Summary: | dev-cpp/folly-2023.06.19.00 fails tests on arm64: SIGSEGV with code: address not mapped to object | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | matoro <matoro_gentoo> |
Component: | Current packages | Assignee: | sin-ack <sin-ack> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | arm64, matoro_gentoo, proxy-maint, sin-ack |
Priority: | Normal | Keywords: | PullRequest, TESTFAILURE |
Version: | unspecified | ||
Hardware: | ARM64 | ||
OS: | Linux | ||
See Also: |
https://github.com/gentoo/gentoo/pull/39294 https://github.com/gentoo/gentoo/pull/39296 |
||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 905054 | ||
Attachments: |
build.log.gz
LastTest.log.gz build.log and emerge --info Fix build on arm64 build.log and emerge --info |
Description
matoro
2023-12-09 20:45:44 UTC
Created attachment 878600 [details]
build.log.gz
Created attachment 878601 [details]
LastTest.log.gz
Can you reproduce this with folly-2024.11.04.00? I ran the tests inside a QEMU chroot and beyond a couple timing-related failures (caused by QEMU) the tests pass. Created attachment 908603 [details] build.log and emerge --info (In reply to sin-ack from comment #3) > Can you reproduce this with folly-2024.11.04.00? I ran the tests inside a > QEMU chroot and beyond a couple timing-related failures (caused by QEMU) the > tests pass. It's quite worse, that version just straight-up fails to compile for me. Log attached. Created attachment 908604 [details, diff] Fix build on arm64 (In reply to matoro from comment #4) > It's quite worse, that version just straight-up fails to compile for me. > Log attached. I hit that issue myself, but forgot to mention it, sorry. I have an -r1 in a branch I'm working on. I've attached a patch that should make the build go through. Created attachment 908615 [details] build.log and emerge --info (In reply to sin-ack from comment #5) > Created attachment 908604 [details, diff] [details, diff] > Fix build on arm64 > > (In reply to matoro from comment #4) > > It's quite worse, that version just straight-up fails to compile for me. > > Log attached. > > I hit that issue myself, but forgot to mention it, sorry. I have an -r1 in a > branch I'm working on. I've attached a patch that should make the build go > through. After that patch, looks like only timeouts remain. (In reply to matoro from comment #6) > After that patch, looks like only timeouts remain. This is on real hardware, correct? Could you do `ctest -R '(futures_retrying_test.RetryingTest.largeRetries|io_async_hh_wheel_timer_test.HHWheelTimerTest.CancelTimeout|concurrent_skip_list_test.ConcurrentSkipList.ConcurrentAdd)' --output-on-errors` in the build directory a couple times to see if they're flaky? We could report and skip them, I suppose, but it's weird that all these timing issues happen on arm64. The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=0856b6850c6326950e69d8490f10307b38a29f59 commit 0856b6850c6326950e69d8490f10307b38a29f59 Author: sin-ack <sin-ack@protonmail.com> AuthorDate: 2024-11-13 00:05:16 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2024-11-23 14:00:57 +0000 dev-cpp/folly: skip some tests on arm64 Only the tests that were found to be failing on hardware are skipped. Closes: https://bugs.gentoo.org/919669 Signed-off-by: sin-ack <sin-ack@protonmail.com> Signed-off-by: Sam James <sam@gentoo.org> dev-cpp/folly/folly-2024.11.04.00-r1.ebuild | 10 ++++++++++ 1 file changed, 10 insertions(+) |