[Debian-med-packaging] Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)

[Chris Chang]

Just a heads-up, today's release contains a fix for an
infinite-loop-instead-of-error bug in the previous release.

On Sun, Feb 20, 2022 at 12:36 PM Andreas Tille <tille at debian.org> wrote:

> Cool!  This finally works.  Thanks a lot for your patience, Andreas.
>
> Am Sun, Feb 20, 2022 at 09:43:10AM -0800 schrieb Chris Chang:
> > Hmm.  Ordinarily, that download link would be fine.  I had not updated it
> > this time because the issue only affects gcc 11+, so it did not affect
> any
> > of the precompiled binaries and I wanted to keep the posted source
> > identical to that used to compile those binaries.
> >
> > On the other hand, the only difference (in the non-gcc-11+ case) between
> > the binaries and the updated source is the version date.  I guess this is
> > insignificant enough that a temporary discordance is less disruptive than
> > making you change your packing process, so I've updated the website
> source
> > link to point to the newer code; let me know if you still run into any
> > problems.
> >
> > On Sat, Feb 19, 2022 at 11:53 PM Andreas Tille <andreas at an3as.eu> wrote:
> >
> > > Hi Chris,
> > >
> > > Am Sat, Feb 19, 2022 at 11:37:34PM -0800 schrieb Chris Chang:
> > > > To elaborate: when I look at
> > > > https://salsa.debian.org/med-team/plink2/-/tree/master , I can see
> that
> > > > plink2.cc, include/plink2_base.h, and include/plink2_base.cc don't
> have
> > > the
> > > > same contents as the v2.00a3-20220218 GitHub tag/release.  So I don't
> > > know
> > > > where you grabbed the code from this time, but it does not appear to
> be
> > > > from GitHub.
> > >
> > > That's true since on Github the source code of plink1.9 and plink2 is
> not
> > > separated.  Thus I download the tarball from
> > >
> > >    https://www.cog-genomics.org/plink/2.0/
> > >
> > > which points to
> > >
> > >    https://www.cog-genomics.org/static/bin/plink2_src_220217.zip
> > >
> > > I confirm that the tag on Github says v2.00a3-20220218 which seems to
> be
> > > later than 220217.  In principle I could point the packaging metadata
> to
> > > Github and exclude everything that I do not need for plink2 packaging
> (so
> > > remove 1.9 directory and code copies of libdeflate, simde and zstd in
> 2.0)
> > >
> > > In case you consider this the safer way to download plink2 code I'll
> do so.
> > >
> > > Kind regards and thanks a lot for comparing the code base
> > >
> > >      Andreas.
> > >
> > > > On Sat, Feb 19, 2022 at 7:54 AM Chris Chang <chrchang523 at gmail.com>
> > > wrote:
> > > >
> > > > > If you build off the current head, the version string should say
> "18
> > > Feb
> > > > > 2022", not "17 Feb 2022".
> > > > >
> > > > > On Sat, Feb 19, 2022 at 7:02 AM Andreas Tille <andreas at an3as.eu>
> > > wrote:
> > > > >
> > > > >> Hi again,
> > > > >>
> > > > >> Am Sat, Feb 19, 2022 at 12:28:21AM -0800 schrieb Chris Chang:
> > > > >> > Ok, new release/tag created.
> > > > >>
> > > > >> I've built the new version now.  The bad news is that it keeps on
> > > > >> SEGFAULTing for me:
> > > > >>
> > > > >> ...
> > > > >> (gdb) run
> > > > >> Starting program: /usr/lib/plink2/plink2-sse3 --debug --pfile
> tmp_data
> > > > >> --export vcf vcf-dosage=DS --out tmp_data2
> > > > >> [Thread debugging using libthread_db enabled]
> > > > >> Using host libthread_db library
> > > "/lib/x86_64-linux-gnu/libthread_db.so.1".
> > > > >> [New Thread 0x7ffff4cc7640 (LWP 985019)]
> > > > >> [New Thread 0x7ffff44c6640 (LWP 985020)]
> > > > >> [New Thread 0x7fffebcc5640 (LWP 985021)]
> > > > >> PLINK v2.00a3 64-bit (17 Feb 2022)
> > > > >> www.cog-genomics.org/plink/2.0/
> > > > >> (C) 2005-2022 Shaun Purcell, Christopher Chang   GNU General
> Public
> > > > >> License v3
> > > > >> Logging to tmp_data2.log.
> > > > >> Options in effect:
> > > > >>   --debug
> > > > >>   --export vcf vcf-dosage=DS
> > > > >>   --out tmp_data2
> > > > >>   --pfile tmp_data
> > > > >>
> > > > >> Start time: Sat Feb 19 16:00:00 2022
> > > > >> 31998 MiB RAM detected; reserving 15999 MiB for main workspace.
> > > > >> Using up to 4 compute threads.
> > > > >> [New Thread 0x7ffff7fc5640 (LWP 985022)]
> > > > >>
> > > > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation
> fault.
> > > > >> plink2::LoadPsam (psamname=psamname at entry=0x7fffffffbe60
> > > > >> "tmp_data.psam", pheno_range_list_ptr=<optimized out>,
> fam_cols=...,
> > > > >> pheno_ct_max=<optimized out>,
> > > > >>     missing_pheno=<optimized out>, affection_01=0,
> max_thread_ct=4,
> > > > >> piip=0x7fffffff8870, sample_include_ptr=0x7fffffff8790,
> > > > >> founder_info_ptr=0x7fffffff87a8, sex_nm_ptr=0x7fffffff8798,
> > > > >>     sex_male_ptr=0x7fffffff87a0, pheno_cols_ptr=0x7fffffff8770,
> > > > >> pheno_names_ptr=0x7fffffff8780, raw_sample_ct_ptr=0x7fffffff8728,
> > > > >> pheno_ct_ptr=0x7fffffff8720,
> > > > >>     max_pheno_name_blen_ptr=0x7fffffff87b0) at
> ../plink2_psam.cc:611
> > > > >> 611             pheno_cols[pheno_idx].nonmiss = nullptr;
> > > > >>
> > > > >>
> > > > >> Please let me know if I can do further checks.
> > > > >>
> > > > >> Kind regards
> > > > >>
> > > > >>       Andreas.
> > > > >>
> > > > >> --
> > > > >> http://fam-tille.de
> > > > >>
> > > > >
> > >
> > > --
> > > http://fam-tille.de
> > >
>
> --
> http://fam-tille.de
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/debian-med-packaging/attachments/20220315/9632d55d/attachment.htm>