__NR_exit should probably be __NR_exit_group for the purposes of that test. But the parent’s not doing waitpid(child, …) so who knows why it’s getting stuck.

None of the other forky tests in that file do a waitpid() either, which is unhygienic at the very least.

I tried adding waitpid to that test earlier today to no avail -- it hung indefinitely at that point rather than at the end of the harness's execution.

Did you try __NR_exit_group?

I hadn't. Did so now, though, and that seems to work. I'll send a CL.

The following revision refers to this bug:
  https://chromium.googlesource.com/breakpad/breakpad/+/16f5b5f5125d4f77a67fc70a0ac09da5e4761bcc

commit 16f5b5f5125d4f77a67fc70a0ac09da5e4761bcc
Author: John Budorick <jbudorick@chromium.org>
Date: Wed May 24 16:05:24 2017

Use __NR_exit_group in MinidumpWriterTest.MinidumpStacksSkippedIfRequested.

Also adds waits for all child processes spawned in MinidumpWriterTest.

Bug:  725754 
Change-Id: I3248925993dede2c113ab1989b322a9d9c8f24bd
Reviewed-on: https://chromium-review.googlesource.com/513480
Reviewed-by: Mark Mentovai <mark@chromium.org>

[modify] https://crrev.com/16f5b5f5125d4f77a67fc70a0ac09da5e4761bcc/src/client/linux/minidump_writer/minidump_writer_unittest.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/breakpad/breakpad/+/dc3ba60f0f7e99f603565ff82ee10a5e2c23476b

commit dc3ba60f0f7e99f603565ff82ee10a5e2c23476b
Author: John Budorick <jbudorick@chromium.org>
Date: Thu May 25 14:44:22 2017

Don't attempt to use PTRACE_GETREGS if it isn't defined.

Follow up to https://chromium-review.googlesource.com/c/484479/, which
does not compile on arm64.

Bug:  chromium:725754 
Change-Id: Iaa6fbc332564909a10e2602a1026c14fb25625f4
Reviewed-on: https://chromium-review.googlesource.com/515044
Reviewed-by: Mark Mentovai <mark@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>

[modify] https://crrev.com/dc3ba60f0f7e99f603565ff82ee10a5e2c23476b/src/client/linux/minidump_writer/linux_ptrace_dumper.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/199d1412c583fb89b734f72a1911d49ecd2dfba5

commit 199d1412c583fb89b734f72a1911d49ecd2dfba5
Author: John Budorick <jbudorick@chromium.org>
Date: Thu May 25 16:12:25 2017

Roll src/breakpad/src/ 47e51739f..97a230bd4 (14 commits)

https://chromium.googlesource.com/breakpad/breakpad/src.git/+log/47e51739fd00..97a230bd4e72

$ git log 47e51739f..97a230bd4 --date=short --no-merges --format='%ad %ae %s'
2017-05-25 jbudorick Don't attempt to use PTRACE_GETREGS if it isn't defined.
2017-05-24 pcc Simplify ELF parser code.
2017-05-24 jbudorick Wrap config.h include in HAVE_CONFIG_H.
2017-05-24 jbudorick Use __NR_exit_group in MinidumpWriterTest.MinidumpStacksSkippedIfRequested.
2017-05-10 lv Only use O_CLOEXEC on platforms that support it
2017-05-10 lv Fix race in ExceptionHandler::GenerateDump()
2017-04-21 andrew Fixing breakpad on old linux kernel
2017-04-11 gsvelto Make minidump name generation on Linux as random as possible
2017-03-29 rsesek Dump MH_DYLINKER images in upload_system_symbols.
2017-03-29 tobiasjs Treat the process stack as the top of memory for free space histograms.
2017-03-24 lv Increase maximum number of threads for minidump_stackwalk.
2017-03-24 vapier minidump_dump: fix up init paths
2016-10-25 vapier minidump_dump: dump stack memory like hexdump
2017-03-20 vapier minidump_dump: add proper cli processing

Created with:
  roll-dep src/breakpad/src

Bug:  725754 
Change-Id: I484f910de13ccea7b02e988bc25afaa41c27f97c
Reviewed-on: https://chromium-review.googlesource.com/514203
Commit-Queue: John Budorick <jbudorick@chromium.org>
Reviewed-by: Mark Mentovai <mark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#474669}
[modify] https://crrev.com/199d1412c583fb89b734f72a1911d49ecd2dfba5/DEPS

breakpad_unittests passing on N as of https://build.chromium.org/p/chromium.android.fyi/builders/Nougat%20Phone%20Tester/builds/2390