mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2025-02-01 05:01:59 +01:00
054071c5ba
Summary:
# Overview
This is somewhat partial.
* Latencies are good {F7371125}
* All of these remaining inconsistencies //appear// to be noise/noisy/flaky.
* NumMicroOps are somewhat good {F7371158}
* Most of the remaining inconsistencies are from `Ld` / `Ld_ReadAfterLd` classes
* Actual unit occupation (pipes, `ResourceCycles`) are undiscovered lands, i did not really look there.
They are basically verbatum copy from `btver2`
* Many `InstRW`. And there are still inconsistencies left...
To be noted:
I think this is the first new schedule profile produced with the new next-gen tools like llvm-exegesis!
# Benchmark
I realize that isn't what was suggested, but i'll start with some "internal" public real-world benchmark i understand - [[ https://github.com/darktable-org/rawspeed | RawSpeed raw image decoding library ]].
Diff (the exact clang from trunk without/with this patch):
```
Comparing /home/lebedevri/rawspeed/build-old/src/utilities/rsbench/rsbench to /home/lebedevri/rawspeed/build-new/src/utilities/rsbench/rsbench
Benchmark Time CPU Time Old Time New CPU Old CPU New
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Canon/EOS 5D Mark II/09.canon.sraw1.cr2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Canon/EOS 5D Mark II/09.canon.sraw1.cr2/threads:8/real_time_mean -0.0607 -0.0604 234 219 233 219
Canon/EOS 5D Mark II/09.canon.sraw1.cr2/threads:8/real_time_median -0.0630 -0.0626 233 219 233 219
Canon/EOS 5D Mark II/09.canon.sraw1.cr2/threads:8/real_time_stddev +0.2581 +0.2587 1 2 1 2
Canon/EOS 5D Mark II/10.canon.sraw2.cr2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Canon/EOS 5D Mark II/10.canon.sraw2.cr2/threads:8/real_time_mean -0.0770 -0.0767 144 133 144 133
Canon/EOS 5D Mark II/10.canon.sraw2.cr2/threads:8/real_time_median -0.0767 -0.0763 144 133 144 133
Canon/EOS 5D Mark II/10.canon.sraw2.cr2/threads:8/real_time_stddev -0.4170 -0.4156 1 0 1 0
Canon/EOS 5DS/2K4A9927.CR2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Canon/EOS 5DS/2K4A9927.CR2/threads:8/real_time_mean -0.0271 -0.0270 463 450 463 450
Canon/EOS 5DS/2K4A9927.CR2/threads:8/real_time_median -0.0093 -0.0093 453 449 453 449
Canon/EOS 5DS/2K4A9927.CR2/threads:8/real_time_stddev -0.7280 -0.7280 13 4 13 4
Canon/EOS 5DS/2K4A9928.CR2/threads:8/real_time_pvalue 0.0004 0.0004 U Test, Repetitions: 25 vs 25
Canon/EOS 5DS/2K4A9928.CR2/threads:8/real_time_mean -0.0065 -0.0065 569 565 569 565
Canon/EOS 5DS/2K4A9928.CR2/threads:8/real_time_median -0.0077 -0.0077 569 564 569 564
Canon/EOS 5DS/2K4A9928.CR2/threads:8/real_time_stddev +1.0077 +1.0068 2 5 2 5
Canon/EOS 5DS/2K4A9929.CR2/threads:8/real_time_pvalue 0.0220 0.0199 U Test, Repetitions: 25 vs 25
Canon/EOS 5DS/2K4A9929.CR2/threads:8/real_time_mean +0.0006 +0.0007 312 312 312 312
Canon/EOS 5DS/2K4A9929.CR2/threads:8/real_time_median +0.0031 +0.0032 311 312 311 312
Canon/EOS 5DS/2K4A9929.CR2/threads:8/real_time_stddev -0.7069 -0.7072 4 1 4 1
Canon/EOS 10D/CRW_7673.CRW/threads:8/real_time_pvalue 0.0004 0.0004 U Test, Repetitions: 25 vs 25
Canon/EOS 10D/CRW_7673.CRW/threads:8/real_time_mean -0.0015 -0.0015 141 141 141 141
Canon/EOS 10D/CRW_7673.CRW/threads:8/real_time_median -0.0010 -0.0011 141 141 141 141
Canon/EOS 10D/CRW_7673.CRW/threads:8/real_time_stddev -0.1486 -0.1456 0 0 0 0
Canon/EOS 40D/_MG_0154.CR2/threads:8/real_time_pvalue 0.6139 0.8766 U Test, Repetitions: 25 vs 25
Canon/EOS 40D/_MG_0154.CR2/threads:8/real_time_mean -0.0008 -0.0005 60 60 60 60
Canon/EOS 40D/_MG_0154.CR2/threads:8/real_time_median -0.0006 -0.0002 60 60 60 60
Canon/EOS 40D/_MG_0154.CR2/threads:8/real_time_stddev -0.1467 -0.1390 0 0 0 0
Canon/EOS 77D/IMG_4049.CR2/threads:8/real_time_pvalue 0.0137 0.0137 U Test, Repetitions: 25 vs 25
Canon/EOS 77D/IMG_4049.CR2/threads:8/real_time_mean +0.0002 +0.0002 275 275 275 275
Canon/EOS 77D/IMG_4049.CR2/threads:8/real_time_median -0.0015 -0.0014 275 275 275 275
Canon/EOS 77D/IMG_4049.CR2/threads:8/real_time_stddev +3.3687 +3.3587 0 2 0 2
Canon/PowerShot G1/crw_1693.crw/threads:8/real_time_pvalue 0.4041 0.3933 U Test, Repetitions: 25 vs 25
Canon/PowerShot G1/crw_1693.crw/threads:8/real_time_mean +0.0004 +0.0004 67 67 67 67
Canon/PowerShot G1/crw_1693.crw/threads:8/real_time_median -0.0000 -0.0000 67 67 67 67
Canon/PowerShot G1/crw_1693.crw/threads:8/real_time_stddev +0.1947 +0.1995 0 0 0 0
Fujifilm/GFX 50S/20170525_0037TEST.RAF/threads:8/real_time_pvalue 0.0074 0.0001 U Test, Repetitions: 25 vs 25
Fujifilm/GFX 50S/20170525_0037TEST.RAF/threads:8/real_time_mean -0.0092 +0.0074 547 542 25 25
Fujifilm/GFX 50S/20170525_0037TEST.RAF/threads:8/real_time_median -0.0054 +0.0115 544 541 25 25
Fujifilm/GFX 50S/20170525_0037TEST.RAF/threads:8/real_time_stddev -0.4086 -0.3486 8 5 0 0
Fujifilm/X-Pro2/_DSF3051.RAF/threads:8/real_time_pvalue 0.3320 0.0000 U Test, Repetitions: 25 vs 25
Fujifilm/X-Pro2/_DSF3051.RAF/threads:8/real_time_mean +0.0015 +0.0204 218 218 12 12
Fujifilm/X-Pro2/_DSF3051.RAF/threads:8/real_time_median +0.0001 +0.0203 218 218 12 12
Fujifilm/X-Pro2/_DSF3051.RAF/threads:8/real_time_stddev +0.2259 +0.2023 1 1 0 0
GoPro/HERO6 Black/GOPR9172.GPR/threads:8/real_time_pvalue 0.0000 0.0001 U Test, Repetitions: 25 vs 25
GoPro/HERO6 Black/GOPR9172.GPR/threads:8/real_time_mean -0.0209 -0.0179 96 94 90 88
GoPro/HERO6 Black/GOPR9172.GPR/threads:8/real_time_median -0.0182 -0.0155 95 93 90 88
GoPro/HERO6 Black/GOPR9172.GPR/threads:8/real_time_stddev -0.6164 -0.2703 2 1 2 1
Kodak/DCS Pro 14nx/D7465857.DCR/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Kodak/DCS Pro 14nx/D7465857.DCR/threads:8/real_time_mean -0.0098 -0.0098 176 175 176 175
Kodak/DCS Pro 14nx/D7465857.DCR/threads:8/real_time_median -0.0126 -0.0126 176 174 176 174
Kodak/DCS Pro 14nx/D7465857.DCR/threads:8/real_time_stddev +6.9789 +6.9157 0 2 0 2
Nikon/D850/Nikon-D850-14bit-lossless-compressed.NEF/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Nikon/D850/Nikon-D850-14bit-lossless-compressed.NEF/threads:8/real_time_mean -0.0237 -0.0238 474 463 474 463
Nikon/D850/Nikon-D850-14bit-lossless-compressed.NEF/threads:8/real_time_median -0.0267 -0.0267 473 461 473 461
Nikon/D850/Nikon-D850-14bit-lossless-compressed.NEF/threads:8/real_time_stddev +0.7179 +0.7178 3 5 3 5
Olympus/E-M1MarkII/Olympus_EM1mk2__HIRES_50MP.ORF/threads:8/real_time_pvalue 0.6837 0.6554 U Test, Repetitions: 25 vs 25
Olympus/E-M1MarkII/Olympus_EM1mk2__HIRES_50MP.ORF/threads:8/real_time_mean -0.0014 -0.0013 1375 1373 1375 1373
Olympus/E-M1MarkII/Olympus_EM1mk2__HIRES_50MP.ORF/threads:8/real_time_median +0.0018 +0.0019 1371 1374 1371 1374
Olympus/E-M1MarkII/Olympus_EM1mk2__HIRES_50MP.ORF/threads:8/real_time_stddev -0.7457 -0.7382 11 3 10 3
Panasonic/DC-G9/P1000476.RW2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Panasonic/DC-G9/P1000476.RW2/threads:8/real_time_mean -0.0080 -0.0289 22 22 10 10
Panasonic/DC-G9/P1000476.RW2/threads:8/real_time_median -0.0070 -0.0287 22 22 10 10
Panasonic/DC-G9/P1000476.RW2/threads:8/real_time_stddev +1.0977 +0.6614 0 0 0 0
Panasonic/DC-GH5/_T012014.RW2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Panasonic/DC-GH5/_T012014.RW2/threads:8/real_time_mean +0.0132 +0.0967 35 36 10 11
Panasonic/DC-GH5/_T012014.RW2/threads:8/real_time_median +0.0132 +0.0956 35 36 10 11
Panasonic/DC-GH5/_T012014.RW2/threads:8/real_time_stddev -0.0407 -0.1695 0 0 0 0
Panasonic/DC-GH5S/P1022085.RW2/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Panasonic/DC-GH5S/P1022085.RW2/threads:8/real_time_mean +0.0331 +0.1307 13 13 6 6
Panasonic/DC-GH5S/P1022085.RW2/threads:8/real_time_median +0.0430 +0.1373 12 13 6 6
Panasonic/DC-GH5S/P1022085.RW2/threads:8/real_time_stddev -0.9006 -0.8847 1 0 0 0
Pentax/645Z/IMGP2837.PEF/threads:8/real_time_pvalue 0.0016 0.0010 U Test, Repetitions: 25 vs 25
Pentax/645Z/IMGP2837.PEF/threads:8/real_time_mean -0.0023 -0.0024 395 394 395 394
Pentax/645Z/IMGP2837.PEF/threads:8/real_time_median -0.0029 -0.0030 395 394 395 393
Pentax/645Z/IMGP2837.PEF/threads:8/real_time_stddev -0.0275 -0.0375 1 1 1 1
Phase One/P65/CF027310.IIQ/threads:8/real_time_pvalue 0.0232 0.0000 U Test, Repetitions: 25 vs 25
Phase One/P65/CF027310.IIQ/threads:8/real_time_mean -0.0047 +0.0039 114 113 28 28
Phase One/P65/CF027310.IIQ/threads:8/real_time_median -0.0050 +0.0037 114 113 28 28
Phase One/P65/CF027310.IIQ/threads:8/real_time_stddev -0.0599 -0.2683 1 1 0 0
Samsung/NX1/2016-07-23-142101_sam_9364.srw/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Samsung/NX1/2016-07-23-142101_sam_9364.srw/threads:8/real_time_mean +0.0206 +0.0207 405 414 405 414
Samsung/NX1/2016-07-23-142101_sam_9364.srw/threads:8/real_time_median +0.0204 +0.0205 405 414 405 414
Samsung/NX1/2016-07-23-142101_sam_9364.srw/threads:8/real_time_stddev +0.2155 +0.2212 1 1 1 1
Samsung/NX30/2015-03-07-163604_sam_7204.srw/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Samsung/NX30/2015-03-07-163604_sam_7204.srw/threads:8/real_time_mean -0.0109 -0.0108 147 145 147 145
Samsung/NX30/2015-03-07-163604_sam_7204.srw/threads:8/real_time_median -0.0104 -0.0103 147 145 147 145
Samsung/NX30/2015-03-07-163604_sam_7204.srw/threads:8/real_time_stddev -0.4919 -0.4800 0 0 0 0
Samsung/NX3000/_3184416.SRW/threads:8/real_time_pvalue 0.0000 0.0000 U Test, Repetitions: 25 vs 25
Samsung/NX3000/_3184416.SRW/threads:8/real_time_mean -0.0149 -0.0147 220 217 220 217
Samsung/NX3000/_3184416.SRW/threads:8/real_time_median -0.0173 -0.0169 221 217 220 217
Samsung/NX3000/_3184416.SRW/threads:8/real_time_stddev +1.0337 +1.0341 1 3 1 3
Sony/DSLR-A350/DSC05472.ARW/threads:8/real_time_pvalue 0.0001 0.0001 U Test, Repetitions: 25 vs 25
Sony/DSLR-A350/DSC05472.ARW/threads:8/real_time_mean -0.0019 -0.0019 194 193 194 193
Sony/DSLR-A350/DSC05472.ARW/threads:8/real_time_median -0.0021 -0.0021 194 193 194 193
Sony/DSLR-A350/DSC05472.ARW/threads:8/real_time_stddev -0.4441 -0.4282 0 0 0 0
Sony/ILCE-7RM2/14-bit-compressed.ARW/threads:8/real_time_pvalue 0.0000 0.4263 U Test, Repetitions: 25 vs 25
Sony/ILCE-7RM2/14-bit-compressed.ARW/threads:8/real_time_mean +0.0258 -0.0006 81 83 19 19
Sony/ILCE-7RM2/14-bit-compressed.ARW/threads:8/real_time_median +0.0235 -0.0011 81 82 19 19
Sony/ILCE-7RM2/14-bit-compressed.ARW/threads:8/real_time_stddev +0.1634 +0.1070 1 1 0 0
```
{F7443905}
If we look at the `_mean`s, the time column, the biggest win is `-7.7%` (`Canon/EOS 5D Mark II/10.canon.sraw2.cr2`),
and the biggest loose is `+3.3%` (`Panasonic/DC-GH5S/P1022085.RW2`);
Overall: mean `-0.7436%`, median `-0.23%`, `cbrt(sum(time^3))` = `-8.73%`
Looks good so far i'd say.
llvm-exegesis details:
{F7371117} {F7371125}
{F7371128} {F7371144} {F7371158}
Reviewers: craig.topper, RKSimon, andreadb, courbet, avt77, spatel, GGanesh
Reviewed By: andreadb
Subscribers: javed.absar, gbedwell, jfb, llvm-commits
Differential Revision: https://reviews.llvm.org/D52779
llvm-svn: 345463
1550 lines
62 KiB
LLVM
1550 lines
62 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=x86-64 -mattr=+sse3 | FileCheck %s --check-prefixes=CHECK,GENERIC
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=atom -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,ATOM
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=slm -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,SLM
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=sandybridge -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,SANDY-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=sandybridge -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,SANDY
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=ivybridge -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,SANDY-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=ivybridge -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,SANDY
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=haswell -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,HASWELL-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=haswell -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,HASWELL
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=broadwell -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,BROADWELL-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=broadwell -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,BROADWELL
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=skylake -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,SKYLAKE-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=skylake -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,SKYLAKE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=skx -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,SKX-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=skx -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,SKX
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=bdver2 -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,BDVER2-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=bdver2 -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,BDVER2
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=btver2 -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,BTVER2-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=btver2 -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,BTVER2
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=znver1 -mattr=-ssse3 | FileCheck %s --check-prefixes=CHECK,ZNVER1-SSE
|
|
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -print-schedule -mcpu=znver1 -mattr=-avx2 | FileCheck %s --check-prefixes=CHECK,ZNVER1
|
|
|
|
define <2 x double> @test_addsubpd(<2 x double> %a0, <2 x double> %a1, <2 x double> *%a2) {
|
|
; GENERIC-LABEL: test_addsubpd:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; GENERIC-NEXT: addsubpd (%rdi), %xmm0 # sched: [9:1.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_addsubpd:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: addsubpd %xmm1, %xmm0 # sched: [6:3.00]
|
|
; ATOM-NEXT: addsubpd (%rdi), %xmm0 # sched: [7:3.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_addsubpd:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: addsubpd (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_addsubpd:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [9:1.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_addsubpd:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; SANDY-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [9:1.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_addsubpd:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [9:1.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_addsubpd:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [9:1.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_addsubpd:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_addsubpd:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_addsubpd:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [10:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_addsubpd:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [10:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_addsubpd:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [10:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_addsubpd:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKX-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [10:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_addsubpd:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [10:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_addsubpd:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [10:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_addsubpd:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_addsubpd:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_addsubpd:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: addsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-SSE-NEXT: addsubpd (%rdi), %xmm0 # sched: [10:1.00]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_addsubpd:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vaddsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-NEXT: vaddsubpd (%rdi), %xmm0, %xmm0 # sched: [10:1.00]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <2 x double> @llvm.x86.sse3.addsub.pd(<2 x double> %a0, <2 x double> %a1)
|
|
%2 = load <2 x double>, <2 x double> *%a2, align 16
|
|
%3 = call <2 x double> @llvm.x86.sse3.addsub.pd(<2 x double> %1, <2 x double> %2)
|
|
ret <2 x double> %3
|
|
}
|
|
declare <2 x double> @llvm.x86.sse3.addsub.pd(<2 x double>, <2 x double>) nounwind readnone
|
|
|
|
define <4 x float> @test_addsubps(<4 x float> %a0, <4 x float> %a1, <4 x float> *%a2) {
|
|
; GENERIC-LABEL: test_addsubps:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; GENERIC-NEXT: addsubps (%rdi), %xmm0 # sched: [9:1.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_addsubps:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: addsubps %xmm1, %xmm0 # sched: [5:5.00]
|
|
; ATOM-NEXT: addsubps (%rdi), %xmm0 # sched: [5:5.00]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_addsubps:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: addsubps (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_addsubps:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [9:1.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_addsubps:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; SANDY-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [9:1.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_addsubps:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [9:1.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_addsubps:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [9:1.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_addsubps:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_addsubps:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_addsubps:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [10:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_addsubps:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [10:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_addsubps:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [10:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_addsubps:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKX-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [10:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_addsubps:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [10:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_addsubps:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [10:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_addsubps:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_addsubps:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_addsubps:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: addsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-SSE-NEXT: addsubps (%rdi), %xmm0 # sched: [10:1.00]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_addsubps:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vaddsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-NEXT: vaddsubps (%rdi), %xmm0, %xmm0 # sched: [10:1.00]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <4 x float> @llvm.x86.sse3.addsub.ps(<4 x float> %a0, <4 x float> %a1)
|
|
%2 = load <4 x float>, <4 x float> *%a2, align 16
|
|
%3 = call <4 x float> @llvm.x86.sse3.addsub.ps(<4 x float> %1, <4 x float> %2)
|
|
ret <4 x float> %3
|
|
}
|
|
declare <4 x float> @llvm.x86.sse3.addsub.ps(<4 x float>, <4 x float>) nounwind readnone
|
|
|
|
define <2 x double> @test_haddpd(<2 x double> %a0, <2 x double> %a1, <2 x double> *%a2) {
|
|
; GENERIC-LABEL: test_haddpd:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: haddpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; GENERIC-NEXT: haddpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_haddpd:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: haddpd %xmm1, %xmm0 # sched: [8:4.00]
|
|
; ATOM-NEXT: haddpd (%rdi), %xmm0 # sched: [9:4.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_haddpd:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: haddpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: haddpd (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_haddpd:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; SANDY-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_haddpd:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; SANDY-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_haddpd:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_haddpd:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_haddpd:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_haddpd:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_haddpd:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_haddpd:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_haddpd:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKX-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_haddpd:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKX-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_haddpd:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [16:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_haddpd:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [16:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_haddpd:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_haddpd:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_haddpd:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: haddpd %xmm1, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: haddpd (%rdi), %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_haddpd:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vhaddpd %xmm1, %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: vhaddpd (%rdi), %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <2 x double> @llvm.x86.sse3.hadd.pd(<2 x double> %a0, <2 x double> %a1)
|
|
%2 = load <2 x double>, <2 x double> *%a2, align 16
|
|
%3 = call <2 x double> @llvm.x86.sse3.hadd.pd(<2 x double> %1, <2 x double> %2)
|
|
ret <2 x double> %3
|
|
}
|
|
declare <2 x double> @llvm.x86.sse3.hadd.pd(<2 x double>, <2 x double>) nounwind readnone
|
|
|
|
define <4 x float> @test_haddps(<4 x float> %a0, <4 x float> %a1, <4 x float> *%a2) {
|
|
; GENERIC-LABEL: test_haddps:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: haddps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; GENERIC-NEXT: haddps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_haddps:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: haddps %xmm1, %xmm0 # sched: [8:4.00]
|
|
; ATOM-NEXT: haddps (%rdi), %xmm0 # sched: [9:4.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_haddps:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: haddps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: haddps (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_haddps:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; SANDY-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_haddps:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; SANDY-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_haddps:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_haddps:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_haddps:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_haddps:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_haddps:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_haddps:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_haddps:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKX-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_haddps:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKX-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_haddps:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [16:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_haddps:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [16:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_haddps:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_haddps:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_haddps:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: haddps %xmm1, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: haddps (%rdi), %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_haddps:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vhaddps %xmm1, %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: vhaddps (%rdi), %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <4 x float> @llvm.x86.sse3.hadd.ps(<4 x float> %a0, <4 x float> %a1)
|
|
%2 = load <4 x float>, <4 x float> *%a2, align 16
|
|
%3 = call <4 x float> @llvm.x86.sse3.hadd.ps(<4 x float> %1, <4 x float> %2)
|
|
ret <4 x float> %3
|
|
}
|
|
declare <4 x float> @llvm.x86.sse3.hadd.ps(<4 x float>, <4 x float>) nounwind readnone
|
|
|
|
define <2 x double> @test_hsubpd(<2 x double> %a0, <2 x double> %a1, <2 x double> *%a2) {
|
|
; GENERIC-LABEL: test_hsubpd:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: hsubpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; GENERIC-NEXT: hsubpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_hsubpd:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: hsubpd %xmm1, %xmm0 # sched: [8:4.00]
|
|
; ATOM-NEXT: hsubpd (%rdi), %xmm0 # sched: [9:4.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_hsubpd:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: hsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: hsubpd (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_hsubpd:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; SANDY-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_hsubpd:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; SANDY-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_hsubpd:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [11:2.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_hsubpd:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_hsubpd:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_hsubpd:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_hsubpd:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_hsubpd:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_hsubpd:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKX-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_hsubpd:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKX-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_hsubpd:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [16:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_hsubpd:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [16:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_hsubpd:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_hsubpd:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_hsubpd:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: hsubpd %xmm1, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: hsubpd (%rdi), %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_hsubpd:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vhsubpd %xmm1, %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: vhsubpd (%rdi), %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <2 x double> @llvm.x86.sse3.hsub.pd(<2 x double> %a0, <2 x double> %a1)
|
|
%2 = load <2 x double>, <2 x double> *%a2, align 16
|
|
%3 = call <2 x double> @llvm.x86.sse3.hsub.pd(<2 x double> %1, <2 x double> %2)
|
|
ret <2 x double> %3
|
|
}
|
|
declare <2 x double> @llvm.x86.sse3.hsub.pd(<2 x double>, <2 x double>) nounwind readnone
|
|
|
|
define <4 x float> @test_hsubps(<4 x float> %a0, <4 x float> %a1, <4 x float> *%a2) {
|
|
; GENERIC-LABEL: test_hsubps:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: hsubps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; GENERIC-NEXT: hsubps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_hsubps:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: hsubps %xmm1, %xmm0 # sched: [8:4.00]
|
|
; ATOM-NEXT: hsubps (%rdi), %xmm0 # sched: [9:4.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_hsubps:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: hsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: hsubps (%rdi), %xmm0 # sched: [6:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_hsubps:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; SANDY-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_hsubps:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; SANDY-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_hsubps:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [11:2.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_hsubps:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; HASWELL-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [11:2.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_hsubps:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_hsubps:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [5:2.00]
|
|
; BROADWELL-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [10:2.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_hsubps:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_hsubps:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKYLAKE-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_hsubps:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [6:2.00]
|
|
; SKX-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [12:2.00]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_hsubps:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [6:2.00]
|
|
; SKX-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [12:2.00]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_hsubps:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [16:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_hsubps:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [11:1.00]
|
|
; BDVER2-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [16:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_hsubps:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [8:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_hsubps:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [8:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_hsubps:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: hsubps %xmm1, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: hsubps (%rdi), %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_hsubps:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vhsubps %xmm1, %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: vhsubps (%rdi), %xmm0, %xmm0 # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <4 x float> @llvm.x86.sse3.hsub.ps(<4 x float> %a0, <4 x float> %a1)
|
|
%2 = load <4 x float>, <4 x float> *%a2, align 16
|
|
%3 = call <4 x float> @llvm.x86.sse3.hsub.ps(<4 x float> %1, <4 x float> %2)
|
|
ret <4 x float> %3
|
|
}
|
|
declare <4 x float> @llvm.x86.sse3.hsub.ps(<4 x float>, <4 x float>) nounwind readnone
|
|
|
|
define <16 x i8> @test_lddqu(i8* %a0) {
|
|
; GENERIC-LABEL: test_lddqu:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: lddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_lddqu:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: lddqu (%rdi), %xmm0 # sched: [3:1.50]
|
|
; ATOM-NEXT: nop # sched: [1:0.50]
|
|
; ATOM-NEXT: nop # sched: [1:0.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_lddqu:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: lddqu (%rdi), %xmm0 # sched: [3:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_lddqu:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_lddqu:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vlddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_lddqu:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_lddqu:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vlddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_lddqu:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [5:0.50]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_lddqu:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vlddqu (%rdi), %xmm0 # sched: [5:0.50]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_lddqu:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_lddqu:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vlddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_lddqu:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_lddqu:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vlddqu (%rdi), %xmm0 # sched: [6:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_lddqu:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [5:0.50]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_lddqu:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vlddqu (%rdi), %xmm0 # sched: [5:0.50]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_lddqu:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [5:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_lddqu:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vlddqu (%rdi), %xmm0 # sched: [5:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_lddqu:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: lddqu (%rdi), %xmm0 # sched: [8:0.50]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_lddqu:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vlddqu (%rdi), %xmm0 # sched: [8:0.50]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = call <16 x i8> @llvm.x86.sse3.ldu.dq(i8* %a0)
|
|
ret <16 x i8> %1
|
|
}
|
|
declare <16 x i8> @llvm.x86.sse3.ldu.dq(i8*) nounwind readonly
|
|
|
|
define void @test_monitor(i8* %a0, i32 %a1, i32 %a2) {
|
|
; GENERIC-LABEL: test_monitor:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: movl %esi, %ecx # sched: [1:0.33]
|
|
; GENERIC-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; GENERIC-NEXT: monitor # sched: [100:0.33]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_monitor:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; ATOM-NEXT: leaq (%rdi), %rax # sched: [1:1.00]
|
|
; ATOM-NEXT: monitor # sched: [45:22.50]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_monitor:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; SLM-NEXT: leaq (%rdi), %rax # sched: [1:1.00]
|
|
; SLM-NEXT: monitor # sched: [100:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_monitor:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: movl %esi, %ecx # sched: [1:0.33]
|
|
; SANDY-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SANDY-SSE-NEXT: monitor # sched: [100:0.33]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_monitor:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: movl %esi, %ecx # sched: [1:0.33]
|
|
; SANDY-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SANDY-NEXT: monitor # sched: [100:0.33]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_monitor:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; HASWELL-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; HASWELL-SSE-NEXT: monitor # sched: [100:0.25]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_monitor:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; HASWELL-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; HASWELL-NEXT: monitor # sched: [100:0.25]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_monitor:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; BROADWELL-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BROADWELL-SSE-NEXT: monitor # sched: [100:0.25]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_monitor:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; BROADWELL-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BROADWELL-NEXT: monitor # sched: [100:0.25]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_monitor:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; SKYLAKE-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SKYLAKE-SSE-NEXT: monitor # sched: [100:0.25]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_monitor:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; SKYLAKE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SKYLAKE-NEXT: monitor # sched: [100:0.25]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_monitor:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; SKX-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SKX-SSE-NEXT: monitor # sched: [100:0.25]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_monitor:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; SKX-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; SKX-NEXT: monitor # sched: [100:0.25]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_monitor:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BDVER2-SSE-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; BDVER2-SSE-NEXT: monitor # sched: [100:0.50]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_monitor:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BDVER2-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; BDVER2-NEXT: monitor # sched: [100:0.50]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_monitor:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: monitor # sched: [100:0.50]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_monitor:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: movl %esi, %ecx # sched: [1:0.50]
|
|
; BTVER2-NEXT: leaq (%rdi), %rax # sched: [1:0.50]
|
|
; BTVER2-NEXT: monitor # sched: [100:0.50]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_monitor:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; ZNVER1-SSE-NEXT: leaq (%rdi), %rax # sched: [1:0.25]
|
|
; ZNVER1-SSE-NEXT: monitor # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_monitor:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: movl %esi, %ecx # sched: [1:0.25]
|
|
; ZNVER1-NEXT: leaq (%rdi), %rax # sched: [1:0.25]
|
|
; ZNVER1-NEXT: monitor # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
tail call void @llvm.x86.sse3.monitor(i8* %a0, i32 %a1, i32 %a2)
|
|
ret void
|
|
}
|
|
declare void @llvm.x86.sse3.monitor(i8*, i32, i32)
|
|
|
|
define <2 x double> @test_movddup(<2 x double> %a0, <2 x double> *%a1) {
|
|
; GENERIC-LABEL: test_movddup:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; GENERIC-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [6:0.50]
|
|
; GENERIC-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_movddup:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; ATOM-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [1:1.00]
|
|
; ATOM-NEXT: subpd %xmm1, %xmm0 # sched: [6:3.00]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_movddup:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: movddup {{.*#+}} xmm1 = mem[0,0] sched: [4:1.00]
|
|
; SLM-NEXT: movddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; SLM-NEXT: subpd %xmm0, %xmm1 # sched: [3:1.00]
|
|
; SLM-NEXT: movapd %xmm1, %xmm0 # sched: [1:0.50]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_movddup:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; SANDY-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [6:0.50]
|
|
; SANDY-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_movddup:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; SANDY-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [6:0.50]
|
|
; SANDY-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_movddup:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; HASWELL-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [5:0.50]
|
|
; HASWELL-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_movddup:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; HASWELL-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [5:0.50]
|
|
; HASWELL-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_movddup:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; BROADWELL-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [5:0.50]
|
|
; BROADWELL-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_movddup:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; BROADWELL-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [5:0.50]
|
|
; BROADWELL-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_movddup:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; SKYLAKE-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [5:0.50]
|
|
; SKYLAKE-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_movddup:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; SKYLAKE-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [5:0.50]
|
|
; SKYLAKE-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_movddup:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:1.00]
|
|
; SKX-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [5:0.50]
|
|
; SKX-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_movddup:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:1.00]
|
|
; SKX-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [5:0.50]
|
|
; SKX-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_movddup:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [2:0.50]
|
|
; BDVER2-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [7:0.50]
|
|
; BDVER2-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_movddup:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [7:0.50]
|
|
; BDVER2-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [2:0.50]
|
|
; BDVER2-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_movddup:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [6:1.00]
|
|
; BTVER2-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_movddup:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [6:1.00]
|
|
; BTVER2-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:0.50]
|
|
; BTVER2-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_movddup:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: movddup {{.*#+}} xmm1 = xmm0[0,0] sched: [1:0.50]
|
|
; ZNVER1-SSE-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] sched: [8:0.50]
|
|
; ZNVER1-SSE-NEXT: subpd %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_movddup:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vmovddup {{.*#+}} xmm1 = mem[0,0] sched: [8:0.50]
|
|
; ZNVER1-NEXT: vmovddup {{.*#+}} xmm0 = xmm0[0,0] sched: [1:0.50]
|
|
; ZNVER1-NEXT: vsubpd %xmm0, %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = shufflevector <2 x double> %a0, <2 x double> undef, <2 x i32> zeroinitializer
|
|
%2 = load <2 x double>, <2 x double> *%a1, align 16
|
|
%3 = shufflevector <2 x double> %2, <2 x double> undef, <2 x i32> zeroinitializer
|
|
%4 = fsub <2 x double> %3, %1 ; Use fsub to stop the movddup from being folded as a broadcast load in avx512vl.
|
|
ret <2 x double> %4
|
|
}
|
|
|
|
define <4 x float> @test_movshdup(<4 x float> %a0, <4 x float> *%a1) {
|
|
; GENERIC-LABEL: test_movshdup:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; GENERIC-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:0.50]
|
|
; GENERIC-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_movshdup:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; ATOM-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [1:1.00]
|
|
; ATOM-NEXT: addps %xmm1, %xmm0 # sched: [5:5.00]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_movshdup:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: movshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [4:1.00]
|
|
; SLM-NEXT: movshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SLM-NEXT: addps %xmm0, %xmm1 # sched: [3:1.00]
|
|
; SLM-NEXT: movaps %xmm1, %xmm0 # sched: [1:0.50]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_movshdup:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SANDY-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SANDY-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_movshdup:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SANDY-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SANDY-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_movshdup:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; HASWELL-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:0.50]
|
|
; HASWELL-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_movshdup:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; HASWELL-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [6:0.50]
|
|
; HASWELL-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_movshdup:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; BROADWELL-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [5:0.50]
|
|
; BROADWELL-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_movshdup:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; BROADWELL-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [5:0.50]
|
|
; BROADWELL-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_movshdup:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SKYLAKE-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SKYLAKE-SSE-NEXT: addps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_movshdup:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SKYLAKE-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SKYLAKE-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_movshdup:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SKX-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SKX-SSE-NEXT: addps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_movshdup:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:1.00]
|
|
; SKX-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [6:0.50]
|
|
; SKX-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_movshdup:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [2:0.50]
|
|
; BDVER2-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [7:0.50]
|
|
; BDVER2-SSE-NEXT: addps %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_movshdup:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [7:0.50]
|
|
; BDVER2-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [2:0.50]
|
|
; BDVER2-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_movshdup:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [6:1.00]
|
|
; BTVER2-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_movshdup:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [6:1.00]
|
|
; BTVER2-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:0.50]
|
|
; BTVER2-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_movshdup:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: movshdup {{.*#+}} xmm1 = xmm0[1,1,3,3] sched: [1:0.50]
|
|
; ZNVER1-SSE-NEXT: movshdup {{.*#+}} xmm0 = mem[1,1,3,3] sched: [8:0.50]
|
|
; ZNVER1-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_movshdup:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vmovshdup {{.*#+}} xmm1 = mem[1,1,3,3] sched: [8:0.50]
|
|
; ZNVER1-NEXT: vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3] sched: [1:0.50]
|
|
; ZNVER1-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = shufflevector <4 x float> %a0, <4 x float> undef, <4 x i32> <i32 1, i32 1, i32 3, i32 3>
|
|
%2 = load <4 x float>, <4 x float> *%a1, align 16
|
|
%3 = shufflevector <4 x float> %2, <4 x float> undef, <4 x i32> <i32 1, i32 1, i32 3, i32 3>
|
|
%4 = fadd <4 x float> %1, %3
|
|
ret <4 x float> %4
|
|
}
|
|
|
|
define <4 x float> @test_movsldup(<4 x float> %a0, <4 x float> *%a1) {
|
|
; GENERIC-LABEL: test_movsldup:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; GENERIC-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:0.50]
|
|
; GENERIC-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_movsldup:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; ATOM-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [1:1.00]
|
|
; ATOM-NEXT: addps %xmm1, %xmm0 # sched: [5:5.00]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_movsldup:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: movsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [4:1.00]
|
|
; SLM-NEXT: movsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SLM-NEXT: addps %xmm0, %xmm1 # sched: [3:1.00]
|
|
; SLM-NEXT: movaps %xmm1, %xmm0 # sched: [1:0.50]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_movsldup:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SANDY-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SANDY-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_movsldup:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SANDY-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SANDY-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_movsldup:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; HASWELL-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:0.50]
|
|
; HASWELL-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_movsldup:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; HASWELL-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [6:0.50]
|
|
; HASWELL-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_movsldup:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; BROADWELL-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [5:0.50]
|
|
; BROADWELL-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_movsldup:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; BROADWELL-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [5:0.50]
|
|
; BROADWELL-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_movsldup:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SKYLAKE-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SKYLAKE-SSE-NEXT: addps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_movsldup:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SKYLAKE-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SKYLAKE-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_movsldup:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SKX-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SKX-SSE-NEXT: addps %xmm1, %xmm0 # sched: [4:0.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_movsldup:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:1.00]
|
|
; SKX-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [6:0.50]
|
|
; SKX-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [4:0.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_movsldup:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [2:0.50]
|
|
; BDVER2-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [7:0.50]
|
|
; BDVER2-SSE-NEXT: addps %xmm1, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_movsldup:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [7:0.50]
|
|
; BDVER2-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [2:0.50]
|
|
; BDVER2-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [5:1.00]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_movsldup:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [6:1.00]
|
|
; BTVER2-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_movsldup:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [6:1.00]
|
|
; BTVER2-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:0.50]
|
|
; BTVER2-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_movsldup:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: movsldup {{.*#+}} xmm1 = xmm0[0,0,2,2] sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: movsldup {{.*#+}} xmm0 = mem[0,0,2,2] sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: addps %xmm1, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_movsldup:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: vmovsldup {{.*#+}} xmm1 = mem[0,0,2,2] sched: [8:0.50]
|
|
; ZNVER1-NEXT: vmovsldup {{.*#+}} xmm0 = xmm0[0,0,2,2] sched: [1:0.50]
|
|
; ZNVER1-NEXT: vaddps %xmm1, %xmm0, %xmm0 # sched: [3:1.00]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
%1 = shufflevector <4 x float> %a0, <4 x float> undef, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
|
|
%2 = load <4 x float>, <4 x float> *%a1, align 16
|
|
%3 = shufflevector <4 x float> %2, <4 x float> undef, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
|
|
%4 = fadd <4 x float> %1, %3
|
|
ret <4 x float> %4
|
|
}
|
|
|
|
define void @test_mwait(i32 %a0, i32 %a1) {
|
|
; GENERIC-LABEL: test_mwait:
|
|
; GENERIC: # %bb.0:
|
|
; GENERIC-NEXT: movl %esi, %eax # sched: [1:0.33]
|
|
; GENERIC-NEXT: movl %edi, %ecx # sched: [1:0.33]
|
|
; GENERIC-NEXT: mwait # sched: [100:0.33]
|
|
; GENERIC-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; ATOM-LABEL: test_mwait:
|
|
; ATOM: # %bb.0:
|
|
; ATOM-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; ATOM-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; ATOM-NEXT: mwait # sched: [46:23.00]
|
|
; ATOM-NEXT: retq # sched: [79:39.50]
|
|
;
|
|
; SLM-LABEL: test_mwait:
|
|
; SLM: # %bb.0:
|
|
; SLM-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; SLM-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; SLM-NEXT: mwait # sched: [100:1.00]
|
|
; SLM-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; SANDY-SSE-LABEL: test_mwait:
|
|
; SANDY-SSE: # %bb.0:
|
|
; SANDY-SSE-NEXT: movl %esi, %eax # sched: [1:0.33]
|
|
; SANDY-SSE-NEXT: movl %edi, %ecx # sched: [1:0.33]
|
|
; SANDY-SSE-NEXT: mwait # sched: [100:0.33]
|
|
; SANDY-SSE-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; SANDY-LABEL: test_mwait:
|
|
; SANDY: # %bb.0:
|
|
; SANDY-NEXT: movl %esi, %eax # sched: [1:0.33]
|
|
; SANDY-NEXT: movl %edi, %ecx # sched: [1:0.33]
|
|
; SANDY-NEXT: mwait # sched: [100:0.33]
|
|
; SANDY-NEXT: retq # sched: [1:1.00]
|
|
;
|
|
; HASWELL-SSE-LABEL: test_mwait:
|
|
; HASWELL-SSE: # %bb.0:
|
|
; HASWELL-SSE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; HASWELL-SSE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; HASWELL-SSE-NEXT: mwait # sched: [20:2.50]
|
|
; HASWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; HASWELL-LABEL: test_mwait:
|
|
; HASWELL: # %bb.0:
|
|
; HASWELL-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; HASWELL-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; HASWELL-NEXT: mwait # sched: [20:2.50]
|
|
; HASWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-SSE-LABEL: test_mwait:
|
|
; BROADWELL-SSE: # %bb.0:
|
|
; BROADWELL-SSE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; BROADWELL-SSE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; BROADWELL-SSE-NEXT: mwait # sched: [100:0.25]
|
|
; BROADWELL-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BROADWELL-LABEL: test_mwait:
|
|
; BROADWELL: # %bb.0:
|
|
; BROADWELL-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; BROADWELL-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; BROADWELL-NEXT: mwait # sched: [100:0.25]
|
|
; BROADWELL-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-SSE-LABEL: test_mwait:
|
|
; SKYLAKE-SSE: # %bb.0:
|
|
; SKYLAKE-SSE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; SKYLAKE-SSE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; SKYLAKE-SSE-NEXT: mwait # sched: [20:2.50]
|
|
; SKYLAKE-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKYLAKE-LABEL: test_mwait:
|
|
; SKYLAKE: # %bb.0:
|
|
; SKYLAKE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; SKYLAKE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; SKYLAKE-NEXT: mwait # sched: [20:2.50]
|
|
; SKYLAKE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-SSE-LABEL: test_mwait:
|
|
; SKX-SSE: # %bb.0:
|
|
; SKX-SSE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; SKX-SSE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; SKX-SSE-NEXT: mwait # sched: [20:2.50]
|
|
; SKX-SSE-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; SKX-LABEL: test_mwait:
|
|
; SKX: # %bb.0:
|
|
; SKX-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; SKX-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; SKX-NEXT: mwait # sched: [20:2.50]
|
|
; SKX-NEXT: retq # sched: [7:1.00]
|
|
;
|
|
; BDVER2-SSE-LABEL: test_mwait:
|
|
; BDVER2-SSE: # %bb.0:
|
|
; BDVER2-SSE-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; BDVER2-SSE-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; BDVER2-SSE-NEXT: mwait # sched: [100:0.50]
|
|
; BDVER2-SSE-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BDVER2-LABEL: test_mwait:
|
|
; BDVER2: # %bb.0:
|
|
; BDVER2-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; BDVER2-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; BDVER2-NEXT: mwait # sched: [100:0.50]
|
|
; BDVER2-NEXT: retq # sched: [5:1.00]
|
|
;
|
|
; BTVER2-SSE-LABEL: test_mwait:
|
|
; BTVER2-SSE: # %bb.0:
|
|
; BTVER2-SSE-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; BTVER2-SSE-NEXT: mwait # sched: [100:0.50]
|
|
; BTVER2-SSE-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; BTVER2-LABEL: test_mwait:
|
|
; BTVER2: # %bb.0:
|
|
; BTVER2-NEXT: movl %esi, %eax # sched: [1:0.50]
|
|
; BTVER2-NEXT: movl %edi, %ecx # sched: [1:0.50]
|
|
; BTVER2-NEXT: mwait # sched: [100:0.50]
|
|
; BTVER2-NEXT: retq # sched: [4:1.00]
|
|
;
|
|
; ZNVER1-SSE-LABEL: test_mwait:
|
|
; ZNVER1-SSE: # %bb.0:
|
|
; ZNVER1-SSE-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; ZNVER1-SSE-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; ZNVER1-SSE-NEXT: mwait # sched: [100:0.25]
|
|
; ZNVER1-SSE-NEXT: retq # sched: [1:0.50]
|
|
;
|
|
; ZNVER1-LABEL: test_mwait:
|
|
; ZNVER1: # %bb.0:
|
|
; ZNVER1-NEXT: movl %esi, %eax # sched: [1:0.25]
|
|
; ZNVER1-NEXT: movl %edi, %ecx # sched: [1:0.25]
|
|
; ZNVER1-NEXT: mwait # sched: [100:0.25]
|
|
; ZNVER1-NEXT: retq # sched: [1:0.50]
|
|
tail call void @llvm.x86.sse3.mwait(i32 %a0, i32 %a1)
|
|
ret void
|
|
}
|
|
declare void @llvm.x86.sse3.mwait(i32, i32)
|