devel/libatomic_ops: Update to 7.4.4
ClosedPublic
Actions

Authored by cpm on Jul 26 2016, 12:38 AM.

Details

Reviewers

AMDmi3
junovitch
rodrigo
feld

Commits

rP420785: - Update to 7.4.4

Summary

Update to 7.4.4
Add patch to fix boehm-gc on sparc

Changes:

Eliminate 'signed-to-unsigned value extension' compiler warning in malloc.
Eliminate 'variable set but not used' Cppcheck warnings in test_stack.
Fix GCC 5.x compatibility for AArch64 double-wide primitives.
Fix makefile preventing AO_pause undefined in libatomic_ops_gpl.
Fix missing casts to match printf format specifier in test_atomic.
Fix missing output folder on making auto-generated test files (Automake).
Fix typo in configure.ac (in description of AO_ASM_X64_AVAILABLE).
Minor fix of code alignment in mips AO_compare_and_swap.
Remove TODO file.
Restore contribution info in ChangeLog for authors not listed in git log.

Diff Detail

Lint

Lint Skipped

Unit

Tests Skipped

Event Timeline

cpm updated this revision to Diff 18744.Jul 26 2016, 12:38 AM

cpm retitled this revision from to devel/libatomic_ops: Update to 7.4.4.

cpm updated this object.

cpm edited the test plan for this revision. (Show Details)

cpm added reviewers: AMDmi3, feld, junovitch, rodrigo.

cpm updated this object.Jul 26 2016, 12:46 AM

All builds are fine via poudriere

# poudriere testport -j 93x32 -o devel/libatomic_ops [OK]
# poudriere testport -j 93x64 -o devel/libatomic_ops [OK]
# poudriere testport -j 102x32 -o devel/libatomic_ops [OK]
# poudriere testport -j 102x64 -o devel/libatomic_ops [OK]
# poudriere testport -j 103x32 -o devel/libatomic_ops [OK]
# poudriere testport -j 103x64 -o devel/libatomic_ops [OK]

Do you have a sparc machine? Curious where you got that patch from and/or how you tested it. Glad to see sparc not being ignored!

Macro lgtm:

This revision is now accepted and ready to land.Jul 26 2016, 1:54 PM

AMDmi3 accepted this revision.Jul 26 2016, 2:08 PM

AMDmi3 edited edge metadata.

AMDmi3 added inline comments.

devel/libatomic_ops/Makefile
5	Though there's little difference, I prefer to use PORTVERSION unless version is in some weird incompatible format

In D7311#152258, @feld wrote:

Do you have a sparc machine? Curious where you got that patch from and/or how you tested it. Glad to see sparc not being ignored!

An user reported an issue to the freebsd-sparc64 mailing-list that affect boehm-gc. The patch [0] was ported from OpenBSD and it fixes boehm-gc on sparc32.

Some changes are needed in boehm-gc port but I'm waiting his feedback.

[0] http://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/devel/boehm-gc/patches/patch-libatomic_ops_src_atomic_ops_sysdeps_gcc_sparc_h

rodrigo@ will take care of this.