Re: [PATCH v2 0/6] ci: replace our Azure Pipeline by GitHub Actions
- From
Danh Doan <congdanhqx@gmail.com>
- Date
- Apr 2, 2020, 00:14 UTC
- Message-ID
- <20200402001408.GA2764@danh.dev>
- In-Reply-To
- <xmqq7dyy6gui.fsf@gitster.c.googlers.com>
On 2020-04-01 14:23:01-0700, Junio C Hamano <gitster@pobox.com> wrote:
Show 37 quoted lines
> Đoàn Trần Công Danh <congdanhqx@gmail.com> writes: > > > This series is based on work started by Dscho, > > I started to work with this series since there're merge conflicts > > with my series at dd/ci-musl-lib, and Dscho said he was busy for a while. > > > > Changes frome Dscho's version: > > * Based on dd/ci-musl-libc > > * Move artifact extraction out of Test phase of windows-test > > * Move ci/print-test-failures.sh out of build-and-test phase > > * set TERM environment variable if not exist > > * add linux-musl job > > * v1 doesn't report failure on linux32, fixed > > * run linux32 directly inside container > > * install development file of curl in documentation job because "curl-config" > > will be called in pu's Makefile > > + Other approach could be call make CURL_CONFIG=: in test-documentation.sh > > > > Sample run: > > * of this series: https://github.com/sgn/git/actions/runs/68291472 > > * pretend compile failure: https://github.com/sgn/git/actions/runs/68292112 > > * pretend test failure: https://github.com/sgn/git/actions/runs/68293056 > > * merged to pu: https://github.com/sgn/git/actions/runs/68301122 > > When I look at > > https://github.com/git/git/actions > > it appears that every time I push[*1*], we are running _two_ sets of > the same tests, one says "Pu nu bu?" and the other says "CI/PR". > > In addition, the former refers to "Pull request #nnn synchronize by > gitster", with #nnn part made clickable, but clicking on it results > in 404 page, as there is no such pull request. > > Does this series need some help from me (i.e. changing my workflow) > to make the CI tests run more smoothly?
I suspect there is a github Pull Request #738 (merging pu to master) in private mode because it's what reported by Github actions.
It seems like the last 2 pushes to pu doesn't trigger it anymore?
If the problem still persists, I think either you or Dscho can login to github and see what is there.
-- Danh