* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, delay: tsc based udelay should have rdtsc_barrier
x86, setup: correct include file in <asm/boot.h>
x86, setup: Fix typo "CONFIG_x86_64" in <asm/boot.h>
x86, mce: percpu mcheck_timer should be pinned
x86: Add sysctl to allow panic on IOCK NMI error
x86: Fix uv bau sending buffer initialization
x86, mce: Fix mce resume on 32bit
x86: Move init_gbpages() to setup_arch()
x86: ensure percpu lpage doesn't consume too much vmalloc space
x86: implement percpu_alloc kernel parameter
x86: fix pageattr handling for lpage percpu allocator and re-enable it
x86: reorganize cpa_process_alias()
x86: prepare setup_pcpu_lpage() for pageattr fix
x86: rename remap percpu first chunk allocator to lpage
x86: fix duplicate free in setup_pcpu_remap() failure path
percpu: fix too lazy vunmap cache flushing
x86: Set cpu_llc_id on AMD CPUs
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
timer stats: Optimize by adding quick check to avoid function calls
timers: Fix timer_migration interface which accepts any number as input
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
ftrace: Fix the output of profile
ring-buffer: Make it generally available
ftrace: Remove duplicate newline
tracing: Fix trace_buf_size boot option
ftrace: Fix t_hash_start()
ftrace: Don't manipulate @pos in t_start()
ftrace: Don't increment @pos in g_start()
tracing: Reset iterator in t_start()
trace_stat: Don't increment @pos in seq start()
tracing_bprintk: Don't increment @pos in t_start()
tracing/events: Don't increment @pos in s_start()
The first entry of the ftrace profile was always skipped when
reading trace_stat/functionX.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A443D59.4080307@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch introduces a new sysctl:
/proc/sys/kernel/panic_on_io_nmi
which defaults to 0 (off).
When enabled, the kernel panics when the kernel receives an NMI
caused by an IO error.
The IO error triggered NMI indicates a serious system
condition, which could result in IO data corruption. Rather
than contiuing, panicing and dumping might be a better choice,
so one can figure out what's causing the IO error.
This could be especially important to companies running IO
intensive applications where corruption must be avoided, e.g. a
bank's databases.
[ SuSE has been shipping it for a while, it was done at the
request of a large database vendor, for their users. ]
Signed-off-by: Kurt Garloff <garloff@suse.de>
Signed-off-by: Roberto Angelino <robertangelino@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
LKML-Reference: <20090624213211.GA11291@kroah.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Yanmin noticed that fault_in_user_writeable() requests 4 pages instead
of one.
That's the result of blindly trusting Linus' proposal :) I even looked
up the prototype to verify the correctness: the argument in question
is confusingly enough named "len" while in reality it means number of
pages.
Pointed-out-by: Yanmin Zhang <yanmin_zhang@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
In hunting down the cause for the hwlat_detector ring buffer spew in
my failed -next builds it became obvious that folks are now treating
ring_buffer as something that is generic independent of tracing and thus,
suitable for public driver consumption.
Given that there are only a few minor areas in ring_buffer that have any
reliance on CONFIG_TRACING or CONFIG_FUNCTION_TRACER, provide stubs for
those and make it generally available.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Cc: Jon Masters <jcm@jonmasters.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <20090625053012.GB19944@linux-sh.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
another race fix in jfs_check_acl()
Get "no acls for this inode" right, fix shmem breakage
inline functions left without protection of ifdef (acl)
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/audit-current:
audit: inode watches depend on CONFIG_AUDIT not CONFIG_AUDIT_SYSCALL
Even though one cannot make use of the audit watch code without
CONFIG_AUDIT_SYSCALL the spaghetti nature of the audit code means that
the audit rule filtering requires that it at least be compiled.
Thus build the audit_watch code when we build auditfilter like it was
before cfcad62c74
Clearly this is a point of potential future cleanup..
Reported-by: Frans Pop <elendil@planet.nl>
Signed-off-by: Eric Paris <eparis@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
commit 64d1304a64 (futex: setup writeable mapping for futex ops which
modify user space data) did address only half of the problem of write
access faults.
The patch was made on two wrong assumptions:
1) access_ok(VERIFY_WRITE,...) would actually check write access.
On x86 it does _NOT_. It's a pure address range check.
2) a RW mapped region can not go away under us.
That's wrong as well. Nobody can prevent another thread to call
mprotect(PROT_READ) on that region where the futex resides. If that
call hits between the get_user_pages_fast() verification and the
actual write access in the atomic region we are toast again.
The solution is to not rely on access_ok and get_user() for any write
access related fault on private and shared futexes. Instead we need to
fault it in with verification of write access.
There is no generic non destructive write mechanism which would fault
the user page in trough a #PF, but as we already know that we will
fault we can as well call get_user_pages() directly and avoid the #PF
overhead.
If get_user_pages() returns -EFAULT we know that we can not fix it
anymore and need to bail out to user space.
Remove a bunch of confusing comments on this issue as well.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@kernel.org
We should be able to specify [KMG] when setting trace_buf_size
boot option, as documented in kernel-parameters.txt
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41F2DB.4020102@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When the kernel is configured with CONFIG_TIMER_STATS but timer
stats are runtime disabled we still get calls to
__timer_stats_timer_set_start_info which initializes some
fields in the corresponding struct timer_list.
So add some quick checks in the the timer stats setup functions
to avoid function calls to __timer_stats_timer_set_start_info
when timer stats are disabled.
In an artificial workload that does nothing but playing ping
pong with a single tcp packet via loopback this decreases cpu
consumption by 1 - 1.5%.
This is part of a modified function trace output on SLES11:
perl-2497 [00] 28630647177732388 [+ 125]: sk_reset_timer <-tcp_v4_rcv
perl-2497 [00] 28630647177732513 [+ 125]: mod_timer <-sk_reset_timer
perl-2497 [00] 28630647177732638 [+ 125]: __timer_stats_timer_set_start_info <-mod_timer
perl-2497 [00] 28630647177732763 [+ 125]: __mod_timer <-mod_timer
perl-2497 [00] 28630647177732888 [+ 125]: __timer_stats_timer_set_start_info <-__mod_timer
perl-2497 [00] 28630647177733013 [+ 93]: lock_timer_base <-__mod_timer
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Mustafa Mesanovic <mustafa.mesanovic@de.ibm.com>
Cc: Arjan van de Ven <arjan@infradead.org>
LKML-Reference: <20090623153811.GA4641@osiris.boeblingen.de.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When the output of set_ftrace_filter is larger than PAGE_SIZE,
t_hash_start() will be called the 2nd time, and then we start
from the head of a hlist, which is wrong and causes some entries
to be outputed twice.
The worse is, if the hlist is large enough, reading set_ftrace_filter
won't stop but in a dead loop.
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41876E.2060407@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's rather confusing that in t_start(), in some cases @pos is
incremented, and in some cases it's decremented and then incremented.
This patch rewrites t_start() in a much more general way.
Thus we fix a bug that if ftrace_filtered == 1, functions have tracer
hooks won't be printed, because the branch is always unreachable:
static void *t_start(...)
{
...
if (!p)
return t_hash_start(m, pos);
return p;
}
Before:
# echo 'sys_open' > /mnt/tracing/set_ftrace_filter
# echo 'sys_write:traceon:4' >> /mnt/tracing/set_ftrace_filter
sys_open
After:
# echo 'sys_open' > /mnt/tracing/set_ftrace_filter
# echo 'sys_write:traceon:4' >> /mnt/tracing/set_ftrace_filter
sys_open
sys_write:traceon:count=4
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A41874B.4090507@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's wrong to increment @pos in g_start(). It causes some entries
lost when reading set_graph_function, if the output of the file
is larger than PAGE_SIZE.
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418738.7090401@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The iterator is m->private, but it's not reset to trace_types in
t_start(). If the output is larger than PAGE_SIZE and t_start()
is called the 2nd time, things will go wrong.
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418728.5020506@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's wrong to increment @pos in stat_seq_start(). It causes some
stat entries lost when reading stat file, if the output of the file
is larger than PAGE_SIZE.
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A418716.90209@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It's wrong to increment @pos in t_start(), otherwise we'll lose
some entries when reading printk_formats, if the output is larger
than PAGE_SIZE.
Reported-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4186FA.1020106@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
While testing syscall tracepoints posted by Jason, I found 3 entries
were missing when reading available_events. The output size of
available_events is < 4 pages, which means we lost 1 entry per page.
The cause is, it's wrong to increment @pos in s_start().
Actually there's another bug here -- reading avaiable_events/set_events
can race with module unload:
# cat available_events |
s_start() |
s_stop() |
| # rmmod foo.ko
s_start() |
call = list_entry(m->private) |
@call might be freed and accessing it will lead to crash.
Reviewed-by: Liming Wang <liming.wang@windriver.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <4A4186DD.6090405@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
If syscall removes the root of subtree being watched, we
definitely do not want the rules refering that subtree
to be destroyed without the syscall in question having
a chance to match them.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
A number of places in the audit system we send an op= followed by a string
that includes spaces. Somehow this works but it's just wrong. This patch
moves all of those that I could find to be quoted.
Example:
Change From: type=CONFIG_CHANGE msg=audit(1244666690.117:31): auid=0 ses=1
subj=unconfined_u:unconfined_r:auditctl_t:s0-s0:c0.c1023 op=remove rule
key="number2" list=4 res=0
Change To: type=CONFIG_CHANGE msg=audit(1244666690.117:31): auid=0 ses=1
subj=unconfined_u:unconfined_r:auditctl_t:s0-s0:c0.c1023 op="remove rule"
key="number2" list=4 res=0
Signed-off-by: Eric Paris <eparis@redhat.com>
audit_get_nd() is only used by audit_watch and could be more cleanly
implemented by having the audit watch functions call it when needed rather
than making the generic audit rule parsing code deal with those objects.
Signed-off-by: Eric Paris <eparis@redhat.com>
In preparation for converting audit to use fsnotify instead of inotify we
seperate the inode watching code into it's own file. This is similar to
how the audit tree watching code is already seperated into audit_tree.c
Signed-off-by: Eric Paris <eparis@redhat.com>
audit_receive_skb is hard to clearly parse what it is doing to the netlink
message. Clean the function up so it is easy and clear to see what is going
on.
Signed-off-by: Eric Paris <eparis@redhat.com>
The audit handling of netlink messages is all over the place. Clean things
up, use predetermined macros, generally make it more readable.
Signed-off-by: Eric Paris <eparis@redhat.com>
audit_update_watch() runs all of the rules for a given watch and duplicates
them, attaches a new watch to them, and then when it finishes that process
and has called free on all of the old rules (ok maybe still inside the rcu
grace period) it proceeds to use the last element from list_for_each_entry_safe()
as if it were a krule rather than being the audit_watch which was anchoring
the list to output a message about audit rules changing.
This patch unfies the audit message from two different places into a helper
function and calls it from the correct location in audit_update_rules(). We
will now get an audit message about the config changing for each rule (with
each rules filterkey) rather than the previous garbage.
Signed-off-by: Eric Paris <eparis@redhat.com>
The audit execve record splitting code estimates the length of the message
generated. But it forgot to include the "" that wrap each string in its
estimation. This means that execve messages with lots of tiny (1-2 byte)
arguments could still cause records greater than 8k to be emitted. Simply
fix the estimate.
Signed-off-by: Eric Paris <eparis@redhat.com>
When an audit watch is added to a parent the temporary watch inside the
original krule from userspace is freed. Yet the original watch is used after
the real watch was created in audit_add_rules()
Signed-off-by: Eric Paris <eparis@redhat.com>
SLAB uses get/put_online_cpus() which use a mutex which is itself only
initialized when cpu_hotplug_init() is called. Currently we hang suring
boot in SLAB due to doing that too late.
Reported by James Bottomley and Sachin Sant (and possibly others).
Debugged by Benjamin Herrenschmidt.
This just removes the dynamic initialization of the data structures, and
replaces it with a static one, avoiding this dependency entirely, and
removing one unnecessary special initcall.
Tested-by: Sachin Sant <sachinp@in.ibm.com>
Tested-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Tested-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
genirq, irq.h: Fix kernel-doc warnings
genirq: fix comment to say IRQ_WAKE_THREAD
* 'perfcounters-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (49 commits)
perfcounter: Handle some IO return values
perf_counter: Push perf_sample_data through the swcounter code
perf_counter tools: Define and use our own u64, s64 etc. definitions
perf_counter: Close race in perf_lock_task_context()
perf_counter, x86: Improve interactions with fast-gup
perf_counter: Simplify and fix task migration counting
perf_counter tools: Add a data file header
perf_counter: Update userspace callchain sampling uses
perf_counter: Make callchain samples extensible
perf report: Filter to parent set by default
perf_counter tools: Handle lost events
perf_counter: Add event overlow handling
fs: Provide empty .set_page_dirty() aop for anon inodes
perf_counter: tools: Makefile tweaks for 64-bit powerpc
perf_counter: powerpc: Add processor back-end for MPC7450 family
perf_counter: powerpc: Make powerpc perf_counter code safe for 32-bit kernels
perf_counter: powerpc: Change how processor-specific back-ends get selected
perf_counter: powerpc: Use unsigned long for register and constraint values
perf_counter: powerpc: Enable use of software counters on 32-bit powerpc
perf_counter tools: Add and use isprint()
...
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: Fix out of scope variable access in sched_slice()
sched: Hide runqueues from direct refer at source code level
sched: Remove unneeded __ref tag
sched, x86: Fix cpufreq + sched_clock() TSC scaling
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (24 commits)
tracing/urgent: warn in case of ftrace_start_up inbalance
tracing/urgent: fix unbalanced ftrace_start_up
function-graph: add stack frame test
function-graph: disable when both x86_32 and optimize for size are configured
ring-buffer: have benchmark test print to trace buffer
ring-buffer: do not grab locks in nmi
ring-buffer: add locks around rb_per_cpu_empty
ring-buffer: check for less than two in size allocation
ring-buffer: remove useless compile check for buffer_page size
ring-buffer: remove useless warn on check
ring-buffer: use BUF_PAGE_HDR_SIZE in calculating index
tracing: update sample event documentation
tracing/filters: fix race between filter setting and module unload
tracing/filters: free filter_string in destroy_preds()
ring-buffer: use commit counters for commit pointer accounting
ring-buffer: remove unused variable
ring-buffer: have benchmark test handle discarded events
ring-buffer: prevent adding write in discarded area
tracing/filters: strloc should be unsigned short
tracing/filters: operand can be negative
...
Fix up kmemcheck-induced conflict in kernel/trace/ring_buffer.c manually
Push the perf_sample_data further outwards to the swcounter interface,
to abstract it away some more.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Prevent from further ftrace_start_up inbalances so that we avoid
future nop patching omissions with dynamic ftrace.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Perfcounter reports the following stats for a wide system
profiling:
#
# (2364 samples)
#
# Overhead Symbol
# ........ ......
#
15.40% [k] mwait_idle_with_hints
8.29% [k] read_hpet
5.75% [k] ftrace_caller
3.60% [k] ftrace_call
[...]
This snapshot has been taken while neither the function tracer nor
the function graph tracer was running.
With dynamic ftrace, such results show a wrong ftrace behaviour
because all calls to ftrace_caller or ftrace_graph_caller (the patched
calls to mcount) are supposed to be patched into nop if none of those
tracers are running.
The problem occurs after the first run of the function tracer. Once we
launch it a second time, the callsites will never be nopped back,
unless you set custom filters.
For example it happens during the self tests at boot time.
The function tracer selftest runs, and then the dynamic tracing is
tested too. After that, the callsites are left un-nopped.
This is because the reset callback of the function tracer tries to
unregister two ftrace callbacks in once: the common function tracer
and the function tracer with stack backtrace, regardless of which
one is currently in use.
It then creates an unbalance on ftrace_start_up value which is expected
to be zero when the last ftrace callback is unregistered. When it
reaches zero, the FTRACE_DISABLE_CALLS is set on the next ftrace
command, triggering the patching into nop. But since it becomes
unbalanced, ie becomes lower than zero, if the kernel functions
are patched again (as in every further function tracer runs), they
won't ever be nopped back.
Note that ftrace_call and ftrace_graph_call are still patched back
to ftrace_stub in the off case, but not the callers of ftrace_call
and ftrace_graph_caller. It means that the tracing is well deactivated
but we waste a useless call into every kernel function.
This patch just unregisters the right ftrace_ops for the function
tracer on its reset callback and ignores the other one which is
not registered, fixing the unbalance. The problem also happens
is .30
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: stable@kernel.org
The bug is ancient.
If we trace the sub-thread of our natural child and this sub-thread exits,
we update parent->signal->cxxx fields. But we should not do this until
the whole thread-group exits, otherwise we account this thread (and all
other live threads) twice.
Add the task_detached() check. No need to check thread_group_empty(),
wait_consider_task()->delay_group_leader() already did this.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Roland McGrath <roland@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Vitaly Mayatskikh <vmayatsk@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
perf_lock_task_context() is buggy because it can return a dead
context.
the RCU read lock in perf_lock_task_context() only guarantees
the memory won't get freed, it doesn't guarantee the object is
valid (in our case refcount > 0).
Therefore we can return a locked object that can get freed the
moment we release the rcu read lock.
perf_pin_task_context() then increases the refcount and does an
unlock on freed memory.
That increased refcount will cause a double free, in case it
started out with 0.
Ammend this by including the get_ctx() functionality in
perf_lock_task_context() (all users already did this later
anyway), and return a NULL context when the found one is
already dead.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The task migrations counter was causing rare and hard to decypher
memory corruptions under load. After a day of debugging and bisection
we found that the problem was introduced with:
3f731ca: perf_counter: Fix cpu migration counter
Turning them off fixes the crashes. Incidentally, the whole
perf_counter_task_migration() logic can be done simpler as well,
by injecting a proper sw-counter event.
This cleanup also fixed the crashes. The precise failure mode is
not completely clear yet, but we are clearly not unhappy about
having a fix ;-)
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
In case gcc does something funny with the stack frames, or the return
from function code, we would like to detect that.
An arch may implement passing of a variable that is unique to the
function and can be saved on entering a function and can be tested
when exiting the function. Usually the frame pointer can be used for
this purpose.
This patch also implements this for x86. Where it passes in the stack
frame of the parent function, and will test that frame on exit.
There was a case in x86_32 with optimize for size (-Os) where, for a
few functions, gcc would align the stack frame and place a copy of the
return address into it. The function graph tracer modified the copy and
not the actual return address. On return from the funtion, it did not go
to the tracer hook, but returned to the parent. This broke the function
graph tracer, because the return of the parent (where gcc did not do
this funky manipulation) returned to the location that the child function
was suppose to. This caused strange kernel crashes.
This test detected the problem and pointed out where the issue was.
This modifies the parameters of one of the functions that the arch
specific code calls, so it includes changes to arch code to accommodate
the new prototype.
Note, I notice that the parsic arch implements its own push_return_trace.
This is now a generic function and the ftrace_push_return_trace should be
used instead. This patch does not touch that code.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
On x86_32, when optimize for size is set, gcc may align the frame pointer
and make a copy of the the return address inside the stack frame.
The return address that is located in the stack frame may not be
the one used to return to the calling function. This will break the
function graph tracer.
The function graph tracer replaces the return address with a jump to a hook
function that can trace the exit of the function. If it only replaces
a copy, then the hook will not be called when the function returns.
Worse yet, when the parent function returns, the function graph tracer
will return back to the location of the child function which will
easily crash the kernel with weird results.
To see the problem, when i386 is compiled with -Os we get:
c106be03: 57 push %edi
c106be04: 8d 7c 24 08 lea 0x8(%esp),%edi
c106be08: 83 e4 e0 and $0xffffffe0,%esp
c106be0b: ff 77 fc pushl 0xfffffffc(%edi)
c106be0e: 55 push %ebp
c106be0f: 89 e5 mov %esp,%ebp
c106be11: 57 push %edi
c106be12: 56 push %esi
c106be13: 53 push %ebx
c106be14: 81 ec 8c 00 00 00 sub $0x8c,%esp
c106be1a: e8 f5 57 fb ff call c1021614 <mcount>
When it is compiled with -O2 instead we get:
c10896f0: 55 push %ebp
c10896f1: 89 e5 mov %esp,%ebp
c10896f3: 83 ec 28 sub $0x28,%esp
c10896f6: 89 5d f4 mov %ebx,0xfffffff4(%ebp)
c10896f9: 89 75 f8 mov %esi,0xfffffff8(%ebp)
c10896fc: 89 7d fc mov %edi,0xfffffffc(%ebp)
c10896ff: e8 d0 08 fa ff call c1029fd4 <mcount>
The compile with -Os will align the stack pointer then set up the
frame pointer (%ebp), and it copies the return address back into
the stack frame. The change to the return address in mcount is done
to the copy and not the real place holder of the return address.
Then compile with -O2 sets up the frame pointer first, this makes
the change to the return address by mcount affect where the function
will jump on exit.
Reported-by: Jake Edge <jake@lwn.net>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Enable gcov profiling of the entire kernel on x86_64. Required changes
include disabling profiling for:
* arch/kernel/acpi/realmode and arch/kernel/boot/compressed:
not linked to main kernel
* arch/vdso, arch/kernel/vsyscall_64 and arch/kernel/hpet:
profiling causes segfaults during boot (incompatible context)
Signed-off-by: Peter Oberparleiter <oberpar@linux.vnet.ibm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Li Wei <W.Li@Sun.COM>
Cc: Michael Ellerman <michaele@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Heiko Carstens <heicars2@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <mschwid2@linux.vnet.ibm.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: WANG Cong <xiyou.wangcong@gmail.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Enable the use of GCC's coverage testing tool gcov [1] with the Linux
kernel. gcov may be useful for:
* debugging (has this code been reached at all?)
* test improvement (how do I change my test to cover these lines?)
* minimizing kernel configurations (do I need this option if the
associated code is never run?)
The profiling patch incorporates the following changes:
* change kbuild to include profiling flags
* provide functions needed by profiling code
* present profiling data as files in debugfs
Note that on some architectures, enabling gcc's profiling option
"-fprofile-arcs" for the entire kernel may trigger compile/link/
run-time problems, some of which are caused by toolchain bugs and
others which require adjustment of architecture code.
For this reason profiling the entire kernel is initially restricted
to those architectures for which it is known to work without changes.
This restriction can be lifted once an architecture has been tested
and found compatible with gcc's profiling. Profiling of single files
or directories is still available on all platforms (see config help
text).
[1] http://gcc.gnu.org/onlinedocs/gcc/Gcov.html
Signed-off-by: Peter Oberparleiter <oberpar@linux.vnet.ibm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Li Wei <W.Li@Sun.COM>
Cc: Michael Ellerman <michaele@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Heiko Carstens <heicars2@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <mschwid2@linux.vnet.ibm.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: WANG Cong <xiyou.wangcong@gmail.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Call constructors (gcc-generated initcall-like functions) during kernel
start and module load. Constructors are e.g. used for gcov data
initialization.
Disable constructor support for usermode Linux to prevent conflicts with
host glibc.
Signed-off-by: Peter Oberparleiter <oberpar@linux.vnet.ibm.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: WANG Cong <xiyou.wangcong@gmail.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Li Wei <W.Li@Sun.COM>
Cc: Michael Ellerman <michaele@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Heiko Carstens <heicars2@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <mschwid2@linux.vnet.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Linus Torvalds
Li Zefan
Kurt Garloff
Thomas Gleixner
Paul Mundt
Eric Paris
Heiko Carstens
Al Viro
Arun R Bharadwaj
Ingo Molnar
Peter Zijlstra
Frederic Weisbecker
Oleg Nesterov
Steven Rostedt
Peter Oberparleiter