maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/X86/splat-for-size.ll

241 lines
9.0 KiB
LLVM
Raw Normal View History

[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=CHECK --check-prefix=AVX
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=CHECK --check-prefix=AVX2
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; Check constant loads of every 128-bit and 256-bit vector type
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
; for size optimization using splat ops available with AVX and AVX2.
; There is no AVX broadcast from double to 128-bit vector because movddup has been around since SSE3 (grrr).
define <2 x double> @splat_v2f64(<2 x double> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-LABEL: splat_v2f64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Force floating point values in constant pool decoding to print in scientific notation so they can't be confused with integers. When the floating point constants are whole numbers they have no decimal point so look like integers, but mean something very different in something like an 'and' instruction. Ideally we would just print a decimal point and a 0, but I couldn't see how to make APFloat::toString do that. llvm-svn: 345488
2018-10-29 12:52:04 +08:00
; CHECK-NEXT: vmovddup {{.*#+}} xmm1 = [1.0E+0,1.0E+0]
[X86] Add 128 MOVDDUP to the constant pool printing in X86AsmPrinter::EmitInstruction. We use this instruction to broadcast a single 64-bit value to a v2i64/v2f64 vector. llvm-svn: 344486
2018-10-15 09:51:53 +08:00
; CHECK-NEXT: # xmm1 = mem[0,0]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-NEXT: vaddpd %xmm1, %xmm0, %xmm0
; CHECK-NEXT: retq
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
%add = fadd <2 x double> %x, <double 1.0, double 1.0>
ret <2 x double> %add
}
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <4 x double> @splat_v4f64(<4 x double> %x) #1 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-LABEL: splat_v4f64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Force floating point values in constant pool decoding to print in scientific notation so they can't be confused with integers. When the floating point constants are whole numbers they have no decimal point so look like integers, but mean something very different in something like an 'and' instruction. Ideally we would just print a decimal point and a 0, but I couldn't see how to make APFloat::toString do that. llvm-svn: 345488
2018-10-29 12:52:04 +08:00
; CHECK-NEXT: vbroadcastsd {{.*#+}} ymm1 = [1.0E+0,1.0E+0,1.0E+0,1.0E+0]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-NEXT: vaddpd %ymm1, %ymm0, %ymm0
; CHECK-NEXT: retq
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
%add = fadd <4 x double> %x, <double 1.0, double 1.0, double 1.0, double 1.0>
ret <4 x double> %add
}
define <4 x float> @splat_v4f32(<4 x float> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-LABEL: splat_v4f32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Force floating point values in constant pool decoding to print in scientific notation so they can't be confused with integers. When the floating point constants are whole numbers they have no decimal point so look like integers, but mean something very different in something like an 'and' instruction. Ideally we would just print a decimal point and a 0, but I couldn't see how to make APFloat::toString do that. llvm-svn: 345488
2018-10-29 12:52:04 +08:00
; CHECK-NEXT: vbroadcastss {{.*#+}} xmm1 = [1.0E+0,1.0E+0,1.0E+0,1.0E+0]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-NEXT: vaddps %xmm1, %xmm0, %xmm0
; CHECK-NEXT: retq
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
%add = fadd <4 x float> %x, <float 1.0, float 1.0, float 1.0, float 1.0>
ret <4 x float> %add
}
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <8 x float> @splat_v8f32(<8 x float> %x) #1 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-LABEL: splat_v8f32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Force floating point values in constant pool decoding to print in scientific notation so they can't be confused with integers. When the floating point constants are whole numbers they have no decimal point so look like integers, but mean something very different in something like an 'and' instruction. Ideally we would just print a decimal point and a 0, but I couldn't see how to make APFloat::toString do that. llvm-svn: 345488
2018-10-29 12:52:04 +08:00
; CHECK-NEXT: vbroadcastss {{.*#+}} ymm1 = [1.0E+0,1.0E+0,1.0E+0,1.0E+0,1.0E+0,1.0E+0,1.0E+0,1.0E+0]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; CHECK-NEXT: vaddps %ymm1, %ymm0, %ymm0
; CHECK-NEXT: retq
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
%add = fadd <8 x float> %x, <float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0>
ret <8 x float> %add
}
; AVX can't do integer splats, so fake it: use vmovddup to splat 64-bit value.
; We also generate vmovddup for AVX2 because it's one byte smaller than vpbroadcastq.
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244460
2015-08-11 00:47:47 +08:00
define <2 x i64> @splat_v2i64(<2 x i64> %x) #1 {
[X86] Cleanup patterns for using VMOVDDUP for broadcasts. -Remove OptForSize. Not all of the backend follows the same rules for creating broadcasts and there is no conflicting pattern. -Don't stop selecting VEX VMOVDDUP when AVX512 is supported. We need VLX for EVEX VMOVDDUP. -Only use VMOVDDUP for v2i64 broadcasts if AVX2 is not supported. llvm-svn: 283020
2016-10-01 15:11:24 +08:00
; AVX-LABEL: splat_v2i64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86] Add 128 MOVDDUP to the constant pool printing in X86AsmPrinter::EmitInstruction. We use this instruction to broadcast a single 64-bit value to a v2i64/v2f64 vector. llvm-svn: 344486
2018-10-15 09:51:53 +08:00
; AVX-NEXT: vmovddup {{.*#+}} xmm1 = [2,2]
; AVX-NEXT: # xmm1 = mem[0,0]
[X86] Cleanup patterns for using VMOVDDUP for broadcasts. -Remove OptForSize. Not all of the backend follows the same rules for creating broadcasts and there is no conflicting pattern. -Don't stop selecting VEX VMOVDDUP when AVX512 is supported. We need VLX for EVEX VMOVDDUP. -Only use VMOVDDUP for v2i64 broadcasts if AVX2 is not supported. llvm-svn: 283020
2016-10-01 15:11:24 +08:00
; AVX-NEXT: vpaddq %xmm1, %xmm0, %xmm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v2i64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastq {{.*#+}} xmm1 = [2,2]
[X86] Cleanup patterns for using VMOVDDUP for broadcasts. -Remove OptForSize. Not all of the backend follows the same rules for creating broadcasts and there is no conflicting pattern. -Don't stop selecting VEX VMOVDDUP when AVX512 is supported. We need VLX for EVEX VMOVDDUP. -Only use VMOVDDUP for v2i64 broadcasts if AVX2 is not supported. llvm-svn: 283020
2016-10-01 15:11:24 +08:00
; AVX2-NEXT: vpaddq %xmm1, %xmm0, %xmm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <2 x i64> %x, <i64 2, i64 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <2 x i64> %add
}
; AVX can't do 256-bit integer ops, so we split this into two 128-bit vectors,
; and then we fake it: use vmovddup to splat 64-bit value.
define <4 x i64> @splat_v4i64(<4 x i64> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v4i64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vextractf128 $1, %ymm0, %xmm1
[X86] Add 128 MOVDDUP to the constant pool printing in X86AsmPrinter::EmitInstruction. We use this instruction to broadcast a single 64-bit value to a v2i64/v2f64 vector. llvm-svn: 344486
2018-10-15 09:51:53 +08:00
; AVX-NEXT: vmovddup {{.*#+}} xmm2 = [2,2]
; AVX-NEXT: # xmm2 = mem[0,0]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddq %xmm2, %xmm1, %xmm1
; AVX-NEXT: vpaddq %xmm2, %xmm0, %xmm0
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v4i64:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastq {{.*#+}} ymm1 = [2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddq %ymm1, %ymm0, %ymm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <4 x i64> %x, <i64 2, i64 2, i64 2, i64 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <4 x i64> %add
}
; AVX can't do integer splats, so fake it: use vbroadcastss to splat 32-bit value.
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <4 x i32> @splat_v4i32(<4 x i32> %x) #1 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v4i32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX-NEXT: vbroadcastss {{.*#+}} xmm1 = [2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddd %xmm1, %xmm0, %xmm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v4i32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastd {{.*#+}} xmm1 = [2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddd %xmm1, %xmm0, %xmm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <4 x i32> %x, <i32 2, i32 2, i32 2, i32 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <4 x i32> %add
}
; AVX can't do integer splats, so fake it: use vbroadcastss to splat 32-bit value.
define <8 x i32> @splat_v8i32(<8 x i32> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v8i32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vextractf128 $1, %ymm0, %xmm1
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX-NEXT: vbroadcastss {{.*#+}} xmm2 = [2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddd %xmm2, %xmm1, %xmm1
; AVX-NEXT: vpaddd %xmm2, %xmm0, %xmm0
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v8i32:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm1 = [2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddd %ymm1, %ymm0, %ymm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <8 x i32> %x, <i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2, i32 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <8 x i32> %add
}
; AVX can't do integer splats, and there's no broadcast fakery for 16-bit. Could use pshuflw, etc?
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <8 x i16> @splat_v8i16(<8 x i16> %x) #1 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v8i16:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddw {{.*}}(%rip), %xmm0, %xmm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v8i16:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastw {{.*#+}} xmm1 = [2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddw %xmm1, %xmm0, %xmm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <8 x i16> %x, <i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <8 x i16> %add
}
; AVX can't do integer splats, and there's no broadcast fakery for 16-bit. Could use pshuflw, etc?
define <16 x i16> @splat_v16i16(<16 x i16> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v16i16:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vextractf128 $1, %ymm0, %xmm1
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
; AVX-NEXT: vmovdqa {{.*#+}} xmm2 = [2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddw %xmm2, %xmm1, %xmm1
; AVX-NEXT: vpaddw %xmm2, %xmm0, %xmm0
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v16i16:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastw {{.*#+}} ymm1 = [2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddw %ymm1, %ymm0, %ymm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <16 x i16> %x, <i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2, i16 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <16 x i16> %add
}
; AVX can't do integer splats, and there's no broadcast fakery for 8-bit. Could use pshufb, etc?
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <16 x i8> @splat_v16i8(<16 x i8> %x) #1 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v16i8:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddb {{.*}}(%rip), %xmm0, %xmm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v16i8:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastb {{.*#+}} xmm1 = [2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddb %xmm1, %xmm0, %xmm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <16 x i8> %x, <i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <16 x i8> %add
}
; AVX can't do integer splats, and there's no broadcast fakery for 8-bit. Could use pshufb, etc?
define <32 x i8> @splat_v32i8(<32 x i8> %x) #0 {
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-LABEL: splat_v32i8:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX: # %bb.0:
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vextractf128 $1, %ymm0, %xmm1
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
; AVX-NEXT: vmovdqa {{.*#+}} xmm2 = [2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX-NEXT: vpaddb %xmm2, %xmm1, %xmm1
; AVX-NEXT: vpaddb %xmm2, %xmm0, %xmm0
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: retq
;
; AVX2-LABEL: splat_v32i8:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; AVX2: # %bb.0:
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX2-NEXT: vpbroadcastb {{.*#+}} ymm1 = [2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2]
[X86][AVX] Regenerate Splat OptSize tests Tidied up triple and regenerate tests using update_llc_test_checks.py llvm-svn: 254060
2015-11-25 17:06:17 +08:00
; AVX2-NEXT: vpaddb %ymm1, %ymm0, %ymm0
; AVX2-NEXT: retq
[x86] adjust test constants to maintain coverage; NFC Increment (add 1) could be transformed to sub -1, and we'd lose coverage for these patterns. llvm-svn: 305644
2017-06-18 22:01:32 +08:00
%add = add <32 x i8> %x, <i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2, i8 2>
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
ret <32 x i8> %add
}
[X86][AVX] Fix failure due to a missing ISel pattern to select VBROADCAST nodes (PR23259). This fixes a regression introduced at revision 218263. On AVX, if we optimize for size, a splat build_vector of a load is lowered into a VBROADCAST node. This is done even if the value type of the splat build_vector node is v2i64. Since AVX doesn't support v2f64/v2i64 broadcasts, revision 218263 added two extra tablegen patterns to allow selecting a VMOVDDUPrm from an X86VBroadcast where the scalar element comes from a loadi64/loadf64. However, revision 218263 forgot to add an extra fallback pattern for the case where we have a X86VBroadcast of a loadi64 with multiple uses. This patch adds the missing tablegen pattern in X86InstrSSE.td. This patch also adds an extra test to 'splat-for-size.ll' to verify that ISel doesn't crash with a 'fatal error in the backend' due to a missing AVX pattern to select v2i64 X86ISD::BROADCAST nodes. llvm-svn: 235509
2015-04-22 22:53:39 +08:00
; PR23259: Verify that ISel doesn't crash with a 'fatal error in backend'
; due to a missing AVX pattern to select a v2i64 X86ISD::BROADCAST of a
Fix comment in test. NFC. llvm-svn: 236262
2015-05-01 05:22:28 +08:00
; loadi64 with multiple uses.
[X86][AVX] Fix failure due to a missing ISel pattern to select VBROADCAST nodes (PR23259). This fixes a regression introduced at revision 218263. On AVX, if we optimize for size, a splat build_vector of a load is lowered into a VBROADCAST node. This is done even if the value type of the splat build_vector node is v2i64. Since AVX doesn't support v2f64/v2i64 broadcasts, revision 218263 added two extra tablegen patterns to allow selecting a VMOVDDUPrm from an X86VBroadcast where the scalar element comes from a loadi64/loadf64. However, revision 218263 forgot to add an extra fallback pattern for the case where we have a X86VBroadcast of a loadi64 with multiple uses. This patch adds the missing tablegen pattern in X86InstrSSE.td. This patch also adds an extra test to 'splat-for-size.ll' to verify that ISel doesn't crash with a 'fatal error in the backend' due to a missing AVX pattern to select v2i64 X86ISD::BROADCAST nodes. llvm-svn: 235509
2015-04-22 22:53:39 +08:00
@A = common global <3 x i64> zeroinitializer, align 32
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
define <8 x i64> @pr23259() #1 {
[X86] Autogenerate complete checks. NFC llvm-svn: 344485
2018-10-15 09:51:50 +08:00
; AVX-LABEL: pr23259:
; AVX: # %bb.0: # %entry
; AVX-NEXT: pushq $1
; AVX-NEXT: .cfi_adjust_cfa_offset 8
; AVX-NEXT: popq %rax
; AVX-NEXT: .cfi_adjust_cfa_offset -8
; AVX-NEXT: vmovq %rax, %xmm0
; AVX-NEXT: vinsertf128 $1, %xmm0, %ymm0, %ymm1
; AVX-NEXT: vperm2f128 {{.*#+}} ymm0 = mem[2,3],ymm0[0,1]
; AVX-NEXT: vunpcklpd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
; AVX-NEXT: vbroadcastsd {{.*#+}} ymm1 = [1,1,1,1]
; AVX-NEXT: retq
;
; AVX2-LABEL: pr23259:
; AVX2: # %bb.0: # %entry
; AVX2-NEXT: vmovdqa {{.*}}(%rip), %ymm0
; AVX2-NEXT: pushq $1
; AVX2-NEXT: .cfi_adjust_cfa_offset 8
; AVX2-NEXT: popq %rax
; AVX2-NEXT: .cfi_adjust_cfa_offset -8
; AVX2-NEXT: vmovq %rax, %xmm1
; AVX2-NEXT: vpunpcklqdq {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[2],ymm1[2]
; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[2,1,1,1]
; AVX2-NEXT: vbroadcastsd {{.*#+}} ymm1 = [1,1,1,1]
; AVX2-NEXT: retq
[X86][AVX] Fix failure due to a missing ISel pattern to select VBROADCAST nodes (PR23259). This fixes a regression introduced at revision 218263. On AVX, if we optimize for size, a splat build_vector of a load is lowered into a VBROADCAST node. This is done even if the value type of the splat build_vector node is v2i64. Since AVX doesn't support v2f64/v2i64 broadcasts, revision 218263 added two extra tablegen patterns to allow selecting a VMOVDDUPrm from an X86VBroadcast where the scalar element comes from a loadi64/loadf64. However, revision 218263 forgot to add an extra fallback pattern for the case where we have a X86VBroadcast of a loadi64 with multiple uses. This patch adds the missing tablegen pattern in X86InstrSSE.td. This patch also adds an extra test to 'splat-for-size.ll' to verify that ISel doesn't crash with a 'fatal error in the backend' due to a missing AVX pattern to select v2i64 X86ISD::BROADCAST nodes. llvm-svn: 235509
2015-04-22 22:53:39 +08:00
entry:
%0 = load <4 x i64>, <4 x i64>* bitcast (<3 x i64>* @A to <4 x i64>*), align 32
%1 = shufflevector <4 x i64> %0, <4 x i64> undef, <3 x i32> <i32 undef, i32 undef, i32 2>
%shuffle = shufflevector <3 x i64> <i64 1, i64 undef, i64 undef>, <3 x i64> %1, <8 x i32> <i32 5, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
ret <8 x i64> %shuffle
}
Use broadcasts to optimize overall size when loading constant splat vectors (x86-64 with AVX or AVX2). We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors. This patch should preserve all existing behavior with regular optimization levels, but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2. The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save at least 8 bytes (up to 31 bytes) of constant pool data. Differential Revision: http://reviews.llvm.org/D5347 llvm-svn: 218263
2014-09-23 02:54:01 +08:00
attributes #0 = { optsize }
fix minsize detection: minsize attribute implies optimizing for size llvm-svn: 244458
2015-08-11 00:43:20 +08:00
attributes #1 = { minsize }