Commit Graph

5 Commits

Author SHA1 Message Date
David Majnemer 12b9e76b62 Update for LLVM changes
InstSimplify has gained the ability to remove needless bitcasts which
perturbed some clang codegen tests.

llvm-svn: 276728
2016-07-26 05:52:37 +00:00
Silviu Baranga 632fdc5919 PR27216: Only define __ARM_FEATURE_FMA when the target has VFPv4
Summary:
According to the ACLE spec, "__ARM_FEATURE_FMA is defined to 1 if
the hardware floating-point architecture supports fused floating-point
multiply-accumulate".

This changes clang's behaviour from emitting this macro for v7-A and v7-R
cores to only emitting it when the target has VFPv4 (and therefore support
for the floating point multiply-accumulate instruction).

Fixes PR27216

Reviewers: t.p.northover, rengolin

Subscribers: aemerson, rengolin, cfe-commits

Differential Revision: http://reviews.llvm.org/D18963

llvm-svn: 267869
2016-04-28 11:29:08 +00:00
Tim Northover 58672974a9 ARM & AArch64: convert asm tests to LLVM IR and restrict optimizations.
This is mostly a one-time autoconversion of tests that checked assembly after
"-Owhatever" compiles to only run "opt -mem2reg" and check the assembly. This
should make them much more stable to changes in LLVM so they won't break on
unrelated changes.

"opt -mem2reg" is a compromise designed to increase the readability of tests
that check dataflow, while minimizing dependency on LLVM. Hopefully mem2reg is
stable enough that no surpises will come along.

Should address http://llvm.org/PR26815.

llvm-svn: 263048
2016-03-09 18:54:42 +00:00
Tim Northover bd7c169664 Fix recent test for more diverse environments.
I think the main issue was the lack of -ffreestanding, which pulled in
the host's stdint.h. After that things went rapidly downhill.

llvm-svn: 172653
2013-01-16 20:35:54 +00:00
Tim Northover 4ef746768b Correct order of operands forwarding NEON vfma to LLVM fma
llvm-svn: 172650
2013-01-16 20:13:15 +00:00