firefox-52.4.0 ppc64 fails to build

Message

rican-linux · Post by **rican-linux** » Fri Jan 12, 2018 6:04 am

I am trying to build firefox 54 on my PowerMac G5. It marked as stable in packages.gentoo.org however it is failing to build. Here is tail end of the build log. Any help to figure out what is wrong is appreciated

Code: Select all

Executing /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/ff/dist/bin/xpcshell -g /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/ff/dist/bin/ -a /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/ff/dist/bin/ -f /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/toolkit/mozapps/installer/precompile_cache.js -e precompile_startupcache("resource://gre/");
[22595] ###!!! ABORT: u_init() failed: file /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/xpcom/build/XPCOMInit.cpp, line 709
[22595] ###!!! ABORT: u_init() failed: file /var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/xpcom/build/XPCOMInit.cpp, line 709
Traceback (most recent call last):
  File "/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/toolkit/mozapps/installer/packager.py", line 415, in <module>
    main()
  File "/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/toolkit/mozapps/installer/packager.py", line 409, in main
    args.source, gre_path, base)
  File "/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/toolkit/mozapps/installer/packager.py", line 166, in precompile_cache
    errors.fatal('Error while running startup cache precompilation')
  File "/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/python/mozbuild/mozpack/errors.py", line 103, in fatal
    self._handle(self.FATAL, msg)
  File "/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/python/mozbuild/mozpack/errors.py", line 98, in _handle
    raise ErrorMessage(msg)
mozpack.errors.ErrorMessage: Error: Error while running startup cache precompilation
make[1]: *** [/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/toolkit/mozapps/installer/packager.mk:41: stage-package] Error 1
make[1]: Leaving directory '/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/ff/browser/installer'
make: *** [/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/browser/build.mk:21: install] Error 2
 [31;01m*[0m ERROR: www-client/firefox-52.4.0::gentoo failed (install phase):
 [31;01m*[0m   emake failed
 [31;01m*[0m 
 [31;01m*[0m If you need support, post the output of `emerge --info '=www-client/firefox-52.4.0::gentoo'`,
 [31;01m*[0m the complete build log and the output of `emerge -pqv '=www-client/firefox-52.4.0::gentoo'`.
 [31;01m*[0m The complete build log is located at '/var/tmp/portage/www-client/firefox-52.4.0/temp/build.log'.
 [31;01m*[0m The ebuild environment file is located at '/var/tmp/portage/www-client/firefox-52.4.0/temp/environment'.
 [31;01m*[0m Working directory: '/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr/ff'
 [31;01m*[0m S: '/var/tmp/portage/www-client/firefox-52.4.0/work/firefox-52.4.0esr'

johker · Post by **johker** » Thu Mar 15, 2018 8:57 am

Hi there, exactly same error here, Kernel 4.14.24, ppc64.
Any help will be much apreciated. Thanks

Post by **fedeliallalinea** » Thu Mar 15, 2018 10:42 am

You enabled debug use flag (see here)?

Post by **John R. Graham** » Thu Mar 15, 2018 11:11 am

Moved from Portage & Programming to Gentoo on PPC where it has a better visibility.

- John

johker · Post by **johker** » Thu Mar 15, 2018 6:20 pm

Thanks, but no success..
Compilation took more to complete, like 4h or more didn't take notice, but failed with same error, trying +custom-optimiztion, as described in https://forums.gentoo.org/viewtopic-t-1 ... 5c7d8dcd06

johker · Post by **johker** » Sat Mar 17, 2018 4:30 pm

Still same output.
neither -debug or +custom-optimization worked.

schwarzygesetzlos · Post by **schwarzygesetzlos** » Mon Mar 26, 2018 4:54 pm

My working USE-flag combination for ppc64, built with gcc-7.3.0:

Code: Select all

www-client/firefox -custom-cflags custom-optimization system-icu system-harfbuzz system-jpeg system-libevent system-libvpx system-sqlite

With gcc-6.4.0 I needed to build with CFLAGS="-O1". Which is generally not a bad idea when building big projects with only 2 cores.

Post by Hu » Tue Mar 27, 2018 1:54 am

schwarzygesetzlos wrote:With gcc-6.4.0 I needed to build with CFLAGS="-O1". Which is generally not a bad idea when building big projects with only 2 cores.

Could you explain your reasoning on this? Reducing the optimization level will reduce the work done by the compiler, which may help during the build, but it comes at the price of building less optimized code, which may run less well. You would probably be better served by reducing the load in other ways, such as reducing MAKEOPTS.

schwarzygesetzlos · Post by **schwarzygesetzlos** » Sun Apr 01, 2018 8:04 pm

@Hu:

Yes that's the point, that using -O1 will reduce the work done by the compiler, which shortens build time greatly on a machine from 2005.

I am aware that the code is less optimized as a result. But to my experience FF needs to be rebuilt not too rarely (when some of its' deps change), which takes rather long on a G4/G5. If I had a Quad-G5 or some POWER6-9 machine, I would of course prefer -O2. Here's some nice comparison of the -O flags and some benchmarks: https://phoronix.com/scan.php?page=news ... imizations

Changing MAKEOPTS to -j2 or -j1 will of course reduce the load but increase build time. I guess running -j3 on this dual core machine is a good choice, and greater values may even increase build time.