Li Zefan [Wed, 29 Oct 2008 09:03:22 +0000 (17:03 +0800)]
sched: cleanup for alloc_rt/fair_sched_group()
Impact: cleanup
Remove checking parent == NULL. It won't be NULLL, because we dynamically
create sub task_group only, and sub task_group always has its parent.
(root task_group is statically defined)
Also replace kmalloc_node(GFP_ZERO) with kzalloc_node().
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Linus Torvalds [Tue, 28 Oct 2008 16:52:25 +0000 (09:52 -0700)]
Merge branch 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (31 commits)
ftrace: fix current_tracer error return
tracing: fix a build error on alpha
ftrace: use a real variable for ftrace_nop in x86
tracing/ftrace: make boot tracer select the sched_switch tracer
tracepoint: check if the probe has been registered
asm-generic: define DIE_OOPS in asm-generic
trace: fix printk warning for u64
ftrace: warning in kernel/trace/ftrace.c
ftrace: fix build failure
ftrace, powerpc, sparc64, x86: remove notrace from arch ftrace file
ftrace: remove ftrace hash
ftrace: remove mcount set
ftrace: remove daemon
ftrace: disable dynamic ftrace for all archs that use daemon
ftrace: add ftrace warn on to disable ftrace
ftrace: only have ftrace_kill atomic
ftrace: use probe_kernel
ftrace: comment arch ftrace code
ftrace: return error on failed modified text.
ftrace: dynamic ftrace process only text section
...
Linus Torvalds [Tue, 28 Oct 2008 16:49:27 +0000 (09:49 -0700)]
Merge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
lockdep: fix irqs on/off ip tracing
lockdep: minor fix for debug_show_all_locks()
x86: restore the old swiotlb alloc_coherent behavior
x86: use GFP_DMA for 24bit coherent_dma_mask
swiotlb: remove panic for alloc_coherent failure
xen: compilation fix of drivers/xen/events.c on IA64
xen: portability clean up and some minor clean up for xencomm.c
xen: don't reload cr3 on suspend
kernel/resource: fix reserve_region_with_split() section mismatch
printk: remove unused code from kernel/printk.c
Linus Torvalds [Tue, 28 Oct 2008 16:46:20 +0000 (09:46 -0700)]
Merge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: fix documentation reference for sched_min_granularity_ns
sched: virtual time buddy preemption
sched: re-instate vruntime based wakeup preemption
sched: weaken sync hint
sched: more accurate min_vruntime accounting
sched: fix a find_busiest_group buglet
sched: add CONFIG_SMP consistency
Linus Torvalds [Tue, 28 Oct 2008 16:42:48 +0000 (09:42 -0700)]
Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev
* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
libata: ahci enclosure management bit mask
libata: ahci enclosure management led sync
pata_ninja32: suspend/resume support
libata: Fix LBA48 on pata_it821x RAID volumes.
libata: clear saved xfer_mode and ncq_enabled on device detach
sata_sil24: configure max read request size to 4k
libata: add missing kernel-doc
libata: fix device iteration bugs
ahci: Add support for Promise PDC42819
ata: Switch all my stuff to a common address
Steven Rostedt [Tue, 28 Oct 2008 14:44:24 +0000 (10:44 -0400)]
ftrace: fix current_tracer error return
The commit (in linux-tip) c2931e05ec5965597cbfb79ad332d4a29aeceb23
( ftrace: return an error when setting a nonexistent tracer )
added useful code that would error when a bad tracer was written into
the current_tracer file.
But this had a bug if the amount written was more than the amount read by
that code. The first iteration would set the tracer correctly, but since
it did not consume the rest of what was written (usually whitespace), the
userspace utility would continue to write what was not consumed. This
second iteration would fail to find a tracer and return -EINVAL. Funny
thing is that the tracer would have already been set.
This patch just consumes all the data that is written to the file.
Signed-off-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Sheng Yang [Wed, 15 Oct 2008 12:15:06 +0000 (20:15 +0800)]
KVM: Fix guest shared interrupt with in-kernel irqchip
Every call of kvm_set_irq() should offer an irq_source_id, which is
allocated by kvm_request_irq_source_id(). Based on irq_source_id, we
identify the irq source and implement logical OR for shared level
interrupts.
The allocated irq_source_id can be freed by kvm_free_irq_source_id().
Currently, we support at most sizeof(unsigned long) different irq sources.
[Amit: - rebase to kvm.git HEAD
- move definition of KVM_USERSPACE_IRQ_SOURCE_ID to common file
- move kvm_request_irq_source_id to the update_irq ioctl]
[Xiantao: - Add kvm/ia64 stuff and make it work for kvm/ia64 guests]
Signed-off-by: Sheng Yang <sheng@linux.intel.com> Signed-off-by: Amit Shah <amit.shah@redhat.com> Signed-off-by: Xiantao Zhang <xiantao.zhang@intel.com> Signed-off-by: Avi Kivity <avi@redhat.com>
Heiko Carstens [Tue, 28 Oct 2008 10:14:58 +0000 (11:14 +0100)]
lockdep: fix irqs on/off ip tracing
Impact: fix lockdep lock-api-caller output when irqsoff tracing is enabled
81d68a96 "ftrace: trace irq disabled critical timings" added wrappers around
trace_hardirqs_on/off_caller. However these functions use
__builtin_return_address(0) to figure out which function actually disabled
or enabled irqs. The result is that we save the ips of trace_hardirqs_on/off
instead of the real caller. Not very helpful.
However since the patch from Steven the ip already gets passed. So use that
and get rid of __builtin_return_address(0) in these two functions.
[S390] s390: Fix build for !CONFIG_S390_GUEST + CONFIG_VIRTIO_CONSOLE
The s390 kernel does not compile if virtio console is enabled, but guest
support is disabled:
LD .tmp_vmlinux1
arch/s390/kernel/built-in.o: In function `setup_arch':
/space/linux-2.5/arch/s390/kernel/setup.c:773: undefined reference to
`s390_virtio_console_init'
The fix is related to
commit 99e65c92f2bbf84f43766a8bf701e36817d62822
Author: Christian Borntraeger <borntraeger@de.ibm.com>
Date: Fri Jul 25 15:50:04 2008 +0200
KVM: s390: Fix guest kconfig
Which changed the build process to build kvm_virtio.c only if CONFIG_S390_GUEST
is set. We must ifdef the prototype in the header file accordingly.
Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Heiko Carstens [Tue, 28 Oct 2008 10:10:21 +0000 (11:10 +0100)]
[S390] No more 4kb stacks.
We got a stack overflow with a small stack configuration on a 32 bit
system. It just looks like as 4kb isn't enough and too dangerous.
So lets get rid of 4kb stacks on 32 bit.
But one thing I completely dislike about the call trace below is that
just for debugging or tracing purposes sprintf gets called (cio_start_key):
/* process condition code */
sprintf(dbf_txt, "ccode:%d", ccode);
CIO_TRACE_EVENT(4, dbf_txt);
But maybe its just me who thinks that this could be done better.
Heiko Carstens [Tue, 28 Oct 2008 10:10:20 +0000 (11:10 +0100)]
[S390] Change default IPL method to IPL_VM.
allyesconfig and allmodconfig built kernels have a tape IPL record.
A the vmreader record makes much more sense, since hardly anybody will
ever IPL a kernel from tape. So change the default.
As I side effect I can test these kernels without fiddling around with
the kernel config ;)
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
irq event stamp: 847
hardirqs last enabled at (847): [<000000000007aca6>] trace_hardirqs_on+0x2a/0x38
hardirqs last disabled at (846): [<0000000000076ca2>] trace_hardirqs_off+0x2a/0x38
softirqs last enabled at (0): [<000000000004909e>] copy_process+0x43e/0x11f4
softirqs last disabled at (0): [<0000000000000000>] 0x0
other info that might help us debug this:
1 lock held by vol_id/2903:
#0: (&bdev->bd_mutex){--..}, at: [<000000000010e0f4>] do_open+0x78/0x358
Heiko Carstens [Tue, 28 Oct 2008 10:10:16 +0000 (11:10 +0100)]
[S390] Fix sysdev class file creation.
Use sysdev_class_create_file() to create create sysdev class attributes
instead of sysfs_create_file(). Using sysfs_create_file() wasn't a very
good idea since the show and store functions have a different amount of
parameters for sysfs files and sysdev class files.
In particular the pointer to the buffer is the last argument and
therefore accesses to random memory regions happened.
Still worked surprisingly well until we got a kernel panic.
Cc: stable@kernel.org Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
[S390] pgtables: Fix race in enable_sie vs. page table ops
The current enable_sie code sets the mm->context.pgstes bit to tell
dup_mm that the new mm should have extended page tables. This bit is also
used by the s390 specific page table primitives to decide about the page
table layout - which means context.pgstes has two meanings. This can cause
any kind of bugs. For example - e.g. shrink_zone can call
ptep_clear_flush_young while enable_sie is running. ptep_clear_flush_young
will test for context.pgstes. Since enable_sie changed that value of the old
struct mm without changing the page table layout ptep_clear_flush_young will
do the wrong thing.
The solution is to split pgstes into two bits
- one for the allocation
- one for the current state
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Jan Glauber [Tue, 28 Oct 2008 10:10:13 +0000 (11:10 +0100)]
[S390] qdio: prevent double qdio shutdown in case of I/O errors
In case of I/O errors on a qdio subchannel qdio_shutdown may be
called twice by the qdio driver and by zfcp. Remove the
superfluous shutdown from qdio and let the upper layer driver
handle the error condition.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When tracing is enabled, some arch have included <linux/irqflags.h>
on their <asm/system.h> but others like alpha or m68k don't.
Build error on alpha:
kernel/trace/trace.c: In function 'tracing_cpumask_write':
kernel/trace/trace.c:2145: error: implicit declaration of function 'raw_local_irq_disable'
kernel/trace/trace.c:2162: error: implicit declaration of function 'raw_local_irq_enable'
Tested on Alpha through a cross-compiler (should correct a similar issue on m68k).
Shaohua Li [Mon, 27 Oct 2008 20:03:18 +0000 (13:03 -0700)]
x86, memory hotplug: remove wrong -1 in calling init_memory_mapping()
Impact: fix crash with memory hotplug
Shuahua Li found:
| I just did some experiments on a desktop for memory hotplug and this bug
| triggered a crash in my test.
|
| Yinghai's suggestion also fixed the bug.
We don't need to round it, just remove that extra -1
Alan Cox [Fri, 17 Oct 2008 18:08:31 +0000 (19:08 +0100)]
pata_ninja32: suspend/resume support
I had assumed that the standard recovery would be sufficient for this
hardware but it isn't. Fix up the other registers on resume as needed. See
bug #11735
Signed-off-by: Alan Cox <alan@redhat.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Tejun Heo [Sun, 26 Oct 2008 06:43:03 +0000 (15:43 +0900)]
libata: clear saved xfer_mode and ncq_enabled on device detach
libata EH saves xfer_mode and ncq_enabled at start to later set
DUBIOUS_XFER flag if it has changed. These values need to be cleared
on device detach such that hot device swap doesn't accidentally miss
DUBIOUS_XFER.
Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Tejun Heo [Sat, 25 Oct 2008 05:26:54 +0000 (14:26 +0900)]
sata_sil24: configure max read request size to 4k
Due to request posting limitations, bandwidth of sil3132 is limited to
around 120MB/s with the minimum pci-e payload size (128bytes) which is
used by most consumer systems. However, write throughput can be
slightly (~3%) increased by increasing the max read requeset size.
Configure it to 4k which is the maximum supported. This optimization
is also done by SIMG's windows driver.
Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Mark Nelson [Thu, 23 Oct 2008 03:08:16 +0000 (14:08 +1100)]
ahci: Add support for Promise PDC42819
Add an appropriate entry for the Promise PDC42819 controller. It has an
AHCI mode and so far works correctly with board_ahci.
This chip is found on Promise's FastTrak TX2650 (2 port) and TX4650 (4 port)
software-based RAID cards (for which there is a binary driver, t3sas) and
can be found on some motherboards, for example the MSI K9A2 Platinum,
which calls the chip a Promise T3 controller.
Although this controller also supports SAS devices, its default bootup mode
is AHCI and the binary driver has to do some magic to get the chip into the
appropriate mode to drive SAS disks.
Seeing as no documentation is provided by Promise, adding this entry to the
ahci driver allows the controller to be useful to people as a SATA
controller (with no ill effects on the system if a SAS disk is connected -
probing of the port just times out with "link online but device
misclassified"), without having to resort to using the binary driver. Users
who require SAS or the proprietary software raid can get this functionality
using the binary driver.
Signed-off-by: Mark Nelson <mdnelson8@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
Linus Torvalds [Mon, 27 Oct 2008 18:41:51 +0000 (11:41 -0700)]
Merge branch 'for-linus' of git://git.kernel.dk/linux-2.6-block
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
libata: fix NCQ devices behind port multipliers
scsi: make sure that scsi_init_shared_tag_map() doesn't overwrite existing map
Jens Axboe [Mon, 27 Oct 2008 18:20:59 +0000 (19:20 +0100)]
scsi: make sure that scsi_init_shared_tag_map() doesn't overwrite existing map
Right now callers have to check whether scsi_host->bqt is already
set up, it's much cleaner to just have scsi_init_shared_tag_map()
does this check on its own.
Linus Torvalds [Mon, 27 Oct 2008 16:36:33 +0000 (09:36 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
ALSA: ASoC: Blackfin: update SPORT0 port selector (v2)
ALSA: hda - Restore default pin configs for realtek codecs
sound: use a common working email address
pci: use pci_ioremap_bar() in sound/
MIPS: TXx9: Make firmware parameter passing more robust
When booting Linux on a txx9 board with VxWorks boot loader, it crashes in
prom_getenv(), as VxWorks doesn't pass firmware parameters in a0-a3 (in my
case, the actual leftover values in these registers were 0x80002000,
0x80001fe0, 0x2000, and 0x20).
Make the parsing of argc, argv, and envp a bit more robust by checking if
argc is a number below CKSEG0, and argv/envp point to CKSEG0.
Remove runtime db_* macros as we don't need them any more. In general,
such helpers are useful for initial porting, but once approved, they are
not indispensable.
- Move EMMA related stuff into arch/mips/emma/Kconfig
- Create CONFIG_SOC_EMMA* to handle more EMMA SoCs effectively
- Rename CONFIG_MARKEINS into CONFIG_NEC_MARKEINS
MIPS: Add CONFIG_CPU_R5500 for NEC VR5500 series processors
We already have sufficient infrastructure to support VR5500 and VR5500A
series processors. Here's a Makefile support to make it selectable by
ports, and enable it for NEC EMMA2RH Markeins board.
This patch also fixes a confused target help, and adds 1Gb PageMask bits
supported by VR5500 and its variants.
Daniel Laird [Mon, 16 Jun 2008 14:49:21 +0000 (15:49 +0100)]
MIPS: Add support for NXP PNX833x (STB222/5) into linux kernel
The following patch add support for the NXP PNX833x SOC. More
specifically it adds support for the STB222/5 variant. It fixes
the vectored interrupt issue.
Yoichi Yuasa [Sat, 25 Oct 2008 08:30:35 +0000 (17:30 +0900)]
MIPS: Fix KGDB build error
In file included from include/linux/ptrace.h:49,
from arch/mips/kernel/kgdb.c:25:
/home/yuasa/src/linux/test/mips/linux/arch/mips/include/asm/ptrace.h:123: error: expected declaration specifiers or '...' before '__s64'
/home/yuasa/src/linux/test/mips/linux/arch/mips/include/asm/ptrace.h:124: error: expected declaration specifiers or '...' before '__s64'
/home/yuasa/src/linux/test/mips/linux/arch/mips/include/asm/ptrace.h:126: error: expected declaration specifiers or '...' before '__u32'
/home/yuasa/src/linux/test/mips/linux/arch/mips/include/asm/ptrace.h:127: error: expected declaration specifiers or '...' before '__u32'
make[1]: *** [arch/mips/kernel/kgdb.o] Error 1
Dmitri Vorobiev [Fri, 24 Oct 2008 22:46:57 +0000 (01:46 +0300)]
INPUT: sgi_btns: Add license specification
The SGI Volume Button interface driver uses GPL-only symbols
platform_driver_unregister and platform_driver_register, but
lacks license specification. Thus, when compiled as a module,
this driver cannot be installed. This patch fixes this by
adding the MODULE_LICENSE() specification.
Signed-off-by: Dmitri Vorobiev <dmitri.vorobiev@movial.fi> Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
...
CC arch/mips/pci/fixup-rc32434.o
arch/mips/pci/fixup-rc32434.c: In function 'pcibios_map_irq':
arch/mips/pci/fixup-rc32434.c:46: error: 'GROUP4_IRQ_BASE' undeclared (first use in this function)
arch/mips/pci/fixup-rc32434.c:46: error: (Each undeclared identifier is reported only once
arch/mips/pci/fixup-rc32434.c:46: error: for each function it appears in.)
make[2]: *** [arch/mips/pci/fixup-rc32434.o] Error 1
<-- snip -->
A virtually identical patch was also submitted by Yoichi Yuasa
<yoichi_yuasa@tripeaks.co.jp>.
David Daney [Sat, 18 Oct 2008 00:48:53 +0000 (17:48 -0700)]
MIPS: Don't unmap the memory for dma_sync*.
We were getting away with this for so long only because the only platform
with a non-empty plat_unmap_dma_mem() doesn't call dma_sync_sg_for_cpu()
and dma_sync_sg_for_device() from its commonly used drivers.
Signed-off-by: David Daney <ddaney@caviumnetworks.com> Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Cliff Cai [Mon, 27 Oct 2008 09:09:25 +0000 (17:09 +0800)]
ALSA: ASoC: Blackfin: update SPORT0 port selector (v2)
- Setting the TFS pin selector for SPORT 0 based on whether the selected
port id F or G. If the port is F then no conflict should exist for the
TFS. When Port G is selected and EMAC then there is a conflict between
the PHY interrupt line and TFS. Current settings prevent the conflict
by ignoring the TFS pin when Port G is selected. This allows both
ssm2602 using Port G and EMAC concurrently.
- some code cleanup
Signed-off-by: Cliff Cai <cliff.cai@analog.com> Signed-off-by: Bryan Wu <cooloney@kernel.org> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de>
Takashi Iwai [Mon, 27 Oct 2008 15:56:24 +0000 (16:56 +0100)]
ALSA: hda - Restore default pin configs for realtek codecs
Some machines have broken BIOS resume that doesn't restore the default
pin configuration properly, which results in a wrong detection of HP
pin. This causes a silent speaker output due to missing HP detection.
Related bug: Novell bug#406101
https://bugzilla.novell.com/show_bug.cgi?id=406101
This patch fixes the issue by saving/restoring the default pin configs
by the driver itself.
Steven Rostedt [Fri, 24 Oct 2008 13:12:17 +0000 (09:12 -0400)]
ftrace: use a real variable for ftrace_nop in x86
Impact: avoid section mismatch warning, clean up
The dynamic ftrace determines which nop is safe to use at start up.
When it finds a safe nop for patching, it sets a pointer called ftrace_nop
to point to the code. All call sites are then patched to this nop.
Later, when tracing is turned on, this ftrace_nop variable is again used
to compare the location to make sure it is a nop before we update it to
an mcount call. If this fails just once, a warning is printed and ftrace
is disabled.
Rakib Mullick noted that the code that sets up the nop is a .init section
where as the nop itself is in the .text section. This is needed because
the nop is used later on after boot up. The problem is that the test of the
nop jumps back to the setup code and causes a "section mismatch" warning.
Rakib first recommended to convert the nop to .init.text, but as stated
above, this would fail since that text is used later.
The real solution is to extend Rabik's patch, and to make the ftrace_nop
into an array, and just save the code from the assembly to this array.
Now the section can stay as an init section, and we have a nop to use
later on.
tracing/ftrace: make boot tracer select the sched_switch tracer
Impact: build fix
If the boot tracer is selected but not the sched_switch,
there will be a build failure:
kernel/built-in.o: In function `boot_trace_init':
trace_boot.c:(.text+0x5ee38): undefined reference to `sched_switch_trace'
kernel/built-in.o: In function `disable_boot_trace':
(.text+0x5eee1): undefined reference to `tracing_stop_cmdline_record'
kernel/built-in.o: In function `enable_boot_trace':
(.text+0x5ef11): undefined reference to `tracing_start_cmdline_record'
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
syncookies: fix inclusion of tcp options in syn-ack
libertas: free sk_buff with kfree_skb
btsdio: free sk_buff with kfree_skb
Phonet: do not reply to indication reset packets
Phonet: include generic link-layer header size in MAX_PHONET_HEADER
Eric Miao [Mon, 27 Oct 2008 09:48:50 +0000 (17:48 +0800)]
leds: da903x: fix the building failure of incomplete type of 'work'
The leds-da903x LED driver was missing the proper #include of
linux/workqueue.h, but happened to compile on ARM due to implied
includes through other header files.
We do need the explict include on other architectures (reported at least
for x86-64).
Reported-tested-and-acked-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Eric Miao <eric.miao@marvell.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Chris Lalancette [Fri, 24 Oct 2008 00:40:25 +0000 (17:40 -0700)]
xen: fix Xen domU boot with batched mprotect
Impact: fix guest kernel boot crash on certain configs
Recent i686 2.6.27 kernels with a certain amount of memory (between
736 and 855MB) have a problem booting under a hypervisor that supports
batched mprotect (this includes the RHEL-5 Xen hypervisor as well as
any 3.3 or later Xen hypervisor).
The problem ends up being that xen_ptep_modify_prot_commit() is using
virt_to_machine to calculate which pfn to update. However, this only
works for pages that are in the p2m list, and the pages coming from
change_pte_range() in mm/mprotect.c are kmap_atomic pages. Because of
this, we can run into the situation where the lookup in the p2m table
returns an INVALID_MFN, which we then try to pass to the hypervisor,
which then (correctly) denies the request to a totally bogus pfn.
The right thing to do is to use arbitrary_virt_to_machine, so that we
can be sure we are modifying the right pfn. This unfortunately
introduces a performance penalty because of a full page-table-walk,
but we can avoid that penalty for pages in the p2m list by checking if
virt_addr_valid is true, and if so, just doing the lookup in the p2m
table.
The attached patch implements this, and allows my 2.6.27 i686 based
guest with 768MB of memory to boot on a RHEL-5 hypervisor again.
Thanks to Jeremy for the suggestions about how to fix this particular
issue.
Signed-off-by: Chris Lalancette <clalance@redhat.com> Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Cc: Chris Lalancette <clalance@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Jonas Bonn [Sat, 25 Oct 2008 09:49:20 +0000 (11:49 +0200)]
asm-generic: define DIE_OOPS in asm-generic
Impact: build fix
DIE_OOPS is now used in the generic trace handling code so it needs to
be defined for all architectures. Define it in asm-generic so that it's
available to all by default and doesn't cause build errors for
architectures that rely on the generic implementation.
Signed-off-by: Jonas Bonn <jonas.bonn@gmail.com> Acked-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Stephen Rothwell [Mon, 27 Oct 2008 06:43:28 +0000 (17:43 +1100)]
trace: fix printk warning for u64
A powerpc ppc64_defconfig build produces these warnings:
kernel/trace/ring_buffer.c: In function 'rb_add_time_stamp':
kernel/trace/ring_buffer.c:969: warning: format '%llu' expects type 'long long unsigned int', but argument 2 has type 'u64'
kernel/trace/ring_buffer.c:969: warning: format '%llu' expects type 'long long unsigned int', but argument 3 has type 'u64'
kernel/trace/ring_buffer.c:969: warning: format '%llu' expects type 'long long unsigned int', but argument 4 has type 'u64'
Just cast the u64s to unsigned long long like we do everywhere else.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Acked-by: Steven Rostedt <srostedt@redhat.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
projects / linux-2.6-omap-h63xx.git / log