Use knowledge about EFLAGS layout (bits 22:63 are always 0) to distingush
EFLAGS word and kernel address in the spin lock stack frame.
Signed-off-by: Andi Kleen <ak@suse.de>
This ports the algorithm from x86-64 (with improvements) to i386.
Previously this only worked for frame pointer enabled kernels.
But spinlocks have a very simple stack frame that can be manually
analyzed. Do this.
Signed-off-by: Andi Kleen <ak@suse.de>
A few trivial spelling and grammar mistakes picked up in
"arch/x86_64/aperture.c", "arch/x86_64/crash.c" and
"arch/x86_64/apic.c". I think all are correct fixes but am ever aware
of my fallibility :o) This is my first patch submission so all
feedback is appreciated, esp. WRT CCing to Linus, Andi and
trivial@kernel.org, is this correct? And which is the most appropriate
kernel version to diff against? If any.
Should apply cleanly to 2.6.18-rc1
Signed-off-by: Adam Henley <adamazing@gmail.com>
Signed-off-by: Andi Kleen <ak@suse.de>
- adam
Hello,
Following my discussion with Andi. Here is a patch that introduces
two new TIF flags to simplify the context switch code in __switch_to().
The idea is to minimize the number of cache lines accessed in the common
case, i.e., when neither the debug registers nor the I/O bitmap are used.
This patch covers the x86-64 modifications. A patch for i386 follows.
Changelog:
- add TIF_DEBUG to track when debug registers are active
- add TIF_IO_BITMAP to track when I/O bitmap is used
- modify __switch_to() to use the new TIF flags
<signed-off-by>: eranian@hpl.hp.com
Signed-off-by: Andi Kleen <ak@suse.de>
For NUMA optimization and some other algorithms it is useful to have a fast
to get the current CPU and node numbers in user space.
x86-64 added a fast way to do this in a vsyscall. This adds a generic
syscall for other architectures to make it a generic portable facility.
I expect some of them will also implement it as a faster vsyscall.
The cache is an optimization for the x86-64 vsyscall optimization. Since
what the syscall returns is an approximation anyways and user space
often wants very fast results it can be cached for some time. The norma
methods to get this information in user space are relatively slow
The vsyscall is in a better position to manage the cache because it has direct
access to a fast time stamp (jiffies). For the generic syscall optimization
it doesn't help much, but enforce a valid argument to keep programs
portable
I only added an i386 syscall entry for now. Other architectures can follow
as needed.
AK: Also added some cleanups from Andrew Morton
Signed-off-by: Andi Kleen <ak@suse.de>
This patch adds a vgetcpu vsyscall, which depending on the CPU RDTSCP
capability uses either the RDTSCP or CPUID to obtain a CPU and node
numbers and pass them to the program.
AK: Lots of changes over Vojtech's original code:
Better prototype for vgetcpu()
It's better to pass the cpu / node numbers as separate arguments
to avoid mistakes when going from SMP to NUMA.
Also add a fast time stamp based cache using a user supplied
argument to speed things more up.
Use fast method from Chuck Ebbert to retrieve node/cpu from
GDT limit instead of CPUID
Made sure RDTSCP init is always executed after node is known.
Drop printk
Signed-off-by: Vojtech Pavlik <vojtech@suse.cz>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch adds initalization of the RDTSCP auxilliary values to CPU numbers
to time.c. If RDTSCP is available, the MSRs are written with the respective
values. It can be later used to initalize per-cpu timekeeping variables.
AK: Some cleanups. Move externs into headers and fix CPU hotplug.
Signed-off-by: Vojtech Pavlik <vojtech@suse.cz>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch adds macros for reading tsc via the RDTSCP instruction, as well
as writing the auxilliary MSR read by RDTSCP to msr.h
[AK: changed rdtscp definition for old binutils]
Signed-off-by: Vojtech Pavlik <vojtech@suse.cz>
Signed-off-by: Andi Kleen <ak@suse.de>
AK: This redoes the changes I temporarily reverted.
Intel now has support for Architectural Performance Monitoring Counters
( Refer to IA-32 Intel Architecture Software Developer's Manual
http://www.intel.com/design/pentium4/manuals/253669.htm ). This
feature is present starting from Intel Core Duo and Intel Core Solo processors.
What this means is, the performance monitoring counters and some performance
monitoring events are now defined in an architectural way (using cpuid).
And there will be no need to check for family/model etc for these architectural
events.
Below is the patch to use this performance counters in nmi watchdog driver.
Patch handles both i386 and x86-64 kernels.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Andi Kleen <ak@suse.de>
I've had good experiences with having this on by default on x86-64.
It turns nasty hangs into easier to debug oopses.
Enable the local APIC wdog by default for systems newer than 2004.
This comes from a strange compromise: according to arjan the reason
it was off by default was some old IBM systems that corrupted
registered when NMI happened in SMI. Can't remember more specific,
but >= 2004 should avoid these. It's probably overly broad
because most older systems should be ok (and the really old systems
won't be supported by the local apic watchdog anyways)
Signed-off-by: Andi Kleen <ak@suse.de>
After a crash we should wait for NMI IPI event and not for external NMI or
NMI watchdog tick.
Signed-off-by: Vivek Goyal <vgoyal@in.ibm.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
After a crash we should wait for NMI IPI event and not for external NMI or
NMI watchdog tick.
Signed-off-by: Vivek Goyal <vgoyal@in.ibm.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
This patch makes the following needlessly global functions static:
- nmi_int.c: profile_exceptions_notify()
- nmi_timer_int.c: profile_timer_exceptions_notify()
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andi Kleen <ak@suse.de>
When a unknown NMI happened the panic would claim a NMI watchdog timeout.
Also it would check the variable set by nmi_watchdog=panic and panic then.
Fix up the panic message to be generic
Unconditionally panic on unknown NMI when panic on unknown nmi is enabled.
Noticed by Jan Beulich
Cc: jbeulich@novell.com
Signed-off-by: Andi Kleen <ak@suse.de>
Making NMI suspend/resume work with SMP. We use CPU hotplug to offline
APs in SMP suspend/resume. Only BSP executes sysdev's .suspend/.resume
method. APs should follow CPU hotplug code path.
And:
+From: Don Zickus <dzickus@redhat.com>
Makes the start/stop paths of nmi watchdog more robust to handle the
suspend/resume cases more gracefully.
AK: I merged the two patches together
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Clean up some of the output messages on the nmi error paths to make more
sense when they are displayed. This is mainly a cosmetic fix and
shouldn't impact any normal code path.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
To quote Alan Cox:
The default Linux behaviour on an NMI of either memory or unknown is to
continue operation. For many environments such as scientific computing
it is preferable that the box is taken out and the error dealt with than
an uncorrected parity/ECC error get propogated.
A small number of systems do generate NMI's for bizarre random reasons
such as power management so the default is unchanged. In other respects
the new proc/sys entry works like the existing panic controls already in
that directory.
This is separate to the edac support - EDAC allows supported chipsets to
handle ECC errors well, this change allows unsupported cases to at least
panic rather than cause problems further down the line.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Adds a new /proc/sys/kernel/nmi_watchdog call that will enable/disable the
nmi watchdog.
By entering a non-zero value here, a user can enable the nmi watchdog to
monitor the online cpus in the system. By entering a zero value here, a
user can disable the nmi watchdog and free up a performance counter which
could then be utilized by the oprofile subsystem, otherwise oprofile may be
short a counter when in use.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Adds a new /proc/sys/kernel/nmi call that will enable/disable the nmi
watchdog.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Removes the un/set_nmi_callback and reserve/release_lapic_nmi functions as
they are no longer needed. The various subsystems are modified to register
with the die_notifier instead.
Also includes compile fixes by Andrew Morton.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
We need TIF_RESTORE_SIGMASK in order to support ppoll() and pselect()
system calls. This patch originally came from Andi, and was based
heavily on David Howells' implementation of same on i386. I fixed a typo
which was causing do_signal() to use the wrong signal mask.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch cleans up the NMI interrupt path. Instead of being gated by if
the 'nmi callback' is set, the interrupt handler now calls everyone who is
registered on the die_chain and additionally checks the nmi watchdog,
reseting it if enabled. This allows more subsystems to hook into the NMI if
they need to (without being block by set_nmi_callback).
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch includes the changes to make the nmi watchdog on i386 SMP aware.
A bunch of code was moved around to make it simpler to read. In addition,
it is now possible to determine if a particular NMI was the result of the
watchdog or not. This feature allows the kernel to filter out unknown NMIs
easier.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
This patch includes the changes to make the nmi watchdog on x86_64 SMP
aware. A bunch of code was moved around to make it simpler to read. In
addition, it is now possible to determine if a particular NMI was the result
of the watchdog or not. This feature allows the kernel to filter out
unknown NMIs easier.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Incorporates the new performance counter reservation system in oprofile.
Also cleans up a lot of the initialization code. The code original zero'd
out every register associated with performance counters regardless if those
registers were used or not. This causes issues with the nmi watchdog.
Now oprofile tries to reserve registers and gives up if it can't get them.
Cc: levon@movementarian.org
Cc: oprofile-list@lists.sf.net
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Adds basic infrastructure to allow subsystems to reserve performance
counters on the x86 chips. Only UP kernels are supported in this patch to
make reviewing easier. The SMP portion makes a lot more changes.
Think of this as a locking mechanism where each bit represents a different
counter. In addition, each subsystem should also reserve an appropriate
event selection register that will correspond to the performance counter it
will be using (this is mainly neccessary for the Pentium 4 chips as they
break the 1:1 relationship to performance counters).
This will help prevent subsystems like oprofile from interfering with the
nmi watchdog.
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andi Kleen <ak@suse.de>
There are some machines around (large xSeries or Unisys ES7000) that
need physical IO-APIC destination mode to access all of their IO
devices. This currently doesn't work in UP kernels as used in
distribution installers.
This patch allows to compile even UP kernels as GENERICARCH which
allows to use physical or clustered APIC mode.
Signed-off-by: Andi Kleen <ak@suse.de>
The klist utility routines currently call _put methods while holding a
spinlock. This is of course illegal; a put routine could try to
unregister a device and hence need to sleep.
No problems have arisen until now because in many cases klist removals
were done synchronously, so the _put methods were never actually used.
In other cases we may simply have been lucky.
This patch (as784) reworks the klist routines so that _put methods are
called only _after_ the klist's spinlock has been released.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch (as783) simplifies the driver core slightly by removing four
unnecessary _get and _put methods.
It is vital that when a driver is removed from its bus's klist of
registered drivers, or when a device is removed from a driver's klist
of bound devices, that the klist updates complete synchronously.
Otherwise the kernel might try binding an unregistered driver to a
newly-registered device, or adding a device to the klist for a new
driver before it has been removed from the old driver's klist.
Since the removals must be synchronous, they don't need to update any
reference counts. Hence the _get and _put methods can be dispensed
with.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
There is a potential deadlock in the driver core. It boils down to
the fact that bus_remove_device() calls klist_remove() instead of
klist_del(), thereby waiting until the reference count of the
klist_node in the bus's klist of devices drops to 0. The refcount
can't reach 0 so long as a modprobe process is trying to bind a new
driver to the device being removed, by calling __driver_attach(). The
problem is that __driver_attach() tries to acquire the device's
parent's semaphore, but the caller of bus_remove_device() is quite
likely to own that semaphore already.
It isn't sufficient just to replace klist_remove() with klist_del().
Doing so runs the risk that the device would remain on the bus's klist
of devices for some time, and so could be bound to another driver even
after it was unregistered. What's needed is a new way to distinguish
whether or not a device is registered, based on a criterion other than
whether its klist_node is linked into the bus's klist of devices. That
way driver binding can fail when the device is unregistered, even if
it is still linked into the klist.
This patch (as782) implements the solution, by adding a new bitflag to
indiate when a struct device is registered, by testing the flag before
allowing a driver to bind a device, and by changing the definition of
the device_is_registered() inline.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This provides a build and run-time option to turn on multhreaded probe
for all PCI drivers. It can cause bad problems on multi-processor
machines that take a while to find their root disks, and play havoc on
machines that don't use persistant device names for block or network
devices.
But it can cause speedups on some machines, my tiny laptop's boot goes
up by 0.4 seconds, and my desktop boots up several seconds faster.
Use at your own risk!!!
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This adds the infrastructure for drivers to do a threaded probe, and
waits at init time for all currently outstanding probes to complete.
A new kernel thread will be created when the probe() function for the
driver is called, if the multithread_probe bit is set in the driver
saying it can support this kind of operation.
I have tested this with USB and PCI, and it works, and shaves off a lot
of time in the boot process, but there are issues with finding root boot
disks, and some USB drivers assume that this can never happen, so it is
currently not enabled for any bus type. Individual drivers can enable
this right now if they wish, and bus authors can selectivly turn it on
as well, once they determine that their subsystem will work properly
with it.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Don't be crufty. Mark it __must_check too.
Cc: "Randy.Dunlap" <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The platform_notify call for Arm and PPC architectures needs to be called
before the driver attaches to the device. The problem only presents itself
when hotplugging certain devices while the driver is already loaded.
Signed-off-by: Brian Walsh <brian@walsh.ws>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
We get hundreds of these:
include/media/v4l2-dev.h:348: warning: ignoring return value of 'class_device_create_file', declared with attribute warn_unused_result
Handle it, and propagate the __must_check back a level.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Those 1500 warnings can be a bit of a pain. Add a config option to shut them
up.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
We're getting a lot of crashes in the sysfs/kobject/device/bus/class code and
they're very hard to diagnose.
I'm suspecting that in some cases this is because drivers aren't checking
return values and aren't handling errors correctly. So the code blithely
blunders on and crashes later in very obscure ways.
There's just no reason to ignore errors which can and do occur. So the patch
sprinkles __must_check all over these APIs.
Causes 1,513 new warnings. Heh.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
When CONFIG_HOTPLUG is n, add_bind_files() definition is wrong.
This patch has fixed it.
Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Make sysfs_remove_bin_file() void. If it detects an error,
printk the file name and call dump_stack().
sysfs_hash_and_remove() now returns an error code indicating
its success or failure so that sysfs_remove_bin_file() can
know success/failure.
Convert the only driver that checked the return value of
sysfs_remove_bin_file().
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Makes it easier for devices to create and remove binary attribute files
so they don't have to call directly into sysfs. This is needed to help
with the conversion from struct class_device to struct device.
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>