[meta-freescale] fsl-image-gui build errors on ubuntu 12.04.2

[Otavio Salvador]

On Tue, Feb 12, 2013 at 10:47 AM, Prasant J <pj0585 at gmail.com> wrote:
> On Tue, Feb 12, 2013 at 6:01 PM, Prasant J <pj0585 at gmail.com> wrote:
>> On Mon, Feb 11, 2013 at 7:31 PM, Otavio Salvador
>> <otavio at ossystems.com.br> wrote:
>>> On Mon, Feb 11, 2013 at 11:57 AM, Prasant J <pj0585 at gmail.com> wrote:
>>>> On Mon, Feb 11, 2013 at 5:51 PM, Otavio Salvador
>>>> <otavio at ossystems.com.br> wrote:
>>>>> On Mon, Feb 11, 2013 at 10:19 AM, Prasant J <pj0585 at gmail.com> wrote:
>>>>>> On Mon, Feb 11, 2013 at 5:37 PM, Otavio Salvador
>>>>>> <otavio at ossystems.com.br> wrote:
>>>>>>> On Mon, Feb 11, 2013 at 10:02 AM, Prasant J <pj0585 at gmail.com> wrote:
>>>>>>>> On Fri, Feb 8, 2013 at 4:28 PM, Otavio Salvador <otavio at ossystems.com.br> wrote:
>>>>>>>>> On Fri, Feb 8, 2013 at 6:52 AM, Prasant J <pj0585 at gmail.com> wrote:
>>>>>>>>>> On Wed, Feb 6, 2013 at 9:17 PM, Otavio Salvador <otavio at ossystems.com.br> wrote:
>>>>>>>>>>> On Wed, Feb 6, 2013 at 11:40 AM, Prasant J <pj0585 at gmail.com> wrote:
>>>>>>>>>>>> Hi,
>>>>>>>>>>>>
>>>>>>>>>>>> I'm trying to build fsl-image-gui but getting the following errors and
>>>>>>>>>>>> no images are prepared in tmp/deploy/
>>>>>>>>>>>
>>>>>>>>>>> This seems your mysql tarball is corrupted. Do:
>>>>>>>>>>>
>>>>>>>>>>> bitbake mysql5 -c cleanall
>>>>>>>>>>>
>>>>>>>>>>> bitbake -c linux-imx -c cleanall
>>>>>>>>>>>
>>>>>>>>>>> It seems the cached tarball is corrupted as well; please do those two
>>>>>>>>>>> and give it a try.
>>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> The linux-imx and linux-libc-headers-imx use same GIT server and it
>>>>>>>>> does take a looooooong time to download as those are huge. powertop
>>>>>>>>> and others are better to cleanall them and give it a new try.
>>>>>>>>>
>>>>>>>>> One thing that might explain the corrupt file is if you press Ctrl-C
>>>>>>>>> many times to force the interruption. So it will kill the transfer and
>>>>>>>>> left a broken file as result. In case you press Ctrl-C *once* it'll
>>>>>>>>> finish after it has done current running tasks, safely.
>>>>>>>>>
>>>>>>>>
>>>>>>>> @Otavio: that was the issue. I had presses Ctrl+C more than once and
>>>>>>>> the big downloads were getting corrupted and the process never
>>>>>>>> finished.
>>>>>>>>
>>>>>>>> I cleaned them all and continued. There were manageable minor issues.
>>>>>>>>
>>>>>>>> Now I'm stuck at this point:
>>>>>>>> parted-native-3.1-r0 do_configure()
>>>>>>>> This process seems to be stuck. Then I tried bitbake fsl-image-gui -v
>>>>>>>> and I see that while configuring parted there is a step "checking for
>>>>>>>> working sleep...".
>>>>>>>> It seems to stuck there. Any idea how long does it sleep?
>>>>>>>
>>>>>>> Weird; this is quite fast ...
>>>>>>>
>>>>>>> Maybe due your broken builds you ended with a broken tmp; I'd remove
>>>>>>> tmp and  give it a new try ... it won't rebuild everything as it's
>>>>>>> going to take them from sstate-cache.
>>>>>>>
>>>>>>
>>>>>> Will it download everything again? (Where is all the downloaded
>>>>>> packages stored? )
>>>>>
>>>>> It will not; check the downloads directory.
>>>>>
>>>>>> Would it be good if I only cleaned parted : bitbake parted -c cleanall ?
>>>>>
>>>>> You might try; I'd go for a clean tmp ;-)
>>>>>
>>>>
>>>>
>>>> bitbake parted -c cleanall did not work :-(
>>>> had to clean tmp/
>>>>
>>>> anyway, going further I'm getting this warning now:
>>>>
>>>>
>>>> WARNING: Setscene task 81
>>>> (/home/prasant/im6-yocto/fsl-community-bsp/sources/meta-fsl-arm/recipes-kernel/linux/linux-imx_3.0.35.bb,
>>>> do_populate_sysroot_setscene) failed with exit code '1' - real task
>>>> will be run instead
>>>>
>>>>
>>>> I was not getting these earlier, but they are appearing now.
>>>> If I have stop and start again, these appear everytime and delays the
>>>> build process (as it will try all these tasks every time ) Any inputs
>>>> on what could be going wrong?
>>>
>>> You sstate-cache seems borked; maybe due your Ctrl-C crazyness ;-)
>>>
>>> The best (but longer) option is to get rid of your tmp and
>>> sstate-cache and start again; keep your downloads directory as it is
>>> safe. In case you need to cancel the build type Ctrl-C once and wait
>>> it finish ... maybe it finish fine but I'd start over and keep a good
>>> sstate-cache around for next builds.
>>>
>>
>> Otavio, I cleaned the sstate-cache & tmp & tried again. Some pkgs were
>> giving errors so I did a cleanall for that pkg & continued. I'm stuck
>> at this point:
>>
>> | In file included from ../../sql/sql_class.h:23:0,
>> |                  from ../../sql/mysql_priv.h:873,
>> |                  from ha_blackhole.cc:22:
>> | ../../sql/log.h:282:3: warning: access declarations are deprecated
>> in favour of using-declarations; suggestion: add the 'using' keyword
>> [-Wdeprecated]
>> | ../../sql/log.h:283:3: warning: access declarations are deprecated
>> in favour of using-declarations; suggestion: add the 'using' keyword
>> [-Wdeprecated]
>> | ../../sql/log.h:421:16: note: the mangling of 'va_list' has changed in GCC 4.4
>> | arm-poky-linux-gnueabi-libtool: compile:  arm-poky-linux-gnueabi-g++
>> -march=armv7-a -mthumb-interwork -mfloat-abi=softfp -mfpu=neon
>> -mtune=cortex-a9
>> --sysroot=/home/prasant/im6-yocto/fsl-community-bsp/build/tmp/sysroots/imx6qsabresd
>> -DHAVE_CONFIG_H -I. -I../../include -I../../include -I../../include
>> -I../../regex -I../../sql -I. -DMYSQL_DYNAMIC_PLUGIN -O2 -pipe -g
>> -feliminate-unused-debug-types -fpermissive
>> -fvisibility-inlines-hidden -fno-implicit-templates -fno-exceptions
>> -fno-rtti -c ha_blackhole.cc -o ha_blackhole_la-ha_blackhole.o
>>>/dev/null 2>&1
>> | make[2]: *** [ha_blackhole_la-ha_blackhole.lo] Error 1
>> | make[2]: Leaving directory
>> `/home/prasant/im6-yocto/fsl-community-bsp/build/tmp/work/armv7a-vfp-neon-poky-linux-gnueabi/mysql5-5.1.40-r10/mysql-5.1.40/storage/blackhole'
>> | make[1]: *** [all-recursive] Error 1
>> | make[1]: Leaving directory
>> `/home/prasant/im6-yocto/fsl-community-bsp/build/tmp/work/armv7a-vfp-neon-poky-linux-gnueabi/mysql5-5.1.40-r10/mysql-5.1.40/storage'
>> | make: *** [all-recursive] Error 1
>> | ERROR: oe_runmake failed
>> | ERROR: Function failed: do_compile (see
>> /home/prasant/im6-yocto/fsl-community-bsp/build/tmp/work/armv7a-vfp-neon-poky-linux-gnueabi/mysql5-5.1.40-r10/temp/log.do_compile.28737
>> for further information)
>>
>> Anything related to by corrupted mysql5 download?
>>
>
> bitbake mysql5 -c clean seem to solve that issue. So it was not a bad
> download but bad uncompressed mysql5 ??

No; it was not a bad download. I am working in a patchset to address
some kernel headers issue that should make it more stable.

-- 
Otavio Salvador                             O.S. Systems
E-mail: otavio at ossystems.com.br  http://www.ossystems.com.br
Mobile: +55 53 9981-7854              http://projetos.ossystems.com.br