|
From: Remus C. <rem...@gm...> - 2018-06-23 23:53:45
|
Hi, I tried to use valgrind's memcheck on a C copy-stack coroutine program, and got many false positive reports about the invalid write/read on the shared stack when memcpy occurred. (I already used the VALGRIND_STACK_REGISTER/VALGRIND_STACK_DEREGISTER.) Valgrind works fine when I use valgrind on a standalone stack coroutine program. So, the question is: Does valgrind support copy-stack coroutine yet? Thank you very much. All best |
|
From: John R. <jr...@bi...> - 2018-06-25 03:36:40
|
On 06/23/2018, Remus Clearwater wrote:
> I tried to use valgrind's memcheck on a C copy-stack coroutine program, and got many false positive reports about the invalid write/read on the shared stack when memcpy occurred.
>
> (I already used the VALGRIND_STACK_REGISTER/VALGRIND_STACK_DEREGISTER.)
>
> Valgrind works fine when I use valgrind on a standalone stack coroutine program.
>
> So, the question is:
>
> Does valgrind support copy-stack coroutine yet?
The errors that you saw when you tried it, say that valgrind does not yet support it.
By the way, which version of valgrind? Which hardware? Which compiler(s)?
Which C run-time library? And what was the exact text of the first relevant
complaint from memcheck?
The surest way to make progress is to submit a New bug report at
https://bugs.kde.org/buglist.cgi?quicksearch=valgrind
and attach two test cases with actual code: one with "standalone stack co-routine",
and one with "copy-stack co-routine". Supplying actual code, and the recipe
to build executables from it, is vital. If the behavior that you saw
cannot be reproduced by a valgrind developer, then it is likely to be
a long time before anything gets fixed.
|
|
From: Remus C. <rem...@gm...> - 2018-06-25 03:57:45
|
Thank you very much, John. It is my fault. I will prepare the test cases and submit the detailed report as soon as possible. All best Remus On Mon, Jun 25, 2018 at 11:36 AM, John Reiser <jr...@bi...> wrote: > On 06/23/2018, Remus Clearwater wrote: > > I tried to use valgrind's memcheck on a C copy-stack coroutine program, >> and got many false positive reports about the invalid write/read on the >> shared stack when memcpy occurred. >> >> (I already used the VALGRIND_STACK_REGISTER/VALGRIND_STACK_DEREGISTER.) >> >> Valgrind works fine when I use valgrind on a standalone stack coroutine >> program. >> >> So, the question is: >> >> Does valgrind support copy-stack coroutine yet? >> > > The errors that you saw when you tried it, say that valgrind does not yet > support it. > By the way, which version of valgrind? Which hardware? Which compiler(s)? > Which C run-time library? And what was the exact text of the first > relevant > complaint from memcheck? > > The surest way to make progress is to submit a New bug report at > https://bugs.kde.org/buglist.cgi?quicksearch=valgrind > and attach two test cases with actual code: one with "standalone stack > co-routine", > and one with "copy-stack co-routine". Supplying actual code, and the > recipe > to build executables from it, is vital. If the behavior that you saw > cannot be reproduced by a valgrind developer, then it is likely to be > a long time before anything gets fixed. > > ------------------------------------------------------------ > ------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! http://sdm.link/slashdot > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > |
|
From: Remus C. <rem...@gm...> - 2018-07-01 14:23:40
|
I had created the bug report: https://bugs.kde.org/show_bug.cgi?id=396053. And the code repository is in: https://github.com/hnes/libaco/tree/valgrind_memcheck_bug_report_on_copy_stack_coroutine . All best Remus On Mon, Jun 25, 2018 at 11:57 AM, Remus Clearwater < rem...@gm...> wrote: > Thank you very much, John. > > It is my fault. I will prepare the test cases and submit the detailed > report as soon as possible. > > All best > > Remus > > On Mon, Jun 25, 2018 at 11:36 AM, John Reiser <jr...@bi...> > wrote: > >> On 06/23/2018, Remus Clearwater wrote: >> >> I tried to use valgrind's memcheck on a C copy-stack coroutine program, >>> and got many false positive reports about the invalid write/read on the >>> shared stack when memcpy occurred. >>> >>> (I already used the VALGRIND_STACK_REGISTER/VALGRIND_STACK_DEREGISTER.) >>> >>> Valgrind works fine when I use valgrind on a standalone stack coroutine >>> program. >>> >>> So, the question is: >>> >>> Does valgrind support copy-stack coroutine yet? >>> >> >> The errors that you saw when you tried it, say that valgrind does not yet >> support it. >> By the way, which version of valgrind? Which hardware? Which >> compiler(s)? >> Which C run-time library? And what was the exact text of the first >> relevant >> complaint from memcheck? >> >> The surest way to make progress is to submit a New bug report at >> https://bugs.kde.org/buglist.cgi?quicksearch=valgrind >> and attach two test cases with actual code: one with "standalone stack >> co-routine", >> and one with "copy-stack co-routine". Supplying actual code, and the >> recipe >> to build executables from it, is vital. If the behavior that you saw >> cannot be reproduced by a valgrind developer, then it is likely to be >> a long time before anything gets fixed. >> >> ------------------------------------------------------------ >> ------------------ >> Check out the vibrant tech community on one of the world's most >> engaging tech sites, Slashdot.org! http://sdm.link/slashdot >> _______________________________________________ >> Valgrind-users mailing list >> Val...@li... >> https://lists.sourceforge.net/lists/listinfo/valgrind-users >> > > |