Andrew Morton wrote:
On Mon, 26 Jul 2010 10:49:50 +0200
Michal Simek <monstr@xxxxxxxxx> wrote:
start/stop_critical_timing function for preemptirqsoff, preemptoff
and irqsoff tracers contains atomic_inc and atomic_dec operations.
Atomic operations used local_irq_save/restore macros to ensure
atomic access but they are traced by the same function which is causing
recursion problem.
The reason is when these tracers are turn ON then local_irq_save/restore
macros are changed in include/linux/irqflags.h to call trace_hardirqs_on/off
which call start/stop_critical_timing.
Microblaze was affected because use generic atomic implementation.
Seems that this will also affect blackfin, mn10300 and score. I guess
they aren't supporting tracing yet?
If they uses asm-generic/atomic.h, then yes. It seems to me that
Blackfin uses own asm/atomic.h. nm10300 and score are the same case as
Microblaze.
In include/linux/irqflags.h is written this commentary.
/*
* The local_irq_*() APIs are equal to the raw_local_irq*()
* if !TRACE_IRQFLAGS.
*/
If architecture doesn't enable TRACE_IRQFLAGS then there is no
difference in behavior. If yes and use asm-generic/atomic.h code, then
IRQFLAG tracer freeze because it is traced part of irqsoff tracer
because of recursion as is describe in patch description.
Signed-off-by: Michal Simek <monstr@xxxxxxxxx>
Acked-by: Steven Rostedt <rostedt@xxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
hm, I wonder how my signoff got there. Doesn't matter.
http://lkml.org/lkml/2010/5/26/364
Michal
--
Michal Simek, Ing. (M.Eng)
w: www.monstr.eu p: +42-0-721842854
Maintainer of Linux kernel 2.6 Microblaze Linux - http://www.monstr.eu/fdt/
Microblaze U-BOOT custodian
--
To unsubscribe from this list: send the line "unsubscribe linux-arch" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html