Arch Linux User Repository

If the build fails:

• Clear your aur helper cache and SRCPKGDEST directory
• Rebuild in clean chroot 1
• If it still fails, use a paste bin 2 to show full build logs

There's a package build already which you can try out from my repo.

FAQ:

• The dependencies are correct. fmt and nlohmann_json are configured as subprojects for ease of development on my end, and it's only natural to statically link C++ projects, as ABI stability with exported C++ libraries isn't a thing (compared to C).
• If you think pod2man is missing, it's a configuration problem on your end. pod2man is part of the perl package, but in a perl-specific PATH handled by /etc/profile.d/perlbin.sh
• I'm only able to test auracle on i686 and x86_64, so that's what I'm willing to commit to in the PKGBUILD. If you want to build this on some other architecture, use makepkg -A. The "any" architecture is reserved for packages with architecture independent files (and compiled C++ is not).

the package doesn't compile anymore with today's update to pacman 6.1.0:

ninja: Entering directory `/home/harvey/abs/auracle-git/src/auracle/build'
ninja: no work to do.
 1/13 auracle:libaur / libaur                      OK              0.06s
 2/13 auracle:libauracle / libauracle              FAIL            0.15s   killed by signal 6 SIGABRT
>>> ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 MALLOC_PERTURB_=145 /home/harvey/abs/auracle-git/src/auracle/build/libauracle_test

 3/13 auracle:auracle / test_outdated              OK              0.24s
 4/13 auracle:auracle / test_info                  OK              0.27s
 5/13 auracle:auracle / test_raw_query             OK              0.29s
 6/13 auracle:auracle / test_show                  OK              0.29s
 7/13 auracle:auracle / test_update                OK              0.29s
 8/13 auracle:auracle / test_custom_format         OK              0.34s
 9/13 auracle:auracle / test_buildorder            OK              0.35s
10/13 auracle:auracle / test_regex_search          FAIL            0.35s   exit status 1
>>> MALLOC_PERTURB_=203 ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 PYTHONDONTWRITEBYTECODE=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 /usr/bin/python3 -W all /home/harvey/abs/auracle-git/src/auracle/tests/test_regex_search.py

11/13 auracle:auracle / test_clone                 OK              0.38s
12/13 auracle:auracle / test_sort                  FAIL            0.63s   exit status 1
>>> MALLOC_PERTURB_=141 ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 PYTHONDONTWRITEBYTECODE=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 /usr/bin/python3 -W all /home/harvey/abs/auracle-git/src/auracle/tests/test_sort.py

13/13 auracle:auracle / test_search                FAIL            0.77s   exit status 1
>>> ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 PYTHONDONTWRITEBYTECODE=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 MALLOC_PERTURB_=127 /usr/bin/python3 -W all /home/harvey/abs/auracle-git/src/auracle/tests/test_search.py


Ok:                 9   
Expected Fail:      0   
Fail:               4   
Unexpected Pass:    0   
Skipped:            0   
Timeout:            0 

slondr: Unfortunately not all aarch64 CPU support the same set of instructions and abseil dependency uses some of those. The fact it's built successful in one aarch64 doesn't mean it will on some other.

Compiled perfectly fine without modification of any sort on my aarch64 pinephone, just using makepkg -A. Is there any appetite for an update to the PKGBUILD? If not, I can throw up an auracle-aarch64 package or something like that.

This package will not be buildable on older AMD processors due to enabling newer CPU features in menson, or is it something else?

Sounds like installing the newer pre-built version to use with my old processor would possibly be problematic as well?

If this is the case, perhaps someone could offer a fork for old CPU version with the appropriate mods. If I knew more, I'd do it...

Model name: AMD Phenom(tm) 9850 Quad-Core Processor

lscpu | awk '/Flags/ {$1=" "; print $0}' | xargs -n1 | sort | grep -Ei 'CMP|XCHG|L-|AHF|POP|SSE|SSS'

cmp_legacy

lahf_lm

misalignsse

popcnt

sse

sse2

sse4a

EDIT: I settled on downloading/installing into/from my local AUR repo, a prebuilt current version binary from chaotic-aur.

Warning! Chaotic AUR is a third-party repository

https://builds.garudalinux.org/repos/chaotic-aur/x86_64/auracle-git-r366.8739929-5-x86_64.pkg.tar.zst

@PS0: that's a limitation of meson. There is an open issue on meson about this and probably won't be fixed. :(

@artafinde It works for Auracle, but if someone would use the random library on an old CPU, it probably will fail as too many files are compiled with -sse4.1.

@PS0 If possible try replacing the PKGBUILD with this and let me know if it works. If yes I Will try to push it upstream.

@PS0 I see that is meson-wrap specific then - raised https://github.com/mesonbuild/wrapdb/issues/544

@artafinde abseil-cpp uses -maes and -msse4.1 only on randen_detect.cc and randen_hwaes.cc. The meson wrap enables these flags globally, which is not what the abseil-cpp developers intended.

@PS0, do you see the flags being used when you compile abseil-cpp with cmake? If yes then raise an issue with abseil-cpp on github. If not then it might be coming from the meson wrap. My AMD CPU has SSE4.1 extension so can't reproduce it.