Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Am Tue, Mar 15, 2022 at 09:01:12AM -0700 schrieb Chris Chang: > Just a heads-up, today's release contains a fix for an > infinite-loop-instead-of-error bug in the previous release. Thanks a lot for the headsup - just uploaded the updated version to Debian. > On Sun, Feb 20, 2022 at 12:36 PM Andreas Tille 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 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 > > > > 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 > > > > 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 0x74cc7640 (LWP 985019)] > > > > > >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > > > > > >> > > > > > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation > > fault. > > > > > >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 > > > > > >> "tmp_data.psam", pheno_range_list_ptr=, > > fam_cols=..., > > > > > >> pheno_ct_max=, > > > > > >> missing_pheno=, affection_01=0, > > max_thread_ct=4, > > > > > >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, > > > > > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > > > > > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > > > > > >> pheno_names_ptr=0x7fff8780,
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 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 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 > > > 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 > > > 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 0x74cc7640 (LWP 985019)] > > > > >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > > > > >> > > > > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation > fault. > > > > >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 > > > > >> "tmp_data.psam", pheno_range_list_ptr=, > fam_cols=..., > > > > >> pheno_ct_max=, > > > > >> missing_pheno=, affection_01=0, > max_thread_ct=4, > > > > >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, > > > > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > > > > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > > > > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > > > > >> pheno_ct_ptr=0x7fff8720, > > > > >> max_pheno_name_blen_ptr=0x7fff87b0) 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. > > > >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 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 > > 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 > > 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 0x74cc7640 (LWP 985019)] > > > >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > > > >> > > > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. > > > >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 > > > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > > > >> pheno_ct_max=, > > > >> missing_pheno=, affection_01=0, max_thread_ct=4, > > > >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, > > > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > > > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > > > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > > > >> pheno_ct_ptr=0x7fff8720, > > > >> max_pheno_name_blen_ptr=0x7fff87b0) 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
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 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 > 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 > 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 0x74cc7640 (LWP 985019)] > > >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > > >> > > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. > > >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 > > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > > >> pheno_ct_max=, > > >> missing_pheno=, affection_01=0, max_thread_ct=4, > > >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, > > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > > >> pheno_ct_ptr=0x7fff8720, > > >> max_pheno_name_blen_ptr=0x7fff87b0) 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 >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 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 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 0x74cc7640 (LWP 985019)] > >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > >> > >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. > >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > >> pheno_ct_max=, > >> missing_pheno=, affection_01=0, max_thread_ct=4, > >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > >> pheno_ct_ptr=0x7fff8720, > >> max_pheno_name_blen_ptr=0x7fff87b0) 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
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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. On Sat, Feb 19, 2022 at 7:54 AM Chris Chang 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 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 0x74cc7640 (LWP 985019)] >> [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] >> >> Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. >> plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., >> pheno_ct_max=, >> missing_pheno=, affection_01=0, max_thread_ct=4, >> piip=0x7fff8870, sample_include_ptr=0x7fff8790, >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, >> pheno_ct_ptr=0x7fff8720, >> max_pheno_name_blen_ptr=0x7fff87b0) 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 >> >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 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 0x74cc7640 (LWP 985019)] > [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] > > Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. > plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 "tmp_data.psam", > pheno_range_list_ptr=, fam_cols=..., pheno_ct_max= out>, > missing_pheno=, affection_01=0, max_thread_ct=4, > piip=0x7fff8870, sample_include_ptr=0x7fff8790, > founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > pheno_ct_ptr=0x7fff8720, > max_pheno_name_blen_ptr=0x7fff87b0) 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 >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
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 0x74cc7640 (LWP 985019)] [New Thread 0x744c6640 (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 0x77fc5640 (LWP 985022)] Thread 1 "plink2-sse3" received signal SIGSEGV, Segmentation fault. plink2::LoadPsam (psamname=psamname@entry=0x7fffbe60 "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., pheno_ct_max=, missing_pheno=, affection_01=0, max_thread_ct=4, piip=0x7fff8870, sample_include_ptr=0x7fff8790, founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, pheno_ct_ptr=0x7fff8720, max_pheno_name_blen_ptr=0x7fff87b0) 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
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Ok, new release/tag created. On Fri, Feb 18, 2022 at 10:55 PM Andreas Tille wrote: > Hi Chris, > > Am Fri, Feb 18, 2022 at 07:27:20PM -0800 schrieb Chris Chang: > > I have confirmed that the original code segfaults when compiled with gcc > > 11.2 on my Debian instance, and that it runs to completion after the > latest > > patch. I have also confirmed that, after the latest patch, all other > tests > > pass. > > Very nice. It would be great if you could tag a new release which would > ease the maintenance. > > Thanks a lot for your great cooperation > > Andreas. > > -- > http://fam-tille.de >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Hi Chris, Am Fri, Feb 18, 2022 at 07:27:20PM -0800 schrieb Chris Chang: > I have confirmed that the original code segfaults when compiled with gcc > 11.2 on my Debian instance, and that it runs to completion after the latest > patch. I have also confirmed that, after the latest patch, all other tests > pass. Very nice. It would be great if you could tag a new release which would ease the maintenance. Thanks a lot for your great cooperation Andreas. -- http://fam-tille.de
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
I have confirmed that the original code segfaults when compiled with gcc 11.2 on my Debian instance, and that it runs to completion after the latest patch. I have also confirmed that, after the latest patch, all other tests pass. On Fri, Feb 18, 2022 at 1:27 PM Chris Chang wrote: > I am installing gcc-11 on my Debian instance now, and will be running more > extensive tests today searching for other things that may have stopped > working for the same reason. > > On Fri, Feb 18, 2022 at 1:22 PM Andreas Tille wrote: > >> I confirm its gcc-11. I'll check tomorrow. Thanks a lot for your quick >> and helpful responses, Andreas. >> >> Am Fri, Feb 18, 2022 at 12:53:58PM -0800 schrieb Chris Chang: >> > I have posted an update under the provisional assumption that it's gcc >> 11's >> > new ipa-modref pass that is causing this code to fail, since it does >> seem >> > to break some similar code. >> > >> > On Fri, Feb 18, 2022 at 11:49 AM Chris Chang >> wrote: >> > >> > > What compiler version are you using? This implies that the pgl_malloc >> > > inline function is not being compiled to the expected code; there is >> an >> > > existing non-inlined version that is used for very old gcc versions, >> but it >> > > looks like it may also be needed here. >> > > >> > > On Fri, Feb 18, 2022 at 11:40 AM Andreas Tille >> wrote: >> > > >> > >> Hi again, >> > >> >> > >> I applied this patch and now I get: >> > >> >> > >> (gdb) run >> > >> Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] >> > >> [New Thread 0x7fffec4c6640 (LWP 4060798)] >> > >> [New Thread 0x7fffebcc5640 (LWP 4060799)] >> > >> PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 >> > >> 31998 MiB RAM detected; reserving 15999 MiB for main workspace. >> > >> Using up to 4 compute threads. >> > >> [New Thread 0x77fc5640 (LWP 4060800)] >> > >> sizeof(PhenoCol): 40 pheno_cols: 0 >> > >> --debug: setting pheno_cols[0].nonmiss. = nullptr >> > >> >> > >> Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. >> > >> 0x556fb82e in plink2::LoadPsam (psamname=psamname@entry >> =0x7fffbe70 >> > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., >> > >> pheno_ct_max=, >> > >> missing_pheno=, affection_01=0, max_thread_ct=4, >> > >> piip=0x7fff8880, sample_include_ptr=0x7fff8790, >> > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, >> > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, >> > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, >> > >> pheno_ct_ptr=0x7fff8720, >> > >> max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 >> > >> warning: Source file is more recent than executable. >> > >> 615 pheno_cols[pheno_idx].nonmiss = nullptr; >> > >> >> > >> Kind regards >> > >> >> > >> Andreas. >> > >> >> > >> Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: >> > >> > Ok, I don't know why that particular line would fail, but I've >> added >> > >> > another debug-print before it on GitHub. >> > >> > >> > >> > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille < >> andr...@fam-tille.de> >> > >> wrote: >> > >> > >> > >> > > Hi Chris, >> > >> > > >> > >> > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: >> > >> > > > I was unable to replicate this issue on a Debian EC2 instance. >> > >> However, >> > >> > > > there are very few things that happen between printing "End >> time:" >> > >> and >> > >> > > > program exit, so I have added a bunch of debug-prints (active >> when >> > >> the >> > >> > > > --debug flag is passed in) to the latest GitHub commit that >> should >> > >> reveal >> > >> > > > which of those few things is triggering the segfault; let me >> know >> > >> if you >> > >> > > > are able to run this build. >> > >> > > >> > >> > > I think the issue is a bit more complex. Debian provides a >> wrapper >> > >> > > which calls the best / most performant plink2. The issue seems >> to >> > >> > > occure for SFX=avx. First I do: >> > >> > > >> > >> > > >> > >> > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 >> > >> dosage-freq=0.1 >> > >> > > --out tmp_data >> > >> > > >> > >> > > This works. In the next step I fire up gdb then which results in >> > >> > > >> > >> > > >> > >> > > (gdb) run >> > >> > > Starting program: /usr/lib/plink2/plink2-avx --debug --pfile >> tmp_data >> > >> > > --export
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
I am installing gcc-11 on my Debian instance now, and will be running more extensive tests today searching for other things that may have stopped working for the same reason. On Fri, Feb 18, 2022 at 1:22 PM Andreas Tille wrote: > I confirm its gcc-11. I'll check tomorrow. Thanks a lot for your quick > and helpful responses, Andreas. > > Am Fri, Feb 18, 2022 at 12:53:58PM -0800 schrieb Chris Chang: > > I have posted an update under the provisional assumption that it's gcc > 11's > > new ipa-modref pass that is causing this code to fail, since it does seem > > to break some similar code. > > > > On Fri, Feb 18, 2022 at 11:49 AM Chris Chang > wrote: > > > > > What compiler version are you using? This implies that the pgl_malloc > > > inline function is not being compiled to the expected code; there is an > > > existing non-inlined version that is used for very old gcc versions, > but it > > > looks like it may also be needed here. > > > > > > On Fri, Feb 18, 2022 at 11:40 AM Andreas Tille > wrote: > > > > > >> Hi again, > > >> > > >> I applied this patch and now I get: > > >> > > >> (gdb) run > > >> Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] > > >> [New Thread 0x7fffec4c6640 (LWP 4060798)] > > >> [New Thread 0x7fffebcc5640 (LWP 4060799)] > > >> PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 > > >> 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > > >> Using up to 4 compute threads. > > >> [New Thread 0x77fc5640 (LWP 4060800)] > > >> sizeof(PhenoCol): 40 pheno_cols: 0 > > >> --debug: setting pheno_cols[0].nonmiss. = nullptr > > >> > > >> Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. > > >> 0x556fb82e in plink2::LoadPsam (psamname=psamname@entry > =0x7fffbe70 > > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > > >> pheno_ct_max=, > > >> missing_pheno=, affection_01=0, max_thread_ct=4, > > >> piip=0x7fff8880, sample_include_ptr=0x7fff8790, > > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > > >> pheno_ct_ptr=0x7fff8720, > > >> max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 > > >> warning: Source file is more recent than executable. > > >> 615 pheno_cols[pheno_idx].nonmiss = nullptr; > > >> > > >> Kind regards > > >> > > >> Andreas. > > >> > > >> Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: > > >> > Ok, I don't know why that particular line would fail, but I've added > > >> > another debug-print before it on GitHub. > > >> > > > >> > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille > > > >> wrote: > > >> > > > >> > > Hi Chris, > > >> > > > > >> > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > > >> > > > I was unable to replicate this issue on a Debian EC2 instance. > > >> However, > > >> > > > there are very few things that happen between printing "End > time:" > > >> and > > >> > > > program exit, so I have added a bunch of debug-prints (active > when > > >> the > > >> > > > --debug flag is passed in) to the latest GitHub commit that > should > > >> reveal > > >> > > > which of those few things is triggering the segfault; let me > know > > >> if you > > >> > > > are able to run this build. > > >> > > > > >> > > I think the issue is a bit more complex. Debian provides a > wrapper > > >> > > which calls the best / most performant plink2. The issue seems to > > >> > > occure for SFX=avx. First I do: > > >> > > > > >> > > > > >> > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 > > >> dosage-freq=0.1 > > >> > > --out tmp_data > > >> > > > > >> > > This works. In the next step I fire up gdb then which results in > > >> > > > > >> > > > > >> > > (gdb) run > > >> > > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] > > >> > > [New Thread 0x744c6640 (LWP 2931409)] > > >> > > [New Thread 0x7fffebcc5640 (LWP 2931411)] > > >> > > PLINK v2.00a3 SSE4.2 (29 Jan 2022) > > >> > > www.cog-genomics.org/plink/2.0/ > > >> > > (C)
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
I confirm its gcc-11. I'll check tomorrow. Thanks a lot for your quick and helpful responses, Andreas. Am Fri, Feb 18, 2022 at 12:53:58PM -0800 schrieb Chris Chang: > I have posted an update under the provisional assumption that it's gcc 11's > new ipa-modref pass that is causing this code to fail, since it does seem > to break some similar code. > > On Fri, Feb 18, 2022 at 11:49 AM Chris Chang wrote: > > > What compiler version are you using? This implies that the pgl_malloc > > inline function is not being compiled to the expected code; there is an > > existing non-inlined version that is used for very old gcc versions, but it > > looks like it may also be needed here. > > > > On Fri, Feb 18, 2022 at 11:40 AM Andreas Tille wrote: > > > >> Hi again, > >> > >> I applied this patch and now I get: > >> > >> (gdb) run > >> Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] > >> [New Thread 0x7fffec4c6640 (LWP 4060798)] > >> [New Thread 0x7fffebcc5640 (LWP 4060799)] > >> PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 > >> 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > >> Using up to 4 compute threads. > >> [New Thread 0x77fc5640 (LWP 4060800)] > >> sizeof(PhenoCol): 40 pheno_cols: 0 > >> --debug: setting pheno_cols[0].nonmiss. = nullptr > >> > >> Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. > >> 0x556fb82e in plink2::LoadPsam > >> (psamname=psamname@entry=0x7fffbe70 > >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > >> pheno_ct_max=, > >> missing_pheno=, affection_01=0, max_thread_ct=4, > >> piip=0x7fff8880, sample_include_ptr=0x7fff8790, > >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > >> pheno_ct_ptr=0x7fff8720, > >> max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 > >> warning: Source file is more recent than executable. > >> 615 pheno_cols[pheno_idx].nonmiss = nullptr; > >> > >> Kind regards > >> > >> Andreas. > >> > >> Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: > >> > Ok, I don't know why that particular line would fail, but I've added > >> > another debug-print before it on GitHub. > >> > > >> > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille > >> wrote: > >> > > >> > > Hi Chris, > >> > > > >> > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > >> > > > I was unable to replicate this issue on a Debian EC2 instance. > >> However, > >> > > > there are very few things that happen between printing "End time:" > >> and > >> > > > program exit, so I have added a bunch of debug-prints (active when > >> the > >> > > > --debug flag is passed in) to the latest GitHub commit that should > >> reveal > >> > > > which of those few things is triggering the segfault; let me know > >> if you > >> > > > are able to run this build. > >> > > > >> > > I think the issue is a bit more complex. Debian provides a wrapper > >> > > which calls the best / most performant plink2. The issue seems to > >> > > occure for SFX=avx. First I do: > >> > > > >> > > > >> > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 > >> dosage-freq=0.1 > >> > > --out tmp_data > >> > > > >> > > This works. In the next step I fire up gdb then which results in > >> > > > >> > > > >> > > (gdb) run > >> > > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] > >> > > [New Thread 0x744c6640 (LWP 2931409)] > >> > > [New Thread 0x7fffebcc5640 (LWP 2931411)] > >> > > PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 > >> > > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > >> > > Using up to 4 compute threads. > >> > > [New Thread 0x77fc5640 (LWP
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
I have posted an update under the provisional assumption that it's gcc 11's new ipa-modref pass that is causing this code to fail, since it does seem to break some similar code. On Fri, Feb 18, 2022 at 11:49 AM Chris Chang wrote: > What compiler version are you using? This implies that the pgl_malloc > inline function is not being compiled to the expected code; there is an > existing non-inlined version that is used for very old gcc versions, but it > looks like it may also be needed here. > > On Fri, Feb 18, 2022 at 11:40 AM Andreas Tille wrote: > >> Hi again, >> >> I applied this patch and now I get: >> >> (gdb) run >> Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] >> [New Thread 0x7fffec4c6640 (LWP 4060798)] >> [New Thread 0x7fffebcc5640 (LWP 4060799)] >> PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 >> 31998 MiB RAM detected; reserving 15999 MiB for main workspace. >> Using up to 4 compute threads. >> [New Thread 0x77fc5640 (LWP 4060800)] >> sizeof(PhenoCol): 40 pheno_cols: 0 >> --debug: setting pheno_cols[0].nonmiss. = nullptr >> >> Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. >> 0x556fb82e in plink2::LoadPsam >> (psamname=psamname@entry=0x7fffbe70 >> "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., >> pheno_ct_max=, >> missing_pheno=, affection_01=0, max_thread_ct=4, >> piip=0x7fff8880, sample_include_ptr=0x7fff8790, >> founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, >> sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, >> pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, >> pheno_ct_ptr=0x7fff8720, >> max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 >> warning: Source file is more recent than executable. >> 615 pheno_cols[pheno_idx].nonmiss = nullptr; >> >> Kind regards >> >> Andreas. >> >> Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: >> > Ok, I don't know why that particular line would fail, but I've added >> > another debug-print before it on GitHub. >> > >> > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille >> wrote: >> > >> > > Hi Chris, >> > > >> > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: >> > > > I was unable to replicate this issue on a Debian EC2 instance. >> However, >> > > > there are very few things that happen between printing "End time:" >> and >> > > > program exit, so I have added a bunch of debug-prints (active when >> the >> > > > --debug flag is passed in) to the latest GitHub commit that should >> reveal >> > > > which of those few things is triggering the segfault; let me know >> if you >> > > > are able to run this build. >> > > >> > > I think the issue is a bit more complex. Debian provides a wrapper >> > > which calls the best / most performant plink2. The issue seems to >> > > occure for SFX=avx. First I do: >> > > >> > > >> > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 >> dosage-freq=0.1 >> > > --out tmp_data >> > > >> > > This works. In the next step I fire up gdb then which results in >> > > >> > > >> > > (gdb) run >> > > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] >> > > [New Thread 0x744c6640 (LWP 2931409)] >> > > [New Thread 0x7fffebcc5640 (LWP 2931411)] >> > > PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 >> > > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. >> > > Using up to 4 compute threads. >> > > [New Thread 0x77fc5640 (LWP 2931412)] >> > > >> > > Thread 1 "plink2-avx" received signal SIGSEGV, Segmentation fault. >> > > plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 >> "tmp_data.psam", >> > > pheno_range_list_ptr=, fam_cols=..., >> pheno_ct_max=> > > out>, >> > > missing_pheno=, affection_01=0, max_thread_ct=4, >> > > piip=0x7fff8880, sample_include_ptr=0x7fff87a0, >> > > founder_info_ptr=0x7fff87b8,
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
What compiler version are you using? This implies that the pgl_malloc inline function is not being compiled to the expected code; there is an existing non-inlined version that is used for very old gcc versions, but it looks like it may also be needed here. On Fri, Feb 18, 2022 at 11:40 AM Andreas Tille wrote: > Hi again, > > I applied this patch and now I get: > > (gdb) run > Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] > [New Thread 0x7fffec4c6640 (LWP 4060798)] > [New Thread 0x7fffebcc5640 (LWP 4060799)] > PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > Using up to 4 compute threads. > [New Thread 0x77fc5640 (LWP 4060800)] > sizeof(PhenoCol): 40 pheno_cols: 0 > --debug: setting pheno_cols[0].nonmiss. = nullptr > > Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. > 0x556fb82e in plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 > "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > pheno_ct_max=, > missing_pheno=, affection_01=0, max_thread_ct=4, > piip=0x7fff8880, sample_include_ptr=0x7fff8790, > founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, > sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, > pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, > pheno_ct_ptr=0x7fff8720, > max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 > warning: Source file is more recent than executable. > 615 pheno_cols[pheno_idx].nonmiss = nullptr; > > Kind regards > > Andreas. > > Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: > > Ok, I don't know why that particular line would fail, but I've added > > another debug-print before it on GitHub. > > > > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille > wrote: > > > > > Hi Chris, > > > > > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > > > > I was unable to replicate this issue on a Debian EC2 instance. > However, > > > > there are very few things that happen between printing "End time:" > and > > > > program exit, so I have added a bunch of debug-prints (active when > the > > > > --debug flag is passed in) to the latest GitHub commit that should > reveal > > > > which of those few things is triggering the segfault; let me know if > you > > > > are able to run this build. > > > > > > I think the issue is a bit more complex. Debian provides a wrapper > > > which calls the best / most performant plink2. The issue seems to > > > occure for SFX=avx. First I do: > > > > > > > > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 > dosage-freq=0.1 > > > --out tmp_data > > > > > > This works. In the next step I fire up gdb then which results in > > > > > > > > > (gdb) run > > > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] > > > [New Thread 0x744c6640 (LWP 2931409)] > > > [New Thread 0x7fffebcc5640 (LWP 2931411)] > > > PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 > > > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > > > Using up to 4 compute threads. > > > [New Thread 0x77fc5640 (LWP 2931412)] > > > > > > Thread 1 "plink2-avx" received signal SIGSEGV, Segmentation fault. > > > plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 > "tmp_data.psam", > > > pheno_range_list_ptr=, fam_cols=..., > pheno_ct_max= > > out>, > > > missing_pheno=, affection_01=0, max_thread_ct=4, > > > piip=0x7fff8880, sample_include_ptr=0x7fff87a0, > > > founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, > > > sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, > > > pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, > > > pheno_ct_ptr=0x7fff8730, > > > max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:611 > > > warning: Source file is more recent than executable. > > > 611
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Hi again, I applied this patch and now I get: (gdb) run Starting program: /usr/lib/plink2/plink2-sse2 --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 0x74cc7640 (LWP 4060797)] [New Thread 0x7fffec4c6640 (LWP 4060798)] [New Thread 0x7fffebcc5640 (LWP 4060799)] PLINK v2.00a3 64-bit (29 Jan 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: Fri Feb 18 19:06:45 2022 31998 MiB RAM detected; reserving 15999 MiB for main workspace. Using up to 4 compute threads. [New Thread 0x77fc5640 (LWP 4060800)] sizeof(PhenoCol): 40 pheno_cols: 0 --debug: setting pheno_cols[0].nonmiss. = nullptr Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. 0x556fb82e in plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., pheno_ct_max=, missing_pheno=, affection_01=0, max_thread_ct=4, piip=0x7fff8880, sample_include_ptr=0x7fff8790, founder_info_ptr=0x7fff87a8, sex_nm_ptr=0x7fff8798, sex_male_ptr=0x7fff87a0, pheno_cols_ptr=0x7fff8770, pheno_names_ptr=0x7fff8780, raw_sample_ct_ptr=0x7fff8728, pheno_ct_ptr=0x7fff8720, max_pheno_name_blen_ptr=0x7fff87b0) at ../plink2_psam.cc:615 warning: Source file is more recent than executable. 615 pheno_cols[pheno_idx].nonmiss = nullptr; Kind regards Andreas. Am Fri, Feb 18, 2022 at 08:45:12AM -0800 schrieb Chris Chang: > Ok, I don't know why that particular line would fail, but I've added > another debug-print before it on GitHub. > > On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille wrote: > > > Hi Chris, > > > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > > > I was unable to replicate this issue on a Debian EC2 instance. However, > > > there are very few things that happen between printing "End time:" and > > > program exit, so I have added a bunch of debug-prints (active when the > > > --debug flag is passed in) to the latest GitHub commit that should reveal > > > which of those few things is triggering the segfault; let me know if you > > > are able to run this build. > > > > I think the issue is a bit more complex. Debian provides a wrapper > > which calls the best / most performant plink2. The issue seems to > > occure for SFX=avx. First I do: > > > > > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 dosage-freq=0.1 > > --out tmp_data > > > > This works. In the next step I fire up gdb then which results in > > > > > > (gdb) run > > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] > > [New Thread 0x744c6640 (LWP 2931409)] > > [New Thread 0x7fffebcc5640 (LWP 2931411)] > > PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 > > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > > Using up to 4 compute threads. > > [New Thread 0x77fc5640 (LWP 2931412)] > > > > Thread 1 "plink2-avx" received signal SIGSEGV, Segmentation fault. > > plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 "tmp_data.psam", > > pheno_range_list_ptr=, fam_cols=..., pheno_ct_max= > out>, > > missing_pheno=, affection_01=0, max_thread_ct=4, > > piip=0x7fff8880, sample_include_ptr=0x7fff87a0, > > founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, > > sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, > > pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, > > pheno_ct_ptr=0x7fff8730, > > max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:611 > > warning: Source file is more recent than executable. > > 611 pheno_cols[pheno_idx].nonmiss = nullptr; > > > > > > I also added some more debug lines in a patch[1]. > > > > It seems that there is actually the weak part of the code since the > > output turns to > > > > ... > > Start time: Fri Feb 18 13:19:13 2022 > > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > > Using up to 4 compute threads. > > [New Thread 0x77fc5640 (LWP 3957711)] > > --debug: setting pheno_cols[0].nonmiss. = nullptr > > > > Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Ok, I don't know why that particular line would fail, but I've added another debug-print before it on GitHub. On Fri, Feb 18, 2022 at 4:24 AM Andreas Tille wrote: > Hi Chris, > > Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > > I was unable to replicate this issue on a Debian EC2 instance. However, > > there are very few things that happen between printing "End time:" and > > program exit, so I have added a bunch of debug-prints (active when the > > --debug flag is passed in) to the latest GitHub commit that should reveal > > which of those few things is triggering the segfault; let me know if you > > are able to run this build. > > I think the issue is a bit more complex. Debian provides a wrapper > which calls the best / most performant plink2. The issue seems to > occure for SFX=avx. First I do: > > >/usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 dosage-freq=0.1 > --out tmp_data > > This works. In the next step I fire up gdb then which results in > > > (gdb) run > Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] > [New Thread 0x744c6640 (LWP 2931409)] > [New Thread 0x7fffebcc5640 (LWP 2931411)] > PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > Using up to 4 compute threads. > [New Thread 0x77fc5640 (LWP 2931412)] > > Thread 1 "plink2-avx" received signal SIGSEGV, Segmentation fault. > plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 "tmp_data.psam", > pheno_range_list_ptr=, fam_cols=..., pheno_ct_max= out>, > missing_pheno=, affection_01=0, max_thread_ct=4, > piip=0x7fff8880, sample_include_ptr=0x7fff87a0, > founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, > sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, > pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, > pheno_ct_ptr=0x7fff8730, > max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:611 > warning: Source file is more recent than executable. > 611 pheno_cols[pheno_idx].nonmiss = nullptr; > > > I also added some more debug lines in a patch[1]. > > It seems that there is actually the weak part of the code since the > output turns to > > ... > Start time: Fri Feb 18 13:19:13 2022 > 31998 MiB RAM detected; reserving 15999 MiB for main workspace. > Using up to 4 compute threads. > [New Thread 0x77fc5640 (LWP 3957711)] > --debug: setting pheno_cols[0].nonmiss. = nullptr > > Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. > 0x556fb6ff in plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 > "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., > pheno_ct_max=, > missing_pheno=, affection_01=0, max_thread_ct=4, > piip=0x7fff8880, sample_include_ptr=0x7fff87a0, > founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, > sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, > pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, > pheno_ct_ptr=0x7fff8730, > max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:614 > warning: Source file is more recent than executable. > 614 pheno_cols[pheno_idx].nonmiss = nullptr; > > > I hope this might help a bit to track down the issue > > Andreas. > > > > [1] > https://salsa.debian.org/med-team/plink2/-/blob/master/debian/patches/debug2.patch > > -- > http://fam-tille.de >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Hi Chris, Am Thu, Feb 17, 2022 at 07:13:49PM -0800 schrieb Chris Chang: > I was unable to replicate this issue on a Debian EC2 instance. However, > there are very few things that happen between printing "End time:" and > program exit, so I have added a bunch of debug-prints (active when the > --debug flag is passed in) to the latest GitHub commit that should reveal > which of those few things is triggering the segfault; let me know if you > are able to run this build. I think the issue is a bit more complex. Debian provides a wrapper which calls the best / most performant plink2. The issue seems to occure for SFX=avx. First I do: /usr/lib/plink2/plink2-avx --debug --dummy 33 65537 0.1 dosage-freq=0.1 --out tmp_data This works. In the next step I fire up gdb then which results in (gdb) run Starting program: /usr/lib/plink2/plink2-avx --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 0x74cc7640 (LWP 2931408)] [New Thread 0x744c6640 (LWP 2931409)] [New Thread 0x7fffebcc5640 (LWP 2931411)] PLINK v2.00a3 SSE4.2 (29 Jan 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: Fri Feb 18 11:58:49 2022 31998 MiB RAM detected; reserving 15999 MiB for main workspace. Using up to 4 compute threads. [New Thread 0x77fc5640 (LWP 2931412)] Thread 1 "plink2-avx" received signal SIGSEGV, Segmentation fault. plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., pheno_ct_max=, missing_pheno=, affection_01=0, max_thread_ct=4, piip=0x7fff8880, sample_include_ptr=0x7fff87a0, founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, pheno_ct_ptr=0x7fff8730, max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:611 warning: Source file is more recent than executable. 611 pheno_cols[pheno_idx].nonmiss = nullptr; I also added some more debug lines in a patch[1]. It seems that there is actually the weak part of the code since the output turns to ... Start time: Fri Feb 18 13:19:13 2022 31998 MiB RAM detected; reserving 15999 MiB for main workspace. Using up to 4 compute threads. [New Thread 0x77fc5640 (LWP 3957711)] --debug: setting pheno_cols[0].nonmiss. = nullptr Thread 1 "plink2-sse2" received signal SIGSEGV, Segmentation fault. 0x556fb6ff in plink2::LoadPsam (psamname=psamname@entry=0x7fffbe70 "tmp_data.psam", pheno_range_list_ptr=, fam_cols=..., pheno_ct_max=, missing_pheno=, affection_01=0, max_thread_ct=4, piip=0x7fff8880, sample_include_ptr=0x7fff87a0, founder_info_ptr=0x7fff87b8, sex_nm_ptr=0x7fff87a8, sex_male_ptr=0x7fff87b0, pheno_cols_ptr=0x7fff8780, pheno_names_ptr=0x7fff8790, raw_sample_ct_ptr=0x7fff8738, pheno_ct_ptr=0x7fff8730, max_pheno_name_blen_ptr=0x7fff87c0) at ../plink2_psam.cc:614 warning: Source file is more recent than executable. 614 pheno_cols[pheno_idx].nonmiss = nullptr; I hope this might help a bit to track down the issue Andreas. [1] https://salsa.debian.org/med-team/plink2/-/blob/master/debian/patches/debug2.patch -- http://fam-tille.de
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
I was unable to replicate this issue on a Debian EC2 instance. However, there are very few things that happen between printing "End time:" and program exit, so I have added a bunch of debug-prints (active when the --debug flag is passed in) to the latest GitHub commit that should reveal which of those few things is triggering the segfault; let me know if you are able to run this build. On Thu, Feb 17, 2022 at 11:47 AM Andreas Tille wrote: > Am Thu, Feb 17, 2022 at 11:39:30AM -0800 schrieb Chris Chang: > > Haven't yet replicated this on any of my usual non-Debian test machines, > > but I will spin up a Debian EC2 instance within the next two days and try > > to replicate the crash there. > > Thanks a lot, Andreas. > > -- > http://fam-tille.de >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Am Thu, Feb 17, 2022 at 11:39:30AM -0800 schrieb Chris Chang: > Haven't yet replicated this on any of my usual non-Debian test machines, > but I will spin up a Debian EC2 instance within the next two days and try > to replicate the crash there. Thanks a lot, Andreas. -- http://fam-tille.de
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Haven't yet replicated this on any of my usual non-Debian test machines, but I will spin up a Debian EC2 instance within the next two days and try to replicate the crash there. On Wed, Feb 16, 2022 at 11:30 PM Andreas Tille wrote: > PS to last mail: Here > >https://salsa.debian.org/med-team/plink2/-/jobs/2479769 > > is a full test log in our Gitlab CI for the latest version which I just > injected into the packaging Git repository but did not upload yet. > > Kind regards > > Andreas. > > -- > http://fam-tille.de >
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
PS to last mail: Here https://salsa.debian.org/med-team/plink2/-/jobs/2479769 is a full test log in our Gitlab CI for the latest version which I just injected into the packaging Git repository but did not upload yet. Kind regards Andreas. -- http://fam-tille.de
Bug#1004037: Segmentation fault in plink2 (Was: src:plink2: fails to migrate to testing for too long: autopkgtest regression)
Control: reopen -1
Control: tags -1 confirmed
Control: tags -1 upstream
Control: forwarded -1 Christopher Chang
Hi Christopher (and Dylan),
I verified the latest version (29 Jan 2022) of plink2 with the same
result for the CI test we are doing in Debian (which was written by
Dylan):
$ plink2 --dummy 33 65537 0.1 dosage-freq=0.1 --out tmp_data
PLINK v2.00a3 SSE4.2 (29 Jan 2022) www.cog-genomics.org/plink/2.0/
(C) 2005-2022 Shaun Purcell, Christopher Chang GNU General Public License v3
Logging to tmp_data.log.
Options in effect:
--dummy 33 65537 0.1 dosage-freq=0.1
--out tmp_data
Start time: Thu Feb 17 06:34:30 2022
31998 MiB RAM detected; reserving 15999 MiB for main workspace.
Using up to 4 compute threads.
Dummy data (33 samples, 65537 SNPs) written to tmp_data.pgen + tmp_data.pvar +
tmp_data.psam .
End time: Thu Feb 17 06:34:30 2022
/usr/bin/plink2: line 8: 156 Segmentation fault "${cmd}" "$@"
Please note that for the build the Debian packaged zstd and libdeflate
are used. This is the same result as for the version released at 11 Oct
2021 which is currently uploaded to Debian unstable. There is a full
log of the CI test available[1].
Kind regards
Andreas.
[1]
https://ci.debian.net/data/autopkgtest/testing/amd64/p/plink2/1921/log.gz
--
http://fam-tille.de
