User Details
- User Since
- May 10 2016, 10:57 AM (314 w, 5 d)
Thu, May 19
nit
Wed, May 18
nit
Tue, May 17
rebase
Hi! I'll push it for you :)
Thu, May 12
lgtm
address comments
Wed, May 11
rebase
- remove the flush parameter
- use perf_event_attr.disabled to set the initial state of the collection
address comments
Tue, May 10
Now the tracedThreads field returned by the GetState request must have the list of all the threads of the process if per-core mode is enabled. This will avoid having to make that field optional.
address comments
add additional test
Mon, May 9
nice!
Thanks for doing this fix. I hope tsan doesn't report this anymore
Thu, May 5
nits
add better error handling
Wed, May 4
Tue, May 3
improve error message
add one test
nits
Mon, May 2
update test
thanks for your review :)
final nits
move procfs functions to Procfs.h
Thu, Apr 28
I made a mistake and I included the second of two commits. This update contains all the code.
nit
Wed, Apr 27
Mon, Apr 25
address comments
Apr 22 2022
address comments
Apr 19 2022
address comments. Still pending response about the iff word.
I think I can do better than this
To catch changes that may increase the memory by a lot?
Apr 18 2022
let me think about it :)
formatting
Apr 13 2022
Committed as bdf3e7e5b886ab690a82e8d34b4502d9c6c5d6a9
Committed as 44103c96fa6b00e7824319de1b10ce26781e3852
Apr 12 2022
Apr 11 2022
Well, not all theoretical trace plugins might have raw trace sizes.
I did a first pass on this diff. I'm asking to refactor a bit the InstructionBlock classes so that they are smarter. Besides that, if you use IDs more ubiquitously and stop using instruction indices everywhere, everything becomes much simpler.
Apr 7 2022
address comments
address comments
thanks for the gotchas
Apr 6 2022
nits
Apr 5 2022
ok! I'll commit it for you
Let me know if you need help upstreaming this patch.
thank you!
great job!
just remove a small comment and good to go!
Apr 4 2022
nits