[sage-release] Re: Sage 10.2.beta3 released
After an upgrade of macOS and Xcode CLT, I also see this problem on my system. There is an open issue https://github.com/mesonbuild/meson/issues/12282 for this. On Thursday, September 21, 2023 at 4:52:56 PM UTC-7 John H Palmieri wrote: > A new problem, after upgrading to OS X 13.6 and the newest version of the > command-line tools: scipy fails to build again, this time with the error: > > ../../meson.build:82:0: ERROR: Unable to detect linker for compiler > `gfortran -Wl,--version > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -Wl,-rpath,/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -Wl,-rpath,/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib` > stdout: > stderr: collect2 version 13.2.0 > /usr/bin/ld -syslibroot > /Library/Developer/CommandLineTools/SDKs/MacOSX13.sdk/ -dynamic -arch arm64 > -platform_version macos 13.0.0 0.0 -o a.out > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -L/opt/homebrew/Cellar/gcc/13.2.0/bin/../lib/gcc/current/gcc/aarch64-apple-darwin22/13 > > -L/opt/homebrew/Cellar/gcc/13.2.0/bin/../lib/gcc/current/gcc > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -L/Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib > -L/opt/homebrew/opt/primesieve/lib -L/opt/homebrew/opt/bdw-gc/lib > -L/opt/homebrew/opt/libpng/lib -L/opt/homebrew/opt/ntl/lib > -L/opt/homebrew/opt/bzip2/lib -L/opt/homebrew/opt/readline/lib > -L/opt/homebrew/lib > -L/opt/homebrew/Cellar/gcc/13.2.0/bin/../lib/gcc/current/gcc/aarch64-apple-darwin22/13/../../.. > > --version -rpath > /Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib -rpath > /Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib -lemutls_w > -lgcc -lSystem -lgcc -no_compact_unwind -rpath @loader_path -rpath > /opt/homebrew/Cellar/gcc/13.2.0/lib/gcc/current/gcc/aarch64-apple-darwin22/13 > -rpath /opt/homebrew/Cellar/gcc/13.2.0/lib/gcc/current/gcc -rpath > /Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib -rpath > /Users/palmieri/Desktop/Sage/TESTING/sage-10.2.beta3/local/lib -rpath > /opt/homebrew/Cellar/primesieve/11.1/lib -rpath > /opt/homebrew/Cellar/bdw-gc/8.2.4/lib -rpath > /opt/homebrew/Cellar/libpng/1.6.40/lib -rpath > /opt/homebrew/Cellar/ntl/11.5.1/lib -rpath > /opt/homebrew/Cellar/bzip2/1.0.8/lib -rpath > /opt/homebrew/Cellar/readline/8.2.1/lib -rpath /opt/homebrew/lib -rpath > /opt/homebrew/Cellar/gcc/13.2.0/lib/gcc/current > ld: unknown options: --version > collect2: error: ld returned 1 exit status > > Suggestions? > > On Sunday, September 17, 2023 at 11:25:02 AM UTC-7 John H Palmieri wrote: > >> Similar problem for me on OS X. I don't understand something: the >> dependencies for scipy include meson_python, but that package is not >> installed before scipy attempts to build, and fails. Running "make >> meson_python" and then "make scipy" succeeds, as does "make". >> >> >> >> On Saturday, September 16, 2023 at 2:59:58 PM UTC-7 Kwankyu Lee wrote: >> >>> Succeeded after sage -pip install meson-python. >>> >>> On Sunday, September 17, 2023 at 6:52:59 AM UTC+9 Kwankyu Lee wrote: >>> Incremental build failed [scipy-1.11.2] [..] [scipy-1.11.2] scipy-1.11.2 [scipy-1.11.2] [sagelib-10.2.beta3] Removing file or directory /Users/kwankyu/GitHub/sage-dev/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/sagemath-standard.egg-link [sagelib-10.2.beta3] Removing pth entries from /Users/kwankyu/GitHub/sage-dev/local/var/lib/sage/venv-python3.10/lib/python3.10/site-packages/easy-install.pth: [sagelib-10.2.beta3] Removing entry: /Users/kwankyu/GitHub/sage-dev/src [sagelib-10.2.beta3] Successfully uninstalled sagemath-standard-10.2b3 [sagelib-10.2.beta3] Running setup.py develop for sagemath-standard [sagelib-10.2.beta3] Running command python setup.py develop [scipy-1.11.2] Setting up build directory for scipy-1.11.2 [scipy-1.11.2] Finished extraction [scipy-1.11.2] Applying patches from ../patches... [scipy-1.11.2] Applying ../patches/numpy-include.patch [scipy-1.11.2] patching file scipy/spatial/setup.py [scipy-1.11.2] Hunk #1 succeeded at 39 (offset -1 lines). [scipy-1.11.2] Applying ../patches/tests_no_internet_dependency.patch [scipy-1.11.2] patching file scipy/datasets/meson.build [scipy-1.11.2] [scipy-1.11.2] Host system: [scipy-1.11.2] Darwin Helios.local 22.6.0 Darwin Kernel Version 22.6.0: Wed Jul 5 22:21:56 PDT 2023; root:xnu-8796.141.3~6/RELEASE_X86_64 x86_64
[sage-release] Sage 10.2.beta4 released
As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html 4d3e807ba54 (tag: 10.2.beta4, github/develop) Updated SageMath version to 10.2.beta4 930d65e684b gh-36308: using itertools.product more often 9ea5127d78f gh-36307: `sage.rings`: Remove code deprecated in #23204, #24483, #24371, #24511, #25848, #26105, #28481, #29010, #29412, #30332, #30372, #31345, #32375, #32606, #32610, #32612, #32641, #32660, #32750, #32869, #33602 fb846668a18 gh-36306: `sage.docs`: Remove; deprecated in #33763 d7404b1b14e gh-36305: `sage.media`, `sage.structure.graphics_file`: Remove; deprecated in #12673, #32988 53cce67ccdf gh-36303: Remove `sage.finance`, deprecated in #32427 (2021) 152cef732bd gh-36300: Remove deprecated files from src/sage/homology (see #31925) 9311c6d46f8 gh-36297: using more libgap instead of gap 461727b4537 gh-36296: remove duplicate args in @options() in region_plot 653bd9fac36 gh-36282: CI: Remove ubuntu-kinetic (defunct) 35c302fb1b2 gh-36281: remove unused vars in matrix/ 979b3688a6e gh-36280: fix E228 and E225 in combinat/*/*.py 0fcf8893590 gh-36279: Support for matplotlib-3.8.0 991a4638d22 gh-36278: `sage.graphs`: some care with return ... else statements in some .pyx files 342e32bf2cf gh-36264: `sage --fixdoctests --update-known-test-failures`; silence modularized distributions in CI 5bd0466757a gh-36252: no generator for S1 5b90f225d5f gh-36239: Disable more packages with various --disable-foo combinations 01a8322c29f gh-36230: build/pkgs/networkx/dependencies: drop decorator cec940d2a3f gh-36220: Python package upgrades, preparation for the Notebook 7 upgrade b4b7c9a2d9e gh-36205: build/pkgs: drop obsolete simplegeneric package ac09279d582 gh-36188: remove some old deprecation warnings in elliptic curves 2cf443a254d gh-36128: Compute the characteristic varieties of a finitely presented group f9541a44e48 gh-36020: Fixes #35804 by computing the DDT of non-square sboxes properly. 80f6d772a15 (tag: 10.2.beta3) Updated SageMath version to 10.2.beta3 -- You received this message because you are subscribed to the Google Groups "sage-release" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-release+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/3d11dca1-921c-41b8-b2bb-ca507346a495n%40googlegroups.com.
Re: [sage-release] Re: Sage 10.2.beta3 released
On Sun, Sep 24, 2023 at 5:53 PM John H Palmieri wrote: > > I've already tried uninstalling and reinstalling every homebrew package, and > it didn't help. I meant removing the whole /opt/homebrew/ and reinstalling homebrew from scratch. > I've also been doing "make distclean" and "./configure" in between attempts. > > I'm confused about your comment about meson_python: I don't think that Sage > ordinarily uses a system package of this (except on gentoo?), so it always > builds its own. It is using the homebrew installation of meson. I do not have > the homebrew meson-python package. It is (or going to) be available with ./configure --enable-system-site-packages (after https://github.com/sagemath/sage/pull/36256 and https://github.com/sagemath/sage/pull/36276) > > On Sunday, September 24, 2023 at 3:28:22 AM UTC-7 Dima Pasechnik wrote: >> >> both meson and meson_python may potentially come from the system (homebrew >> in your case). >> Make sure they are up to date, and you don't have these packages installed >> in Sage. >> And run ./configure >> >> Perhaps the whole homebrew must be reinstated after an OS update, too. >> >> On Sun, 24 Sept 2023, 03:07 John H Palmieri, wrote: >>> >>> Building Sage's own gfortran worked to build scipy (and this is an argument >>> to keep the gfortran package around, by the way). >>> >>> Alternatively, building all of scipy's dependencies and then using `./sage >>> --python3 -m pip install scipy` also seems to have worked. I then "touch"ed >>> the appropriate file to convince Sage that scipy had built, and the rest of >>> the build succeeded, except for some extra warnings during doctests: "ld: >>> warning: duplicate -rpath [...SAGE_ROOT/local/lib...] ignored" >>> >>> On Saturday, September 23, 2023 at 3:59:40 PM UTC-7 John H Palmieri wrote: If by "they" you mean scipy, then: I have no problems building scipy on another OS X machine: an Intel machine running OS X 13.5.2 (rather than 13.6). "xcode-select --version" reports the same for both machines, but I don't know how informative this is. "gfortran --version" says GNU Fortran (Homebrew GCC 13.2.0) 13.2.0 Copyright (C) 2023 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. and this is the one that scipy finds. On the M2 machine, Sage's scipy built back in August with earlier versions of the OS (probably 13.5.2) and command-line tools and gfortran (13.1.0). On Saturday, September 23, 2023 at 3:34:29 PM UTC-7 Dima Pasechnik wrote: > > do they even support gcc/gfortran 13? > > On Sat, 23 Sept 2023, 23:04 John H Palmieri, wrote: >> >> Same result if I do `pip3 install scipy --no-binary scipy` (another of >> their suggested ways of building from source). I guess something is >> broken with my fortran compiler, but I don't know how to troubleshoot it. >> >> On Saturday, September 23, 2023 at 2:35:52 PM UTC-7 John H Palmieri >> wrote: >>> >>> It didn't work for me. At least directly attempting a scipy build left >>> an intact log file. Any clues about what's broken on this machine? >>> >>> Build started at 2023-09-23T14:31:26.723020 >>> Main binary: /Users/palmieri/Downloads/scipy-1.11.2/venv/bin/python3 >>> Build Options: >>> -Dprefix=/Users/palmieri/Downloads/scipy-1.11.2/build-install >>> Python system: Darwin >>> The Meson build system >>> Version: 1.2.1 >>> Source dir: /Users/palmieri/Downloads/scipy-1.11.2 >>> Build dir: /Users/palmieri/Downloads/scipy-1.11.2/build >>> Build type: native build >>> Project name: SciPy >>> Project version: 1.11.2 >>> --- >>> Detecting compiler via: `cc --version` -> 0 >>> stdout: >>> Apple clang version 15.0.0 (clang-1500.0.40.1) >>> Target: arm64-apple-darwin22.6.0 >>> Thread model: posix >>> InstalledDir: /Library/Developer/CommandLineTools/usr/bin >>> --- >>> Running command: cc -E -dM - >>> - >>> --- >>> Detecting linker via: `cc -Wl,--version` -> 1 >>> stderr: >>> ld: unknown options: --version >>> clang: error: linker command failed with exit code 1 (use -v to see >>> invocation) >>> --- >>> --- >>> Detecting Apple linker via: `cc -Wl,-v` -> 1 >>> stderr: >>> @(#)PROGRAM:ld PROJECT:dyld-1015.7 >>> BUILD 18:48:48 Aug 22 2023 >>> configured to support archs: armv6 armv7 armv7s arm64 arm64e arm64_32 >>> i386 x86_64 x86_64h armv6m armv7k armv7m armv7em >>> will use ld-classic for: armv6 armv7 armv7s arm64_32 i386 armv6m armv7k >>> armv7m armv7em >>> LTO support using: LLVM version 15.0.0 (static support for 29, runtime >>> is 29) >>> TAPI support using: Apple TAPI version 15.0.0
Re: [sage-release] Re: Sage 10.2.beta3 released
I've already tried uninstalling and reinstalling every homebrew package, and it didn't help. I've also been doing "make distclean" and "./configure" in between attempts. I'm confused about your comment about meson_python: I don't think that Sage ordinarily uses a system package of this (except on gentoo?), so it always builds its own. It is using the homebrew installation of meson. I do not have the homebrew meson-python package. On Sunday, September 24, 2023 at 3:28:22 AM UTC-7 Dima Pasechnik wrote: > both meson and meson_python may potentially come from the system (homebrew > in your case). > Make sure they are up to date, and you don't have these packages installed > in Sage. > And run ./configure > > Perhaps the whole homebrew must be reinstated after an OS update, too. > > On Sun, 24 Sept 2023, 03:07 John H Palmieri, wrote: > >> Building Sage's own gfortran worked to build scipy (and this is an >> argument to keep the gfortran package around, by the way). >> >> Alternatively, building all of scipy's dependencies and then using >> `./sage --python3 -m pip install scipy` also seems to have worked. I then >> "touch"ed the appropriate file to convince Sage that scipy had built, and >> the rest of the build succeeded, except for some extra warnings during >> doctests: "ld: warning: duplicate -rpath [...SAGE_ROOT/local/lib...] >> ignored" >> >> On Saturday, September 23, 2023 at 3:59:40 PM UTC-7 John H Palmieri wrote: >> >>> If by "they" you mean scipy, then: I have no problems building scipy on >>> another OS X machine: an Intel machine running OS X 13.5.2 (rather than >>> 13.6). "xcode-select --version" reports the same for both machines, but I >>> don't know how informative this is. "gfortran --version" says >>> >>> GNU Fortran (Homebrew GCC 13.2.0) 13.2.0 >>> Copyright (C) 2023 Free Software Foundation, Inc. >>> This is free software; see the source for copying conditions. There is >>> NO >>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR >>> PURPOSE. >>> >>> and this is the one that scipy finds. On the M2 machine, Sage's scipy >>> built back in August with earlier versions of the OS (probably 13.5.2) and >>> command-line tools and gfortran (13.1.0). >>> >>> >>> On Saturday, September 23, 2023 at 3:34:29 PM UTC-7 Dima Pasechnik wrote: >>> do they even support gcc/gfortran 13? On Sat, 23 Sept 2023, 23:04 John H Palmieri, wrote: > Same result if I do `pip3 install scipy --no-binary scipy` (another of > their suggested ways of building from source). I guess something is > broken > with my fortran compiler, but I don't know how to troubleshoot it. > > On Saturday, September 23, 2023 at 2:35:52 PM UTC-7 John H Palmieri > wrote: > >> It didn't work for me. At least directly attempting a scipy build >> left an intact log file. Any clues about what's broken on this machine? >> >> Build started at 2023-09-23T14:31:26.723020 >> Main binary: /Users/palmieri/Downloads/scipy-1.11.2/venv/bin/python3 >> Build Options: >> -Dprefix=/Users/palmieri/Downloads/scipy-1.11.2/build-install >> Python system: Darwin >> The Meson build system >> Version: 1.2.1 >> Source dir: /Users/palmieri/Downloads/scipy-1.11.2 >> Build dir: /Users/palmieri/Downloads/scipy-1.11.2/build >> Build type: native build >> Project name: SciPy >> Project version: 1.11.2 >> --- >> Detecting compiler via: `cc --version` -> 0 >> stdout: >> Apple clang version 15.0.0 (clang-1500.0.40.1) >> Target: arm64-apple-darwin22.6.0 >> Thread model: posix >> InstalledDir: /Library/Developer/CommandLineTools/usr/bin >> --- >> Running command: cc -E -dM - >> - >> --- >> Detecting linker via: `cc -Wl,--version` -> 1 >> stderr: >> ld: unknown options: --version >> clang: error: linker command failed with exit code 1 (use -v to see >> invocation) >> --- >> --- >> Detecting Apple linker via: `cc -Wl,-v` -> 1 >> stderr: >> @(#)PROGRAM:ld PROJECT:dyld-1015.7 >> BUILD 18:48:48 Aug 22 2023 >> configured to support archs: armv6 armv7 armv7s arm64 arm64e arm64_32 >> i386 x86_64 x86_64h armv6m armv7k armv7m armv7em >> will use ld-classic for: armv6 armv7 armv7s arm64_32 i386 armv6m >> armv7k armv7m armv7em >> LTO support using: LLVM version 15.0.0 (static support for 29, >> runtime is 29) >> TAPI support using: Apple TAPI version 15.0.0 (tapi-1500.0.12.3) >> Library search paths: >> /opt/homebrew/opt/primesieve/lib >> /opt/homebrew/opt/bdw-gc/lib >> /opt/homebrew/opt/libpng/lib >> /opt/homebrew/opt/ntl/lib >> /opt/homebrew/opt/bzip2/lib >> /opt/homebrew/opt/readline/lib >> /opt/homebrew/lib >> /usr/local/lib >> Framework search paths: >> ld: Undefined symbols: >> _main, referenced from: >>
Re: [sage-release] Re: Sage 10.2.beta3 released
both meson and meson_python may potentially come from the system (homebrew in your case). Make sure they are up to date, and you don't have these packages installed in Sage. And run ./configure Perhaps the whole homebrew must be reinstated after an OS update, too. On Sun, 24 Sept 2023, 03:07 John H Palmieri, wrote: > Building Sage's own gfortran worked to build scipy (and this is an > argument to keep the gfortran package around, by the way). > > Alternatively, building all of scipy's dependencies and then using `./sage > --python3 -m pip install scipy` also seems to have worked. I then "touch"ed > the appropriate file to convince Sage that scipy had built, and the rest of > the build succeeded, except for some extra warnings during doctests: "ld: > warning: duplicate -rpath [...SAGE_ROOT/local/lib...] ignored" > > On Saturday, September 23, 2023 at 3:59:40 PM UTC-7 John H Palmieri wrote: > >> If by "they" you mean scipy, then: I have no problems building scipy on >> another OS X machine: an Intel machine running OS X 13.5.2 (rather than >> 13.6). "xcode-select --version" reports the same for both machines, but I >> don't know how informative this is. "gfortran --version" says >> >> GNU Fortran (Homebrew GCC 13.2.0) 13.2.0 >> Copyright (C) 2023 Free Software Foundation, Inc. >> This is free software; see the source for copying conditions. There is NO >> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR >> PURPOSE. >> >> and this is the one that scipy finds. On the M2 machine, Sage's scipy >> built back in August with earlier versions of the OS (probably 13.5.2) and >> command-line tools and gfortran (13.1.0). >> >> >> On Saturday, September 23, 2023 at 3:34:29 PM UTC-7 Dima Pasechnik wrote: >> >>> do they even support gcc/gfortran 13? >>> >>> On Sat, 23 Sept 2023, 23:04 John H Palmieri, >>> wrote: >>> Same result if I do `pip3 install scipy --no-binary scipy` (another of their suggested ways of building from source). I guess something is broken with my fortran compiler, but I don't know how to troubleshoot it. On Saturday, September 23, 2023 at 2:35:52 PM UTC-7 John H Palmieri wrote: > It didn't work for me. At least directly attempting a scipy build left > an intact log file. Any clues about what's broken on this machine? > > Build started at 2023-09-23T14:31:26.723020 > Main binary: /Users/palmieri/Downloads/scipy-1.11.2/venv/bin/python3 > Build Options: > -Dprefix=/Users/palmieri/Downloads/scipy-1.11.2/build-install > Python system: Darwin > The Meson build system > Version: 1.2.1 > Source dir: /Users/palmieri/Downloads/scipy-1.11.2 > Build dir: /Users/palmieri/Downloads/scipy-1.11.2/build > Build type: native build > Project name: SciPy > Project version: 1.11.2 > --- > Detecting compiler via: `cc --version` -> 0 > stdout: > Apple clang version 15.0.0 (clang-1500.0.40.1) > Target: arm64-apple-darwin22.6.0 > Thread model: posix > InstalledDir: /Library/Developer/CommandLineTools/usr/bin > --- > Running command: cc -E -dM - > - > --- > Detecting linker via: `cc -Wl,--version` -> 1 > stderr: > ld: unknown options: --version > clang: error: linker command failed with exit code 1 (use -v to see > invocation) > --- > --- > Detecting Apple linker via: `cc -Wl,-v` -> 1 > stderr: > @(#)PROGRAM:ld PROJECT:dyld-1015.7 > BUILD 18:48:48 Aug 22 2023 > configured to support archs: armv6 armv7 armv7s arm64 arm64e arm64_32 > i386 x86_64 x86_64h armv6m armv7k armv7m armv7em > will use ld-classic for: armv6 armv7 armv7s arm64_32 i386 armv6m > armv7k armv7m armv7em > LTO support using: LLVM version 15.0.0 (static support for 29, runtime > is 29) > TAPI support using: Apple TAPI version 15.0.0 (tapi-1500.0.12.3) > Library search paths: > /opt/homebrew/opt/primesieve/lib > /opt/homebrew/opt/bdw-gc/lib > /opt/homebrew/opt/libpng/lib > /opt/homebrew/opt/ntl/lib > /opt/homebrew/opt/bzip2/lib > /opt/homebrew/opt/readline/lib > /opt/homebrew/lib > /usr/local/lib > Framework search paths: > ld: Undefined symbols: > _main, referenced from: > > clang: error: linker command failed with exit code 1 (use -v to see > invocation) > --- > Sanity testing C compiler: cc > Is cross compiler: False. > Sanity check compiler command line: cc sanitycheckc.c -o > sanitycheckc.exe > Sanity check compile stdout: > > - > Sanity check compile stderr: > > - > Running test binary command: > > /Users/palmieri/Downloads/scipy-1.11.2/build/meson-private/sanitycheckc.exe > C compiler for the host machine: cc (clang 15.0.0 "Apple clang version > 15.0.0 (clang-1500.0.40.1)") > C linker for the host machine: cc ld64 1015.7 >
