commit c903f0456b upstream.
At the moment the MSR driver only relies upon file system
checks. This means that anything as root with any capability set
can write to MSRs. Historically that wasn't very interesting but
on modern processors the MSRs are such that writing to them
provides several ways to execute arbitary code in kernel space.
Sample code and documentation on doing this is circulating and
MSR attacks are used on Windows 64bit rootkits already.
In the Linux case you still need to be able to open the device
file so the impact is fairly limited and reduces the security of
some capability and security model based systems down towards
that of a generic "root owns the box" setup.
Therefore they should require CAP_SYS_RAWIO to prevent an
elevation of capabilities. The impact of this is fairly minimal
on most setups because they don't have heavy use of
capabilities. Those using SELinux, SMACK or AppArmor rules might
want to consider if their rulesets on the MSR driver could be
tighter.
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9174adbee4 upstream.
This fixes CVE-2013-0190 / XSA-40
There has been an error on the xen_failsafe_callback path for failed
iret, which causes the stack pointer to be wrong when entering the
iret_exc error path. This can result in the kernel crashing.
In the classic kernel case, the relevant code looked a little like:
popl %eax # Error code from hypervisor
jz 5f
addl $16,%esp
jmp iret_exc # Hypervisor said iret fault
5: addl $16,%esp
# Hypervisor said segment selector fault
Here, there are two identical addls on either option of a branch which
appears to have been optimised by hoisting it above the jz, and
converting it to an lea, which leaves the flags register unaffected.
In the PVOPS case, the code looks like:
popl_cfi %eax # Error from the hypervisor
lea 16(%esp),%esp # Add $16 before choosing fault path
CFI_ADJUST_CFA_OFFSET -16
jz 5f
addl $16,%esp # Incorrectly adjust %esp again
jmp iret_exc
It is possible unprivileged userspace applications to cause this
behaviour, for example by loading an LDT code selector, then changing
the code selector to be not-present. At this point, there is a race
condition where it is possible for the hypervisor to return back to
userspace from an interrupt, fault on its own iret, and inject a
failsafe_callback into the kernel.
This bug has been present since the introduction of Xen PVOPS support
in commit 5ead97c84 (xen: Core Xen implementation), in 2.6.23.
Signed-off-by: Frediano Ziglio <frediano.ziglio@citrix.com>
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a9acc5365d upstream.
SNB graphics devices have a bug that prevent them from accessing certain
memory ranges, namely anything below 1M and in the pages listed in the
table. So reserve those at boot if set detect a SNB gfx device on the
CPU to avoid GPU hangs.
Stephane Marchesin had a similar patch to the page allocator awhile
back, but rather than reserving pages up front, it leaked them at
allocation time.
[ hpa: made a number of stylistic changes, marked arrays as static
const, and made less verbose; use "memblock=debug" for full
verbosity. ]
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: CAI Qian <caiqian@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2bbf0a1427 upstream.
The Way Access Filter in recent AMD CPUs may hurt the performance of
some workloads, caused by aliasing issues in the L1 cache.
This patch disables it on the affected CPUs.
The issue is similar to that one of last year:
http://lkml.indiana.edu/hypermail/linux/kernel/1107.3/00041.html
This new patch does not replace the old one, we just need another
quirk for newer CPUs.
The performance penalty without the patch depends on the
circumstances, but is a bit less than the last year's 3%.
The workloads affected would be those that access code from the same
physical page under different virtual addresses, so different
processes using the same libraries with ASLR or multiple instances of
PIE-binaries. The code needs to be accessed simultaneously from both
cores of the same compute unit.
More details can be found here:
http://developer.amd.com/Assets/SharedL1InstructionCacheonAMD15hCPU.pdf
CPUs affected are anything with the core known as Piledriver.
That includes the new parts of the AMD A-Series (aka Trinity) and the
just released new CPUs of the FX-Series (aka Vishera).
The model numbering is a bit odd here: FX CPUs have model 2,
A-Series has model 10h, with possible extensions to 1Fh. Hence the
range of model ids.
Signed-off-by: Andre Przywara <osp@andrep.de>
Link: http://lkml.kernel.org/r/1351700450-9277-1-git-send-email-osp@andrep.de
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: CAI Qian <caiqian@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2bbf0a1427 upstream.
The Way Access Filter in recent AMD CPUs may hurt the performance of
some workloads, caused by aliasing issues in the L1 cache.
This patch disables it on the affected CPUs.
The issue is similar to that one of last year:
http://lkml.indiana.edu/hypermail/linux/kernel/1107.3/00041.html
This new patch does not replace the old one, we just need another
quirk for newer CPUs.
The performance penalty without the patch depends on the
circumstances, but is a bit less than the last year's 3%.
The workloads affected would be those that access code from the same
physical page under different virtual addresses, so different
processes using the same libraries with ASLR or multiple instances of
PIE-binaries. The code needs to be accessed simultaneously from both
cores of the same compute unit.
More details can be found here:
http://developer.amd.com/Assets/SharedL1InstructionCacheonAMD15hCPU.pdf
CPUs affected are anything with the core known as Piledriver.
That includes the new parts of the AMD A-Series (aka Trinity) and the
just released new CPUs of the FX-Series (aka Vishera).
The model numbering is a bit odd here: FX CPUs have model 2,
A-Series has model 10h, with possible extensions to 1Fh. Hence the
range of model ids.
Signed-off-by: Andre Przywara <osp@andrep.de>
Link: http://lkml.kernel.org/r/1351700450-9277-1-git-send-email-osp@andrep.de
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: CAI Qian <caiqian@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 29e9bf1841 upstream.
Thermal throttle and power limit events are not defined as MCE errors in x86
architecture and should not generate MCE errors in mcelog.
Current kernel generates fake software defined MCE errors for these events.
This may confuse users because they may think the machine has real MCE errors
while actually only thermal throttle or power limit events happen.
To make it worse, buggy firmware on some platforms may falsely generate
the events. Therefore, kernel reports MCE errors which users think as real
hardware errors. Although the firmware bugs should be fixed, on the other hand,
kernel should not report MCE errors either.
So mcelog is not a good mechanism to report these events. To report the events, we count them in respective counters (core_power_limit_count,
package_power_limit_count, core_throttle_count, and package_throttle_count) in
/sys/devices/system/cpu/cpu#/thermal_throttle/. Users can check the counters
for each event on each CPU. Please note that all CPU's on one package report
duplicate counters. It's user application's responsibity to retrieve a package
level counter for one package.
This patch doesn't report package level power limit, core level power limit, and
package level thermal throttle events in mcelog. When the events happen, only
report them in respective counters in sysfs.
Since core level thermal throttle has been legacy code in kernel for a while and
users accepted it as MCE error in mcelog, core level thermal throttle is still
reported in mcelog. In the mean time, the event is counted in a counter in sysfs
as well.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Acked-by: Borislav Petkov <bp@amd64.org>
Acked-by: Tony Luck <tony.luck@intel.com>
Link: http://lkml.kernel.org/r/20111215001945.GA21009@linux-os.sc.intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: maximilian attems <max@stro.at>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f6365201d8 upstream.
The X86_32-only disable_hlt/enable_hlt mechanism was used by the
32-bit floppy driver. Its effect was to replace the use of the
HLT instruction inside default_idle() with cpu_relax() - essentially
it turned off the use of HLT.
This workaround was commented in the code as:
"disable hlt during certain critical i/o operations"
"This halt magic was a workaround for ancient floppy DMA
wreckage. It should be safe to remove."
H. Peter Anvin additionally adds:
"To the best of my knowledge, no-hlt only existed because of
flaky power distributions on 386/486 systems which were sold to
run DOS. Since DOS did no power management of any kind,
including HLT, the power draw was fairly uniform; when exposed
to the much hhigher noise levels you got when Linux used HLT
caused some of these systems to fail.
They were by far in the minority even back then."
Alan Cox further says:
"Also for the Cyrix 5510 which tended to go castors up if a HLT
occurred during a DMA cycle and on a few other boxes HLT during
DMA tended to go astray.
Do we care ? I doubt it. The 5510 was pretty obscure, the 5520
fixed it, the 5530 is probably the oldest still in any kind of
use."
So, let's finally drop this.
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Josh Boyer <jwboyer@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Acked-by: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Stephen Hemminger <shemminger@vyatta.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/n/tip-3rhk9bzf0x9rljkv488tloib@git.kernel.org
[ If anyone cares then alternative instruction patching could be
used to replace HLT with a one-byte NOP instruction. Much simpler. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a349e23d1c upstream.
In 32 bit guests, if a userspace process has %eax == -ERESTARTSYS
(-512) or -ERESTARTNOINTR (-513) when it is interrupted by an event
/and/ the process has a pending signal then %eip (and %eax) are
corrupted when returning to the main process after handling the
signal. The application may then crash with SIGSEGV or a SIGILL or it
may have subtly incorrect behaviour (depending on what instruction it
returned to).
The occurs because handle_signal() is incorrectly thinking that there
is a system call that needs to restarted so it adjusts %eip and %eax
to re-execute the system call instruction (even though user space had
not done a system call).
If %eax == -514 (-ERESTARTNOHAND (-514) or -ERESTART_RESTARTBLOCK
(-516) then handle_signal() only corrupted %eax (by setting it to
-EINTR). This may cause the application to crash or have incorrect
behaviour.
handle_signal() assumes that regs->orig_ax >= 0 means a system call so
any kernel entry point that is not for a system call must push a
negative value for orig_ax. For example, for physical interrupts on
bare metal the inverse of the vector is pushed and page_fault() sets
regs->orig_ax to -1, overwriting the hardware provided error code.
xen_hypervisor_callback() was incorrectly pushing 0 for orig_ax
instead of -1.
Classic Xen kernels pushed %eax which works as %eax cannot be both
non-negative and -RESTARTSYS (etc.), but using -1 is consistent with
other non-system call entry points and avoids some of the tests in
handle_signal().
There were similar bugs in xen_failsafe_callback() of both 32 and
64-bit guests. If the fault was corrected and the normal return path
was used then 0 was incorrectly pushed as the value for orig_ax.
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Acked-by: Jan Beulich <JBeulich@suse.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1bbbbe779a upstream.
On systems with very large memory (1 TB in our case), BIOS may report a
reserved region or a hole in the E820 map, even above the 4 GB range. Exclude
these from the direct mapping.
[ hpa: this should be done not just for > 4 GB but for everything above the legacy
region (1 MB), at the very least. That, however, turns out to require significant
restructuring. That work is well underway, but is not suitable for rc/stable. ]
Signed-off-by: Jacob Shin <jacob.shin@amd.com>
Link: http://lkml.kernel.org/r/1319145326-13902-1-git-send-email-jacob.shin@amd.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 49d859d78c upstream.
If the CPU declares that RDRAND is available, go through a guranteed
reseed sequence, and make sure that it is actually working (producing
data.) If it does not, disable the CPU feature flag.
Allow RDRAND to be disabled on the command line (as opposed to at
compile time) for a user who has special requirements with regards to
random numbers.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: Matt Mackall <mpm@selenic.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a129a7c845 upstream.
When running on 32bit the mce handler could misinterpret
vm86 mode as ring 0. This can affect whether it does recovery
or not; it was possible to panic when recovery was actually
possible.
Fix this by always forcing vm86 to look like ring 3.
[ Backport to 3.0 notes:
Things changed there slightly:
- move mce_get_rip() up. It fills up m->cs and m->ip values which
are evaluated in mce_severity(). Therefore move it up right before
the mce_severity call. This seem to be another bug in 3.0?
- Place the backport (fix m->cs in V86 case) to where m->cs gets
filled which is mce_get_rip() in 3.0
]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Thomas Renninger <trenn@suse.de>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c9fc3f778a upstream.
Microcode reloading in a per-core manner is a very bad idea for both
major x86 vendors. And the thing is, we have such interface with which
we can end up with different microcode versions applied on different
cores of an otherwise homogeneous wrt (family,model,stepping) system.
So turn off the possibility of doing that per core and allow it only
system-wide.
This is a minimal fix which we'd like to see in stable too thus the
more-or-less arbitrary decision to allow system-wide reloading only on
the BSP:
$ echo 1 > /sys/devices/system/cpu/cpu0/microcode/reload
...
and disable the interface on the other cores:
$ echo 1 > /sys/devices/system/cpu/cpu23/microcode/reload
-bash: echo: write error: Invalid argument
Also, allowing the reload only from one CPU (the BSP in
that case) doesn't allow the reload procedure to degenerate
into an O(n^2) deal when triggering reloads from all
/sys/devices/system/cpu/cpuX/microcode/reload sysfs nodes
simultaneously.
A more generic fix will follow.
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1340280437-7718-2-git-send-email-bp@amd64.org
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 141168c36c and
commit 3f806e5098 upstream.
Several fields in struct cpuinfo_x86 were not defined for the
!SMP case, likely to save space. However, those fields still
have some meaning for UP, and keeping them allows some #ifdef
removal from other files. The additional size of the UP kernel
from this change is not significant enough to worry about
keeping up the distinction:
text data bss dec hex filename
4737168 506459 972040 6215667 5ed7f3 vmlinux.o.before
4737444 506459 972040 6215943 5ed907 vmlinux.o.after
for a difference of 276 bytes for an example UP config.
If someone wants those 276 bytes back badly then it should
be implemented in a cleaner way.
Signed-off-by: Kevin Winchester <kjwinchester@gmail.com>
Cc: Steffen Persvold <sp@numascale.com>
Link: http://lkml.kernel.org/r/1324428742-12498-1-git-send-email-kjwinchester@gmail.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7f68b4c2e1 upstream.
Current WARN msg is only for the ati_ixp4x0 board, while this function
is used by mulitple platforms. So this one board specific warning
is not appropriate any more.
Signed-off-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ae10ccdc30 upstream.
Currently when acpi_skip_timer_override is set, it only cover the
(source_irq == 0 && global_irq == 2) cases. While there is also
platform which need use this option and its global_irq is not 2.
This patch will extend acpi_skip_timer_override to cover all
timer overriding cases as long as the source irq is 0.
This is the first part of a fix to kernel bug bugzilla 40002:
"IRQ 0 assigned to VGA"
https://bugzilla.kernel.org/show_bug.cgi?id=40002
Reported-and-tested-by: Szymon Kowalczyk <fazerxlo@o2.pl>
Signed-off-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4ad3341130 upstream.
It makes sense to label "Digital Thermal Sensor" as "DTS", but
unfortunately the string "dts" was already used for "Debug Store", and
/proc/cpuinfo is a user space ABI.
Therefore, rename this to "dtherm".
This conflict went into mainline via the hwmon tree without any x86
maintainer ack, and without any kind of hint in the subject.
a4659053 x86/hwmon: fix initialization of coretemp
Reported-by: Jean Delvare <khali@linux-fr.org>
Link: http://lkml.kernel.org/r/4FE34BCB.5050305@linux.intel.com
Cc: Jan Beulich <JBeulich@suse.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
[bwh: Backported to 3.2: drop the coretemp device table change]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f6b54f083c upstream.
This is the 2nd part of fix for kernel bugzilla 40002:
"IRQ 0 assigned to VGA"
https://bugzilla.kernel.org/show_bug.cgi?id=40002
The root cause is the buggy FW, whose ACPI tables assign the GSI 16
to 2 irqs 0 and 16(VGA), and the VGA is the right owner of GSI 16.
So add a quirk to ignore the irq0 overriding GSI 16 for the
FUJITSU SIEMENS AMILO PRO V2030 platform will solve this issue.
Reported-and-tested-by: Szymon Kowalczyk <fazerxlo@o2.pl>
Signed-off-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f227d4306c upstream.
Currently, the APIC LVT interrupt for error thresholding is implicitly
enabled. However, there are models in the F15h range which do not enable
it. Make the code machinery which sets up the APIC interrupt support
an optional setting and add an ->interrupt_capable member to the bank
representation mirroring that capability and enable the interrupt offset
programming only if it is true.
Simplify code and fixup comment style while at it.
Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: Robert Richter <robert.richter@amd.com>
commit 875e26648c upstream.
Linus pointed out that there was no value is checking whether m->ip
was zero - because zero is a legimate value. If we have a reliable
(or faked in the VM86 case) "m->cs" we can use it to tell whether we
were in user mode or kernelwhen the machine check hit.
Reported-by: Linus Torvalds <torvalds@linuxfoundation.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d5e28005a1 upstream.
With the embed percpu first chunk allocator, x86 uses either PAGE_SIZE
or PMD_SIZE for atom_size. PMD_SIZE is used when CPU supports PSE so
that percpu areas are aligned to PMD mappings and possibly allow using
PMD mappings in vmalloc areas in the future. Using larger atom_size
doesn't waste actual memory; however, it does require larger vmalloc
space allocation later on for !first chunks.
With reasonably sized vmalloc area, PMD_SIZE shouldn't be a problem
but x86_32 at this point is anything but reasonable in terms of
address space and using larger atom_size reportedly leads to frequent
percpu allocation failures on certain setups.
As there is no reason to not use PMD_SIZE on x86_64 as vmalloc space
is aplenty and most x86_64 configurations support PSE, fix the issue
by always using PMD_SIZE on x86_64 and PAGE_SIZE on x86_32.
v2: drop cpu_has_pse test and make x86_64 always use PMD_SIZE and
x86_32 PAGE_SIZE as suggested by hpa.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Yanmin Zhang <yanmin.zhang@intel.com>
Reported-by: ShuoX Liu <shuox.liu@intel.com>
Acked-by: H. Peter Anvin <hpa@zytor.com>
LKML-Reference: <4F97BA98.6010001@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit cbf2829b61 upstream.
Current APIC code assumes MSR_IA32_APICBASE is present for all systems.
Pentium Classic P5 and friends didn't have this MSR. MSR_IA32_APICBASE
was introduced as an architectural MSR by Intel @ P6.
Code paths that can touch this MSR invalidly are when vendor == Intel &&
cpu-family == 5 and APIC bit is set in CPUID - or when you simply pass
lapic on the kernel command line, on a P5.
The below patch stops Linux incorrectly interfering with the
MSR_IA32_APICBASE for P5 class machines. Other code paths exist that
touch the MSR - however those paths are not currently reachable for a
conformant P5.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linux.intel.com>
Link: http://lkml.kernel.org/r/4F8EEDD3.1080404@linux.intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9ddd592a19 upstream.
Unfortunatly the interrupts for the event log and the
peripheral page-faults are only enabled at boot but not
re-enabled at resume. Fix that.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
[bwh: Backport to 3.0:
- Drop change to PPR log which was added in 3.3
- Source is under arch/x86/kernel]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9993bc635d upstream.
When a machine boots up, the TSC generally gets reset. However,
when kexec is used to boot into a kernel, the TSC value would be
carried over from the previous kernel. The computation of
cycns_offset in set_cyc2ns_scale is prone to an overflow, if the
machine has been up more than 208 days prior to the kexec. The
overflow happens when we multiply *scale, even though there is
enough room to store the final answer.
We fix this issue by decomposing tsc_now into the quotient and
remainder of division by CYC2NS_SCALE_FACTOR and then performing
the multiplication separately on the two components.
Refactor code to share the calculation with the previous
fix in __cycles_2_ns().
Signed-off-by: Salman Qazi <sqazi@google.com>
Acked-by: John Stultz <john.stultz@linaro.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Turner <pjt@google.com>
Cc: john stultz <johnstul@us.ibm.com>
Link: http://lkml.kernel.org/r/20120310004027.19291.88460.stgit@dungbeetle.mtv.corp.google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 3751d3e85c upstream.
There has long been a limitation using software breakpoints with a
kernel compiled with CONFIG_DEBUG_RODATA going back to 2.6.26. For
this particular patch, it will apply cleanly and has been tested all
the way back to 2.6.36.
The kprobes code uses the text_poke() function which accommodates
writing a breakpoint into a read-only page. The x86 kgdb code can
solve the problem similarly by overriding the default breakpoint
set/remove routines and using text_poke() directly.
The x86 kgdb code will first attempt to use the traditional
probe_kernel_write(), and next try using a the text_poke() function.
The break point install method is tracked such that the correct break
point removal routine will get called later on.
Cc: x86@kernel.org
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Inspried-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 57779dc2b3 upstream.
While running the latest Linux as guest under VMware in highly
over-committed situations, we have seen cases when the refined TSC
algorithm fails to get a valid tsc_start value in
tsc_refine_calibration_work from multiple attempts. As a result the
kernel keeps on scheduling the tsc_irqwork task for later. Subsequently
after several attempts when it gets a valid start value it goes through
the refined calibration and either bails out or uses the new results.
Given that the kernel originally read the TSC frequency from the
platform, which is the best it can get, I don't think there is much
value in refining it.
So for systems which get the TSC frequency from the platform we
should skip the refined tsc algorithm.
We can use the TSC_RELIABLE cpu cap flag to detect this, right now it is
set only on VMware and for Moorestown Penwell both of which have there
own TSC calibration methods.
Signed-off-by: Alok N Kataria <akataria@vmware.com>
Cc: John Stultz <johnstul@us.ibm.com>
Cc: Dirk Brandewie <dirk.brandewie@gmail.com>
Cc: Alan Cox <alan@linux.intel.com>
[jstultz: Reworked to simply not schedule the refining work,
rather then scheduling the work and bombing out later]
Signed-off-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 29a2e2836f upstream.
The problem occurs on !CONFIG_VM86 kernels [1] when a kernel-mode task
returns from a system call with a pending signal.
A real-life scenario is a child of 'khelper' returning from a failed
kernel_execve() in ____call_usermodehelper() [ kernel/kmod.c ].
kernel_execve() fails due to a pending SIGKILL, which is the result of
"kill -9 -1" (at least, busybox's init does it upon reboot).
The loop is as follows:
* syscall_exit_work:
- work_pending: // start_of_the_loop
- work_notify_sig:
- do_notify_resume()
- do_signal()
- if (!user_mode(regs)) return;
- resume_userspace // TIF_SIGPENDING is still set
- work_pending // so we call work_pending => goto
// start_of_the_loop
More information can be found in another LKML thread:
http://www.serverphorums.com/read.php?12,457826
[1] the problem was also seen on MIPS.
Signed-off-by: Dmitry Adamushko <dmitry.adamushko@gmail.com>
Link: http://lkml.kernel.org/r/1332448765.2299.68.camel@dimm
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Roland McGrath <roland@hack.frob.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1a5a9906d4 upstream.
In some cases it may happen that pmd_none_or_clear_bad() is called with
the mmap_sem hold in read mode. In those cases the huge page faults can
allocate hugepmds under pmd_none_or_clear_bad() and that can trigger a
false positive from pmd_bad() that will not like to see a pmd
materializing as trans huge.
It's not khugepaged causing the problem, khugepaged holds the mmap_sem
in write mode (and all those sites must hold the mmap_sem in read mode
to prevent pagetables to go away from under them, during code review it
seems vm86 mode on 32bit kernels requires that too unless it's
restricted to 1 thread per process or UP builds). The race is only with
the huge pagefaults that can convert a pmd_none() into a
pmd_trans_huge().
Effectively all these pmd_none_or_clear_bad() sites running with
mmap_sem in read mode are somewhat speculative with the page faults, and
the result is always undefined when they run simultaneously. This is
probably why it wasn't common to run into this. For example if the
madvise(MADV_DONTNEED) runs zap_page_range() shortly before the page
fault, the hugepage will not be zapped, if the page fault runs first it
will be zapped.
Altering pmd_bad() not to error out if it finds hugepmds won't be enough
to fix this, because zap_pmd_range would then proceed to call
zap_pte_range (which would be incorrect if the pmd become a
pmd_trans_huge()).
The simplest way to fix this is to read the pmd in the local stack
(regardless of what we read, no need of actual CPU barriers, only
compiler barrier needed), and be sure it is not changing under the code
that computes its value. Even if the real pmd is changing under the
value we hold on the stack, we don't care. If we actually end up in
zap_pte_range it means the pmd was not none already and it was not huge,
and it can't become huge from under us (khugepaged locking explained
above).
All we need is to enforce that there is no way anymore that in a code
path like below, pmd_trans_huge can be false, but pmd_none_or_clear_bad
can run into a hugepmd. The overhead of a barrier() is just a compiler
tweak and should not be measurable (I only added it for THP builds). I
don't exclude different compiler versions may have prevented the race
too by caching the value of *pmd on the stack (that hasn't been
verified, but it wouldn't be impossible considering
pmd_none_or_clear_bad, pmd_bad, pmd_trans_huge, pmd_none are all inlines
and there's no external function called in between pmd_trans_huge and
pmd_none_or_clear_bad).
if (pmd_trans_huge(*pmd)) {
if (next-addr != HPAGE_PMD_SIZE) {
VM_BUG_ON(!rwsem_is_locked(&tlb->mm->mmap_sem));
split_huge_page_pmd(vma->vm_mm, pmd);
} else if (zap_huge_pmd(tlb, vma, pmd, addr))
continue;
/* fall through */
}
if (pmd_none_or_clear_bad(pmd))
Because this race condition could be exercised without special
privileges this was reported in CVE-2012-1179.
The race was identified and fully explained by Ulrich who debugged it.
I'm quoting his accurate explanation below, for reference.
====== start quote =======
mapcount 0 page_mapcount 1
kernel BUG at mm/huge_memory.c:1384!
At some point prior to the panic, a "bad pmd ..." message similar to the
following is logged on the console:
mm/memory.c:145: bad pmd ffff8800376e1f98(80000000314000e7).
The "bad pmd ..." message is logged by pmd_clear_bad() before it clears
the page's PMD table entry.
143 void pmd_clear_bad(pmd_t *pmd)
144 {
-> 145 pmd_ERROR(*pmd);
146 pmd_clear(pmd);
147 }
After the PMD table entry has been cleared, there is an inconsistency
between the actual number of PMD table entries that are mapping the page
and the page's map count (_mapcount field in struct page). When the page
is subsequently reclaimed, __split_huge_page() detects this inconsistency.
1381 if (mapcount != page_mapcount(page))
1382 printk(KERN_ERR "mapcount %d page_mapcount %d\n",
1383 mapcount, page_mapcount(page));
-> 1384 BUG_ON(mapcount != page_mapcount(page));
The root cause of the problem is a race of two threads in a multithreaded
process. Thread B incurs a page fault on a virtual address that has never
been accessed (PMD entry is zero) while Thread A is executing an madvise()
system call on a virtual address within the same 2 MB (huge page) range.
virtual address space
.---------------------.
| |
| |
.-|---------------------|
| | |
| | |<-- B(fault)
| | |
2 MB | |/////////////////////|-.
huge < |/////////////////////| > A(range)
page | |/////////////////////|-'
| | |
| | |
'-|---------------------|
| |
| |
'---------------------'
- Thread A is executing an madvise(..., MADV_DONTNEED) system call
on the virtual address range "A(range)" shown in the picture.
sys_madvise
// Acquire the semaphore in shared mode.
down_read(¤t->mm->mmap_sem)
...
madvise_vma
switch (behavior)
case MADV_DONTNEED:
madvise_dontneed
zap_page_range
unmap_vmas
unmap_page_range
zap_pud_range
zap_pmd_range
//
// Assume that this huge page has never been accessed.
// I.e. content of the PMD entry is zero (not mapped).
//
if (pmd_trans_huge(*pmd)) {
// We don't get here due to the above assumption.
}
//
// Assume that Thread B incurred a page fault and
.---------> // sneaks in here as shown below.
| //
| if (pmd_none_or_clear_bad(pmd))
| {
| if (unlikely(pmd_bad(*pmd)))
| pmd_clear_bad
| {
| pmd_ERROR
| // Log "bad pmd ..." message here.
| pmd_clear
| // Clear the page's PMD entry.
| // Thread B incremented the map count
| // in page_add_new_anon_rmap(), but
| // now the page is no longer mapped
| // by a PMD entry (-> inconsistency).
| }
| }
|
v
- Thread B is handling a page fault on virtual address "B(fault)" shown
in the picture.
...
do_page_fault
__do_page_fault
// Acquire the semaphore in shared mode.
down_read_trylock(&mm->mmap_sem)
...
handle_mm_fault
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma))
// We get here due to the above assumption (PMD entry is zero).
do_huge_pmd_anonymous_page
alloc_hugepage_vma
// Allocate a new transparent huge page here.
...
__do_huge_pmd_anonymous_page
...
spin_lock(&mm->page_table_lock)
...
page_add_new_anon_rmap
// Here we increment the page's map count (starts at -1).
atomic_set(&page->_mapcount, 0)
set_pmd_at
// Here we set the page's PMD entry which will be cleared
// when Thread A calls pmd_clear_bad().
...
spin_unlock(&mm->page_table_lock)
The mmap_sem does not prevent the race because both threads are acquiring
it in shared mode (down_read). Thread B holds the page_table_lock while
the page's map count and PMD table entry are updated. However, Thread A
does not synchronize on that lock.
====== end quote =======
[akpm@linux-foundation.org: checkpatch fixes]
Reported-by: Ulrich Obergfell <uobergfe@redhat.com>
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dave Jones <davej@redhat.com>
Acked-by: Larry Woodman <lwoodman@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Mark Salter <msalter@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 34ddc81a23 upstream.
After all the FPU state cleanups and finally finding the problem that
caused all our FPU save/restore problems, this re-introduces the
preloading of FPU state that was removed in commit b3b0870ef3 ("i387:
do not preload FPU state at task switch time").
However, instead of simply reverting the removal, this reimplements
preloading with several fixes, most notably
- properly abstracted as a true FPU state switch, rather than as
open-coded save and restore with various hacks.
In particular, implementing it as a proper FPU state switch allows us
to optimize the CR0.TS flag accesses: there is no reason to set the
TS bit only to then almost immediately clear it again. CR0 accesses
are quite slow and expensive, don't flip the bit back and forth for
no good reason.
- Make sure that the same model works for both x86-32 and x86-64, so
that there are no gratuitous differences between the two due to the
way they save and restore segment state differently due to
architectural differences that really don't matter to the FPU state.
- Avoid exposing the "preload" state to the context switch routines,
and in particular allow the concept of lazy state restore: if nothing
else has used the FPU in the meantime, and the process is still on
the same CPU, we can avoid restoring state from memory entirely, just
re-expose the state that is still in the FPU unit.
That optimized lazy restore isn't actually implemented here, but the
infrastructure is set up for it. Of course, older CPU's that use
'fnsave' to save the state cannot take advantage of this, since the
state saving also trashes the state.
In other words, there is now an actual _design_ to the FPU state saving,
rather than just random historical baggage. Hopefully it's easier to
follow as a result.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f94edacf99 upstream.
This moves the bit that indicates whether a thread has ownership of the
FPU from the TS_USEDFPU bit in thread_info->status to a word of its own
(called 'has_fpu') in task_struct->thread.has_fpu.
This fixes two independent bugs at the same time:
- changing 'thread_info->status' from the scheduler causes nasty
problems for the other users of that variable, since it is defined to
be thread-synchronous (that's what the "TS_" part of the naming was
supposed to indicate).
So perfectly valid code could (and did) do
ti->status |= TS_RESTORE_SIGMASK;
and the compiler was free to do that as separate load, or and store
instructions. Which can cause problems with preemption, since a task
switch could happen in between, and change the TS_USEDFPU bit. The
change to TS_USEDFPU would be overwritten by the final store.
In practice, this seldom happened, though, because the 'status' field
was seldom used more than once, so gcc would generally tend to
generate code that used a read-modify-write instruction and thus
happened to avoid this problem - RMW instructions are naturally low
fat and preemption-safe.
- On x86-32, the current_thread_info() pointer would, during interrupts
and softirqs, point to a *copy* of the real thread_info, because
x86-32 uses %esp to calculate the thread_info address, and thus the
separate irq (and softirq) stacks would cause these kinds of odd
thread_info copy aliases.
This is normally not a problem, since interrupts aren't supposed to
look at thread information anyway (what thread is running at
interrupt time really isn't very well-defined), but it confused the
heck out of irq_fpu_usable() and the code that tried to squirrel
away the FPU state.
(It also caused untold confusion for us poor kernel developers).
It also turns out that using 'task_struct' is actually much more natural
for most of the call sites that care about the FPU state, since they
tend to work with the task struct for other reasons anyway (ie
scheduling). And the FPU data that we are going to save/restore is
found there too.
Thanks to Arjan Van De Ven <arjan@linux.intel.com> for pointing us to
the %esp issue.
Cc: Arjan van de Ven <arjan@linux.intel.com>
Reported-and-tested-by: Raphael Prevost <raphael@buro.asia>
Acked-and-tested-by: Suresh Siddha <suresh.b.siddha@intel.com>
Tested-by: Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4903062b54 upstream.
The AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception is
pending. In order to not leak FIP state from one process to another, we
need to do a floating point load after the fxsave of the old process,
and before the fxrstor of the new FPU state. That resets the state to
the (uninteresting) kernel load, rather than some potentially sensitive
user information.
We used to do this directly after the FPU state save, but that is
actually very inconvenient, since it
(a) corrupts what is potentially perfectly good FPU state that we might
want to lazy avoid restoring later and
(b) on x86-64 it resulted in a very annoying ordering constraint, where
"__unlazy_fpu()" in the task switch needs to be delayed until after
the DS segment has been reloaded just to get the new DS value.
Coupling it to the fxrstor instead of the fxsave automatically avoids
both of these issues, and also ensures that we only do it when actually
necessary (the FP state after a save may never actually get used). It's
simply a much more natural place for the leaked state cleanup.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b3b0870ef3 upstream.
Yes, taking the trap to re-load the FPU/MMX state is expensive, but so
is spending several days looking for a bug in the state save/restore
code. And the preload code has some rather subtle interactions with
both paravirtualization support and segment state restore, so it's not
nearly as simple as it should be.
Also, now that we no longer necessarily depend on a single bit (ie
TS_USEDFPU) for keeping track of the state of the FPU, we migth be able
to do better. If we are really switching between two processes that
keep touching the FP state, save/restore is inevitable, but in the case
of having one process that does most of the FPU usage, we may actually
be able to do much better than the preloading.
In particular, we may be able to keep track of which CPU the process ran
on last, and also per CPU keep track of which process' FP state that CPU
has. For modern CPU's that don't destroy the FPU contents on save time,
that would allow us to do a lazy restore by just re-enabling the
existing FPU state - with no restore cost at all!
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6d59d7a9f5 upstream.
This creates three helper functions that do the TS_USEDFPU accesses, and
makes everybody that used to do it by hand use those helpers instead.
In addition, there's a couple of helper functions for the "change both
CR0.TS and TS_USEDFPU at the same time" case, and the places that do
that together have been changed to use those. That means that we have
fewer random places that open-code this situation.
The intent is partly to clarify the code without actually changing any
semantics yet (since we clearly still have some hard to reproduce bug in
this area), but also to make it much easier to use another approach
entirely to caching the CR0.TS bit for software accesses.
Right now we use a bit in the thread-info 'status' variable (this patch
does not change that), but we might want to make it a full field of its
own or even make it a per-cpu variable.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 15d8791cae upstream.
Commit 5b1cbac377 ("i387: make irq_fpu_usable() tests more robust")
added a sanity check to the #NM handler to verify that we never cause
the "Device Not Available" exception in kernel mode.
However, that check actually pinpointed a (fundamental) race where we do
cause that exception as part of the signal stack FPU state save/restore
code.
Because we use the floating point instructions themselves to save and
restore state directly from user mode, we cannot do that atomically with
testing the TS_USEDFPU bit: the user mode access itself may cause a page
fault, which causes a task switch, which saves and restores the FP/MMX
state from the kernel buffers.
This kind of "recursive" FP state save is fine per se, but it means that
when the signal stack save/restore gets restarted, it will now take the
'#NM' exception we originally tried to avoid. With preemption this can
happen even without the page fault - but because of the user access, we
cannot just disable preemption around the save/restore instruction.
There are various ways to solve this, including using the
"enable/disable_page_fault()" helpers to not allow page faults at all
during the sequence, and fall back to copying things by hand without the
use of the native FP state save/restore instructions.
However, the simplest thing to do is to just allow the #NM from kernel
space, but fix the race in setting and clearing CR0.TS that this all
exposed: the TS bit changes and the TS_USEDFPU bit absolutely have to be
atomic wrt scheduling, so while the actual state save/restore can be
interrupted and restarted, the act of actually clearing/setting CR0.TS
and the TS_USEDFPU bit together must not.
Instead of just adding random "preempt_disable/enable()" calls to what
is already excessively ugly code, this introduces some helper functions
that mostly mirror the "kernel_fpu_begin/end()" functionality, just for
the user state instead.
Those helper functions should probably eventually replace the other
ad-hoc CR0.TS and TS_USEDFPU tests too, but I'll need to think about it
some more: the task switching functionality in particular needs to
expose the difference between the 'prev' and 'next' threads, while the
new helper functions intentionally were written to only work with
'current'.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5b1cbac377 upstream.
Some code - especially the crypto layer - wants to use the x86
FP/MMX/AVX register set in what may be interrupt (typically softirq)
context.
That *can* be ok, but the tests for when it was ok were somewhat
suspect. We cannot touch the thread-specific status bits either, so
we'd better check that we're not going to try to save FP state or
anything like that.
Now, it may be that the TS bit is always cleared *before* we set the
USEDFPU bit (and only set when we had already cleared the USEDFP
before), so the TS bit test may actually have been sufficient, but it
certainly was not obviously so.
So this explicitly verifies that we will not touch the TS_USEDFPU bit,
and adds a few related sanity-checks. Because it seems that somehow
AES-NI is corrupting user FP state. The cause is not clear, and this
patch doesn't fix it, but while debugging it I really wanted the code to
be more obviously correct and robust.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit be98c2cdb1 upstream.
It was marked asmlinkage for some really old and stale legacy reasons.
Fix that and the equally stale comment.
Noticed when debugging the irq_fpu_usable() bugs.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5b68edc91c upstream.
We've decided to provide CPU family specific container files
(starting with CPU family 15h). E.g. for family 15h we have to
load microcode_amd_fam15h.bin instead of microcode_amd.bin
Rationale is that starting with family 15h patch size is larger
than 2KB which was hard coded as maximum patch size in various
microcode loaders (not just Linux).
Container files which include patches larger than 2KB cause
different kinds of trouble with such old patch loaders. Thus we
have to ensure that the default container file provides only
patches with size less than 2KB.
Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
Cc: Borislav Petkov <borislav.petkov@amd.com>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/r/20120120164412.GD24508@alberich.amd.com
[ documented the naming convention and tidied the code a bit. ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
