www-client/firefox-67.0 does not emerge. Full build log and emerge info at https://bpaste.net/show/1e2192eaa3c2 .
(In reply to Pistos from comment #0) > www-client/firefox-67.0 does not emerge. > > Full build log and emerge info at https://bpaste.net/show/1e2192eaa3c2 . Please attach those to this bug report directly.
Created attachment 578116 [details] Firefox 67.0 build log
Created attachment 578118 [details] emerge --info
Rust team please have a look, I am not seeing any failure with 67.0 or 68.0b* on an x86_64 musl setup.
I do hit the same bug, but on armv7a. Right now I'm trying to finish the build with ebuild /usr/portage/www-client/firefox-67.0 install to see wether this "low level" approach may be helpfull I opened an issue at rusts upstream https://github.com/rust-lang/rust/issues/61468 and will upload my emerge --info and build.log in a minute
Created attachment 578522 [details] output of emerge --info this is from arm, but same error
Created attachment 578526 [details] compressed build log same, this from arm but with the same error
When I changed from MAKEOPTS="-s -j1 -l 2" to MAKEOPTS="-s -j2 -l 3", the emerge finished successfully. No idea why that would make a difference, but thank you to @tt_1 for suggesting this. I will update the bug title accordingly.
I was able to reproduce on amd64 today, using MAKEOPTS="-j1" to trigger. Also its possible to resume the compile with ebuild install/merge command, which produces a fully working firefox binary.
MAKEOPTS="-j2 -l 1" seems to be working as a workaround, if you're working with a memory restrained system and can't go beyond one job at a time.
I have the same issue with firefox-68.0.1. The workaround WFM: MAKEOPTS="-j5 -l1" It seems that "-l1" is the critical part here. "-j5" or "-j1" only did not help.
*** Bug 692396 has been marked as a duplicate of this bug. ***
can you guys try building it without makeopts workaround, but with FEATURES="-pid-sandbox" ?
@Georgy, I've tried to follow your advice and removed MAKEOPTS introducing FEATURES="-pid-sandbox". Unfortunately, re-building firefox now is failing with different error: ``` 18:42.54 [style 0.0.1] cargo:rerun-if-changed=/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/dist/include/js/Utility.h 18:42.54 [style 0.0.1] cargo:rerun-if-changed=/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/dist/include/mozilla/Range.h 18:42.54 thread 'main' panicked at 'stack backtrace: 18:42.54 0: 0x556b6ac0f4 - <unknown> 18:42.54 1: 0x556b6b1d14 - <unknown> 18:42.54 2: 0x556b6b1adc - <unknown> 18:42.54 3: 0x556b6b2440 - <unknown> 18:42.54 4: 0x556b6b203c - <unknown> 18:42.55 5: 0x556b6b1fa0 - <unknown> 18:42.55 6: 0x556b6aa764 - <unknown> 18:42.55 7: 0x556b2c9804 - <unknown> 18:42.55 8: 0x556b2f4158 - <unknown> 18:42.55 9: 0x556b6b8ad0 - <unknown> 18:42.55 10: 0x556b390ffc - <unknown> 18:42.55 11: 0x556b2c7e54 - <unknown> 18:42.55 12: 0x556b35e350 - <unknown> 18:42.55 13: 0x556b1181a0 - <unknown> 18:42.55 14: 0x556b1170f8 - <unknown> 18:42.56 15: 0x556b0b84e4 - <unknown> 18:42.56 16: 0x556b0a84a8 - <unknown> 18:42.56 17: 0x556b0e1550 - <unknown> 18:42.56 18: 0x556b0b922c - <unknown> 18:42.56 19: 0x556b6b1ebc - <unknown> 18:42.56 20: 0x556b6b8ad0 - <unknown> 18:42.56 21: 0x556b6b65f0 - <unknown> 18:42.56 22: 0x556b0e3894 - <unknown> 18:42.56 23: 0x7fa62e14f0 - __libc_start_main 18:42.56 at ../csu/libc-start.c:308 18:42.56 24: 0x556b09c47c - <unknown> 19:32.02 [style 0.0.1] cargo:rerun-if-changed=/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/dist/include/nsIRunnable.h 19:32.03 gmake[4]: *** [/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/config/makefiles/rust.mk:240: force-cargo-library-build] Error 101 19:32.03 gmake[4]: Leaving directory '/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/toolkit/library/rust' 19:32.03 gmake[3]: *** [/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/config/recurse.mk:74: toolkit/library/rust/target] Error 2 19:32.03 gmake[3]: Leaving directory '/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff' 19:32.03 gmake[2]: *** [/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/config/recurse.mk:34: compile] Error 2 19:32.04 gmake[2]: Leaving directory '/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff' 19:32.04 gmake[1]: *** [/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/config/rules.mk:413: default] Error 2 19:32.04 gmake[1]: Leaving directory '/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff' 19:32.04 gmake: *** [client.mk:125: build] Error 2 19:32.06 11 compiler warnings present. ```
The issue above has gone after reboot. Now the old thing is back (when trying FEATURES="-pid-sandbox"): 18:38.92 [style 0.0.1] cargo:rerun-if-changed=/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/dist/include/js/CallArgs.h 19:38.81 thread 'main' panicked at 'failed printing to stdout: Resource temporarily unavailable (os error 11)', [style 0.0.1] cargo:rerun-if-changed=/var/tmp/portage/www-client/firefox-68.0.2/work/firefox-68.0.2/ff/dist/include/mozilla/ReverseIterator.h
thanks, I just had an idea and since you still hit it with pid-sandbox disabled my idea was wrong. still can't reproduce on my machines. EAGAIN with stdout usually means some program in process tree set non-blocking mode on the inherited file descriptor. maybe cargo does something like that if -j1 passed, will try poking it.
I have the same issue with www-client/firefox-68.1.0. The workaround with MAKEOPTS="-j2 -l1" helped.
*** Bug 695874 has been marked as a duplicate of this bug. ***
Only MAKEOPTS="-j1" is broken on arm, aarch64 and even amd64. If you can afford it memorywise, go for MAKEOPTS="-j2", it should work.
*** Bug 696920 has been marked as a duplicate of this bug. ***
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=dfda5d7a4371ec0a092cd8c37ccf0035318bc503 commit dfda5d7a4371ec0a092cd8c37ccf0035318bc503 Author: Thomas Deutschmann <whissi@gentoo.org> AuthorDate: 2019-10-08 21:09:36 +0000 Commit: Thomas Deutschmann <whissi@gentoo.org> CommitDate: 2019-10-08 21:13:57 +0000 mail-client/thunderbird: force at least MAKEOPTS=-j2 Bug: https://bugs.gentoo.org/687028 Package-Manager: Portage-2.3.76, Repoman-2.3.17 Signed-off-by: Thomas Deutschmann <whissi@gentoo.org> mail-client/thunderbird/thunderbird-60.9.0.ebuild | 8 +++++++- mail-client/thunderbird/thunderbird-68.1.1.ebuild | 8 +++++++- 2 files changed, 14 insertions(+), 2 deletions(-) https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=97644d196d7bd5dba9ab1526df934648b281750d commit 97644d196d7bd5dba9ab1526df934648b281750d Author: Thomas Deutschmann <whissi@gentoo.org> AuthorDate: 2019-10-08 21:06:26 +0000 Commit: Thomas Deutschmann <whissi@gentoo.org> CommitDate: 2019-10-08 21:13:55 +0000 www-client/firefox: force at least MAKEOPTS=-j2 Bug: https://bugs.gentoo.org/687028 Package-Manager: Portage-2.3.76, Repoman-2.3.17 Signed-off-by: Thomas Deutschmann <whissi@gentoo.org> www-client/firefox/firefox-60.9.0.ebuild | 9 ++++++++- www-client/firefox/firefox-68.1.0.ebuild | 8 +++++++- www-client/firefox/firefox-69.0.2.ebuild | 8 +++++++- 3 files changed, 22 insertions(+), 3 deletions(-)
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/proj/mozilla.git/commit/?id=7145ef907d11a39d6cd06a239d06040ddbdeb2ea commit 7145ef907d11a39d6cd06a239d06040ddbdeb2ea Author: Thomas Deutschmann <whissi@gentoo.org> AuthorDate: 2019-10-08 21:16:28 +0000 Commit: Thomas Deutschmann <whissi@gentoo.org> CommitDate: 2019-10-08 21:16:28 +0000 www-client/firefox: force at least MAKEOPTS=-j2 Bug: https://bugs.gentoo.org/687028 Package-Manager: Portage-2.3.76, Repoman-2.3.17 Signed-off-by: Thomas Deutschmann <whissi@gentoo.org> www-client/firefox/firefox-70.0_beta12.ebuild | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
I got a tip over here: https://github.com/rust-lang/rust/issues/57632#issuecomment-546614184 that this could be solved by the patch in https://bugzilla.mozilla.org/show_bug.cgi?id=1500436 you might want to try?
I've started a test emerge, and got over the point of failure. So this most likely is your fix.
the fix is included from firefox-69.0 on, so the hack can be removed from >=firefox-70.0.
I annoyed a few people at upstream, they uplifted the fix to the upcoming firefox-68.3.0 ; can the reporter please state wether this is also fixing his problem?
Trying now.
glad it was found finally. re-assigning bug to mozilla guys, leaving rust@ in cc.
It's working for me with firefox-68.3.0, and did so too with >=firefox-70.0.1 so, can the hack be removed, please? I'm in need of -j1 on a low end machine.
The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=d2c5a3099cccc05a7e25c396b768d9bd1057b423 commit d2c5a3099cccc05a7e25c396b768d9bd1057b423 Author: Thomas Deutschmann <whissi@gentoo.org> AuthorDate: 2019-12-02 18:55:27 +0000 Commit: Thomas Deutschmann <whissi@gentoo.org> CommitDate: 2019-12-02 18:55:42 +0000 www-client/firefox: dropping -j1 workaround Closes: https://bugs.gentoo.org/687028 Package-Manager: Portage-2.3.80, Repoman-2.3.19 Signed-off-by: Thomas Deutschmann <whissi@gentoo.org> www-client/firefox/firefox-68.3.0.ebuild | 8 +------- www-client/firefox/firefox-71.0.ebuild | 8 +------- 2 files changed, 2 insertions(+), 14 deletions(-)
