Bug#1011322: waypipe: ftbfs on riscv64 (Timeout: 2)
Bo YU writes: > On Fri, May 20, 2022 at 02:35:45PM +0200, Gard Spreemann wrote: >>Hello, and thanks you the bug report and patch. >> >>Bo YU writes: >> >>> […] >>> >>> The attached patch is simple to disable tests like mips do that. >>> >>> If you need me to do more test, please let me know. >> >>I believe upstream has made changes that look like they might deal with >>the test failures on all architectures. From >>https://gitlab.freedesktop.org/mstoeckl/waypipe/ it looks like commits >> >> ffd2d4379e481372ca6240f4b3bcedbd943dec4c >> >> 15aab5c344636336f5327d91d3e6a92733503312 >> >>might do the trick (on all architectures). I was waiting for upstream to >>make a new release with those changes, but they haven't yet. If you >>could cherry-pick those two commits and see if they fix the bug on >>riscv64, then that would be most appreciated. Thanks! > > Ok, very happy to do it:) > > But there is some trickes for me: how to cherry-pick these commits > into package that is come from `apt source`? > > One way is put the two commits into one patch and apply it into > d/patches? I think the way you're suggesting is a good idea. That's anyway how I'd represent the commits when applying the fix to the Debian package. Thanks! PS: This shouldn't be a high priority. At some point, upstream will surely release a new version with these fixes in place, obsoleting the patch :-) Best, Gard signature.asc Description: PGP signature
Bug#1011322: waypipe: ftbfs on riscv64 (Timeout: 2)
On Fri, May 20, 2022 at 02:35:45PM +0200, Gard Spreemann wrote: Hello, and thanks you the bug report and patch. Bo YU writes: […] The attached patch is simple to disable tests like mips do that. If you need me to do more test, please let me know. I believe upstream has made changes that look like they might deal with the test failures on all architectures. From https://gitlab.freedesktop.org/mstoeckl/waypipe/ it looks like commits ffd2d4379e481372ca6240f4b3bcedbd943dec4c 15aab5c344636336f5327d91d3e6a92733503312 might do the trick (on all architectures). I was waiting for upstream to make a new release with those changes, but they haven't yet. If you could cherry-pick those two commits and see if they fix the bug on riscv64, then that would be most appreciated. Thanks! Ok, very happy to do it:) But there is some trickes for me: how to cherry-pick these commits into package that is come from `apt source`? One way is put the two commits into one patch and apply it into d/patches? I build package with sbuild normally. Thank you~ Bo Best, Gard
Bug#1011322: waypipe: ftbfs on riscv64 (Timeout: 2)
Hello, and thanks you the bug report and patch. Bo YU writes: > Package: waypipe > Version: 0.8.2-3 > Severity: normal > Tags: ftbfs, patch > User: debian-ri...@lists.debian.org > Usertags: riscv64 > > Dear Maintainer, > > The waypipe package has a ftbfs issue on riscv64 arch: > > […] > > The attached patch is simple to disable tests like mips do that. > > If you need me to do more test, please let me know. I believe upstream has made changes that look like they might deal with the test failures on all architectures. From https://gitlab.freedesktop.org/mstoeckl/waypipe/ it looks like commits ffd2d4379e481372ca6240f4b3bcedbd943dec4c 15aab5c344636336f5327d91d3e6a92733503312 might do the trick (on all architectures). I was waiting for upstream to make a new release with those changes, but they haven't yet. If you could cherry-pick those two commits and see if they fix the bug on riscv64, then that would be most appreciated. Thanks! Best, Gard signature.asc Description: PGP signature
Bug#1011322: waypipe: ftbfs on riscv64 (Timeout: 2)
Package: waypipe Version: 0.8.2-3 Severity: normal Tags: ftbfs, patch User: debian-ri...@lists.debian.org Usertags: riscv64 Dear Maintainer, The waypipe package has a ftbfs issue on riscv64 arch: ``` Summary of Failures: 7/8 How well buffers are replicatedTIMEOUT40.09s killed by signal 15 SIGTERM 8/8 Whether diff operations successfully roundtrip TIMEOUT60.12s killed by signal 15 SIGTERM Ok: 6 Expected Fail: 0 Fail: 0 Unexpected Pass:0 Skipped:0 Timeout:2 dh_auto_test: error: cd obj-riscv64-linux-gnu && LC_ALL=C.UTF-8 MESON_TESTTHREADS=4 meson test returned exit code 2 make: *** [debian/rules:8: binary-arch] Error 25 dpkg-buildpackage: error: debian/rules binary-arch subprocess returned exit status 2 ``` The full buildd log is here: https://buildd.debian.org/status/fetch.php?pkg=waypipe&arch=riscv64&ver=0.8.2-3&stamp=1650239386&raw=0 The attached patch is simple to disable tests like mips do that. If you need me to do more test, please let me know. BR, Bo diff -Nru waypipe-0.8.2/debian/rules waypipe-0.8.2/debian/rules --- waypipe-0.8.2/debian/rules 2021-12-01 15:39:26.0 + +++ waypipe-0.8.2/debian/rules 2021-12-01 15:39:26.0 + @@ -16,7 +16,7 @@ dh_auto_configure -- -Dwith_avx2=false -Dwith_avx512f=false -Dwith_neon_opts=false -Dwith_sse3=false endif -ifneq (, $(filter $(DEB_HOST_ARCH),mipsel mips64el)) +ifneq (, $(filter $(DEB_HOST_ARCH),mipsel mips64el riscv64)) override_dh_auto_test: @echo "Skipping tests on $(DEB_HOST_ARCH)." endif