Todo
From Perf Wiki
(Difference between revisions)
Line 17: | Line 17: | ||
* Forward port the page fault tracepoints and use it in 'trace': | * Forward port the page fault tracepoints and use it in 'trace': | ||
** http://git.kernel.org/?p=linux/kernel/git/acme/linux.git;a=commitdiff;h=d53b11976093b6d8afeb8181db53aaffc754920d;hp=32ba4abf60ae1b710d22a75725491815de649bc5 | ** http://git.kernel.org/?p=linux/kernel/git/acme/linux.git;a=commitdiff;h=d53b11976093b6d8afeb8181db53aaffc754920d;hp=32ba4abf60ae1b710d22a75725491815de649bc5 | ||
+ | |||
+ | * Use the highest precision level available by default, e.g.: cycles:pp |
Revision as of 18:45, 29 January 2013
Perf tools
- Factorize the multidimensional sorting between perf report and annotate (will be used by perf trace)
- Implement a perf cmp (profile comparison between two perf.data) (DONE, its called 'perf diff')
- Implement a perf view (GUI)
- Enhance perf trace:
- Handle the cpu field
- Handle the timestamp
- Use the in-perf ip -> symbol resolving
- Use the in-perf pid -> cmdline resolving
- Implement multidimensional sorting by field name
- The feature tests performed only when a file that needs those tests, or at least only when some .c or .h file will be rebuilt
- Packages needed for the build should be checked before we start building object files, such as bison (bpetkov)
- Forward port the page fault tracepoints and use it in 'trace':
- Use the highest precision level available by default, e.g.: cycles:pp