From: Maynard J. <may...@us...> - 2014-01-02 14:57:18
|
On 12/23/2013 09:38 AM, David Flater wrote: > On 12/19/2013 12:26 PM, Maynard Johnson wrote: >> David, In a different posting thread, Michael Rutter already gave me some positive feedback on this "false recursion" patch. I would appreciate if you could run with this patch, too, to make sure it addresses your original issue. > > On the simple test case that reproduced the failure, it does make the failure go away. I have not tested extensively to determine if there was any collateral damage. My only question about the patch would be if anyone has inspected the implementation of perf in the tools/perf subdirectory of the Linux kernel source tree to determine how perf itself deals with the duplicated addresses. Thanks for testing, David. I've pushed the patch to oprofile's upstream repo. As for perf, that's outside of my scope. You can post to the linux-perf-users mailing list to pursue a change in perf. -Maynard > > -- > David Flater, National Institute of Standards and Technology, U.S.A. > |