This fixes threshold_bank4 support on multi-node processors.
The correct mask to use is llc_shared_map, representing an internal
node on Magny-Cours.
We need to create 2 sets of symlinks for sibling shared banks -- one
set for each internal node, symlinks of each set should target the
first core on same internal node.
Currently only one set is created where all symlinks are targeting
the first core of the entire socket.
Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
An older test-box started hanging at the following point during
bootup:
[ 0.022996] Mount-cache hash table entries: 512
[ 0.024996] Initializing cgroup subsys debug
[ 0.025996] Initializing cgroup subsys cpuacct
[ 0.026995] Initializing cgroup subsys devices
[ 0.027995] Initializing cgroup subsys freezer
[ 0.028995] mce: CPU supports 5 MCE banks
I've bisected it down to commit 4efc0670 ("x86, mce: use 64bit
machine check code on 32bit"), which utilizes the MCE code on
32-bit systems too.
The problem is caused by this detail in my config:
# CONFIG_CPU_SUP_INTEL is not set
This disables the quirks in mce_cpu_quirks() but still enables
MCE support - which then hangs due to the missing quirk
workaround needed on this CPU:
if (c->x86 == 6 && c->x86_model < 0x1A && banks > 0)
mce_banks[0].init = 0;
The safe solution is to not initialize MCEs if we dont know on
what CPU we are running (or if that CPU's support code got
disabled in the config).
Also be a bit more defensive on 32-bit systems: dont do a
boot-time dump of pending MCEs not just on the specific system
that we found a problem with (Pentium-M), but earlier ones as
well.
Now this problem is probably not common and disabling CPU
support is rare - but still being more defensive in something
we turned on for a wide range of CPUs is prudent.
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
LKML-Reference: Message-ID: <4A88E3E4.40506@jp.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
On my legacy Pentium M laptop (Acer Extensa 2900) I get bogus MCE on a cold
boot with CONFIG_X86_NEW_MCE enabled, i.e. (after decoding it with mcelog):
MCE 0
HARDWARE ERROR. This is *NOT* a software problem!
Please contact your hardware vendor
CPU 0 BANK 1 MCG status:
MCi status:
Error overflow
Uncorrected error
Error enabled
Processor context corrupt
MCA: Data CACHE Level-1 UNKNOWN Error
STATUS f200000000000195 MCGSTATUS 0
[ The other STATUS values observed: f2000000000001b5 (... UNKNOWN error)
and f200000000000115 (... READ Error).
To verify that this is not a CONFIG_X86_NEW_MCE bug I also modified
the CONFIG_X86_OLD_MCE code (which doesn't log any MCEs) to dump
content of STATUS MSR before it is cleared during initialization. ]
Since the bogus MCE results in a kernel taint (which in turn disables
lockdep support) don't log boot MCEs on Pentium M (model == 13) CPUs
by default ("mce=bootlog" boot parameter can be be used to get the old
behavior).
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>
Reviewed-by: Andi Kleen <andi@firstfloor.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
My Latitude d630 seems to be handling thermal events in SMI by
lowering the max frequency of the CPU till it cools down but
still leaks the "everything is normal" events.
This spams the console and with high priority printks.
Adjust therm_throt driver to only print messages about the fact
that temperatire returned back to normal when leaving the
throttling state.
Also lower the severity of "back to normal" message from
KERN_CRIT to KERN_INFO.
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Acked-by: H. Peter Anvin <hpa@zytor.com>
LKML-Reference: <20090810051513.0558F526EC9@mailhub.coreip.homeip.net>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix the condition checking the result of strchr() (which previously
could result in an oops), and make the function return the number of
bytes actively used.
[ Impact: fix oops ]
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: Andi Kleen <andi@firstfloor.org>
LKML-Reference: <4A5F04B7020000780000AB59@vpn.id2.novell.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Commit 5fd29d6ccb ("printk: clean up
handling of log-levels and newlines") changed printk semantics. printk
lines with multiple KERN_<level> prefixes are no longer emitted as
before the patch.
<level> is now included in the output on each additional use.
Remove all uses of multiple KERN_<level>s in formats.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If CONFIG_NO_HZ + CONFIG_SMP, timer added via add_timer() might
be migrated on other cpu. Use add_timer_on() instead.
Avoids the following failure:
Maciej Rutecki wrote:
> > After normal boot I try:
> >
> > echo 1 > /sys/devices/system/machinecheck/machinecheck0/check_interval
> >
> > I found this in dmesg:
> >
> > [ 141.704025] ------------[ cut here ]------------
> > [ 141.704039] WARNING: at arch/x86/kernel/cpu/mcheck/mce.c:1102
> > mcheck_timer+0xf5/0x100()
Reported-by: Maciej Rutecki <maciej.rutecki@gmail.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Tested-by: Maciej Rutecki <maciej.rutecki@gmail.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
We need a cleared cpu_mask to record if mce is initialized, especially
when MAXSMP is used.
used zalloc_... instead
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Reviewed-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: stable@kernel.org
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The sysfs attribute cmci_disabled was accidentall turned into a
duplicate of ignore_ce, breaking all other attributes.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
mce_intel.c uses apic_write() and lapic_get_maxlvt(), and so it needs
<asm/apic.h>.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
If APIC was disabled (for some reason) and as result
it's not even mapped we should not try to enable thermal
interrupts at all.
Reported-by: Simon Holm Thøgersen <odie@cs.aau.dk>
Tested-by: Simon Holm Thøgersen <odie@cs.aau.dk>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
LKML-Reference: <20090615182633.GA7606@lenovo>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Rename files that are no longer 64bit specific:
mce_amd_64.c => mce_amd.c
mce_intel_64.c => mce_intel.c
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Now all symbols in the header are static. Remove the header.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
move intel_init_thermal() into therm_throt.c
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Put common functions into therm_throt.c, modify Makefile.
unexpected_thermal_interrupt
intel_thermal_interrupt
smp_thermal_interrupt
intel_set_thermal_handler
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Break smp_thermal_interrupt() into two functions.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Remove unused argument regs from handlers, and use inc_irq_stat.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The mce_disabled on 32bit is a tristate variable [1,0,-1],
while 64bit version is boolean [0,1].
This patch makes mce_disabled always boolean, and use mce_p5_enabled
to indicate the third state instead.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
There are 2 headers:
arch/x86/include/asm/mce.h
arch/x86/kernel/cpu/mcheck/mce.h
and in the latter small header:
#include <asm/mce.h>
This patch move all contents in the latter header into the former,
and fix all files using the latter to include the former instead.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Add sysfs interface for admins who want to tweak these options without
rebooting the system.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
"trigger" is not straight forward name for valiable that holds name
of user mode helper program which triggered by machine check events.
This patch renames this valiable and kins to more recognizable names.
trigger => mce_helper
trigger_argv => mce_helper_argv
notify_user => mce_need_notify
No functional changes.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Add __read_mostly to data written during setup.
Suggested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Simplify interface of mce_start():
- no_way_out = mce_start(no_way_out, &order);
+ order = mce_start(&no_way_out);
Now Monarch and Subjects share same exit(return) in usual path.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
In mce_cpu_restart, mce_init_timer is called unconditionally.
If !mce_available (e.g. mce is disabled), there are no useful work
for timer. Stop running it.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
If one CPU has no_way_out == 1, all other CPUs should have no_way_out
== 1. But despite global_nwo is read after mce_callin, global_nwo is
updated after mce_callin too. So it is possible that some CPU read
global_nwo before some other CPU update global_nwo, so that no_way_out
== 1 for some CPU, while no_way_out == 0 for some other CPU.
This patch fixes this race condition via moving mce_callin updating
after global_nwo updating, with a smp_wmb in between. A smp_rmb is
added between their reading too.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
* 'kvm-updates/2.6.31' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (138 commits)
KVM: Prevent overflow in largepages calculation
KVM: Disable large pages on misaligned memory slots
KVM: Add VT-x machine check support
KVM: VMX: Rename rmode.active to rmode.vm86_active
KVM: Move "exit due to NMI" handling into vmx_complete_interrupts()
KVM: Disable CR8 intercept if tpr patching is active
KVM: Do not migrate pending software interrupts.
KVM: inject NMI after IRET from a previous NMI, not before.
KVM: Always request IRQ/NMI window if an interrupt is pending
KVM: Do not re-execute INTn instruction.
KVM: skip_emulated_instruction() decode instruction if size is not known
KVM: Remove irq_pending bitmap
KVM: Do not allow interrupt injection from userspace if there is a pending event.
KVM: Unprotect a page if #PF happens during NMI injection.
KVM: s390: Verify memory in kvm run
KVM: s390: Sanity check on validity intercept
KVM: s390: Unlink vcpu on destroy - v2
KVM: s390: optimize float int lock: spin_lock_bh --> spin_lock
KVM: s390: use hrtimer for clock wakeup from idle - v2
KVM: s390: Fix memory slot versus run - v3
...
This patch introduces three boot options (no_cmci, dont_log_ce
and ignore_ce) to control handling for corrected errors.
The "mce=no_cmci" boot option disables the CMCI feature.
Since CMCI is a new feature so having boot controls to disable
it will be a help if the hardware is misbehaving.
The "mce=dont_log_ce" boot option disables logging for corrected
errors. All reported corrected errors will be cleared silently.
This option will be useful if you never care about corrected
errors.
The "mce=ignore_ce" boot option disables features for corrected
errors, i.e. polling timer and cmci. All corrected events are
not cleared and kept in bank MSRs.
Usually this disablement is not recommended, however it will be
a help if there are some conflict with the BIOS or hardware
monitoring applications etc., that clears corrected events in
banks instead of OS.
[ And trivial cleanup (space -> tab) for doc is included. ]
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
LKML-Reference: <4A30ACDF.5030408@jp.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch:
- Adds print_mce_head() instead of first flag
- Makes the header to be printed always
- Stops double printing of corrected errors
[ This portion originates from Huang Ying's patch ]
Originally-From: Huang Ying <ying.huang@intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
LKML-Reference: <4A30AC83.5010708@jp.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'x86-cleanups-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, nmi: Use predefined numbers instead of hardcoded one
x86: asm/processor.h: remove double declaration
x86, mtrr: replace MTRRdefType_MSR with msr-index's MSR_MTRRdefType
x86, mtrr: replace MTRRfix4K_C0000_MSR with msr-index's MSR_MTRRfix4K_C0000
x86, mtrr: remove mtrr MSRs double declaration
x86, mtrr: replace MTRRfix16K_80000_MSR with msr-index's MSR_MTRRfix16K_80000
x86, mtrr: replace MTRRfix64K_00000_MSR with msr-index's MSR_MTRRfix64K_00000
x86, mtrr: replace MTRRcap_MSR with msr-index's MSR_MTRRcap
x86: mce: remove duplicated #include
x86: msr-index.h remove duplicate MSR C001_0015 declaration
x86: clean up arch/x86/kernel/tsc_sync.c a bit
x86: use symbolic name for VM86_SIGNAL when used as vm86 default return
x86: added 'ifndef _ASM_X86_IOMAP_H' to iomap.h
x86: avoid multiple declaration of kstack_depth_to_print
x86: vdso/vma.c declare vdso_enabled and arch_setup_additional_pages before they get used
x86: clean up declarations and variables
x86: apic/x2apic_cluster.c x86_cpu_to_logical_apicid should be static
x86 early quirks: eliminate unused function
VT-x needs an explicit MC vector intercept to handle machine checks in the
hyper visor.
It also has a special option to catch machine checks that happen
during VT entry.
Do these interceptions and forward them to the Linux machine check
handler. Make it always look like user space is interrupted because
the machine check handler treats kernel/user space differently.
Thanks to Jiang Yunhong for help and testing.
Cc: stable@kernel.org
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
These are defined as static cpumask_var_t so if MAXSMP is not used,
they are cleared already. Avoid surprises when MAXSMP is enabled.
Signed-off-by: Yinghai Lu <yinghai.lu@kernel.org>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Newer Intel CPUs support a new class of machine checks called recoverable
action optional.
Action Optional means that the CPU detected some form of corruption in
the background and tells the OS about using a machine check
exception. The OS can then take appropiate action, like killing the
process with the corrupted data or logging the event properly to disk.
This is done by the new generic high level memory failure handler added
in a earlier patch. The high level handler takes the address with the
failed memory and does the appropiate action, like killing the process.
In this version of the patch the high level handler is stubbed out
with a weak function to not create a direct dependency on the hwpoison
branch.
The high level handler cannot be directly called from the machine check
exception though, because it has to run in a defined process context to
be able to sleep when taking VM locks (it is not expected to sleep for a
long time, just do so in some exceptional cases like lock contention)
Thus the MCE handler has to queue a work item for process context,
trigger process context and then call the high level handler from there.
This patch adds two path to process context: through a per thread kernel
exit notify_user() callback or through a high priority work item.
The first runs when the process exits back to user space, the other when
it goes to sleep and there is no higher priority process.
The machine check handler will schedule both, and whoever runs first
will grab the event. This is done because quick reaction to this
event is critical to avoid a potential more fatal machine check
when the corruption is consumed.
There is a simple lock less ring buffer to queue the corrupted
addresses between the exception handler and the process context handler.
Then in process context it just calls the high level VM code with
the corrupted PFNs.
The code adds the required code to extract the failed address from
the CPU's machine check registers. It doesn't try to handle all
possible cases -- the specification has 6 different ways to specify
memory address -- but only the linear address.
Most of the required checking has been already done earlier in the
mce_severity rule checking engine. Following the Intel
recommendations Action Optional errors are only enabled for known
situations (encoded in MCACODs). The errors are ignored otherwise,
because they are action optional.
v2: Improve comment, disable preemption while processing ring buffer
(reported by Ying Huang)
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Rename the mce_notify_user function to mce_notify_irq. The next
patch will split the wakeup handling of interrupt context
and of process context and it's better to give it a clearer
name for this.
Contains a fix from Ying Huang
[ Impact: cleanup ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Huang Ying <ying.huang@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The MCE severity judgement code is data-driven, so code coverage tools
such as gcov can not be used for measuring coverage. Instead a dedicated
coverage mechanism is implemented. The kernel keeps track of rules
executed and reports them in debugfs.
This is useful for increasing coverage of the mce-test testsuite.
Right now it's unconditionally enabled because it's very little code.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The x86 architecture recently added some new machine check status bits:
S(ignalled) and AR (Action-Required). Signalled allows to check
if a specific event caused an exception or was just logged through CMCI.
AR allows the kernel to decide if an event needs immediate action
or can be delayed or ignored.
Implement support for these new status bits. mce_severity() uses
the new bits to grade the machine check correctly and decide what
to do. The exception handler uses AR to decide to kill or not.
The S bit is used to separate events between the poll/CMCI handler
and the exception handler.
Classical UC always leads to panic. That was true before anyways
because the existing CPUs always passed a PCC with it.
Also corrects the rules whether to kill in user or kernel context
and how to handle missing RIPV.
The machine check handler largely uses the mce-severity grading
engine now instead of making its own decisions. This means the logic
is centralized in one place. This is useful because it has to be
evaluated multiple times.
v2: Some rule fixes; Add AO events
Fix RIPV, RIPV|EIPV order (Ying Huang)
Fix UCNA with AR=1 message (Ying Huang)
Add comment about panicing in m_c_p.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
When multiple MCEs are printed print the "HARDWARE ERROR" header
and "This is not a software error" footer only once. This
makes the output much more compact with many CPUs.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Fatal machine checks can be logged to disk after boot, but only if
the system did a warm reboot. That's unfortunately difficult with the
default panic behaviour, which waits forever and the admin has to
press the power button because modern systems usually miss a reset button.
This clears the machine checks in the registers and make
it impossible to log them.
This patch changes the default for machine check panic to always
reboot after 30s. Then the mce can be successfully logged after
reboot.
I believe this will improve machine check experience for any
system running the X server.
This is dependent on successfull boot logging of MCEs. This currently
only works on Intel systems, on AMD there are quite a lot of systems
around which leave junk in the machine check registers after boot,
so it's disabled here. These systems will continue to default
to endless waiting panic.
v2: Only force panic timeout when it's shorter (H.Seto)
v3: Only force timeout when there is no timeout
(based on comment H.Seto)
[ Fix changelog - HS ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Assume IP on the stack is valid when either EIPV or RIPV are set.
This influences whether the machine check exception handler decides
to return or panic.
This fixes a test case in the mce-test suite and is more compliant
to the specification.
This currently only makes a difference in a artificial testing
scenario with the mce-test test suite.
Also in addition do not force the EIPV to be valid with the exact
register MSRs, and keep in trust the CS value on stack even if MSR
is available.
[AK: combination of patches from Huang Ying and Hidetoshi Seto, with
new description by me]
[add some description, no code changed - HS]
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
... instead of "Machine check". This is for consistency with the Monarch
panic message.
Based on a report from Ying Huang.
v2: But add a descriptive postfix so that the test suite can distingush.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
On Intel platforms machine check exceptions are always broadcast to
all CPUs. This patch makes the machine check handler synchronize all
these machine checks, elect a Monarch to handle the event and collect
the worst event from all CPUs and then process it first.
This has some advantages:
- When there is a truly data corrupting error the system panics as
quickly as possible. This improves containment of corrupted
data and makes sure the corrupted data never hits stable storage.
- The panics are synchronized and do not reenter the panic code
on multiple CPUs (which currently does not handle this well).
- All the errors are reported. Currently it often happens that
another CPU happens to do the panic first, but reports useless
information (empty machine check) because the real error
happened on another CPU which came in later.
This is a big advantage on Nehalem where the 8 threads per CPU
lead to often the wrong CPU winning the race and dumping
useless information on a machine check. The problem also occurs
in a less severe form on older CPUs.
- The system can detect when no CPUs detected a machine check
and shut down the system. This can happen when one CPU is so
badly hung that that it cannot process a machine check anymore
or when some external agent wants to stop the system by
asserting the machine check pin. This follows Intel hardware
recommendations.
- This matches the recommended error model by the CPU designers.
- The events can be output in true severity order
- When a panic happens on another CPU it makes sure to be actually
be able to process the stop IPI by enabling interrupts.
The code is extremly careful to handle timeouts while waiting
for other CPUs. It can't rely on the normal timing mechanisms
(jiffies, ktime_get) because of its asynchronous/lockless nature,
so it uses own timeouts using ndelay() and a "SPINUNIT"
The timeout is configurable. By default it waits for upto one
second for the other CPUs. This can be also disabled.
From some informal testing AMD systems do not see to broadcast
machine checks, so right now it's always disabled by default on
non Intel CPUs or also on very old Intel systems.
Includes fixes from Ying Huang
Fixed a "ecception" in a comment (H.Seto)
Moved global_nwo reset later based on suggestion from H.Seto
v2: Avoid duplicate messages
[ Impact: feature, fixes long standing problems. ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
In some circumstances multiple CPUs can enter mce_panic() in parallel.
This gives quite confused output because they will all dump the same
machine check buffer.
The other problem is that they would all panic in parallel, but not
process each other's shutdown IPIs because interrupts are disabled.
Detect this situation early on in mce_panic(). On the first CPU
entering will do the panic, the others will just wait to be killed.
For paranoia reasons in case the other CPU dies during the MCE I added
a 5 seconds timeout. If it expires each CPU will panic on its own again.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Machine checks support waking up the mcelog daemon quickly.
The original wake up code for this was pretty ugly, relying on
a idle notifier and a special process flag. The reason it did
it this way is that the machine check handler is not subject
to normal interrupt locking rules so it's not safe
to call wake_up(). Instead it set a process flag
and then either did the wakeup in the syscall return
or in the idle notifier.
This patch adds a new "bootstraping" method as replacement.
The idea is that the handler checks if it's in a state where
it is unsafe to call wake_up(). If it's safe it calls it directly.
When it's not safe -- that is it interrupted in a critical
section with interrupts disables -- it uses a new "self IPI" to trigger
an IPI to its own CPU. This can be done safely because IPI
triggers are atomic with some care. The IPI is raised
once the interrupts are reenabled and can then safely call
wake_up().
When APICs are disabled the event is just queued and will be picked up
eventually by the next polling timer. I think that's a reasonable
compromise, since it should only happen quite rarely.
Contains fixes from Ying Huang.
[ solve conflict on irqinit, make it work on 32bit (entry_arch.h) - HS ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
The exception handler should behave differently if the exception is
fatal versus one that can be returned from. In the first case it should
never clear any registers because these need to be preserved
for logging after the next boot. Otherwise it should clear them
on each CPU step by step so that other CPUs sharing the same bank don't
see duplicate events. Otherwise we risk reporting events multiple
times on any CPUs which have shared machine check banks, which
is a common problem on Intel Nehalem which has both SMT (two
CPU threads sharing banks) and shared machine check banks in the uncore.
Determine early in a special pass if any event requires a panic.
This uses the mce_severity() function added earlier.
This is needed for the next patch.
Also fixes a problem together with an earlier patch
that corrected events weren't logged on a fatal MCE.
[ Impact: Feature ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
