[Bug testsuite/31312] attach-many-short-lived-threads gives inconsistent results

["thiago.bauermann at linaro dot org" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=31312

--- Comment #40 from Thiago Jung Bauermann <thiago.bauermann at linaro dot org> ---
(In reply to Carl E Love from comment #37)
> +FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 10: break at
> break_fn: 1 (the program is no longer running)

As I mentioned in comments 19 and 23, attach-many-short-lived-threads.exp
exposed three problems in aarch64-linux-gnu and powerpc64le-linux-gnu.

I proposed that we use this bugzilla to track the issue of GDB hanging with
100% CPU use because it wasn't able to deal with zombie threads. This is the
problem I described in comment 17 and fixed in the patch series that I
committed. This bug would appear in gdb.sum as timeout failures, because GDB
would be effectively hung and DejaGNU wouldn't be able to communicate with it
anymore. So the FAIL messages above would indicate a different kind of problem.

A separate issue I described there was GDB thinking that it had found all
inferior threads, and leaving the breakpoint instructions in the inferior while
inferior threads unknown to it were still running. This would cause the
inferior to exit with a SIGTRAP as it hit those unexpected breakpoint
instructions. The FAIL message above is consistent with this bug, which as Tom
de Vries mentioned is tracked in PR26286. We would need gdb.log to confirm
whether it is really this problem that is happening.

So if one accepts my suggestion of using this PR to track the problem of GDB
entering an infinite loop in the presence of zombie inferior threads, then
based on the gdb.sum that Carl posted I'd argue that there's no evidence that
it's still happening.

-- 
You are receiving this mail because:
You are on the CC list for the bug.