Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Ximin Luo: > Ximin Luo: >> Julien Puydt: >>> Le 19/10/2017 à 22:24, Ximin Luo a écrit : >>> I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the previous stuff in a cleaner way. I've rebuilt all of the reverse dependencies () and they all succeeded including tests. So I think we're good to go ahead with the transition, whenever Julien is ready. >>> >>> Can you remind me what you want to transition exactly? I think most of >>> my packages related to debian-science are up to date in git ; they >>> should just need sponsorship to experimental... >>> >> >> We'd like to transition NTL to version 10.5.0, it's been sitting at 9.9.1 >> for too long. It shouldn't need any source uploads (except NTL itself), we >> just need to request a transition slot to do binary-only rebuilds from the >> release team: >> >> https://wiki.debian.org/Teams/ReleaseTeam/Transitions >> >> I can sponsor your upload once you feel 10.5.0 is ready, just let me know. >> (Perhaps now would be a good time to make it multi-arch compatible? Not sure >> how hard it is, upstream does not seem to do it explicitly.) >> > > I've uploaded this to experimental, it's sitting in NEW. > Transition is being tracked here: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=886319 For some reason singular and sagemath rebuilds have been scheduled before flint's rebuilds completed, so they are failing due to pulling in both libnt27 (via the old flint) and libntl35 (via the rebuilt flint). I've noted this on the bug report. Hopefully the release team will fix this by themselves. X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Ximin Luo: > Julien Puydt: >> Le 19/10/2017 à 22:24, Ximin Luo a écrit : >> >>> I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the >>> previous stuff in a cleaner way. I've rebuilt all of the reverse >>> dependencies () and they all succeeded including tests. So I think we're >>> good to go ahead with the transition, whenever Julien is ready. >> >> Can you remind me what you want to transition exactly? I think most of >> my packages related to debian-science are up to date in git ; they >> should just need sponsorship to experimental... >> > > We'd like to transition NTL to version 10.5.0, it's been sitting at 9.9.1 for > too long. It shouldn't need any source uploads (except NTL itself), we just > need to request a transition slot to do binary-only rebuilds from the release > team: > > https://wiki.debian.org/Teams/ReleaseTeam/Transitions > > I can sponsor your upload once you feel 10.5.0 is ready, just let me know. > (Perhaps now would be a good time to make it multi-arch compatible? Not sure > how hard it is, upstream does not seem to do it explicitly.) > I've uploaded this to experimental, it's sitting in NEW. X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Julien Puydt: > Le 19/10/2017 à 22:24, Ximin Luo a écrit : > >> I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the >> previous stuff in a cleaner way. I've rebuilt all of the reverse >> dependencies () and they all succeeded including tests. So I think we're >> good to go ahead with the transition, whenever Julien is ready. > > Can you remind me what you want to transition exactly? I think most of > my packages related to debian-science are up to date in git ; they > should just need sponsorship to experimental... > We'd like to transition NTL to version 10.5.0, it's been sitting at 9.9.1 for too long. It shouldn't need any source uploads (except NTL itself), we just need to request a transition slot to do binary-only rebuilds from the release team: https://wiki.debian.org/Teams/ReleaseTeam/Transitions I can sponsor your upload once you feel 10.5.0 is ready, just let me know. (Perhaps now would be a good time to make it multi-arch compatible? Not sure how hard it is, upstream does not seem to do it explicitly.) X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Le 19/10/2017 à 22:24, Ximin Luo a écrit : > I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the > previous stuff in a cleaner way. I've rebuilt all of the reverse dependencies > () and they all succeeded including tests. So I think we're good to go ahead > with the transition, whenever Julien is ready. Can you remind me what you want to transition exactly? I think most of my packages related to debian-science are up to date in git ; they should just need sponsorship to experimental... Snark on #debian-science
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Ximin Luo: > [..] > > I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the > previous stuff in a cleaner way. I've rebuilt all of the reverse dependencies > () and they all succeeded including tests. So I think we're good to go ahead > with the transition, whenever Julien is ready. > To be clear, the issue and fix are to the NTL Debian packaging, FLINT does not need any fixes, and NTL upstream does not need any fixes (AFAICT). The reverse-dependencies I rebuilt are: eclib flint giac linbox singular pynac (I did not rebuild sagemath yet.) X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Julien Puydt: > > > Le 01/09/2017 à 00:03, Ximin Luo a écrit : >> Thanks for that. I forced the rebuild to continue by skipping the flint >> tests with DEB_BUILD_OPTIONS=nocheck sbuild --profiles=nocheck , and am >> pleased to report that singular and pynac built (including tests) >> successfully. > > That is good news, but since FLINT saw problems, they may be hidden. > >> If the flint maintainer doesn't fix this "soon" then perhaps we can just >> disable that one test for the short-term to allow this transition to go >> forward. > > I'm not sure it is a good idea to push things forward: Bill (main > developer) said those tests only pass data to NTL, let it do the > calculation and get the result back. If those conversions are > problematic, the root can be: > - NTL changed something and it wasn't intentional, so NTL will need a fix ; > - NTL changed something and it was intentional, so FLINT will need a fix. > > In either case, there is a good chance that this change will break > things also higher up the stack : FLINT is just an early warning sign, > and ignoring it is wrong. > > I suggest to wait for a diagnosis before planning anything. > Your cautiousness was well-placed, I think I nailed down the cause and fixed it. It was actually very serious. I did not drill down into the *exact* underlying cause but saw enough to be fairly certain it was due to MAKEFLAGS plus incompletely-declared dependencies in d/rules. Those interested can read "recursive Make considered harmful" for a description of the generic problem. In summary, when you incompletely-duplicate dependency declarations from child Makefiles, the top-level make process (here, the one running debian/rules) does not have enough dependency information to always correctly determine the order in which targets should be rebuilt. Sometimes this gets overlooked because compilation "looks like" it worked but actually things were miscompiled (I admit I don't know the exact details of how the binary stuff works on this level). These errors are much more likely when running builds in parallel, because here make will aggressively parallelise non-dependent targets, so if two targets actually do depend on each other in the child Makefiles but you didn't declare this in the top-level Makefile, you get screwed. Several observations pointed to this as being the culprit, e.g. rebuilding NTL using `dpkg-buildpackage -J1` or `debian/rules build && fakeroot debian/rules binary` directly (with no DEB_BUILD_OPTIONS=parallel=X in the env) would make the FLINT test failures go away. I fixed things in git by removing MAKEFLAGS and rewriting d/rules to do the previous stuff in a cleaner way. I've rebuilt all of the reverse dependencies () and they all succeeded including tests. So I think we're good to go ahead with the transition, whenever Julien is ready. X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Le 01/09/2017 à 00:03, Ximin Luo a écrit : > Thanks for that. I forced the rebuild to continue by skipping the flint tests > with DEB_BUILD_OPTIONS=nocheck sbuild --profiles=nocheck , and am > pleased to report that singular and pynac built (including tests) > successfully. That is good news, but since FLINT saw problems, they may be hidden. > If the flint maintainer doesn't fix this "soon" then perhaps we can just > disable that one test for the short-term to allow this transition to go > forward. I'm not sure it is a good idea to push things forward: Bill (main developer) said those tests only pass data to NTL, let it do the calculation and get the result back. If those conversions are problematic, the root can be: - NTL changed something and it wasn't intentional, so NTL will need a fix ; - NTL changed something and it was intentional, so FLINT will need a fix. In either case, there is a good chance that this change will break things also higher up the stack : FLINT is just an early warning sign, and ignoring it is wrong. I suggest to wait for a diagnosis before planning anything. Snark
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Julien Puydt: > Hi, > >> [..] > > I could reproduce the matter, asked upstream if it was known, and since > it wasn't: > https://github.com/wbhart/flint2/issues/372 > Thanks for that. I forced the rebuild to continue by skipping the flint tests with DEB_BUILD_OPTIONS=nocheck sbuild --profiles=nocheck , and am pleased to report that singular and pynac built (including tests) successfully. If the flint maintainer doesn't fix this "soon" then perhaps we can just disable that one test for the short-term to allow this transition to go forward. X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Hi, Le 31/08/2017 à 14:48, Ximin Luo a écrit : > eclib, giac, linbox succeeded but flint fails: > > [..] > make[3]: Entering directory '/<>' > g++ -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong > -Wformat -Werror=format-security -I/<> -c > interfaces/NTL-interface.cpp -o build/interfaces/NTL-interface.o > g++ -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong > -Wformat -Werror=format-security -I/<> > interfaces/test/t-NTL-interface.cpp build/interfaces/NTL-interface.o -o > build/interfaces/test/t-NTL-interface -L/<> -L/usr/local/lib > -L/usr/local/lib -L/usr/local/lib -lflint -lmpfr -lgmp -lm -lntl -lpthread > make[3]: Leaving directory '/<>' > ZZ_to_fmpzPASS > ZZX_to_fmpz_polyPASS > ZZ_pX_to_fmpz_mod_polyPASS > ZZ_pE_to_fqPASS > ZZ_pEX_to_fq_polyPASS > zz_pX_to_fmpz_mod_polyFAIL: > f_poly1 = 61 40559 2065 31588 35949 5581 18565 23027 17733 40355 21583 29748 > 34876 9541 13193 37235 25453 21973 26716 1397 32001 6909 21509 13786 20347 > 3693 31431 1471 14890 19641 12608 1894 24849 8969 15639 30973 20096 8110 > 30791 35078 29416 16789 35010 10460 34287 20737 768 35601 22276 17792 3619 > 10427 1950 5016 21067 39969 274 29530 39547 13397 8480 1868 19224 > f_poly2 = 61 40559 -19790727 -303309173 -345202259 -53532299 -178238240 > -221104641 -170248949 -387500890 -207234907 -285627289 -334901346 -91572681 > -126652564 -357530909 -244383081 -210965945 -256508959 -13342514 -307283542 > -66307056 -206504919 -132330231 -195352356 -35404314 -301808647 -14072502 > -142955585 -188579709 -121015448 -18127979 -238583748 -86097788 -150133779 > -297388174 -192959626 -77824611 -295644319 -336847976 -282464019 -161164677 > -336158541 -100413624 -329223675 -199083394 -7299852 -341836210 -213885890 > -170816716 -34714885 -100089185 -18655190 -48097958 -202287225 -383810407 > -2554943 -283558998 -379754929 -128599192 -81393433 -17884651 -184564785 > zz_pE_to_fqFAIL: > p = 31 > mod = x^2+19*x+12 > f1 = 20*a+16 - 2 > zzpe:[-418 -538] > f2 = -538*a-418 - 2 > zz_pEX_to_fq_polyAborted > Makefile:211: recipe for target 'check' failed > make[2]: *** [check] Error 134 > [..] I could reproduce the matter, asked upstream if it was known, and since it wasn't: https://github.com/wbhart/flint2/issues/372 Snark
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Ximin Luo: > Julien Puydt: >> [..] >> >> I still pushed this work-in-progress, as it should be good enought for >> some tests already. >> > > I will have a go at testing it tomorrow/soon. Thanks! > eclib, giac, linbox succeeded but flint fails: [..] make[3]: Entering directory '/<>' g++ -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong -Wformat -Werror=format-security -I/<> -c interfaces/NTL-interface.cpp -o build/interfaces/NTL-interface.o g++ -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong -Wformat -Werror=format-security -I/<> interfaces/test/t-NTL-interface.cpp build/interfaces/NTL-interface.o -o build/interfaces/test/t-NTL-interface -L/<> -L/usr/local/lib -L/usr/local/lib -L/usr/local/lib -lflint -lmpfr -lgmp -lm -lntl -lpthread make[3]: Leaving directory '/<>' ZZ_to_fmpzPASS ZZX_to_fmpz_polyPASS ZZ_pX_to_fmpz_mod_polyPASS ZZ_pE_to_fqPASS ZZ_pEX_to_fq_polyPASS zz_pX_to_fmpz_mod_polyFAIL: f_poly1 = 61 40559 2065 31588 35949 5581 18565 23027 17733 40355 21583 29748 34876 9541 13193 37235 25453 21973 26716 1397 32001 6909 21509 13786 20347 3693 31431 1471 14890 19641 12608 1894 24849 8969 15639 30973 20096 8110 30791 35078 29416 16789 35010 10460 34287 20737 768 35601 22276 17792 3619 10427 1950 5016 21067 39969 274 29530 39547 13397 8480 1868 19224 f_poly2 = 61 40559 -19790727 -303309173 -345202259 -53532299 -178238240 -221104641 -170248949 -387500890 -207234907 -285627289 -334901346 -91572681 -126652564 -357530909 -244383081 -210965945 -256508959 -13342514 -307283542 -66307056 -206504919 -132330231 -195352356 -35404314 -301808647 -14072502 -142955585 -188579709 -121015448 -18127979 -238583748 -86097788 -150133779 -297388174 -192959626 -77824611 -295644319 -336847976 -282464019 -161164677 -336158541 -100413624 -329223675 -199083394 -7299852 -341836210 -213885890 -170816716 -34714885 -100089185 -18655190 -48097958 -202287225 -383810407 -2554943 -283558998 -379754929 -128599192 -81393433 -17884651 -184564785 zz_pE_to_fqFAIL: p = 31 mod = x^2+19*x+12 f1 = 20*a+16 - 2 zzpe:[-418 -538] f2 = -538*a-418 - 2 zz_pEX_to_fq_polyAborted Makefile:211: recipe for target 'check' failed make[2]: *** [check] Error 134 [..] I did not yet try singular or pynac because they depend on flint. X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Hi, Le 31/08/2017 à 00:10, Ximin Luo a écrit : > Julien Puydt: >> Hi, >> >> Le 29/08/2017 à 14:25, Tobias Hansen a écrit : >>> Now is a good time for the NTL transition. Julien, could you update the >>> package to 10.3.0? Then we can test-build the reverse dependencies and >>> ask for a transition. >> >> Indeed there was a soname version bump from 27 to 35, so a transition is >> in order. >> >> There also was a change of license, so I changed d/copyright -- I would >> be glad if someone could check the result. >> >> Finally, lintian complains about hardening-no-fortify-functions and >> hardening-no-bindnow, which means there's more digging to do for me. >> > > I'd ignore both: > > 1. hardening-no-fortify-functions happens naturally when the library does not > use any libc functions that change under -D_FORTIFY_SOURCE=2 and this is > probably a false positive > > 2. hardening-no-bindnow is just due to dpkg-buildflags not defaulting that to > on. If you really want to fix it you can > > export DEB_BUILD_MAINT_OPTIONS=hardening=+bindnow Won't work : the configure script isn't an autotools script -- it calls a DoConfig perl script, which does its own work in its own way. And of course it won't respect what we fix... Snark
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Ximin Luo: > Julien Puydt: >> Hi, >> >> Le 29/08/2017 à 14:25, Tobias Hansen a écrit : >>> Now is a good time for the NTL transition. Julien, could you update the >>> package to 10.3.0? Then we can test-build the reverse dependencies and >>> ask for a transition. >> >> Indeed there was a soname version bump from 27 to 35, so a transition is >> in order. >> >> There also was a change of license, so I changed d/copyright -- I would >> be glad if someone could check the result. >> >> Finally, lintian complains about hardening-no-fortify-functions and >> hardening-no-bindnow, which means there's more digging to do for me. >> > > I'd ignore both: > > 1. hardening-no-fortify-functions happens naturally when the library does not > use any libc functions that change under -D_FORTIFY_SOURCE=2 and this is > probably a false positive > > 2. hardening-no-bindnow is just due to dpkg-buildflags not defaulting that to > on. If you really want to fix it you can > > export DEB_BUILD_MAINT_OPTIONS=hardening=+bindnow > > but IMO this is better fixed in dpkg-buildflags itself as a matter of Debian > policy. > >> I still pushed this work-in-progress, as it should be good enought for >> some tests already. >> > > I will have a go at testing it tomorrow/soon. Thanks! > BTW, please don't forget about the other bug involving gf2x, #872711 X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Julien Puydt: > Hi, > > Le 29/08/2017 à 14:25, Tobias Hansen a écrit : >> Now is a good time for the NTL transition. Julien, could you update the >> package to 10.3.0? Then we can test-build the reverse dependencies and >> ask for a transition. > > Indeed there was a soname version bump from 27 to 35, so a transition is > in order. > > There also was a change of license, so I changed d/copyright -- I would > be glad if someone could check the result. > > Finally, lintian complains about hardening-no-fortify-functions and > hardening-no-bindnow, which means there's more digging to do for me. > I'd ignore both: 1. hardening-no-fortify-functions happens naturally when the library does not use any libc functions that change under -D_FORTIFY_SOURCE=2 and this is probably a false positive 2. hardening-no-bindnow is just due to dpkg-buildflags not defaulting that to on. If you really want to fix it you can export DEB_BUILD_MAINT_OPTIONS=hardening=+bindnow but IMO this is better fixed in dpkg-buildflags itself as a matter of Debian policy. > I still pushed this work-in-progress, as it should be good enought for > some tests already. > I will have a go at testing it tomorrow/soon. Thanks! X -- GPG: ed25519/56034877E1F87C35 GPG: rsa4096/1318EFAC5FBBDBCE https://github.com/infinity0/pubkeys.git
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Hi, Le 29/08/2017 à 14:25, Tobias Hansen a écrit : > Now is a good time for the NTL transition. Julien, could you update the > package to 10.3.0? Then we can test-build the reverse dependencies and > ask for a transition. Uh, latest upstream is 10.5.0, and that's what I'm now working on. Apparently quite a few things changed. Snark on #debian-science
Bug#864831: [Debian-science-sagemath] NTL update to 10.3.0
Hi, Le 29/08/2017 à 14:25, Tobias Hansen a écrit : > Now is a good time for the NTL transition. Julien, could you update the > package to 10.3.0? Then we can test-build the reverse dependencies and > ask for a transition. Indeed there was a soname version bump from 27 to 35, so a transition is in order. There also was a change of license, so I changed d/copyright -- I would be glad if someone could check the result. Finally, lintian complains about hardening-no-fortify-functions and hardening-no-bindnow, which means there's more digging to do for me. I still pushed this work-in-progress, as it should be good enought for some tests already. Snark on #debian-science
