Gerrit: Reliability and fidelity of verification results

[David Ostrovsky]

On Mon, 2015-07-13 at 16:51 -0500, Norbert Thiebaud wrote:
> On Mon, Jul 13, 2015 at 1:51 PM, David Ostrovsky <d.ostrovsky at gmx.de> wrote:
> >> And your patch 8 would have failed the same way on tb58/tb59/tb60
> >
[...]

> The release builder have the so-called 'stale' tool chain.
> Just like in real-life and like other platform there is a diversity of
> platform and we usually do not drop
> support for 'older' platform unless there is an imperative motivation.
> 

I have another discrepancy with verification: [1]. This time tinderbox
TB63 compiler version is ahead of mine: [2]

TB63: GCC 4.8.3
Mine: GCC 4.8.1

On patch set 8 of this change, it's issuing -Werror=return-type: [3]
even though the last statement is assert(false);

control reaches end of non-void function [-Werror=return-type]
         }

I wrote this test program:

  $ cat foo.cxx
  #include <assert.h>
  int bar() {
    assert(false);
  }

And tried to compile it with and without debug option:

* $ g++ -Wreturn-type -ggdb2 -c foo.cxx
* $ g++ -Wreturn-type -c foo.cxx

The same story (no warning) with the most recent clang (3.7) version, 
with and without debug symbol option:

* clang++ -Wreturn-type -c foo.cxx
* clang++ -Wreturn-type -g -c foo.cxx

Can you try to compile this test program on TB63?
Any clue what is going on here?

[1] https://gerrit.libreoffice.org/17056
[2]
$ gcc --version
gcc 4.8.1 20130909 [gcc-4_8-branch revision 202388]

$ clang++ --version
clang version 3.7.0 (trunk 238405)
Target: x86_64-unknown-linux-gnu
Thread model: posix
[3]
http://ci.libreoffice.org/job/lo_gerrit_master/4355/Gerrit=Gerrit,Platform=Linux/consoleFull