This is an archive of the discontinued LLVM Phabricator instance.

Differential D9178

Fix correctness check for test_vec_fpextend_double
ClosedPublic

Authored by pirama on Apr 21 2015, 3:59 PM.

Details

Reviewers
dsanders
srhines
Commits
rG67e82482c0f6: Fix correctness check for test_vec_fpextend_double
rL235530: Fix correctness check for test_vec_fpextend_double
Summary

Remove the CHECK-DAG calls introduced in r235341, and add a comment that
this test may break due to scheduling variations.

This patch completes the fix discussed in http://reviews.llvm.org/D8804

Diff Detail

Repository
rL LLVM

Event Timeline

pirama updated this revision to Diff 24183.Apr 21 2015, 3:59 PM
pirama retitled this revision from to Fix correctness check for test_vec_fpextend_double.
pirama updated this object.
pirama edited the test plan for this revision. (Show Details)
pirama added reviewers: dsanders, srhines.
pirama added a subscriber: Unknown Object (MLST).
dsanders accepted this revision.Apr 22 2015, 3:55 AM
dsanders edited edge metadata.

LGTM, thanks.

This revision is now accepted and ready to land.Apr 22 2015, 3:55 AM
Closed by commit rL235530: Fix correctness check for test_vec_fpextend_double (authored by pirama). · Explain WhyApr 22 2015, 11:07 AM
This revision was automatically updated to reflect the committed changes.

Revision Contents

PathSize
llvm/
trunk/
test/
CodeGen/
Mips/
16 lines

Diff 24246

llvm/trunk/test/CodeGen/Mips/fp16-promote.ll

Show First 20 Lines • Show All 51 Lines • ▼ Show 20 Lines
; CHECK-LIBCALL: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
; CHECK-LIBCALL: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
define <4 x float> @test_vec_fpext_float(<4 x half>* %p) #0 { define <4 x float> @test_vec_fpext_float(<4 x half>* %p) #0 {
%a = load <4 x half>, <4 x half>* %p, align 8 %a = load <4 x half>, <4 x half>* %p, align 8
%b = fpext <4 x half> %a to <4 x float> %b = fpext <4 x half> %a to <4 x float>
ret <4 x float> %b ret <4 x float> %b
} }
; This test is not robust against variations in instruction scheduling.
; See the discussion in http://reviews.llvm.org/D8804
; CHECK-LIBCALL-LABEL: test_vec_fpext_double: ; CHECK-LIBCALL-LABEL: test_vec_fpext_double:
; CHECK-LIBCALL-DAG: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
; CHECK-LIBCALL-DAG: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
; CHECK-LIBCALL-DAG: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
; CHECK-LIBCALL-DAG: %call16(__gnu_h2f_ieee) ; CHECK-LIBCALL: cvt.d.s
; CHECK-LIBCALL-DAG: cvt.d.s ; CHECK-LIBCALL: cvt.d.s
; CHECK-LIBCALL-DAG: cvt.d.s ; CHECK-LIBCALL: cvt.d.s
; CHECK-LIBCALL-DAG: cvt.d.s ; CHECK-LIBCALL: %call16(__gnu_h2f_ieee)
; CHECK-LIBCALL: cvt.d.s ; CHECK-LIBCALL: cvt.d.s
define <4 x double> @test_vec_fpext_double(<4 x half>* %p) #0 { define <4 x double> @test_vec_fpext_double(<4 x half>* %p) #0 {
%a = load <4 x half>, <4 x half>* %p, align 8 %a = load <4 x half>, <4 x half>* %p, align 8
%b = fpext <4 x half> %a to <4 x double> %b = fpext <4 x half> %a to <4 x double>
ret <4 x double> %b ret <4 x double> %b
} }
; CHECK-LIBCALL-LABEL: test_vec_fptrunc_float: ; CHECK-LIBCALL-LABEL: test_vec_fptrunc_float:
Show All 21 Lines