[PATCH] run-tests: test result shows when a failed test could not start a server

[Simon Heimberg]

On 2014-03-01 11:07, Simon Heimberg wrote:
> On 28.02.2014 20:57, Augie Fackler wrote:
>> On Jan 16, 2014, at 3:58 PM, Simon Heimberg <simohe at besonet.ch> wrote:
>>
>>> # HG changeset patch
>>> # User Simon Heimberg <simohe at besonet.ch>
>>> # Date 1385413246 -3600
>>> # Node ID afdcb3c8d7aae39b02ea208c180e36a2a6602bf0
>>> # Parent  e2068d5d4289fe61f2e5739d62d92fa3c0c2f285
>>> run-tests: test result shows when a failed test could not start a server
>> This looks mostly reasonable to me. Does anyone else have feelings?
> Not sure if we still need this when
> [PATCH stable] win32: spawndetached returns pid of detached process and
> not of cmd.exe [1]
> is accepted (plus buildbot is rebooted). Server start on windows should
> not fail after this.
>
> [1]
> http://www.selenic.com/pipermail/mercurial-devel/2014-February/056358.html
>

Well, not sure any more about my statement.
In windows build 131 [2] test-wireproto.t starts to have a serve 
failure. And I do not have an idea why this has happened. The previous 
run did not report any problem (and kill should do this now). No run of 
(windows) stable did run between the two.

So probably this patch would still be helpful.

[2] 
http://hgbuildbot.kublai.com/builders/Windows%202008%20R2%20hg%20tests/builds/131 


>>> Failing to start a server happens regularly, at least on windows
>>> buildbot.
>>> Such a failure often has nothing to do with the test, but with the
>>> environment.
>>> But half the test output can change because some data is missing.
>>> Therefore this
>>> is worth an extended error message.
>>>
>>>
> <snip>