[Buildroot] [autobuild.buildroot.net] Build results for 2014-07-11

[Buildroot] [autobuild.buildroot.net] Build results for 2014-07-11

[Thomas Petazzoni]

Build statistics for 2014-07-11
===============================

        success : 37 
       failures : 7  
       timeouts : 0  
          TOTAL : 44 

Classification of failures by reason
====================================

            host-python3-3.4.1 | 2 
                 webkit-1.11.5 | 1 
            kismet-2013-03-R1b | 1 
          perl-net-ssleay-1.64 | 1 
                  python-2.7.8 | 1 
                oprofile-0.9.9 | 1 

Detail of failures
===================

      i686 |             host-python3-3.4.1 | NOK | http://autobuild.buildroot.net/results/c294e3e4842637fb07efe2d8bfeb6c6cde78ad33/
   powerpc |             host-python3-3.4.1 | NOK | http://autobuild.buildroot.net/results/ec60fe3d507f47c7ab80d1a34de86b67dbbafa82/
   powerpc |             kismet-2013-03-R1b | NOK | http://autobuild.buildroot.net/results/d99dbb16aaeb22a319533cf6517bbca054daa1ac/
       arm |                 oprofile-0.9.9 | NOK | http://autobuild.buildroot.net/results/4f887c4bb46fcc1f7966c9f8652e47682fb4d30b/
       arm |           perl-net-ssleay-1.64 | NOK | http://autobuild.buildroot.net/results/c7787542651e047631213e7be781008c60d48e0f/
       arm |                   python-2.7.8 | NOK | http://autobuild.buildroot.net/results/10f268cd55f16ee53f4c5e17e1248b0a1820f6ea/
       arm |                  webkit-1.11.5 | NOK | http://autobuild.buildroot.net/results/baa98dcc3a29c5184fef2a9e447037d8ba5cf6ea/


-- 
http://autobuild.buildroot.net

[Buildroot] [autobuild.buildroot.net] Build results for 2014-07-11

[Yann E. MORIN]

Thomas, All,

On 2014-07-12 08:30 +0200, Thomas Petazzoni spake thusly:
>       i686 |             host-python3-3.4.1 | NOK | http://autobuild.buildroot.net/results/c294e3e4842637fb07efe2d8bfeb6c6cde78ad33/
>    powerpc |             host-python3-3.4.1 | NOK | http://autobuild.buildroot.net/results/ec60fe3d507f47c7ab80d1a34de86b67dbbafa82/

Unable to reproduce here, neither on my host system (Ubuntu 14.04,
32-bit) or in a Squeeze 32-bit chroot.

>        arm |                   python-2.7.8 | NOK | http://autobuild.buildroot.net/results/10f268cd55f16ee53f4c5e17e1248b0a1820f6ea/

Reproduced, investigating...

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

[Buildroot] [autobuild.buildroot.net] Build results for 2014-07-11

[Yann E. MORIN]

Thomas, All,

On 2014-07-12 11:55 +0200, Yann E. MORIN spake thusly:
> On 2014-07-12 08:30 +0200, Thomas Petazzoni spake thusly:
> >        arm |                   python-2.7.8 | NOK | http://autobuild.buildroot.net/results/10f268cd55f16ee53f4c5e17e1248b0a1820f6ea/
> 
> Reproduced, investigating...

OK, the sqlite3 extension has a bug. It has code like:

    ---8<---
    void _pysqlite_final_callback(sqlite3_context* context)
    {
        PyObject* function_result;
        PyObject** aggregate_instance;
        int ok;

    #ifdef WITH_THREAD
        PyGILState_STATE threadstate;

        threadstate = PyGILState_Ensure();
    #endif

        aggregate_instance = (PyObject**)sqlite3_aggregate_context(context, sizeof(PyObject*));
        if (!*aggregate_instance) {
            goto error;
        }

        [......]

    error:
    #ifdef WITH_THREAD
        PyGILState_Release(threadstate);
    #endif
    }
    ---8<---

So, they are doing right to cleanup in case of error. But when threads
are not available, the function is like:

    ---8<---
    void _pysqlite_final_callback(sqlite3_context* context)
    {
        PyObject* function_result;
        PyObject** aggregate_instance;
        int ok;

        aggregate_instance = (PyObject**)sqlite3_aggregate_context(context, sizeof(PyObject*));
        if (!*aggregate_instance) {
            goto error;
        }

        [......]

    error:
    }
    ---8<---

And indeed, there's nothing after the 'error' label. And gcc does not
like that:
    connection.c:785:1: error: label@end of compound statement

So, the solution is relatively easy: add a dummy statement after the
label:

    error:
        ;
    }

And it works, now. I'll submit a patch soonish...

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

[Buildroot] [autobuild.buildroot.net] Build results for 2014-07-11

[Yann E. MORIN]

Thomas, All,

On 2014-07-12 08:30 +0200, Thomas Petazzoni spake thusly:
>        arm |           perl-net-ssleay-1.64 | NOK | http://autobuild.buildroot.net/results/c7787542651e047631213e7be781008c60d48e0f/

Brain-damage.

The build system for perl-net-ssleay is completely f*ked-up for
cross-compilation. It tries to execute the openssl binary, and *runs* it
to verify a few things about openssl: availability, version...

Then it looks for openssl/ssl.h in host-system locations:
    $prefix/include  $prefix/inc32  /usr/kerberos/include

Of course $prefix is /usr, so it picks the host system openssl.

Ditto for the -L paths.

Totally brain-damaged.

And it is happy to just state:
    *** Found OpenSSL-1.0.1f installed in /usr

Yeah, we're packaging openssl-1.0.1h in Buildroot. 1.0.1f is on my host
system's openssl...

Grrr...

/me reaches for his chainsaw...

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'