Tejun Heo [Tue, 10 Mar 2009 07:27:48 +0000 (16:27 +0900)]
percpu: generalize embedding first chunk setup helper
Impact: code reorganization
Separate out embedding first chunk setup helper from x86 embedding
first chunk allocator and put it in mm/percpu.c. This will be used by
the default percpu first chunk allocator and possibly by other archs.
Tejun Heo [Tue, 10 Mar 2009 07:27:48 +0000 (16:27 +0900)]
percpu: more flexibility for @dyn_size of pcpu_setup_first_chunk()
Impact: cleanup, more flexibility for first chunk init
Non-negative @dyn_size used to be allowed iff @unit_size wasn't auto.
This restriction stemmed from implementation detail and made things a
bit less intuitive. This patch allows @dyn_size to be specified
regardless of @unit_size and swaps the positions of @dyn_size and
@unit_size so that the parameter order makes more sense (static,
reserved and dyn sizes followed by enclosing unit_size).
While at it, add @unit_size >= PCPU_MIN_UNIT_SIZE sanity check.
There's nothing arch specific about x86 __addr_to_pcpu_ptr() and
__pcpu_ptr_to_addr(). With proper __per_cpu_load and __per_cpu_start
defined, they'll do the right thing regardless of actual layout.
Move these macros from arch/x86/include/asm/percpu.h to mm/percpu.c
and allow archs to override it as necessary.
Tejun Heo [Tue, 10 Mar 2009 07:27:48 +0000 (16:27 +0900)]
linker script: define __per_cpu_load on all SMP capable archs
Impact: __per_cpu_load available on all SMP capable archs
Percpu now requires three symbols to be defined - __per_cpu_load,
__per_cpu_start and __per_cpu_end. There were three archs which
didn't have it. Update them as follows.
* powerpc: can use generic PERCPU() macro. Compile tested for
powerpc32, compile/boot tested for powerpc64.
* ia64: can use generic PERCPU_VADDR() macro. __phys_per_cpu_start is
identical to __per_cpu_load. Compile tested and symbol table looks
identical after the change except for the additional __per_cpu_load.
* arm: added explicit __per_cpu_load definition. Currently uses
unified .init output section so can't use the generic macro. Dunno
whether the unified .init ouput section is required by arch
peculiarity so I left it alone. Please break it up and use PERCPU()
if possible.
Signed-off-by: Tejun Heo <tj@kernel.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Pat Gefre <pfg@sgi.com> Cc: Russell King <rmk@arm.linux.org.uk>
Cliff Wickman [Fri, 6 Mar 2009 23:30:56 +0000 (17:30 -0600)]
x86: UV: remove uv_flush_tlb_others() WARN_ON
In uv_flush_tlb_others() (arch/x86/kernel/tlb_uv.c),
the "WARN_ON(!in_atomic())" fails if CONFIG_PREEMPT is not enabled.
And CONFIG_PREEMPT is not enabled by default in the distribution that
most UV owners will use.
We could #ifdef CONFIG_PREEMPT the warning, but that is not good form.
And there seems to be no suitable fix to in_atomic() when CONFIG_PREMPT
is not on.
As Ingo commented:
> and we have no proper primitive to test for atomicity. (mainly
> because we dont know about atomicity on a non-preempt kernel)
Tejun Heo [Fri, 6 Mar 2009 15:44:13 +0000 (00:44 +0900)]
percpu: finer grained locking to break deadlock and allow atomic free
Impact: fix deadlock and allow atomic free
Percpu allocation always uses GFP_KERNEL and whole alloc/free paths
were protected by single mutex. All percpu allocations have been from
GFP_KERNEL-safe context and the original allocator had this assumption
too. However, by protecting both alloc and free paths with the same
mutex, the new allocator creates free -> alloc -> GFP_KERNEL
dependency which the original allocator didn't have. This can lead to
deadlock if free is called from FS or IO paths. Also, in general,
allocators are expected to allow free to be called from atomic
context.
This patch implements finer grained locking to break the deadlock and
allow atomic free. For details, please read the "Synchronization
rules" comment.
While at it, also add CONTEXT: to function comments to describe which
context they expect to be called from and what they do to it.
This problem was reported by Thomas Gleixner and Peter Zijlstra.
http://thread.gmane.org/gmane.linux.kernel/802384
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Thomas Gleixner <tglx@linutronix.de> Reported-by: Peter Zijlstra <peterz@infradead.org>
Tejun Heo [Fri, 6 Mar 2009 15:44:09 +0000 (00:44 +0900)]
percpu: move chunk area map extension out of area allocation
Impact: code reorganization for later changes
Separate out chunk area map extension into a separate function -
pcpu_extend_area_map() - and call it directly from pcpu_alloc() such
that pcpu_alloc_area() is guaranteed to have enough area map slots on
invocation.
With this change, pcpu_alloc_area() does only area allocation and the
only failure mode is when the chunk doens't have enough room, so
there's no need to distinguish it from memory allocation failures.
Make it return -1 on such cases instead of hacky -ENOSPC.
Tejun Heo [Fri, 6 Mar 2009 15:44:09 +0000 (00:44 +0900)]
percpu: replace pcpu_realloc() with pcpu_mem_alloc() and pcpu_mem_free()
Impact: code reorganization for later changes
With static map handling moved to pcpu_split_block(), pcpu_realloc()
only clutters the code and it's also unsuitable for scheduled locking
changes. Implement and use pcpu_mem_alloc/free() instead.
Tejun Heo [Fri, 6 Mar 2009 05:33:59 +0000 (14:33 +0900)]
x86, percpu: setup reserved percpu area for x86_64
Impact: fix relocation overflow during module load
x86_64 uses 32bit relocations for symbol access and static percpu
symbols whether in core or modules must be inside 2GB of the percpu
segement base which the dynamic percpu allocator doesn't guarantee.
This patch makes x86_64 reserve PERCPU_MODULE_RESERVE bytes in the
first chunk so that module percpu areas are always allocated from the
first chunk which is always inside the relocatable range.
This problem exists for any percpu allocator but is easily triggered
when using the embedding allocator because the second chunk is located
beyond 2GB on it.
This patch also changes the meaning of PERCPU_DYNAMIC_RESERVE such
that it only indicates the size of the area to reserve for dynamic
allocation as static and dynamic areas can be separate. New
PERCPU_DYNAMIC_RESERVED is increased by 4k for both 32 and 64bits as
the reserved area separation eats away some allocatable space and
having slightly more headroom (currently between 4 and 8k after
minimal boot sans module area) makes sense for common case
performance.
x86_32 can address anywhere from anywhere and doesn't need reserving.
Mike Galbraith first reported the problem first and bisected it to the
embedding percpu allocator commit.
Tejun Heo [Fri, 6 Mar 2009 05:33:59 +0000 (14:33 +0900)]
percpu, module: implement reserved allocation and use it for module percpu variables
Impact: add reserved allocation functionality and use it for module
percpu variables
This patch implements reserved allocation from the first chunk. When
setting up the first chunk, arch can ask to set aside certain number
of bytes right after the core static area which is available only
through a separate reserved allocator. This will be used primarily
for module static percpu variables on architectures with limited
relocation range to ensure that the module perpcu symbols are inside
the relocatable range.
If reserved area is requested, the first chunk becomes reserved and
isn't available for regular allocation. If the first chunk also
includes piggy-back dynamic allocation area, a separate chunk mapping
the same region is created to serve dynamic allocation. The first one
is called static first chunk and the second dynamic first chunk.
Although they share the page map, their different area map
initializations guarantee they serve disjoint areas according to their
purposes.
If arch doesn't setup reserved area, reserved allocation is handled
like any other allocation.
Tejun Heo [Fri, 6 Mar 2009 05:33:59 +0000 (14:33 +0900)]
x86: make embedding percpu allocator return excessive free space
Impact: reduce unnecessary memory usage on certain configurations
Embedding percpu allocator allocates unit_size *
smp_num_possible_cpus() bytes consecutively and use it for the first
chunk. However, if the static area is small, this can result in
excessive prellocated free space in the first chunk due to
PCPU_MIN_UNIT_SIZE restriction.
This patch makes embedding percpu allocator preallocate only what's
necessary as described by PERPCU_DYNAMIC_RESERVE and return the
leftover to the bootmem allocator.
Tejun Heo [Fri, 6 Mar 2009 05:33:59 +0000 (14:33 +0900)]
percpu: use negative for auto for pcpu_setup_first_chunk() arguments
Impact: argument semantic cleanup
In pcpu_setup_first_chunk(), zero @unit_size and @dyn_size meant
auto-sizing. It's okay for @unit_size as 0 doesn't make sense but 0
dynamic reserve size is valid. Alos, if arch @dyn_size is calculated
from other parameters, it might end up passing in 0 @dyn_size and
malfunction when the size is automatically adjusted.
This patch makes both @unit_size and @dyn_size ssize_t and use -1 for
auto sizing.
Tejun Heo [Fri, 6 Mar 2009 05:33:59 +0000 (14:33 +0900)]
percpu: improve first chunk initial area map handling
Impact: no functional change
When the first chunk is created, its initial area map is not allocated
because kmalloc isn't online yet. The map is allocated and
initialized on the first allocation request on the chunk. This works
fine but the scattering of initialization logic between the init
function and allocation path is a bit confusing.
This patch makes the first chunk initialize and use minimal statically
allocated map from pcpu_setpu_first_chunk(). The map resizing path
still needs to handle this specially but it's more straight-forward
and gives more latitude to the init path. This will ease future
changes.
arch/x86/mm/highmem_32.c:187: error: static declaration of 'set_highmem_pages_init' follows non-static declaration
arch/x86/include/asm/numa_32.h:8: error: previous declaration of 'set_highmem_pages_init' was here
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
LKML-Reference: <1236082212.2675.24.camel@penberg-laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Pekka Enberg [Tue, 3 Mar 2009 12:10:12 +0000 (14:10 +0200)]
x86: set_highmem_pages_init() cleanup
Impact: cleanup
This patch moves set_highmem_pages_init() to arch/x86/mm/highmem_32.c.
The declaration of the function is kept in asm/numa_32.h because
asm/highmem.h is included only if CONFIG_HIGHMEM is enabled so we
can't put the empty static inline function there.
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
LKML-Reference: <1236082212.2675.24.camel@penberg-laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Pekka Enberg [Tue, 3 Mar 2009 11:15:04 +0000 (13:15 +0200)]
x86: make sure initmem is writable on 64-bit
Impact: unification
This patch ports commit 3c1df68b848b39270752ff8d4b956cc4a4dce0f6 ("x86: make
sure initmem is writable") to the 64-bit version to unify implementations of
free_init_pages().
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Arjan van de Ven <arjan@linux.intel.com>
LKML-Reference: <1236078904.2675.17.camel@penberg-laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Pekka Enberg [Tue, 3 Mar 2009 10:55:05 +0000 (12:55 +0200)]
x86: use roundup() instead of PAGE_ALIGN() in find_early_table_space()
Impact: cleanup
This patch changes find_early_table_space() to use roundup() for rounding up
tables to page size to unify the common parts of the 32-bit and 64-bit
implementations.
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
LKML-Reference: <1236077705.2675.6.camel@penberg-laptop> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Yinghai Lu [Tue, 3 Mar 2009 07:36:13 +0000 (23:36 -0800)]
x86: fix init_memory_mapping() to handle small ranges
Impact: fix failed EFI bootup in certain circumstances
Ying Huang found init_memory_mapping() has problem with small ranges
less than 2M when he tried to direct map the EFI runtime code out of
max_low_pfn_mapped.
It turns out we never considered that case and didn't check the range...
Reported-by: Ying Huang <ying.huang@intel.com> Signed-off-by: Yinghai Lu <yinghai@kernel.org> Cc: Brian Maly <bmaly@redhat.com>
LKML-Reference: <49ACDDED.1060508@kernel.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Linus Torvalds [Tue, 3 Mar 2009 00:23:33 +0000 (16:23 -0800)]
Revert "menu: fix embedded menu snafu"
This reverts commit 155b25bcc28631a5b5230191aa3f56c40dfffa3f, which was
totally wrong - the "embedded" options still exists (very much so) even
on non-embedded platforms.
It's just that we don't bother with actually asking about them when
we're not embedded, we just take their default values (which is usually
'y' - the options add features that may not be worth it in a constrained
environment).
Noticed-by: Andrew Morton <akpm@linux-foundation.org> Cc: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Tue, 3 Mar 2009 00:11:36 +0000 (16:11 -0800)]
Merge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6
* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
drm/i915: Fix use-before-null-check in i915_irq_emit().
drm: Avoid client deadlocks when the master disappears.
drm: Wake up all lock waiters when the master disappears.
drm: Don't return ERESTARTSYS to user-space.
drm: Avoid client deadlocks when the master disappears.
This is done by
1) Wake up lock waiters when we close the master file descriptor.
Not when the master structure is removed, since the latter
requires the waiters themselves to release the refcount on the
master structure -> Deadlock.
2) Send a SIGTERM to all clients waiting for the lock.
Normally these clients will get a SIGPIPE when the X server dies,
but clients may also spin trying to grab the DRM lock, without
getting any sort of notification.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@linux.ie>
Randy Dunlap [Mon, 2 Mar 2009 22:14:06 +0000 (14:14 -0800)]
menu: fix embedded menu snafu
The COMPAT_BRK kconfig symbol does not depend on EMBEDDED, but it is in
the midst of the EMBEDDED menu symbols, so it mucks up the EMBEDDED
menu. Fix by moving it to just after all of the EMBEDDED menu symbols.
Also, surround all of the EMBEDDED symbols with "if EMBEDDED"/"endif" so
that this EMBEDDED block is clearer.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds [Mon, 2 Mar 2009 23:48:00 +0000 (15:48 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/drzeus/mmc:
sdhci: Add NO_BUSY_IRQ quirk for Marvell CAFE host chip
sdhci: Add quirk for controllers with no end-of-busy IRQ
Linus Torvalds [Mon, 2 Mar 2009 23:47:19 +0000 (15:47 -0800)]
Merge branch 'fix/hda' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
* 'fix/hda' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
ALSA: hda - Add probe_mask default for Toshiba laptop with ALC268
ALSA: hda - Add quirk for new HP xw series
ALSA: hda - Fix digital mic on dell-m4-1 and dell-m4-3
Linus Torvalds [Mon, 2 Mar 2009 23:47:01 +0000 (15:47 -0800)]
Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
fix warning in io_mapping_map_wc()
x86: i915 needs pgprot_writecombine() and is_io_mapping_possible()
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (29 commits)
zaurus: add usb id for motomagx phones
usbnet: make usbnet_get_link() fall back to ethtool_op_get_link()
veth: Fix carrier detect
cdc_ether: add usb id for Ericsson F3507g
r8169: read MAC address from EEPROM on init (2nd attempt)
tcp: fix retrans_out leaks
net headers: export dcbnl.h
net headers: cleanup dcbnl.h
netpoll: Add drop checks to all entry points
gianfar: Do right check on num_txbdfree
pkt_sched: sch_drr: Fix oops in drr_change_class.
b44: Disable device on shutdown
b44: Unconditionally enable interrupt routing on reset
net: fix hp-plus build error
libertas: fix misuse of netdev_priv() and dev->ml_priv
ipv6: don't use tw net when accounting for recycled tw
asix: new device ids
tcp_scalable: Update malformed & dead url
netfilter: xt_recent: fix proc-file addition/removal of IPv4 addresses
netxen: handle pci bar 0 mapping failure
...
Linus Torvalds [Mon, 2 Mar 2009 23:43:03 +0000 (15:43 -0800)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: elantech - touchpad driver miss-recognising logitech mice
Input: synaptics - ensure we reset the device on resume
Input: usbtouchscreen - fix eGalax HID ignoring
Input: ambakmi - fix timeout handling in amba_kmi_write()
Input: pxa930_trkball - fix write timeout handling
Input: struct device - replace bus_id with dev_name(), dev_set_name()
Input: bf54x-keys - fix debounce time validation
Input: spitzkbd - mark probe function as __devinit
Input: omap-keypad - mark probe function as __devinit
Input: corgi_ts - mark probe function as __devinit
Input: corgikbd - mark probe function as __devinit
Input: uvc - the button on the camera is KEY_CAMERA
Input: psmouse - make MOUSE_PS2_LIFEBOOK depend on X86
Input: atkbd - make forced_release_keys[] static
Input: usbtouchscreen - allow reporting calibrated data
Linus Torvalds [Mon, 2 Mar 2009 23:42:26 +0000 (15:42 -0800)]
Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
ext4: don't call jbd2_journal_force_commit_nested without journal
ext4: Reorder fs/Makefile so that ext2 root fs's are mounted using ext2
ext4: Remove duplicate call to ext4_commit_super() in ext4_freeze()
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6:
[SCSI] mpt: fix disable lsi sas to use msi as default
[SCSI] fix ABORTED_COMMAND looping forever problem
[SCSI] sd: revive sd_index_lock
[SCSI] cxgb3i: update the driver version to 1.0.1
[SCSI] cxgb3i: Fix spelling errors in documentation
[SCSI] cxgb3i: added missing include in cxgb3i_ddp.h
[SCSI] cxgb3i: Outgoing pdus need to observe skb's MAX_SKB_FRAGS
[SCSI] cxgb3i: added per-task data to track transmit progress
[SCSI] cxgb3i: transmit work-request fixes
[SCSI] hptiop: Add new PCI device ID
Roland McGrath [Sat, 28 Feb 2009 07:25:54 +0000 (23:25 -0800)]
x86-64: seccomp: fix 32/64 syscall hole
On x86-64, a 32-bit process (TIF_IA32) can switch to 64-bit mode with
ljmp, and then use the "syscall" instruction to make a 64-bit system
call. A 64-bit process make a 32-bit system call with int $0x80.
In both these cases under CONFIG_SECCOMP=y, secure_computing() will use
the wrong system call number table. The fix is simple: test TS_COMPAT
instead of TIF_IA32. Here is an example exploit:
/* test case for seccomp circumvention on x86-64
There are two failure modes: compile with -m64 or compile with -m32.
The -m64 case is the worst one, because it does "chmod 777 ." (could
be any chmod call). The -m32 case demonstrates it was able to do
stat(), which can glean information but not harm anything directly.
A buggy kernel will let the test do something, print, and exit 1; a
fixed kernel will make it exit with SIGKILL before it does anything.
*/
Signed-off-by: Roland McGrath <roland@redhat.com>
[ I don't know if anybody actually uses seccomp, but it's enabled in
at least both Fedora and SuSE kernels, so maybe somebody is. - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Sat, 28 Feb 2009 03:03:24 +0000 (19:03 -0800)]
x86-64: syscall-audit: fix 32/64 syscall hole
On x86-64, a 32-bit process (TIF_IA32) can switch to 64-bit mode with
ljmp, and then use the "syscall" instruction to make a 64-bit system
call. A 64-bit process make a 32-bit system call with int $0x80.
In both these cases, audit_syscall_entry() will use the wrong system
call number table and the wrong system call argument registers. This
could be used to circumvent a syscall audit configuration that filters
based on the syscall numbers or argument details.
Signed-off-by: Roland McGrath <roland@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andres Salomon [Mon, 2 Mar 2009 20:48:20 +0000 (21:48 +0100)]
sdhci: Add NO_BUSY_IRQ quirk for Marvell CAFE host chip
As described here: http://lkml.org/lkml/2009/2/20/265
The CAFE chip is broken due to commit e809517f6fa5803a5a1cd5602.
Anton added a quirk here: http://lkml.org/lkml/2009/2/20/279 that fixes
CAFE's problem. This adds the quirk for CAFE.
Signed-off-by: Andres Salomon <dilinger@debian.org> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
Ben Dooks [Fri, 20 Feb 2009 17:33:08 +0000 (20:33 +0300)]
sdhci: Add quirk for controllers with no end-of-busy IRQ
The Samsung SDHCI (and FSL eSDHC) controller block seems to fail
to generate an INT_DATA_END after the transfer has completed and
the bus busy state finished.
Signed-off-by: Ben Dooks <ben-linux@fluff.org> Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
The virtually mapped percpu space causes us two problems:
- for hypercalls which take an mfn, we need to do a full pagetable
walk to convert the percpu va into an mfn, and
- when a hypercall requires a page to be mapped RO via all its aliases,
we need to make sure its RO in both the percpu mapping and in the
linear mapping
This primarily affects the gdt and the vcpu info structure.
With x86-32 and -64 using the same mechanism for managing the
tss io permissions bitmap, large chunks of process*.c are
trivially unifyable, including:
- exit_thread
- flush_thread
- __switch_to_xtra (along with tsc enable/disable)
and as bonus pickups:
- sys_fork
- sys_vfork
(Note: asmlinkage expands to empty on x86-64)
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: remove 32-bit optimization to prepare unification
x86-32 and -64 differ in the way they context-switch tasks
with io permission bitmaps. x86-64 simply copies the next
tasks io bitmap into place (if any) on context switch. x86-32
invalidates the bitmap on context switch, so that the next
IO instruction will fault; at that point it installs the
appropriate IO bitmap.
This makes context switching IO-bitmap-using tasks a bit more
less expensive, at the cost of making the next IO instruction
slower due to the extra fault. This tradeoff only makes sense
if IO-bitmap-using processes are relatively common, but they
don't actually use IO instructions very often.
However, in a typical desktop system, the only process likely
to be using IO bitmaps is the X server, and nothing at all on
a server. Therefore the lazy context switch doesn't really win
all that much, and its just a gratuitious difference from
64-bit code.
This patch removes the lazy context switch, with a view to
unifying this code in a later change.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Remove es7000_cpu_mask_to_apicid_cluster completely, because it's
almost the same as es7000_cpu_mask_to_apicid except 2 code paths.
One of them is about to be removed soon, the another should be
BAD_APICID (it's a fail path).
The _cluster one was not invoked on apic->cpu_mask_to_apicid_and
anyway, since there was no _cluster_and variant.
Ingo Molnar [Mon, 2 Mar 2009 10:00:57 +0000 (11:00 +0100)]
x86, mm: dont use non-temporal stores in pagecache accesses
Impact: standardize IO on cached ops
On modern CPUs it is almost always a bad idea to use non-temporal stores,
as the regression in this commit has shown it:
30d697f: x86: fix performance regression in write() syscall
The kernel simply has no good information about whether using non-temporal
stores is a good idea or not - and trying to add heuristics only increases
complexity and inserts fragility.
The regression on cached write()s took very long to be found - over two
years. So dont take any chances and let the hardware decide how it makes
use of its caches.
The only exception is drivers/gpu/drm/i915/i915_gem.c: there were we are
absolutely sure that another entity (the GPU) will pick up the dirty
data immediately and that the CPU will not touch that data before the
GPU will.
Also, keep the _nocache() primitives to make it easier for people to
experiment with these details. There may be more clear-cut cases where
non-cached copies can be used, outside of filemap.c.
Takashi Iwai [Sun, 1 Mar 2009 09:55:44 +0000 (10:55 +0100)]
ALSA: hda - Add probe_mask default for Toshiba laptop with ALC268
Some Toshiba laptops have another ALC268 codec on slot#3 that conflicts
with the primary codec. The codec#3 is for the digital I/O, and should
be fixed by the driver, but it'd need a bunch of changes.
So, let's fix the probe problem temporarily by setting the default
probe_mask value.
The Motorola MOTOMAGX phones (Z6, E8, Zn5 so far) are providing
combined ACM/BLAN USB configuration. Since it has Vendor Specific
class, the corresponding drivers (cdc-acm, zaurus) can't find it just
by interface info. This patch adds usb id so the zaurus driver can
properly handle this combined device.
Signed-off-by: Dmitriy Taychenachev <dimichxp@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Bjørn Mork [Mon, 2 Mar 2009 04:45:40 +0000 (20:45 -0800)]
usbnet: make usbnet_get_link() fall back to ethtool_op_get_link()
Make usbnet_get_link() fall back to ethtool_op_get_link() instead of
defaulting to 1.
This makes usbnet_get_link return valid results without the need for a
driver specific check_connect or mii ops as long as the driver calls
netif_carrier_{on,off}() as appropriate. cdc_ether is an example of
such a driver.
Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
The current implementation of carrier detect in veth is broken.
It reports the link is down until both sides of the veth pair
are administatively up and then forever after it reports link up.
So fix veth so that it only reports link up when both interfaces
of the pair are administratively up.
Signed-off-by: Eric Biederman <ebiederm@aristanetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Bjørn Mork [Wed, 25 Feb 2009 04:33:58 +0000 (04:33 +0000)]
cdc_ether: add usb id for Ericsson F3507g
The Ericsson F3507g wireless broadband module provides a CDC Ethernet
compliant interface, but identifies it as a "Mobile Direct Line" CDC
subclass, thereby preventing the CDC Ethernet class driver from picking
it up. This patch adds the device id to cdc_ether.c as a workaround.
Ericsson has provided a "class" driver for this device:
http://kerneltrap.org/mailarchive/linux-net/2008/10/28/3832094
But closer inspection of that driver reveals that it adds little more
than duplication of code from cdc_ether.c. See also
http://marc.info/?l=linux-usb&m=123334979706403&w=2
Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
Ivan Vecera [Mon, 2 Mar 2009 04:34:48 +0000 (20:34 -0800)]
r8169: read MAC address from EEPROM on init (2nd attempt)
This is 2nd attempt to implement the initialization/reading of MAC address
from EEPROM. The first used PCI's VPD and there were some problems, some
devices are not able to read EEPROM content by VPD. The 2nd one uses direct
access to EEPROM through bit-banging interface and my testing results seem
to be much better.
I tested 5 systems each with different Realtek NICs and I didn't find any
problem. AFAIK Francois's NICs also works fine.
Original description:
This fixes the problem when MAC address is set by ifconfig or by
ip link commands and this address is stored in the device after
reboot. The power-off is needed to get right MAC address.
This is problem when Xen daemon is running because it renames the device
name from ethX to pethX and sets its MAC address to FE:FF:FF:FF:FF:FF.
After reboot the device is still using FE:FF:FF:FF:FF:FF.
Signed-off-by: Ivan Vecera <ivecera@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Paul Moore [Fri, 27 Feb 2009 20:00:03 +0000 (15:00 -0500)]
selinux: Fix a panic in selinux_netlbl_inode_permission()
Rick McNeal from LSI identified a panic in selinux_netlbl_inode_permission()
caused by a certain sequence of SUNRPC operations. The problem appears to be
due to the lack of NULL pointer checking in the function; this patch adds the
pointer checks so the function will exit safely in the cases where the socket
is not completely initialized.
Signed-off-by: Paul Moore <paul.moore@hp.com> Signed-off-by: James Morris <jmorris@namei.org>
Ilpo Järvinen [Sun, 1 Mar 2009 08:21:36 +0000 (00:21 -0800)]
tcp: fix retrans_out leaks
There's conflicting assumptions in shifting, the caller assumes
that dupsack results in S'ed skbs (or a part of it) for sure but
never gave a hint to tcp_sacktag_one when dsack is actually in
use. Thus DSACK retrans_out -= pcount was not taken and the
counter became out of sync. Remove obstacle from that information
flow to get DSACKs accounted in tcp_sacktag_one as expected.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Tested-by: Denys Fedoryshchenko <denys@visp.net.lb> Signed-off-by: David S. Miller <davem@davemloft.net>
Chris Leech [Fri, 27 Feb 2009 10:01:36 +0000 (10:01 +0000)]
net headers: cleanup dcbnl.h
1) add an include for <linux/types.h>
2) change dcbmsg.dcb_family from unsigned char to __u8 to be more
consistent with use of kernel types
Signed-off-by: Chris Leech <christopher.leech@intel.com> Acked-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Herbert Xu [Sun, 1 Mar 2009 08:11:52 +0000 (00:11 -0800)]
netpoll: Add drop checks to all entry points
The netpoll entry checks are required to ensure that we don't
receive normal packets when invoked via netpoll. Unfortunately
it only ever worked for the netif_receive_skb/netif_rx entry
points. The VLAN (and subsequently GRO) entry point didn't
have the check and therefore can trigger all sorts of weird
problems.
This patch adds the netpoll check to all entry points.
I'm still uneasy with receiving at all under netpoll (which
apparently is only used by the out-of-tree kdump code). The
reason is it is perfectly legal to receive all data including
headers into highmem if netpoll is off, but if you try to do
that with netpoll on and someone gets a printk in an IRQ handler
you're going to get a nice BUG_ON.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Tejun Heo [Sun, 1 Mar 2009 07:06:56 +0000 (16:06 +0900)]
bootmem, x86: further fixes for arch-specific bootmem wrapping
Impact: fix new breakages introduced by previous fix
Commit c132937556f56ee4b831ef4b23f1846e05fde102 tried to clean up
bootmem arch wrapper but it wasn't quite correct. Before the commit,
the followings were broken.
* reserve_bootmem(...) can't be mapped into
reserve_bootmem_node(NODE_DATA(0)->bdata, ...) because the node is
not preference here. The region specified MUST fall into the
specified region; otherwise, it will panic.
After the commit,
* If allocation fails for the arch preferred node, it should fallback
to whatever is available. Instead, it simply failed allocation.
There are too many internal details to allow generic wrapping and
still keep things simple for archs. Plus, all that arch wants is a
way to prefer certain node over another.
This patch drops the generic wrapping around alloc_bootmem_core() and
add alloc_bootmem_core() instead. If necessary, arch can define
bootmem_arch_referred_node() macro or function which takes all
allocation information and returns the preferred node. bootmem
generic code will always try the preferred node first and then
fallback to other nodes as usual.
Breakages noted and changes reviewed by Johannes Weiner.
Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Tejun Heo [Sun, 1 Mar 2009 06:42:36 +0000 (15:42 +0900)]
percpu: kill compile warning in pcpu_populate_chunk()
Impact: remove compile warning
Mark local variable map_end in pcpu_populate_chunk() with
uninitialized_var(). The variable is always used in tandem with
map_start and guaranteed to be initialized before use but gcc doesn't
understand that.
Some Logitech mice react to the magic knock like Elantech touchpad would.
This leads to those mice being misdetected as Elantech touchpads. Add a
version query to elantech_detect() to distinguish the two.
[dtor@mail.ru:
- lower severity of some messages - when we are not sure yet if
device is Elantech or not not responding to knock is not an error.
]
Andy Whitcroft [Sat, 28 Feb 2009 20:51:01 +0000 (12:51 -0800)]
Input: synaptics - ensure we reset the device on resume
When resuming from suspend newer Synaptics touchpads do not recover
correctly. Analysis of the resume sequence as applied in Linux was
compared to that of other operating systems. This indicated that the
other OSs were resetting the mouse before attempting to detect it (for
all Synaptics touchpads, old and new). Applying this same modification
fixes these newer Synaptics touchpads and brings the driver into line
with common OS reset behaviour.
Signed-off-by: Andy Whitcroft <apw@canonical.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Devices confirmed to be real HID have the class set to HID and the protocol
set to 'mouse'. Some have HID class but protocol set to 'none'. Those are
not HID and should be driven by usbtouchscreen.
Fix the device ignoring macro by adding match for the protocol too.
Signed-off-by: Daniel Ritz <daniel.ritz@gmx.ch> Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
This patch unify fixmap_32.h and fixmap_64.h into fixmap.h.
Things that we can't merge now are using CONFIG_X86_{32,64}
(e.g.:vsyscall and EFI)
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Just prepare fixmap for later mechanic unification.
No real modification on code.
text data bss dec hex filename 3831152 353188 372736 4557076 458914 vmlinux-32.after 3831152 353188 372736 4557076 458914 vmlinux-32.before
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Just prepare fixmap for later mechanic unification.
No real modification on code.
text data bss dec hex filename 4312362 527192 421924 5261478 5048a6 vmlinux-64.after 4312362 527192 421924 5261478 5048a6 vmlinux-64.before
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
FIX_EFI_IO_MAP_FIRST_PAGE is used only when EFI is enabled.
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Add CONFIG_X86_{LOCAL,IO}_APIC to enum fixed_address.
FIX_APIC_BASE is used only when CONFIG_X86_LOCAL_APIC is
enabled and FIX_IO_APIC_BASE_* are used only when
CONFIG_X86_IO_APIC is enabled.
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
x86, fixmap: define reserve_top_address for x86_64
Impact: new interface (not yet use)
Define reserve_top_address for x86_64; only for later x86 integration.
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
x86, fixmap: define FIXADDR_BOOT_* and redefine FIX_ADDR_SIZE
Impact: new interface, not yet used
Now, with these macros, x86_64 code can know where start the
permanent and non-permanent fixed mapped address.
This patch make these macros equal fixmap_32.h for future
x86 integration.
Signed-off-by: Gustavo F. Padovan <gustavo@las.ic.unicamp.br> Acked-by: Glauber Costa <gcosta@redhat.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
projects / linux-beck.git / log