Summary: | sys-apps/openrc-0.4.3-r1 - src_test() - Missing hidden defs: rc_parse_service_state rc_service_state_names [ !! ] | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Jeroen Roovers (RETIRED) <jer> |
Component: | [OLD] baselayout | Assignee: | OpenRC Team <openrc> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | kentnl |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | HPPA | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
sys-apps:openrc-0.4.3-r1:20090214-002344.log [hppa,fails tests]
readelf -a /lib/librc.so readelf -a openrc-0.4.3/src/librc/librc.so readelf -Wr src/librc/*.o 0001-tests-filter-valid-parisc-relocs.patch |
Description
Jeroen Roovers (RETIRED)
2009-02-14 00:37:48 UTC
Created attachment 181918 [details]
sys-apps:openrc-0.4.3-r1:20090214-002344.log [hppa,fails tests]
I guess that's an hppa issue as both rc_parse_service_state and rc_service_state_names are static function and array respectively, so should not be in public API. post the output as an attachment of running `readelf -a` on librc.so Created attachment 181938 [details]
readelf -a /lib/librc.so
well the test runs on the unstripped version, so i'll need the readelf output on that file, not the installed stripped one ... I have something that looks like the same failure happening on amd64 here, I'm doing it under paludis though, not sure if that's relevant or not. There are some discrepancies in the failure cause though, mine also dies directly after is_older_than , but *doesn't* have the * Missing hidden defs: rc_parse_service_state rc_service_state_names part in the output. But to me, the fact he gets as far as *running* the is_older_than test indicates the preceding failures were not deemed 'failures' by the test script. Also, this may be relevant, but I'm building on a tmpfs. if you dont see the hidden errors, then your issue is unrelated (In reply to comment #7) > if you dont see the hidden errors, then your issue is unrelated > Right you are, I didn't notice -r1 was available, and it solved my issue. I didn't search enough the first time. :< ( was #258688 ) Created attachment 182016 [details]
readelf -a openrc-0.4.3/src/librc/librc.so
ok, the test is looking at the .o objects ... post the output of doing: readelf -Wr src/librc/*.o Created attachment 182321 [details]
readelf -Wr src/librc/*.o
this has to do with pic handling on hppa ... it isnt as efficient as other arches the symbols in question are fine, but the grep assumes a clean link (which hppa cannot provide currently) Is openrc still failing tests on hppa? (In reply to comment #13) > Is openrc still failing tests on hppa? Yes, all the way up to 0.7.0 with the same error. Created attachment 261669 [details, diff]
0001-tests-filter-valid-parisc-relocs.patch
guess we'll just have to filter the valid relocs. try this patch.
oh well, committed that now |