Bug 474326

Summary:	illumos: gdbserver tests hanging
Product:	[Developer tools] valgrind	Reporter:	Paul Floyd <pjfloyd>
Component:	general	Assignee:	Paul Floyd <pjfloyd>
Status:	RESOLVED FIXED
Severity:	normal
Priority:	NOR
Version First Reported In:	3.22 GIT
Target Milestone:	---
Platform:	Compiled Sources
OS:	Unspecified
Latest Commit:		Version Fixed/Implemented In:
Sentry Crash Report:

Description Paul Floyd 2023-09-09 10:12:05 UTC

OI-hipster, probably 3-6 months old, but this has been happening for quite some time.

 SunOS openindiana 5.11 illumos-ed093b41a9 i86pc i386 i86pc


0000000058057da6 pollsys  (5b203bf0, 1, 0, 0)
 000000005804639a vgPlain_poll () + 6a
 0000000058102bd5 vgPlain_poll_no_eintr () + 65
 00000000581037be readchar.part.0 () + 2e
 0000000058104148 getpkt () + 28
 0000000058094b17 server_main () + 67
 00000000580d79b8 call_gdbserver () + 118
 00000000580d84ba vgPlain_gdbserver () + 7a
 000000005809b873 vgPlain_scheduler () + 83
 00000000580acdc1 run_a_thread_NORETURN () + b1
paulf@openindiana:~$

Comment 1 Paul Floyd 2023-09-28 06:09:36 UTC

Looks like a GDB problem

(gdb) target remote | ./vgdb --wait=60 --vgdb-prefix=./vgdb-prefix-mcleak
/jenkins/jobs/oi-userland/workspace/components/developer/gdb/gdb-13.2/gdbsupport/filestuff.cc:381: internal-error: gdb_socketpair_cloexec: socketpair not available on this host
A problem internal to GDB has been detected,
further debugging may prove unreliable.
----- Backtrace -----
0x7b05c4 ???
0xb6524c ???
0xb65558 ???
0xca19fc ???
0xca2f46 ???
0xa8e281 ???
0xa8f6d2 ???
0xa79a96 ???
0xb153de ???
0x7e472c ???
0xb26158 ???
0x8cc40b ???
0x8cd8e8 ???
0x8cce40 ???
0xbaa251 ???
0x8cbd9d ???
0x8ccc54 ???
0x8cb849 ???
0xca2105 ???
0xca27f2 ???
0x987bf1 ???
0x989594 ???
0x70a596 ???
0x70a4a6 ???
0x70a407 ???
---------------------
/jenkins/jobs/oi-userland/workspace/components/developer/gdb/gdb-13.2/gdbsupport/filestuff.cc:381: internal-error: gdb_socketpair_cloexec: socketpair not available on this host
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n) 

I'll log an issue with the illumos bugtracker, but that's usually pretty hopeless.

Comment 2 Paul Floyd 2023-10-05 13:59:42 UTC

The first part to fixing this is on the way upstream in gdb.

https://sourceware.org/bugzilla/show_bug.cgi?id=30927

That should fix the hangs.

There are still several issues, including

https://sourceware.org/bugzilla/show_bug.cgi?id=29315

Comment 3 Paul Floyd 2024-11-17 10:28:23 UTC

I no longer get any hangs. We don't have a 'fixed indirectly' category so this is just 'fixed'.

Most of the remaining errors seem to be signal related.

== 25 tests, 6 stderr failures, 0 stdout failures, 2 stderrB failures, 6 stdoutB failures, 0 post failures ==
gdbserver_tests/mcinvokeWS               (stdoutB)
gdbserver_tests/mcinvokeWS               (stderrB)
gdbserver_tests/mcsignopass              (stderr)
gdbserver_tests/mcsignopass              (stdoutB)
gdbserver_tests/mcsigpass                (stderr)
gdbserver_tests/mcsigpass                (stdoutB)
gdbserver_tests/nlgone_abrt              (stderr)
gdbserver_tests/nlpasssigalrm            (stderr)
gdbserver_tests/nlpasssigalrm            (stdoutB)
gdbserver_tests/nlpasssigalrm            (stderrB)
gdbserver_tests/nlvgdbsigqueue           (stderr)
gdbserver_tests/nlvgdbsigqueue           (stdoutB)
gdbserver_tests/solaris/nlcontrolc       (stderr)
gdbserver_tests/solaris/nlcontrolc       (stdoutB)