This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFA 4/8] Make strip_bg_char return a unique_xmalloc_ptr


>>>>> "Yao" == Yao Qi <qiyaoltc@gmail.com> writes:

Yao> Yes, that is one approach I though of, but it needs changing code.  I am
Yao> looking at running gdb testsuite with valgrind, but there are too many
Yao> warnings on leak, so it is not that useful to find new leaks.

That sounds pretty bad.
A valgrind or ASAN buildbot would be good...

Yao> The patch is the good direction to go, so I don't want to block it.
Yao> Meanwhile, I am still looking at memory leaks reported by valgrind.

I don't know if the cleanup checker still works, but that would be one
route.  I don't believe there are all that many dangling cleanups any
more, but I do agree that it's hard to know.

It's maybe worth pointing out that, at least in the past (I didn't look
recently), commands were always run with a cleanup installed, and this
cleanup was run after the command returned.  So, before the cleanup
checker work, it was normal to see dangling cleanups in command
implementations.  If this still exists then I think it would reduce the
risk of changing strip_bg_char.

Tom


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]