amehsan closed this revision.
amehsan added a comment.
https://reviews.llvm.org/rL287872
https://reviews.llvm.org/D26544
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
kbarton accepted this revision.
kbarton added a comment.
This revision is now accepted and ready to land.
LGTM
https://reviews.llvm.org/D26544
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
amehsan updated the summary for this revision.
amehsan updated this revision to Diff 78376.
https://reviews.llvm.org/D26544
Files:
lib/Headers/altivec.h
test/CodeGen/builtins-ppc-altivec.c
test/CodeGen/builtins-ppc-p8vector.c
test/CodeGen/builtins-ppc-quadword.c
amehsan added inline comments.
Comment at: lib/Headers/altivec.h:350
+__tempc = __tempc & 0x0001;
+unsigned long long __longa = (unsigned long long) __tempa;
+unsigned long long __longb = (unsigned long long) __tempb;
kbarton wrote:
> nemanjai
kbarton added a comment.
There are several inline comments that need to be addressed.
I also think it's worthwhile putting a comment at the top of the review
indicating the (assumed) semantics for the vec_sube and vec_subec instructions
that are being implemented (i.e., the behaviour mimics the
nemanjai added inline comments.
Comment at: lib/Headers/altivec.h:314
+ vector signed int __carry = __c & __mask;
+ return vec_add(vec_add(__a, __b), __mask);
+}
I don't understand why we're adding `__mask` to the sum of `__a` and `__b`.
Shouldn't that be
