[regression/native-gdbserver][buildbot] Python testscases get staled

[Pedro Alves]

On 01/12/2015 04:43 AM, Sergio Durigan Junior wrote:
> 
> Buildbot (which is running only internally so far, but will hopefully be
> deployed this week) "caught" this when building in x86_64 (Fedora 21)
> and testing the native-gdbserver variant (both x86_64 and x86).  When
> you run the test on the gdb.python/ directory, you see that it stales
> when it reaches the gdb.python/py-section-script.exp testcase.
> Buildbot's gdb.log specifically has:
> 
>   Running ../../../binutils-gdb/gdb/testsuite/gdb.python/py-section-script.exp ...
>   ERROR: (timeout) GDB never initialized after 10 seconds.
>   ERROR: no fileid for gdbuild
>   ERROR: Couldn't send python print ('test') to GDB.
>   ERROR: no fileid for gdbuild
>   ERROR: Couldn't send python print (sys.version_info[0]) to GDB.
>   ERROR: no fileid for gdbuild
>   ERROR: Couldn't send python print (sys.version_info[1]) to GDB.
>   ERROR: no fileid for gdbuild
>   ERROR: no fileid for gdbuild
>   ...
>   (this goes on and on, for several testcases)
> 

Here's what I get:

$ make check RUNTESTFLAGS="--target_board=native-gdbserver gdb.python/py-section-script.exp"
...
Running /home/pedro/gdb/mygit/src/gdb/testsuite/gdb.python/py-section-script.exp ...

                === gdb Summary ===

# of expected passes            7
...

> I still did not debug this (intend to do so tomorrow, if you don't see
> this before I start my day), but running a simple git-bisect showed me
> that this specific commit is the culprit.

Odd, my commit did not touch this file.  Actually, the test does not
use "attach" at all.

Thanks,
Pedro Alves