Re: [PATCH] gnu: Add ninja.

[ludo@gnu.org (Ludovic Courtès)]

宋文武 <iyzsong@gmail.com> skribis:

> Ludovic Courtès <ludo@gnu.org> writes:
>
>> 宋文武 <iyzsong@gmail.com> skribis:
>>
>>> From 712f161a358a4bbcf005c5875e6d91eff21bdef4 Mon Sep 17 00:00:00 2001
>>> From: =?UTF-8?q?=E5=AE=8B=E6=96=87=E6=AD=A6?= <iyzsong@gmail.com>
>>> Date: Fri, 9 Jan 2015 21:37:30 +0800
>>> Subject: [PATCH] gnu: Add ninja.
>>>
>>> * gnu/packages/ninja.scm: New file.
>>> * gnu-system.am (GNU_SYSTEM_MODULES): Add it.
>>
>> Yes, OK to push, thanks!
> Even with 'system', the test InterruptChild failed on hydra for x86_64,
> but it passed for i686, mips64el and my x86_64 locally by 'guix build'.

I don’t see any failures actually:

  http://hydra.gnu.org/job/gnu/master/ninja-1.5.3.i686-linux
  http://hydra.gnu.org/job/gnu/master/ninja-1.5.3.mips64el-linux
  http://hydra.gnu.org/job/gnu/master/ninja-1.5.3.x86_64-linux

Are you sure it failed?

> Perhaps it's a random issue?

It looks like it.  It builds fine on my machine (x86_64).

I’ve looked at the test in question and it’s not obvious what could go
wrong:

--8<---------------cut here---------------start------------->8---
TEST_F(SubprocessTest, InterruptChild) {
  Subprocess* subproc = subprocs_.Add("kill -INT $$");
  ASSERT_NE((Subprocess *) 0, subproc);

  while (!subproc->Done()) {
    subprocs_.DoWork();
  }

  EXPECT_EQ(ExitInterrupted, subproc->Finish());
}
--8<---------------cut here---------------end--------------->8---

I think the idea is that ‘subproc_.Add’ launches a shell that gets
interrupted by the ‘kill’ subprocess, which ‘Finish’ must reflect.

Thanks,
Ludo’.