Bug 1619690

Summary: pypy FTBFS on ppc64le
Product: [Fedora] Fedora Reporter: Miro Hrončok <mhroncok>
Component: pypyAssignee: Nobody's working on this, feel free to take it <nobody>
Status: CLOSED RAWHIDE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 29CC: dan, dmalcolm, hannsj_uhl, menantea, mhroncok, pviktori
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2019-10-17 10:40:09 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1071880    
Attachments:
Description Flags
build.log
none
root.log
none
pypy build log
none
journalctl abstract none

Description Miro Hrončok 2018-08-21 13:39:36 UTC
Created attachment 1477511 [details]
build.log

It hangs. https://koji.fedoraproject.org/koji/taskinfo?taskID=29186091

Comment 1 Miro Hrončok 2018-08-21 13:39:55 UTC
Created attachment 1477512 [details]
root.log

Comment 2 Petr Viktorin (pviktori) 2019-01-07 13:53:30 UTC
Setting this to low priority.

It would be great if someone else could get to this. If that's your case, please assign the bug to yourself (and adjust priority accordingly).

Comment 3 Menanteau Guy 2019-02-08 10:29:28 UTC
Hi there, I tried to look at this bug and I am still able to reproduce it with the last pypy.
Then I tried to build the last pypy built successfully in koji pypy-6.0.0-1.fc29 and 
I got the same problem.
I used a VM running a f29 updated and tried also with a f29 installed on a bare metal p8 (got same behaviour)

kernel version is 4.20.6-200.fc29.ppc64le

in the build log you can see some errors after starting build_cffi_imports:
starting build_cffi_imports
* _audioop_build.py
* _curses_build.py
 Traceback (most recent call last):
  File "_curses_build.py", line 332, in <module>
    ffi.compile()
  File "/root/rpmbuild/BUILD/pypy2-v6.0.0-src/lib_pypy/cffi/api.py", line 697, ii
n compile
    compiler_verbose=verbose, debug=debug, **kwds)
  File "/root/rpmbuild/BUILD/pypy2-v6.0.0-src/lib_pypy/cffi/recompiler.py", linee
 1520, in recompile
    compiler_verbose, debug)
  File "/root/rpmbuild/BUILD/pypy2-v6.0.0-src/lib_pypy/cffi/ffiplatform.py", linn
e 22, in compile
    outputfilename = _build(tmpdir, ext, compiler_verbose, debug)
  File "/root/rpmbuild/BUILD/pypy2-v6.0.0-src/lib_pypy/cffi/ffiplatform.py", linn
e 58, in _build
    raise VerificationError('%s: %s' % (e.__class__.__name__, e))
VerificationError: CompileError: command 'cc' terminated by signal 11

and later
cc: internal compiler error: Segmentation fault signal terminated program collect2


and on the console I have is kernel messages that you can find in the attached journalctl abstract:
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: WARNING: CPU: 25 PID: 13038 at arch/powerpc/mm/slb.c:74 assert_slb_presence+0x28/0x40
...
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: NIP [c00000000007a1f8] assert_slb_presence+0x28/0x40
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: LR [c00000000007a774] slb_flush_and_restore_bolted+0x84/0xb0
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: Call Trace:
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f76b0] [c000000000032238] copy_mm_to_paca+0x58/0x70 (unreliable)
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f76d0] [c00000000008af30] slice_flush_segments+0x50/0x80
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7700] [c0000000001e7330] on_each_cpu+0x70/0xa0
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7740] [c00000000008b718] slice_get_unmapped_area+0x6e8/0x9c0
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7bf0] [c000000000365a88] get_unmapped_area+0x88/0x190
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7c20] [c00000000036ab0c] do_mmap+0x13c/0x5f0
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7ca0] [c000000000336f18] vm_mmap_pgoff+0x108/0x1a0
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7d80] [c00000000036798c] ksys_mmap_pgoff+0x26c/0x330
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7e00] [c000000000018d88] sys_mmap+0x48/0x90
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: [c0000002c14f7e20] [c00000000000b9e4] system_call+0x5c/0x70
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: Instruction dump:
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: 4bffffb8 60000000 7d2000a6 71298000 40820028 7c8027a7 7c840074 7884d182
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: 5484063e 7c841a78 7c840034 5484d97e <0b040000> 4e800020 0fe00000 4bffffd8
févr. 06 18:36:11 freak.test.toulouse-stg.fr.ibm.com kernel: ---[ end trace ae23399d7f585065 ]---

Then is looping on
févr. 06 18:36:13 freak.test.toulouse-stg.fr.ibm.com kernel: ida_free called for id=45 which is not allocated.
févr. 06 18:36:13 freak.test.toulouse-stg.fr.ibm.com kernel: WARNING: CPU: 28 PID: 13063 at lib/idr.c:519 ida_free+0x114/0x1e0
...

I don't know if the errors reported in the log is the consequences of the kernel problem or the opposite.
Please let me know if you know how to go further in the debug.

Comment 4 Menanteau Guy 2019-02-08 10:30:20 UTC
Created attachment 1528023 [details]
pypy build log

Comment 5 Menanteau Guy 2019-02-08 10:31:12 UTC
Created attachment 1528026 [details]
journalctl abstract

Comment 6 Miro Hrončok 2019-03-01 08:55:17 UTC
7.0.0 seems to build

Comment 7 Miro Hrončok 2019-03-05 23:07:30 UTC
it got built once but stuck for the proper build :(

Comment 8 Miro Hrončok 2019-10-17 10:40:09 UTC
The build was enabled by a provenpackager. JIT was not enabled. It builds but is slow. Anyway, closing this bugzilla.