ppc October 2017

ppc@lists.fedoraproject.org

3 participants
4 discussions

ppc64le: c++: internal compiler error: Killed (program cc1plus)
by Sérgio Basto 10 Mar '19

10 Mar '19

Hello ppc list , I don't know where I should report this issue, if this is not the correct place, please tell me where it is, thanks. I can't build opencv-3.1 on ppc64le the builds always ends with : {standard input}: Assembler messages: {standard input}:36678: Warning: end of file not at end of a line; newline inserted {standard input}:39048: Error: unknown pseudo-op: `.l295' {standard input}: Error: open CFI at the end of file; missing .cfi_endproc directive c++: internal compiler error: Killed (program cc1plus) Please submit a full bug report, with preprocessed source if appropriate. See <http://bugzilla.redhat.com/bugzilla> for instructions. make[2]: *** [modules/cvv/CMakeFiles/opencv_cvv.dir/src/view/rawview.cpp.o] Error 4 make[2]: *** Waiting for unfinished jobs.... Build logs: https://copr.fedorainfracloud.org/coprs/sergiomb/opencv/build/183181/ https://copr-be.cloud.fedoraproject.org/results/sergiomb/opencv/fedora- 23-ppc64le/00183181-opencv/build.log.gz https://copr-be.cloud.fedoraproject.org/results/sergiomb/opencv/fedora- rawhide-ppc64le/00181899-opencv/build.log.gz Best regards, -- Sérgio M. B.

4 7

Switching Fedora ppc64 glibc to IFUNC-only optimizations
by Florian Weimer 13 Dec '17

13 Dec '17

I would like to tweak glibc so that it only builds one POWER 620/970 libc for ppc64, and not multiple libcs with POWER6/POWER7/POWER8 optimizations. This does not affect ppc64le in any way. The reason for this change is that the ppc64 builders are awfully slow as of late, and the multiple builds (four times currently) currently make ppc64 the slowest architecture to build by quite a margin. Any comments? Thanks, Florian

9 31

Random segfaults on Fedora builders for POWER
by Florian Weimer 03 Nov '17

03 Nov '17

I'm seeing random-looking segfaults on Fedora builders: buildvm-ppc64-08.ppc.fedoraproject.org <https://koji.fedoraproject.org/koji/taskinfo?taskID=22351226> “ This might take a while Testing GB18030run-iconv-test.sh: line 59: 27196 Segmentation fault (core dumped) $PROG -t ../localedata/charmaps/$from -f ../localedata/charmaps/$tc -o $temp2 $temp1 < /dev/null Testing EUC-JISX0213 ” I don't have logs for the other failed tasks because I reused the NVR, sorry. All were for glibc-2.26-13.fc27, and most of them where in fact on ppc64le. The place varies somewhat between runs. So far, I've only seen it for the Fedora 27. Could it be a hardware issue? Thanks, Florian

2 2

building upstream Firefox
by Dan Horák 20 Oct '17

20 Oct '17

Hi, there is official mercurial repo [1] and official git clone [2] for FF I've started with mercurial one (next time I will use the git one), so I do # updates SKIA for big endians - you can skip for ppc64le hg import -u jenkins -m "build-big-endian" https://src.fedoraproject.org/fork/sharkcz/rpms/firefox/raw/next/f/build-bi… # curl - seems needed even for ppc64le, I plan to upstream it after fixing hg import -u jenkins -m "curl" https://src.fedoraproject.org/fork/sharkcz/rpms/firefox/raw/next/f/build-pp… and I start the build with (you can drop the MAKE_FLAGS) # build (serial) with everything bundled make -f client.mk MOZ_MAKE_FLAGS="-j1" before the start I've installed the dependencies for the FF rpm with "dnf builddep firefox.spec", but not all of them are used for the "default" build. I use F-26 as the build environment. October 3rd seems to the last date in the period of green builds, later some issues appeared. [1] https://developer.mozilla.org/en-US/docs/Mozilla/Developer_guide/Build_Inst… [2] https://github.com/mozilla/gecko-dev Dan

2 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

ppc October 2017