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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
This patch fixes several bugs in the gianfar driver, including a major one
where spinlocks were horribly broken:
* Split gianfar locks into two types: TX and RX
* Made it so gfar_start() now clears RHALT
* Fixed a bug where calling gfar_start_xmit() with interrupts off would
corrupt the interrupt state
* Fixed a bug where a frame could potentially arrive, and never be handled
(if no more frames arrived
* Fixed a bug where the rx_work_limit would never be observed by the rx
completion code
* Fixed a bug where the interrupt handlers were not actually protected by
their spinlocks
Signed-off-by: Andy Fleming <afleming@freescale.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
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>
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>
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>
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>
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>
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>
- fix mddev_lock() usage bugs in md_attr_show() and md_attr_store().
[they did not anticipate the possibility of getting a signal]
- remove mddev_lock_uninterruptible() [unused]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Lee Schermerhorn