linux/kernel/trace
Steven Rostedt eb4a03780d function-graph: disable when both x86_32 and optimize for size are configured
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>
2009-06-18 18:39:30 -04:00
..
blktrace.c Merge branch 'for-2.6.31' of git://git.kernel.dk/linux-2.6-block 2009-06-11 11:10:35 -07:00
ftrace.c Merge branch 'tracing-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2009-06-10 19:58:10 -07:00
Kconfig function-graph: disable when both x86_32 and optimize for size are configured 2009-06-18 18:39:30 -04:00
kmemtrace.c cpumask: use new operators in kernel/trace 2009-06-15 11:36:42 -04:00
Makefile tracing/events: convert block trace points to TRACE_EVENT() 2009-06-09 12:34:23 -04:00
ring_buffer_benchmark.c ring-buffer: have benchmark test print to trace buffer 2009-06-17 17:01:09 -04:00
ring_buffer.c ring-buffer: do not grab locks in nmi 2009-06-17 14:16:27 -04:00
trace_boot.c tracing: use macros to denote usec and nsec per second 2009-04-07 14:43:06 +02:00
trace_branch.c Merge branch 'linus' into tracing/core 2009-05-07 11:17:34 +02:00
trace_clock.c tracing: fix four sparse warnings 2009-03-22 18:16:54 +01:00
trace_event_profile.c tracing/events: fix concurrent access to ftrace_events list 2009-05-06 10:38:19 +02:00
trace_event_types.h tracing/filters: distinguish between signed and unsigned fields 2009-04-29 14:06:03 +02:00
trace_events_filter.c tracing/filters: fix race between filter setting and module unload 2009-06-16 16:25:37 -04:00
trace_events.c tracing: add protection around module events unload 2009-06-09 17:29:07 -04:00
trace_export.c tracing/filters: distinguish between signed and unsigned fields 2009-04-29 14:06:03 +02:00
trace_functions_graph.c Merge branch 'tracing-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2009-06-10 19:58:10 -07:00
trace_functions.c tracing/core: use appropriate waiting on trace_pipe 2009-02-18 01:40:20 +01:00
trace_hw_branches.c Merge branch 'tracing/hw-branch-tracing' into tracing/core 2009-05-07 13:36:22 +02:00
trace_irqsoff.c tracing: have latency tracers set the latency format 2009-03-04 22:15:30 -05:00
trace_mmiotrace.c tracing: use macros to denote usec and nsec per second 2009-04-07 14:43:06 +02:00
trace_nop.c tracing/ftrace: make nop-tracer use polling wait for events on pipe 2009-03-23 09:22:15 +01:00
trace_output.c tracing: add protection around module events unload 2009-06-09 17:29:07 -04:00
trace_output.h tracing: add protection around module events unload 2009-06-09 17:29:07 -04:00
trace_power.c Merge branch 'linus' into tracing/core 2009-05-07 11:17:34 +02:00
trace_printk.c tracing/ftrace: factorize the tracing files creation 2009-04-07 14:43:07 +02:00
trace_sched_switch.c tracing/events: move trace point headers into include/trace/events 2009-04-14 22:05:43 -04:00
trace_sched_wakeup.c tracing/wakeup: move access to wakeup_cpu into spinlock 2009-04-23 23:01:36 -04:00
trace_selftest_dynamic.c ftrace: fix dynamic ftrace selftest 2008-05-23 21:13:23 +02:00
trace_selftest.c x86, hw-branch-tracer: allocate selftest iterator on heap 2009-04-07 13:36:21 +02:00
trace_stack.c tracing/trace_stack: fix the number of entries in the header 2009-06-03 10:24:44 -04:00
trace_stat.c tracing/stat: remove unappropriate safe walk on list 2009-06-02 01:18:33 +02:00
trace_stat.h tracing: add handler to trace_stat 2009-03-24 23:22:58 -04:00
trace_syscalls.c tracing/syscalls: use a dedicated file header 2009-04-09 05:43:32 +02:00
trace_sysprof.c tracing/ftrace: factorize the tracing files creation 2009-04-07 14:43:07 +02:00
trace_workqueue.c trace_workqueue: remove blank line between each cpu 2009-06-02 01:14:26 +02:00
trace.c tracing: replace a GFP_ATOMIC with GFP_KERNEL allocation 2009-06-15 11:37:14 -04:00
trace.h Merge branch 'tracing/hw-branch-tracing' into tracing/core 2009-05-07 13:36:22 +02:00