[X86] vector-reduce-or-cmp.ll - add v4i64 signbit test coverage

[RKSimon]

Missing fold to make use of VTESTPD

[llvmbot]

@llvm/pr-subscribers-backend-x86

Author: Simon Pilgrim (RKSimon)

Changes

Missing fold to make use of VTESTPD

---

Full diff: https://github.com/llvm/llvm-project/pull/165588.diff

1 Files Affected:

• (modified) llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll (+57-12)

diff --git a/llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll b/llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll
index ab1feba98b008..9816fa7c83560 100644
--- a/llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll
+++ b/llvm/test/CodeGen/X86/vector-reduce-or-cmp.ll
@@ -992,6 +992,51 @@ define i1 @signtest_v8i32(<8 x i32> %a0) {
   ret i1 %2
 }
 
+define i1 @signtest_v4i64(<4 x i64> %a0) {
+; SSE2-LABEL: signtest_v4i64:
+; SSE2:       # %bb.0:
+; SSE2-NEXT:    por %xmm1, %xmm0
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
+; SSE2-NEXT:    por %xmm0, %xmm1
+; SSE2-NEXT:    movq %xmm1, %rax
+; SSE2-NEXT:    testq %rax, %rax
+; SSE2-NEXT:    setns %al
+; SSE2-NEXT:    retq
+;
+; SSE41-LABEL: signtest_v4i64:
+; SSE41:       # %bb.0:
+; SSE41-NEXT:    por %xmm1, %xmm0
+; SSE41-NEXT:    ptest {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
+; SSE41-NEXT:    sete %al
+; SSE41-NEXT:    retq
+;
+; AVX1-LABEL: signtest_v4i64:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vptest {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0
+; AVX1-NEXT:    sete %al
+; AVX1-NEXT:    vzeroupper
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: signtest_v4i64:
+; AVX2:       # %bb.0:
+; AVX2-NEXT:    vpbroadcastq {{.*#+}} ymm1 = [9223372036854775808,9223372036854775808,9223372036854775808,9223372036854775808]
+; AVX2-NEXT:    vptest %ymm1, %ymm0
+; AVX2-NEXT:    sete %al
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
+;
+; AVX512-LABEL: signtest_v4i64:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vpbroadcastq {{.*#+}} ymm1 = [9223372036854775808,9223372036854775808,9223372036854775808,9223372036854775808]
+; AVX512-NEXT:    vptest %ymm1, %ymm0
+; AVX512-NEXT:    sete %al
+; AVX512-NEXT:    vzeroupper
+; AVX512-NEXT:    retq
+  %1 = call i64 @llvm.vector.reduce.or.v4i64(<4 x i64> %a0)
+  %2 = icmp sgt i64 %1, -1
+  ret i1 %2
+}
+
 define i1 @trunc_v16i16(<16 x i16> %a0) {
 ; SSE2-LABEL: trunc_v16i16:
 ; SSE2:       # %bb.0:
@@ -1162,11 +1207,11 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; SSE2-NEXT:    movd %xmm0, %eax
 ; SSE2-NEXT:    orl %ecx, %eax
 ; SSE2-NEXT:    testb $1, %al
-; SSE2-NEXT:    je .LBB29_2
+; SSE2-NEXT:    je .LBB30_2
 ; SSE2-NEXT:  # %bb.1:
 ; SSE2-NEXT:    xorl %eax, %eax
 ; SSE2-NEXT:    retq
-; SSE2-NEXT:  .LBB29_2:
+; SSE2-NEXT:  .LBB30_2:
 ; SSE2-NEXT:    movl $1, %eax
 ; SSE2-NEXT:    retq
 ;
@@ -1181,11 +1226,11 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; SSE41-NEXT:    pextrd $2, %xmm1, %eax
 ; SSE41-NEXT:    orl %ecx, %eax
 ; SSE41-NEXT:    testb $1, %al
-; SSE41-NEXT:    je .LBB29_2
+; SSE41-NEXT:    je .LBB30_2
 ; SSE41-NEXT:  # %bb.1:
 ; SSE41-NEXT:    xorl %eax, %eax
 ; SSE41-NEXT:    retq
-; SSE41-NEXT:  .LBB29_2:
+; SSE41-NEXT:  .LBB30_2:
 ; SSE41-NEXT:    movl $1, %eax
 ; SSE41-NEXT:    retq
 ;
@@ -1200,11 +1245,11 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; AVX1OR2-NEXT:    vpextrd $2, %xmm0, %eax
 ; AVX1OR2-NEXT:    orl %ecx, %eax
 ; AVX1OR2-NEXT:    testb $1, %al
-; AVX1OR2-NEXT:    je .LBB29_2
+; AVX1OR2-NEXT:    je .LBB30_2
 ; AVX1OR2-NEXT:  # %bb.1:
 ; AVX1OR2-NEXT:    xorl %eax, %eax
 ; AVX1OR2-NEXT:    retq
-; AVX1OR2-NEXT:  .LBB29_2:
+; AVX1OR2-NEXT:  .LBB30_2:
 ; AVX1OR2-NEXT:    movl $1, %eax
 ; AVX1OR2-NEXT:    retq
 ;
@@ -1219,12 +1264,12 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; AVX512F-NEXT:    korw %k0, %k1, %k0
 ; AVX512F-NEXT:    kmovw %k0, %eax
 ; AVX512F-NEXT:    testb $1, %al
-; AVX512F-NEXT:    je .LBB29_2
+; AVX512F-NEXT:    je .LBB30_2
 ; AVX512F-NEXT:  # %bb.1:
 ; AVX512F-NEXT:    xorl %eax, %eax
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
-; AVX512F-NEXT:  .LBB29_2:
+; AVX512F-NEXT:  .LBB30_2:
 ; AVX512F-NEXT:    movl $1, %eax
 ; AVX512F-NEXT:    vzeroupper
 ; AVX512F-NEXT:    retq
@@ -1240,12 +1285,12 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; AVX512BW-NEXT:    korw %k0, %k1, %k0
 ; AVX512BW-NEXT:    kmovd %k0, %eax
 ; AVX512BW-NEXT:    testb $1, %al
-; AVX512BW-NEXT:    je .LBB29_2
+; AVX512BW-NEXT:    je .LBB30_2
 ; AVX512BW-NEXT:  # %bb.1:
 ; AVX512BW-NEXT:    xorl %eax, %eax
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
-; AVX512BW-NEXT:  .LBB29_2:
+; AVX512BW-NEXT:  .LBB30_2:
 ; AVX512BW-NEXT:    movl $1, %eax
 ; AVX512BW-NEXT:    vzeroupper
 ; AVX512BW-NEXT:    retq
@@ -1259,11 +1304,11 @@ define i32 @mask_v3i1(<3 x i32> %a, <3 x i32> %b) {
 ; AVX512BWVL-NEXT:    korw %k0, %k1, %k0
 ; AVX512BWVL-NEXT:    kmovd %k0, %eax
 ; AVX512BWVL-NEXT:    testb $1, %al
-; AVX512BWVL-NEXT:    je .LBB29_2
+; AVX512BWVL-NEXT:    je .LBB30_2
 ; AVX512BWVL-NEXT:  # %bb.1:
 ; AVX512BWVL-NEXT:    xorl %eax, %eax
 ; AVX512BWVL-NEXT:    retq
-; AVX512BWVL-NEXT:  .LBB29_2:
+; AVX512BWVL-NEXT:  .LBB30_2:
 ; AVX512BWVL-NEXT:    movl $1, %eax
 ; AVX512BWVL-NEXT:    retq
   %1 = icmp ne <3 x i32> %a, %b

[llvm-ci]

LLVM Buildbot has detected a new failure on builder llvm-clang-x86_64-darwin running on doug-worker-3 while building llvm at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/23/builds/15097

Here is the relevant piece of the build log for the reference

Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'Clang :: Index/pch-with-module.m' FAILED ********************
Exit Code: 127

Command Output (stdout):
--
# RUN: at line 2
rm -rf /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.cache
# executed command: rm -rf /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.cache
# note: command had no output on stdout or stderr
# RUN: at line 3
rm -rf /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.h.pch
# executed command: rm -rf /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.h.pch
# note: command had no output on stdout or stderr
# RUN: at line 4
/Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/bin/c-index-test -write-pch /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.h.pch /Volumes/RAMDisk/buildbot-root/x86_64-darwin/llvm-project/clang/test/Index/pch-with-module.m -target x86_64-apple-macosx10.7 -fobjc-arc -fmodules-cache-path=/Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.cache -fmodules -F /Volumes/RAMDisk/buildbot-root/x86_64-darwin/llvm-project/clang/test/Index/../Modules/Inputs -Xclang -fdisable-module-hash
# executed command: /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/bin/c-index-test -write-pch /Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.h.pch /Volumes/RAMDisk/buildbot-root/x86_64-darwin/llvm-project/clang/test/Index/pch-with-module.m -target x86_64-apple-macosx10.7 -fobjc-arc -fmodules-cache-path=/Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/tools/clang/test/Index/Output/pch-with-module.m.tmp.cache -fmodules -F /Volumes/RAMDisk/buildbot-root/x86_64-darwin/llvm-project/clang/test/Index/../Modules/Inputs -Xclang -fdisable-module-hash
# .---command stderr------------
# | Could not create process (/Volumes/RAMDisk/buildbot-root/x86_64-darwin/build/bin/c-index-test) due to [Errno 24] Too many open files
# `-----------------------------
# error: command failed with exit status: 127

--

********************