| Summary: | sys-apps/portage-2.2.18: portage fails with AssertionError on emerge with some prebuilt packages (dev-libs/intel-common, dev-lang/icc) | ||
|---|---|---|---|
| Product: | Portage Development | Reporter: | Peter Foley <pefoley2> |
| Component: | Unclassified | Assignee: | Portage team <dev-portage> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | amadio, bircoph, gentoo, jkt, mozestygart, powerman-asdf, rose, tsigarid, vapier |
| Priority: | Normal | Keywords: | InVCS, REGRESSION |
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Bug Depends on: | |||
| Bug Blocks: | 694246, 484436 | ||
| Attachments: | Emerge log | ||
|
Description
Peter Foley
2015-03-19 20:14:25 UTC
Created attachment 399272 [details]
Emerge log
Emerge output from failure.
The package emerges successfully with portage 2.2.14. Works with sys-apps/portage-2.2.17. Without any background knowledge and a bit of google usage, the following commit f1c1b8a77eebf7713b32e5f9945690f60f4f46de Generate soname dependency metadata (bug 282639) seems to introduce the problem. This is where I deem it better to leave this to more capable hands and people who know what that change is about. Keep up the good work and have fun. :-) It looks like it's installing some odd binaries that are probably not even intended to execute on your processor. I'll fix it to generate a "QA Notice" in this case. There's a patch in the following branch: https://github.com/zmedico/portage/tree/bug_543818 I've posted it for review here: http://thread.gmane.org/gmane.linux.gentoo.portage.devel/5385 With portage installed from the bug_543818 branch, the build succeeds with the below output:
* QA Notice: Unrecognized ELF file(s):
*
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libpdbx.so.5;libpdbx.so.5;;libpthread.so.0,libstdc++.so.6,libm.so.6,libgcc_s.so.1,libc.so.6,ld-linux-k1om.so.2,libdl.so.2;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libsvml.so;libsvml.so;;libintlc.so.5,libc.so.6;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libirc.so;libirc.so;;;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libintlc.so.5;libintlc.so.5;;;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libirng.so;;;libintlc.so.5,libc.so.6;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/offload_main;;;libimf.so,liboffload.so.5,libcoi_device.so.0,libmyo-service.so,libm.so.6,libiomp5.so,libstdc++.so.6,libgcc_s.so.1,libpthread.so.0,libc.so.6,libdl.so.2;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/liboffload.so.5;liboffload.so.5;;libcoi_device.so.0,libmyo-service.so,libimf.so,libsvml.so,libirng.so,libm.so.6,libiomp5.so,libstdc++.so.6,libgcc_s.so.1,libintlc.so.5,libpthread.so.0,libc.so.6,ld-linux-k1om.so.2,libdl.so.2;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libiompstubs5.so;libiompstubs5.so;;libc.so.6;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libicaf.so;libicaf.so;;libm.so.6,libmpi_mt.so.4,libintlc.so.5,libc.so.6;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libiomp5.so;libiomp5.so;;libpthread.so.0,libc.so.6,libdl.so.2;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libcilkrts.so.5;libcilkrts.so.5;;libpthread.so.0,libm.so.6,libstdc++.so.6,libgcc_s.so.1,libc.so.6,libdl.so.2;
* NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libimf.so;libimf.so;;libc.so.6;
*
>>> Installing (1 of 1) dev-libs/intel-common-15.0.2.164::gentoo
* NOWN_TYPE;/opt/intel/co
looks like a big typo
KNOWN_TYPE
;)
But I don't see that as having come from Zac's patch
(In reply to Brian Dolbec from comment #7) > * NOWN_TYPE;/opt/intel/co > > looks like a big typo > > KNOWN_TYPE > > ;) > > But I don't see that as having come from Zac's patch That's scanelf output. It's actuallly UNKNOWN_TYPE, missing the first 3 characters. Normally, the first 3 characters are always "EM_", so they are trimmed off (the NEEDED.ELF.2 file format has always been this way). (In reply to Peter Foley from comment #6) > With portage installed from the bug_543818 branch, the build succeeds with > the below output: > > * QA Notice: Unrecognized ELF file(s): > * > * > NOWN_TYPE;/opt/intel/composerxe-2015_update2.2.164/compiler/lib/mic/libpdbx. > so.5;libpdbx.so.5;;libpthread.so.0,libstdc++.so.6,libm.so.6,libgcc_s.so.1, > libc.so.6,ld-linux-k1om.so.2,libdl.so.2; Some flawed logic in the patch is fixed now, so the warning for this package will be suppressed because it sets QA_PREBUILT="${INTEL_SDP_DIR}/*" in intel-sdb.eclass. This is in the master branch now: http://gitweb.gentoo.org/proj/portage.git/commit/?id=c9aebae4633a972b78f7f999aa755fa6d25c8451 *** Bug 544274 has been marked as a duplicate of this bug. *** *** Bug 544974 has been marked as a duplicate of this bug. *** (In reply to Zac Medico from comment #10) > This is in the master branch now: > > http://gitweb.gentoo.org/proj/portage.git/commit/ > ?id=c9aebae4633a972b78f7f999aa755fa6d25c8451 How can I install the patched portage? (In reply to Juergen Rose from comment #13) > How can I install the patched portage? ACCEPT_KEYWORDS="**" emerge -1 =sys-apps/portage-9999 (In reply to Zac Medico from comment #14) > (In reply to Juergen Rose from comment #13) > > How can I install the patched portage? > > ACCEPT_KEYWORDS="**" emerge -1 =sys-apps/portage-9999 Thanks, worked. (In reply to Juergen Rose from comment #15) > (In reply to Zac Medico from comment #14) > > (In reply to Juergen Rose from comment #13) > > > How can I install the patched portage? > > > > ACCEPT_KEYWORDS="**" emerge -1 =sys-apps/portage-9999 > > Thanks, worked. BTW., when can I return to the normal portage version? *** Bug 547036 has been marked as a duplicate of this bug. *** if a new version isn't slated soon, probably should cherry pick this into an -r1 I'm thinking next weekend. I want to see if I can make a few changes for the squashfs sync module's release. Rollback from portage-2.2.18 to 2.2.14 solve this issue. Released in portage-2.2.19 Just fell on this bug with portage 2.2.18. Is this bug really fixed, or only on ~ versions? (In reply to Kostas from comment #22) > Just fell on this bug with portage 2.2.18. Is this bug really fixed, or only > on ~ versions? Yes, it's only fixed in ~ versions (greater than 2.2.18). *** Bug 553216 has been marked as a duplicate of this bug. *** |
