Prev: [COUNTERPATCH] mm: avoid overflowing preempt_count() in mmu_take_all_locks()
Next: mxser: spin_lock() => spin_lock_irq()
From: Jiri Olsa on 1 Apr 2010 12:50
hi,

ok, saga continues ;)

I'm sending reworked version of the graph output support for
preemptirqsoff/preemptoff/irqsoff tracers.

v5:
- function graph events are back in game
- changed function graph tracer interface to be able to
pass custom tracer flags
- changed calls to trace_function to trace graph records
when the display-graph otion is enabled. As function tracer
and graph tracer have different spacing during the record
output, this way it looks better, and the info stays.

This could be also done via TRACE_PRINT, then it'd be displayed
as comment.. I chose the first way, though it could be easily
changed.

v4:
- fixed header to display all the info as for
current irqsoff tracer
- fixed reset of the max_tr buffer during the
display-graph switch

v3:
- the function_graph tracer should have the same output
as current version even with events enabled,
unless I missed something again ;)
- I gave up the idea of function_graph events,
since it was unsuitable for outputing other
events as comments "/* */"
- I'm not including the raw function_graph for now


attached patches:
- 1/4 adding ftrace events for graph tracer
- 2/4 graph tracer - adding support for passing flags to ouput functions
- 3/4 graph output support for irqsoff tracer
- 4/4 graph output support for preemptirqsoff/preemptoff tracers


plz let me know what you think

wbr,
jirka
---
kernel/trace/trace.c | 35 +++--
kernel/trace/trace.h | 27 +++-
kernel/trace/trace_functions_graph.c | 164 +++++++++++++--------
kernel/trace/trace_irqsoff.c | 267 ++++++++++++++++++++++++++++++++--
4 files changed, 408 insertions(+), 85 deletions(-)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: [COUNTERPATCH] mm: avoid overflowing preempt_count() in mmu_take_all_locks()
Next: mxser: spin_lock() => spin_lock_irq()