Commit graph

26186 commits

Author SHA1 Message Date
Mike Waychison
5b20192727 [PATCH] x86_64: Fix a race in the free_iommu path
We do this by removing a micro-optimization that tries to avoid grabbing
the iommu_bitmap_lock spinlock and using a bus-locked operation.

This still races with other simultaneous alloc_iommu or free_iommu(size >
1) which both use bus-unlocked operations.

The end result of this race is eventually ending up with an
iommu_gart_bitmap that has bits errornously set all over, making large
contiguous iommu space allocations fail with 'PCI-DMA: Out of IOMMU space'.

Signed-off-by: Mike Waychison <mikew@google.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-22 09:19:52 -07:00
Andi Kleen
ddbe3eec59 [PATCH] x86_64: Pass -32 to the assembler when compiling the 32bit vsyscall pages
This quietens warnings and actually fixes a bug.  The unwind tables would
come out wrong without -32, causing pthread cancellation during them to
crash in the gcc runtime.

The problem seems to only happen with newer binutils (it doesn't happen
with 2.16.91.0.2 but happens wit 2.16.91.0.5)

Thanks to David Altobelli <david.altobelli@hp.com> and Brian Baker
<Brian.B@hp.com> for test case and initial analysis.

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-22 09:19:52 -07:00
Lee Schermerhorn
304dbdb7a4 [PATCH] add migratepage address space op to shmem
Basic problem: pages of a shared memory segment can only be migrated once.

In 2.6.16 through 2.6.17-rc1, shared memory mappings do not have a
migratepage address space op.  Therefore, migrate_pages() falls back to
default processing.  In this path, it will try to pageout() dirty pages.
Once a shared memory page has been migrated it becomes dirty, so
migrate_pages() will try to page it out.  However, because the page count
is 3 [cache + current + pte], pageout() will return PAGE_KEEP because
is_page_cache_freeable() returns false.  This will abort all subsequent
migrations.

This patch adds a migratepage address space op to shared memory segments to
avoid taking the default path.  We use the "migrate_page()" function
because it knows how to migrate dirty pages.  This allows shared memory
segment pages to migrate, subject to other conditions such as # pte's
referencing the page [page_mapcount(page)], when requested.

I think this is safe.  If we're migrating a shared memory page, then we
found the page via a page table, so it must be in memory.

Can be verified with memtoy and the shmem-mbind-test script, both
available at:  http://free.linux.hp.com/~lts/Tools/

Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-22 09:19:52 -07:00
Andy Whitcroft
3b5fd59fdd [PATCH] x86_64: sparsemem does not need node_mem_map
Seems we are trying to init the node_mem_map when we don't need to, for
example when SPARSEMEM is enabled.  This causes the error below during
compilation.  Use CONFIG_FLAT_NODE_MEM_MAP to gate allocation and init.

  arch/x86_64/mm/numa.c: In function `setup_node_zones':
  arch/x86_64/mm/numa.c:191: error: structure has no member
                                                  named `node_mem_map'

Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Acked-by: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-22 09:19:52 -07:00
Tilman Schmidt
73a8881454 [PATCH] isdn4linux: Siemens Gigaset base driver: fix disconnect handling
Fix a possible Oops in the Siemens Gigaset base driver when the device is
unplugged while an ISDN connection is still active, and makes sure that the
isdn4linux link level (LL) is properly informed if a connection is broken
by the USB cable being unplugged.

- Avoid unsafe checks of URB status fields outside the URB completion
  handlers, keep track of in-use URBs myself instead.

- If an isochronous transfer URB completes with status==0, also check the
  status of the frame descriptors.

- Verify length of interrupt messages received from the device.

- Align the length limit on transmitted AT commands with the device
  documentation.

- In case of AT response receive overrun, keep newly arrived instead of old
  unread data.

- Remove redundant check of device ID in the USB probe function.

- Correct and improve some comments and formatting.

Signed-off-by: Tilman Schmidt <tilman@imap.cc>
Acked-by: Hansjoerg Lipp <hjlipp@web.de>
Cc: Karsten Keil <kkeil@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-22 09:19:52 -07:00
Steve French
60808233f3 [CIFS] Readdir fixes to allow search to start at arbitrary position
in directory

Also includes first part of fix to compensate for servers which forget
to return . and .. as well as updates to changelog and cifs readme.

Signed-off-by: Steve French <sfrench@us.ibm.com>
2006-04-22 15:53:05 +00:00
Paul Mackerras
7c8ce71b09 Merge branch 'misc' of master.kernel.org:/pub/scm/linux/kernel/git/galak/powerpc into merge 2006-04-22 19:19:06 +10:00
Steve French
45af7a0f2e [CIFS] Use the kthread_ API instead of opencoding lots of hairy code for kernel
thread creation and teardown.

It does not move the cifsd thread handling to kthread due to problems
found in testing with wakeup of threads blocked in the socket peek api,
but the other cifs kernel threads now use kthread.
Also cleanup cifs_init to properly unwind when thread creation fails.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Steve French <sfrench@us.ibm.com>
2006-04-21 22:52:25 +00:00
Daniel Ritz
48b950ff24 [PATCH] pcmcia/pcmcia_resource.c: fix crash when using Cardbus cards
Using the old ioctl interface together with cardbus card gives a NULL
pointer dereference since cardbus devices don't have a struct pcmcia_device.
also s->io[0].res can be NULL as well.

Fix is to move the pcmcia code after the cardbus code and to check for a null
pointer.

Signed-off-by: Daniel Ritz <daniel.ritz@gmx.ch>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:19 +02:00
Yoichi Yuasa
daaeb72bdf [PATCH] vrc4171: update config
This patch updates "depends on" for PCMCIA_VRC4171.
CONFIG_VRC4171 has been removed, so replace it with CPU_VR41XX && ISA.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:18 +02:00
Dominik Brodowski
2aff541c69 [PATCH] pcmcia: fix oops in static mapping case
As static maps do not have IO resources, this setting oopses. However, as
we do not ever read this value, we can safely remove it.

Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:17 +02:00
Andrew Morton
80a55e923c [PATCH] pcmcia: remove unneeded forward declarations
Also remove a couple of unneeded typecasts.

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:16 +02:00
Dominik Brodowski
a0aab14322 [PATCH] pcmcia: do not set dev_node to NULL too early
If we set dev_node to NULL too early, some drivers which used this to
determine whether unregister_netdev() needs to be called fail when removing
a PCMCIA card.

Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:15 +02:00
Komuro
90ff87008d [PATCH] pcmcia: fix comment for pcmcia_load_firmware
The comment of "pcmcia_load_firmware" is wrong: the
firmware(*.cis) files reside in /lib/firmware/ _not_
/lib/firmware/cis/ .

Signed-off-by: komurojun-mbn@nifty.com
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:14 +02:00
Komuro
6171b88b43 [PATCH] pcmcia: unload second device first
Use list_add instead of list_add_tail for pcmcia_device_add
so that second device of multi-function-card will be unloaded first.

Signed-off-by: komurojun-mbn@nifty.com
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:13 +02:00
Dominik Brodowski
6542729809 [PATCH] pcmcia: add new ID to pcnet_cs
This adds a new ID to pcnet_cs, as noted by Kuro Moji.

Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
2006-04-22 00:49:12 +02:00
Kyle McMartin
223232de06 [PARISC] MAINTAINERS
Add myself to maintainers and add the parisc trees.

Acked-by: Matthew Wilcox <willy@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:36 +00:00
Kyle McMartin
1b52d7c221 [PARISC] Make ioremap default to _nocache
Since it is way more work to change most drivers to comply with parisc, take
the easy way out and make ioremap _NO_CACHE by default. This is in line with
what powerpc does.

Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:35 +00:00
Kyle McMartin
6ca773cf8b [PARISC] Add new entries to the syscall table
Most are easy, but sync_file_range needed special handling when entering
through the 32-bit syscall table.

Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:35 +00:00
Helge Deller
2fd8303816 [PARISC] Further work for multiple page sizes
More work towards supporing multiple page sizes on 64-bit. Convert
some assumptions that 64bit uses 3 level page tables into testing
PT_NLEVELS. Also some BUG() to BUG_ON() conversions and some cleanups
to assembler.

Signed-off-by: Helge Deller <deller@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:34 +00:00
Matthew Wilcox
d668da80d6 [PARISC] Fix up hil_kbd.c mismerge
Signed-off-by: Matthew Wilcox <willy@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:34 +00:00
Helge Deller
8ea6947b4b [PARISC] defconfig updates
Make the defconfig more generally useful. Turn on IPv6, modules,
cardbus, etc. Boots 32bit on 715 with HIL, B160L with sound,
PrecisionBook, and C3000.

Signed-off-by: Helge Deller <deller@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:33 +00:00
Grant Grundler
b312c33e36 [PARISC] Document that we tolerate "Relaxed Ordering"
This means "DMA Read returns" can bypass "MMIO Writes".
Violating the PCI specs in this case improves outbound DMA "flows"
and is currently not required by any drivers.

This is NOT a new behavior. Previous chipsets did this
already and I believe ZX1 PDC was already setting this
for hpux. I just want to further document the behavior.

Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:33 +00:00
Helge Deller
67a5a59d33 [PARISC] Misc. janitorial work
Fix a spelling mistake, add a KERN_INFO flag, and fix some whitespace
uglies.

Signed-off-by: Helge Deller <deller@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:32 +00:00
Helge Deller
b2d6b9fb35 [PARISC] EISA regions must be mapped NO_CACHE
Signed-off-by: Helge Deller <deller@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:32 +00:00
Stuart Brady
e74eb808ba [PARISC] OSS ad1889: Match register names with ALSA driver
Signed-off-by: Stuart Brady <sdb@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
2006-04-21 22:20:31 +00:00
Steve French
296034f7de [CIFS] Don't allow a backslash in a path component
Unless Posix paths have been negotiated, the backslash, "\", is not a valid
character in a path component.

Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
Signed-off-by: Steve French  <sfrench@us.ibm.com>
2006-04-21 18:18:37 +00:00
Steve French
0bd4fa977f [CIFS] [CIFS] Do not take rename sem on most path based calls (during
building of full path) to avoid hang rename/readdir hang

Reported by Alan Tyson

Signed-off-by: Steve French <sfrench@us.ibm.com>
2006-04-21 18:17:42 +00:00
Benjamin Herrenschmidt
4d6c58899c [PATCH] powerpc: fix oops in alsa powermac driver
This fixes an oops in 2.6.16.X when loading the snd_powermac module. The
name of the requested module changed during the 2.6.16 development cycle
from i2c-keylargo to i2c-powermac.

Signed-off-by: Guido Guenther <agx@sigxcpu.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:34:26 +10:00
Will Schmidt
5ad1bfeaf7 [PATCH] powerpc: update {g5,iseries,pseries}_defconfigs
Default choices for all.

Signed-off-by: Will Schmidt <willschm@us.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:34:03 +10:00
Becky Bruce
ea1e847cc2 [PATCH] ppc: Fix powersave code on arch/ppc
Fix asm_offsets.c and entry.S to work with the new power save code.
Changes in arch/powerpc needed to exist in arch/ppc as well since the
idle code is shared by both ppc and powerpc..

Signed-off-by: Becky Bruce <becky.bruce@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:33:22 +10:00
Arnd Bergmann
23b2527d5e [PATCH] powerpc/cell: remove BUILD_BUG_ON and add sys_tee to spu_syscall_table
Every time a new syscall gets added, a BUILD_BUG_ON in
arch/powerpc/platforms/cell/spu_callbacks.c gets triggered.
Since the addition of a new syscall is rather harmless,
the error should just be removed.

While we're here, add sys_tee to the list and add a comment
to systbl.S to remind people that there is another list
on powerpc.

Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:32:50 +10:00
Benjamin Herrenschmidt
60162e498e [PATCH] powermac: Fix i2c on keywest based chips
The new i2c implementation for PowerMac has a regression that causes the
hardware to go out of state when probing non-existent devices. While
fixing that, I also found & fixed a couple of other corner cases. This
fixes booting with a pbbuttons version that scans the i2c bus for an LMU
controller among others. Tested on a dual G5 with thermal control (which
has heavy i2c activity) with no problem so far.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:29:46 +10:00
Olof Johansson
2889773131 [PATCH] powerpc: Lower threshold for DART enablement to 1GB
Turn on the DART already at 1GB. This is needed because of crippled
devices in some systems, i.e. Airport Extreme cards, only supporting
30-bit DMA addresses.

Otherwise, users with between 1 and 2GB of memory will need to manually
enable it with iommu=force, and that's no good.

Some simple performance tests show that there's a slight impact of
enabling DART, but it's in the 1-3% range (kernel build with disk I/O
as well as over NFS).

iommu=off can still be used for those who don't want to deal with the
overhead (and don't need it for any devices).

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:29:37 +10:00
Olof Johansson
7daa411b81 [PATCH] powerpc: IOMMU support for honoring dma_mask
Some devices don't support full 32-bit DMA address space, which we currently
assume. Add the required mask-passing to the IOMMU allocators.

Signed-off-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2006-04-21 22:28:55 +10:00
Linus Torvalds
f4ffaa452e Merge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6
* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6: (21 commits)
  [PATCH] wext: Fix RtNetlink ENCODE security permissions
  [PATCH] bcm43xx: iw_priv_args names should be <16 characters
  [PATCH] bcm43xx: sysfs code cleanup
  [PATCH] bcm43xx: fix pctl slowclock limit calculation
  [PATCH] bcm43xx: fix dyn tssi2dbm memleak
  [PATCH] bcm43xx: fix config menu alignment
  [PATCH] bcm43xx wireless: fix printk format warnings
  [PATCH] softmac: report when scanning has finished
  [PATCH] softmac: fix event sending
  [PATCH] softmac: handle iw_mode properly
  [PATCH] softmac: dont send out packets while scanning
  [PATCH] softmac: return -EAGAIN from getscan while scanning
  [PATCH] bcm43xx: set trans_start on TX to prevent bogus timeouts
  [PATCH] orinoco: fix truncating commsquality RID with the latest Symbol firmware
  [PATCH] softmac: fix spinlock recursion on reassoc
  [PATCH] Revert NET_RADIO Kconfig title change
  [PATCH] wext: Fix IWENCODEEXT security permissions
  [PATCH] wireless/atmel: send WEXT scan completion events
  [PATCH] wireless/airo: clean up WEXT association and scan events
  [PATCH] softmac uses Wiress Ext.
  ...
2006-04-20 15:26:25 -07:00
Linus Torvalds
d3d3cf05ed Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [EBTABLES]: Clean up vmalloc usage in net/bridge/netfilter/ebtables.c
  [NET]: Add skb->truesize assertion checking.
  [TCP]: Account skb overhead in tcp_fragment
  [SUNGEM]: Marvell PHY suspend.
  [LLC]: Use pskb_trim_rcsum() in llc_fixup_skb().
  [NET]: sockfd_lookup_light() returns random error for -EBADFD
2006-04-20 15:25:37 -07:00
Linus Torvalds
a6db522c93 Merge master.kernel.org:/home/rmk/linux-2.6-arm
* master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] for_each_possible_cpu
  [ARM] add_memory() build fix
  [ARM] 3483/1: ixp23xx: update defconfig to 2.6.17-rc2
  [ARM] 3482/1: ixp2000: update defconfig to 2.6.17-rc2
  [ARM] 3481/1: ep93xx: update defconfig to 2.6.17-rc2
  [ARM] 3480/1: ixp4xx: fix irq2gpio array type
2006-04-20 15:22:41 -07:00
Jeff Garzik
f18b95c3e2 Merge branch 'upstream-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-2.6 2006-04-20 17:36:10 -04:00
Sergei Shtylyov
17c281ab3e [PATCH] NEx000: fix RTL8019AS base address for RBTX4938
Correct the base address of the Realtek RTL8019AS chip on the Toshiba RBTX4938
board -- this should make the driver work at least when CONFIG_PCI is enabled.

Signed-off-by: Yuri Shpilevsky <yshpilevsky@ru.mvista.com>
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-04-20 17:28:57 -04:00
KAMEZAWA Hiroyuki
036999d501 [ARM] for_each_possible_cpu
for_each_cpu() actually iterates across all possible CPUs.  We've had mistakes
in the past where people were using for_each_cpu() where they should have been
iterating across only online or present CPUs.  This is inefficient and
possibly buggy.

We're renaming for_each_cpu() to for_each_possible_cpu() to avoid this in the
future.

This patch replaces for_each_cpu with for_each_possible_cpu.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:48:02 +01:00
Andrew Morton
1c97b73e89 [ARM] add_memory() build fix
This is back again.  Offending patch is x86_64-mm-hotadd-reserve.patch

arch/arm/kernel/setup.c:435: error: conflicting types for 'add_memory'
include/linux/memory_hotplug.h:102: error: previous declaration of 'add_memory' was here
arch/arm/kernel/setup.c:435: error: conflicting types for 'add_memory'
include/linux/memory_hotplug.h:102: error: previous declaration of 'add_memory' was here

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:41:18 +01:00
Lennert Buytenhek
4b91ff4f42 [ARM] 3483/1: ixp23xx: update defconfig to 2.6.17-rc2
Patch from Lennert Buytenhek

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:24:41 +01:00
Lennert Buytenhek
d2f9caa1be [ARM] 3482/1: ixp2000: update defconfig to 2.6.17-rc2
Patch from Lennert Buytenhek

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:24:40 +01:00
Lennert Buytenhek
67c6024cb4 [ARM] 3481/1: ep93xx: update defconfig to 2.6.17-rc2
Patch from Lennert Buytenhek

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:24:39 +01:00
Lennert Buytenhek
6cc1b65886 [ARM] 3480/1: ixp4xx: fix irq2gpio array type
Patch from Lennert Buytenhek

The irq2gpio array was recently converted from an array of ints to an
array of chars (by patch 3368/1.)  However, this array contains elements
that are -1, and on ARM, the char type is unsigned by default, so this
patch broke the GPIO check in ixp4xx_set_irq_type.

Change the 'char' to be a 'signed char' to fix this.

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-04-20 21:24:38 +01:00
Linus Torvalds
402a26f0c0 Merge branch 'for-linus' of git://brick.kernel.dk/data/git/linux-2.6-block
* 'for-linus' of git://brick.kernel.dk/data/git/linux-2.6-block:
  [PATCH] block/elevator.c: remove unused exports
  [PATCH] splice: fix smaller sized splice reads
  [PATCH] Don't inherit ->splice_pipe across forks
  [patch] cleanup: use blk_queue_stopped
  [PATCH] Document online io scheduler switching
2006-04-20 08:17:04 -07:00
Linus Torvalds
bf7cf6ee1b Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6
* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6:
  [SPARC]: __NR_sys_splice --> __NR_splice
2006-04-20 07:58:58 -07:00
Eric Dumazet
0b699e36b2 [PATCH] x86_64: bring back __read_mostly support to linux-2.6.17-rc2
It seems latest kernel has a wrong/missing __read_mostly implementation
for x86_64

__read_mostly macro should be declared outside of #if CONFIG_X86_VSMP block

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-20 07:58:11 -07:00
Andi Kleen
18bd057b14 [PATCH] i386/x86-64: Fix x87 information leak between processes
AMD K7/K8 CPUs only save/restore the FOP/FIP/FDP x87 registers in FXSAVE
when an exception is pending.  This means the value leak through
context switches and allow processes to observe some x87 instruction
state of other processes.

This was actually documented by AMD, but nobody recognized it as
being different from Intel before.

The fix first adds an optimization: instead of unconditionally
calling FNCLEX after each FXSAVE test if ES is pending and skip
it when not needed. Then do a x87 load from a kernel variable to
clear FOP/FIP/FDP.

This means other processes always will only see a constant value
defined by the kernel in their FP state.

I took some pain to make sure to chose a variable that's already
in L1 during context switch to make the overhead of this low.

Also alternative() is used to patch away the new code on CPUs
who don't need it.

Patch for both i386/x86-64.

The problem was discovered originally by Jan Beulich. Richard
Brunner provided the basic code for the workarounds, with contribution
from Jan.

This is CVE-2006-1056

Cc: richard.brunner@amd.com
Cc: jbeulich@novell.com

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-20 07:58:11 -07:00