Summary: | sys-devel/automake-1.16-r1 fails with -j>1: help2man: can't get `--help' info from aclocal-1.16 | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Alex Xu (Hello71) <alex_y_xu> |
Component: | Current packages | Assignee: | Gentoo's Team for Core System packages <base-system> |
Status: | UNCONFIRMED --- | ||
Severity: | normal | CC: | bkohler, lperkins |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
build log.
emerge --info build.log using -j31 |
Description
Alex Xu (Hello71)
2018-03-11 17:54:15 UTC
Please attach build.log and `emerge --info` output. Created attachment 568902 [details]
build log.
Created attachment 568914 [details]
emerge --info
Just ran into this on my system. As far as I can tell it's a race condition in the automake build process. I'm not sure exactly where though. After discussion on IRC I am going to drop "CONFIRMED" status. We still don't have a reproducer. Ben noticed that > ln bin/aclocal bin/aclocal-1.16 and > ln bin/automake bin/automake-1.16 are missing in attached build.log which looks like > install-exec-hook: > @$(POST_INSTALL) > @for p in $(bin_SCRIPTS); do \ > f=`echo $$p | sed -e 's,.*/,,' -e '$(transform)'`; \ > fv="$$f-$(APIVERSION)"; \ > rm -f "$(DESTDIR)$(bindir)/$$fv"; \ > echo " $(LN) '$(DESTDIR)$(bindir)/$$f' '$(DESTDIR)$(bindir)/$$fv'"; \ > $(LN) "$(DESTDIR)$(bindir)/$$f" "$(DESTDIR)$(bindir)/$$fv"; \ > done from $src/bin/local.mk didn't run for some reason... Created attachment 568918 [details]
build.log using -j31
maybe it would be worth adding a check if make fails and if so, print a warning to retry the build with instructions to post build log to this bug? pressed submit too early. I think it's probably not worth forcing -j1 for the apparent majority of users that have no problem, even if the difference is only a few seconds. I can confirm the failure and the fix (MAKEOPTS=-j1) on one of my system. |