri, 28 Mar 2003 11:19:09 +0000 (03:19 -0800)]
[PATCH] handle oom in tmpfs

From: Hugh Dickins <hugh@veritas.com>

move_from_swap_cache and add_to_page_cache_lru are using GFP_ATOMIC,
which can easily fail in an intermittent way.  Rude if shmem_getpage
then fails with -ENOMEM: use blk_congestion_wait() to let kswapd in,
and repeat.

23 years ago[PATCH] tmpfs truncation fix
Andrew Morton [Fri, 28 Mar 2003 11:19:02 +0000 (03:19 -0800)]
[PATCH] tmpfs truncation fix

From: Hugh Dickins <hugh@veritas.com>

Recent testing has shown that swapoff can sneak a page back into the
tmpfs page cache after truncate_inode_pages has cleaned it, before
shmem_truncate resets next_index to stop that: BUG_ON(inode->i_blocks)
in shmem_delete_inode.  So call truncate_inode_pages again to be safe.

23 years ago[PATCH] make add_to_swap_cache() static
Andrew Morton [Fri, 28 Mar 2003 11:18:54 +0000 (03:18 -0800)]
[PATCH] make add_to_swap_cache() static

From: Hugh Dickins <hugh@veritas.com>

Make add_to_swap_cache static, it's only used by read_swap_cache_async;
and since that has just done a GFP_HIGHUSER allocation, surely it's
better for add_to_swap_cache to use GFP_KERNEL than GFP_ATOMIC.

23 years ago[PATCH] permit page unmapping if !CONFIG_SWAP
Andrew Morton [Fri, 28 Mar 2003 11:18:47 +0000 (03:18 -0800)]
[PATCH] permit page unmapping if !CONFIG_SWAP

From: Hugh Dickins <hugh@veritas.com>

Raised #endif CONFIG_SWAP in shrink_list, it was excluding
try_to_unmap of file pages.  Suspect !CONFIG_MMU relied on
that to suppress try_to_unmap, added SWAP_FAIL stub for it.

23 years ago[PATCH] remove SWAP_ERROR
Andrew Morton [Fri, 28 Mar 2003 11:18:40 +0000 (03:18 -0800)]
[PATCH] remove SWAP_ERROR

From: Hugh Dickins <hugh@veritas.com>

Delete unused SWAP_ERROR and non-existent page_over_rsslimit().

23 years ago[PATCH] posix timers: fix double-reporting of timer expiration
Andrew Morton [Fri, 28 Mar 2003 11:18:32 +0000 (03:18 -0800)]
[PATCH] posix timers: fix double-reporting of timer expiration

From: george anzinger <george@mvista.com>

Timer expirations are being reported twice.

23 years ago[PATCH] add flush_cache_page() to install_page()
Andrew Morton [Fri, 28 Mar 2003 11:18:25 +0000 (03:18 -0800)]
[PATCH] add flush_cache_page() to install_page()

install_page() needs to run flush_cache_page() prior to overwriting an
already-established pte.

23 years ago[PATCH] slab: fix off-by-one in size calculation
Andrew Morton [Fri, 28 Mar 2003 11:18:18 +0000 (03:18 -0800)]
[PATCH] slab: fix off-by-one in size calculation

From: Manfred Spraul <manfred@colorfullife.com>

Brian spotted a stupid bug in the slab initialization:

If multiple objects fit into one cacheline, then the allocator ignores
SLAB_HWCACHE_ALIGN and squeezes the objects into the same cacheline.  The
implementation contains an off by one error and thus doesn't work correctly:
For Athlon optimized kernels, the 32-byte slab uses 64 byte of memory.

23 years ago[PATCH] POSIX timers interface long/int cleanup
Andrew Morton [Fri, 28 Mar 2003 11:18:11 +0000 (03:18 -0800)]
[PATCH] POSIX timers interface long/int cleanup

From: Eric Piel <Eric.Piel@Bull.Net>

Fixes some long/int confusion on 64-bit machines which was causing failures
on ia64 - we end up trying to set bits in the 32-63 range on an int and the
kernel locks up.

Also cleans up idr.h.

George has acked this change.

23 years ago[PATCH] initcall debug code
Andrew Morton [Fri, 28 Mar 2003 11:18:04 +0000 (03:18 -0800)]
[PATCH] initcall debug code

The patch is designed to help locate where the kernel is dying during the
startup sequence.

- Boot parameter "initcall_debug" causes the kernel to print out the
  address of each initcall before calling it.

  The kallsyms tables do not cover __init sections, so printing the
  symbolic version of these symbols doesn't work.  They need to be looked up
  in System.map.

- Detect whether an initcall returns with interrupts disabled or with a
  locking imbalance.  If it does, complain and then try to fix it up.

23 years ago[PATCH] aha152x pcmcia updates
Christoph Hellwig [Fri, 28 Mar 2003 03:29:03 +0000 (21:29 -0600)]
[PATCH] aha152x pcmcia updates

(1) use new pcmcia_register_driver() & co API
(2) use scsi_add_host & co.  This needed some restructuring in aha152x.c
(3) add a bunch of missing statics to aha152x.c
(4) move prototypes for functions used by aha152x_stub.c to aha152x.h

23 years agoMerge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
Linus Torvalds [Thu, 27 Mar 2003 13:29:17 +0000 (05:29 -0800)]
Merge bk://bk.arm.linux.org.uk/linux-2.5-pcmcia
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years ago[PCMCIA] fix compilation with PCMCIA_DEBUG on
Dominik Brodowski [Fri, 28 Mar 2003 01:14:05 +0000 (01:14 +0000)]
[PCMCIA] fix compilation with PCMCIA_DEBUG on

23 years ago[PCMCIA] fix pcmcia_bind_driver
Dominik Brodowski [Fri, 28 Mar 2003 01:09:49 +0000 (01:09 +0000)]
[PCMCIA] fix pcmcia_bind_driver

Don't allow "bind_request" to be called before "register_pccard_driver".

23 years agoMerge penguin:v2.5/linux
Linus Torvalds [Thu, 27 Mar 2003 13:14:18 +0000 (05:14 -0800)]
Merge penguin:v2.5/linux
into home.transmeta.com:/home/torvalds/v2.5/linux

23 years agoMerge bk://ldm@bkbits.net/linux-2.5-core
Patrick Mochel [Thu, 27 Mar 2003 11:39:17 +0000 (05:39 -0600)]
Merge bk://ldm@bkbits.net/linux-2.5-core
into osdl.org:/home/mochel/src/kernel/devel/linux-2.5-core

23 years agoMerge osdl.org:/home/mochel/src/kernel/devel/linux-2.5-virgin
Patrick Mochel [Thu, 27 Mar 2003 10:56:52 +0000 (04:56 -0600)]
Merge osdl.org:/home/mochel/src/kernel/devel/linux-2.5-virgin
into osdl.org:/home/mochel/src/kernel/devel/linux-2.5-core

23 years agoMerge bk://linux.bkbits.net/linux-2.5
Patrick Mochel [Thu, 27 Mar 2003 10:31:38 +0000 (04:31 -0600)]
Merge bk://linux.bkbits.net/linux-2.5
into osdl.org:/home/mochel/src/kernel/devel/linux-2.5-virgin

23 years agosysfs: Fix file removal
Patrick Mochel [Thu, 27 Mar 2003 10:27:09 +0000 (04:27 -0600)]
sysfs: Fix file removal

Turns out we do need to do d_delete() on individual files (and symlinks)
when removing them the tree, to make sure they actually get removed from
the hierarchy.

23 years agoMerge hostme.bitkeeper.com:/ua/repos/l/ldm/linux-2.5
Patrick Mochel [Thu, 27 Mar 2003 08:55:03 +0000 (00:55 -0800)]
Merge hostme.bitkeeper.com:/ua/repos/l/ldm/linux-2.5
into hostme.bitkeeper.com:/ua/repos/l/ldm/linux-2.5-core

23 years ago[PATCH] fix ipmi_devintf.c compilation
Corey Minyard [Thu, 27 Mar 2003 07:30:10 +0000 (23:30 -0800)]
[PATCH] fix ipmi_devintf.c compilation

This fixes ipmi compilation (with some documentation updates and another
minor fix, too).

It also fixes a problem with the state machine getting stuck in a
certain error condition.

23 years ago[SPARC64]: Use GFP_ATOMIC in request_irq.
David S. Miller [Thu, 27 Mar 2003 06:00:51 +0000 (22:00 -0800)]
[SPARC64]: Use GFP_ATOMIC in request_irq.

23 years ago[SPARC64]: Implement dump stack and handle dumping currents stack properly.
David S. Miller [Thu, 27 Mar 2003 05:57:08 +0000 (21:57 -0800)]
[SPARC64]: Implement dump stack and handle dumping currents stack properly.

23 years agoMerge raven.il.steeleye.com:/home/jejb/BK/scsi-qlogicfc-2.5
James Bottomley [Thu, 27 Mar 2003 05:35:18 +0000 (23:35 -0600)]
Merge raven.il.steeleye.com:/home/jejb/BK/scsi-qlogicfc-2.5
into raven.il.steeleye.com:/home/jejb/BK/scsi-misc-2.5

23 years agofix scsi/qlogicfc.c stack problems
James Bottomley [Thu, 27 Mar 2003 05:33:43 +0000 (23:33 -0600)]
fix scsi/qlogicfc.c stack problems

Revert Randy's stack fix and apply the one from the -dj tree
instead.

23 years ago[PATCH] block scsi_ioctl.c lk 2.5.66
Douglas Gilbert [Thu, 27 Mar 2003 03:40:15 +0000 (21:40 -0600)]
[PATCH] block scsi_ioctl.c lk 2.5.66

This is a multi-part message in MIME format.

23 years ago[PATCH] scsi_ioctl.c for SEND_DIAGNOSTIC
Douglas Gilbert [Thu, 27 Mar 2003 03:39:43 +0000 (21:39 -0600)]
[PATCH] scsi_ioctl.c for SEND_DIAGNOSTIC

This is a multi-part message in MIME format.

23 years ago[PATCH] Complete support for PC-9800 sub-arch (9/9) SCSI
Osamu Tomita [Thu, 27 Mar 2003 03:38:14 +0000 (21:38 -0600)]
[PATCH] Complete support for PC-9800 sub-arch (9/9) SCSI

This is the patch to support NEC PC-9800 subarchitecture
against 2.5.65-ac4. (9/9)

SCSI host adapter support.
 - BIOS parameter change for PC98.
 - Add pc980155 driver for old PC98.
 - wd33c93.c update error handler for eh_*.
 - wd33c93.h register to int for PIO mode.

Regards,
Osamu Tomita

23 years ago[PATCH] some more entries for the largelun list
Christoph Hellwig [Thu, 27 Mar 2003 03:26:58 +0000 (21:26 -0600)]
[PATCH] some more entries for the largelun list

23 years ago[PATCH] update qlogic and fdomain drivers to use pcmcia_register_driver
Christoph Hellwig [Thu, 27 Mar 2003 03:26:11 +0000 (21:26 -0600)]
[PATCH] update qlogic and fdomain drivers to use pcmcia_register_driver

23 years agoMerge bk://ppc.bkbits.net/for-linus-ppc
Linus Torvalds [Thu, 27 Mar 2003 03:23:55 +0000 (19:23 -0800)]
Merge bk://ppc.bkbits.net/for-linus-ppc
into penguin.transmeta.com:/home/penguin/torvalds/repositories/kernel/linux