stable: please queue commit 0231bb5 ("perf: Fix event group context move")

DISCUSSION Applied Li Zefan 4 participants 4 messages Aug 5, 2013 – Aug 31, 2013 Stable branch

Share & tools

Li Zefan

author

2013-08-05 08:18 UTC #0 lore raw

Plese consider adding commit 0231bb5336758426b44ccd798ccd3c5419c95d58 to stable tree. It fixes a real bug, which was explained here:

	http://thread.gmane.org/gmane.linux.kernel.perf.user/1144

and I've managed to reproduce the bug in 3.4 kernel with the test program attached in the above email, and I've confirmed the bug fix.

The effect of the bug is, if we open a group of mixed events(sw/hw) and the "disabled" flag is not set and sw events come first, then some of the perf events won't be started after the call to perf_event_open().

The bug fix should be applicable to kernels >= 3.0 && <= 3.7.

--
Li Zefan

Ingo Molnar

Acked-by

2013-08-05 08:21 UTC #1 lore raw

Quoted 5 lines · Li Zefan

Plese consider adding commit 0231bb5336758426b44ccd798ccd3c5419c95d58 to stable tree. It fixes a real bug, which was explained here:
	http://thread.gmane.org/gmane.linux.kernel.perf.user/1144

Quoted 8 lines

and I've managed to reproduce the bug in 3.4 kernel with the test program attached in the above email, and I've confirmed the bug fix.
The effect of the bug is, if we open a group of mixed events(sw/hw) and the "disabled" flag is not set and sw events come first, then some of the perf events won't be started after the call to perf_event_open().
The bug fix should be applicable to kernels >= 3.0 && <= 3.7.

Acked-by Ingo Molnar <[email protected]>

Thanks,

Ingo

Ben Hutchings

Applied

2013-08-31 20:47 UTC #2 lore raw

Quoted 10 lines · Ingo Molnar

Quoted 8 lines · Li Zefan
Plese consider adding commit 0231bb5336758426b44ccd798ccd3c5419c95d58 to stable tree. It fixes a real bug, which was explained here:
	http://thread.gmane.org/gmane.linux.kernel.perf.user/1144
and I've managed to reproduce the bug in 3.4 kernel with the test program attached in the above email, and I've confirmed the bug fix.

Quoted 7 lines

Quoted 5 lines
The effect of the bug is, if we open a group of mixed events(sw/hw) and the "disabled" flag is not set and sw events come first, then some of the perf events won't be started after the call to perf_event_open().
The bug fix should be applicable to kernels >= 3.0 && <= 3.7.
Acked-by Ingo Molnar <[email protected]>

Queued up for 3.2, thanks.

Ben.

Signature · 3 lines

Ben Hutchings
If God had intended Man to program,
we'd have been born with serial I/O ports.

Luis Henriques

2013-08-05 10:57 UTC #3 lore raw

In reply to Li Zefan

Quoted 12 lines · Li Zefan

Plese consider adding commit 0231bb5336758426b44ccd798ccd3c5419c95d58 to stable tree. It fixes a real bug, which was explained here:
	http://thread.gmane.org/gmane.linux.kernel.perf.user/1144
and I've managed to reproduce the bug in 3.4 kernel with the test program attached in the above email, and I've confirmed the bug fix.
The effect of the bug is, if we open a group of mixed events(sw/hw) and the "disabled" flag is not set and sw events come first, then some of the perf events won't be started after the call to perf_event_open().

Quoted 8 lines

The bug fix should be applicable to kernels >= 3.0 && <= 3.7.
--
Li Zefan
--
To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Thanks, I'm queuing it for the 3.5 kernel.

Cheers,

Signature · 1 line

Luis