[Bug 274927] Toolchain fails on the __sync_val_compare_and_swap function without -march=native (port biology/seqwish)

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=274927

--- Comment #5 from Mark Millard <marklmi26-fbsd@yahoo.com> ---
Not reported about the build were the related warnings about
16 bytes exceeding the lock-free (text from my local reproduction
of the error via poudriere-devel), also tied to transclosure.cpp .
This might be of help for an investigation.

--- CMakeFiles/seqwish.dir/src/transclosure.cpp.o ---
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/src/transclosure.cpp:1:
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/src/transclosure.hpp:13:
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/mmmulti/src/mmiitree.hpp:19:
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o.hpp:38:
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/ips4o.hpp:45:
In file included from
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/memory.hpp:47:
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/bucket_pointers.hpp:106:28:
warning: large atomic operation may incur significant performance penalty; the
access size (16 bytes) exceeds the max lock-free size (8  bytes)
[-Watomic-alignment]
            const auto p = __atomic_fetch_sub(&all_,
                           ^
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/bucket_pointers.hpp:86:28:
warning: large atomic operation may incur significant performance penalty; the
access size (16 bytes) exceeds the max lock-free size (8  bytes)
[-Watomic-alignment]
            const auto p = __atomic_fetch_add(&all_, Cfg::kBlockSize,
__ATOMIC_RELAXED);
                           ^
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/bucket_pointers.hpp:106:28:
warning: large atomic operation may incur significant performance penalty; the
access size (16 bytes) exceeds the max lock-free size (8  bytes)
[-Watomic-alignment]
            const auto p = __atomic_fetch_sub(&all_,
                           ^
/wrkdirs/usr/ports/biology/seqwish/work/seqwish-v0.7.9/deps/ips4o/ips4o/bucket_pointers.hpp:86:28:
warning: large atomic operation may incur significant performance penalty; the
access size (16 bytes) exceeds the max lock-free size (8  bytes)
[-Watomic-alignment]
            const auto p = __atomic_fetch_add(&all_, Cfg::kBlockSize,
__ATOMIC_RELAXED);
                           ^

-- 
You are receiving this mail because:
You are the assignee for the bug.