Hi, git maintainers:
Paul Gevers writes ("Bug#1146518: git breaks dgit autopkgtest"):
> With a recent upload of git the autopkgtest of dgit fails in testing
> when that autopkgtest is run with the binary packages of git from
> unstable. It passes when run with only packages from testing. In tabular
> form:
I looked at the logs and saw this
1859s remote: %apiquery dgit -dtest-dummy $DGIT_TEST_OPTS -D archive-api-query
suites
1859s remote:
1859s remote: dgit: error: On entry to dgit, SIGPIPE is set to something other
than SIG_DFL
1859s remote: dgit: This is a bug produced by something in your execution
environment.
1859s remote: dgit: Giving up.
I haven't investigated properly, but I think this means git is
newly invoking hook scripts with SIGPIPE ignored. If I am right, that
would IMO be a regression bug in git. (Executing a nonconsenting
program with non-default signal configuration is, in general, a bug.
This margin is too small to deal with the edge cases but I don't think
this is one of them.)
I'm travelling and it's not easy for me to investigate this properly
now, but I hope this guess is helpful.
Ian.
--
Ian Jackson <[email protected]> These opinions are my own.
Pronouns: they/he. If I emailed you from @fyvzl.net or @evade.org.uk,
that is a private address which bypasses my fierce spamfilter.
