RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-21 12:02:58 +02:00
Code Issues Projects Releases Wiki Activity
65b90c2fa8
llvm-mirror/test/CodeGen/X86/recip-fastmath.ll

110 lines
2.6 KiB
LLVM
Raw Normal View History

Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=core2 | FileCheck %s
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=btver2 | FileCheck %s --check-prefix=BTVER2
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+use-recip-est,+avx -x86-recip-refinement-steps=2 | FileCheck %s --check-prefix=REFINE
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
; If the target's divss/divps instructions are substantially
; slower than rcpss/rcpps with a Newton-Raphson refinement,
; we should generate the estimate sequence.
; See PR21385 ( http://llvm.org/bugs/show_bug.cgi?id=21385 )
; for details about the accuracy, speed, and implementation
; differences of x86 reciprocal estimates.
define float @reciprocal_estimate(float %x) #0 {
%div = fdiv fast float 1.0, %x
ret float %div
; CHECK-LABEL: reciprocal_estimate:
; CHECK: movss
; CHECK-NEXT: divss
; CHECK-NEXT: movaps
; CHECK-NEXT: retq
; BTVER2-LABEL: reciprocal_estimate:
; BTVER2: vrcpss
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; BTVER2: vmulss
; BTVER2: vsubss
; BTVER2: vmulss
; BTVER2: vaddss
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
; BTVER2-NEXT: retq
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; REFINE-LABEL: reciprocal_estimate:
; REFINE: vrcpss
; REFINE: vmulss
; REFINE: vsubss
; REFINE: vmulss
; REFINE: vaddss
; REFINE: vmulss
; REFINE: vsubss
; REFINE: vmulss
; REFINE: vaddss
; REFINE-NEXT: retq
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
}
define <4 x float> @reciprocal_estimate_v4f32(<4 x float> %x) #0 {
%div = fdiv fast <4 x float> <float 1.0, float 1.0, float 1.0, float 1.0>, %x
ret <4 x float> %div
; CHECK-LABEL: reciprocal_estimate_v4f32:
; CHECK: movaps
; CHECK-NEXT: divps
; CHECK-NEXT: movaps
; CHECK-NEXT: retq
; BTVER2-LABEL: reciprocal_estimate_v4f32:
; BTVER2: vrcpps
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; BTVER2: vmulps
; BTVER2: vsubps
; BTVER2: vmulps
; BTVER2: vaddps
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
; BTVER2-NEXT: retq
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; REFINE-LABEL: reciprocal_estimate_v4f32:
; REFINE: vrcpps
; REFINE: vmulps
; REFINE: vsubps
; REFINE: vmulps
; REFINE: vaddps
; REFINE: vmulps
; REFINE: vsubps
; REFINE: vmulps
; REFINE: vaddps
; REFINE-NEXT: retq
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
}
define <8 x float> @reciprocal_estimate_v8f32(<8 x float> %x) #0 {
%div = fdiv fast <8 x float> <float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0, float 1.0>, %x
ret <8 x float> %div
; CHECK-LABEL: reciprocal_estimate_v8f32:
; CHECK: movaps
; CHECK: movaps
; CHECK-NEXT: divps
; CHECK-NEXT: divps
; CHECK-NEXT: movaps
; CHECK-NEXT: movaps
; CHECK-NEXT: retq
; BTVER2-LABEL: reciprocal_estimate_v8f32:
; BTVER2: vrcpps
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; BTVER2: vmulps
; BTVER2: vsubps
; BTVER2: vmulps
; BTVER2: vaddps
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
; BTVER2-NEXT: retq
Expose the number of Newton-Raphson iterations applied to the hardware's reciprocal estimate as a parameter (x86). This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385. This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests. llvm-svn: 221819
2014-11-12 22:39:01 +01:00
; REFINE-LABEL: reciprocal_estimate_v8f32:
; REFINE: vrcpps
; REFINE: vmulps
; REFINE: vsubps
; REFINE: vmulps
; REFINE: vaddps
; REFINE: vmulps
; REFINE: vsubps
; REFINE: vmulps
; REFINE: vaddps
; REFINE-NEXT: retq
Use rcpss/rcpps (X86) to speed up reciprocal calcs (PR21385). This is a first step for generating SSE rcp instructions for reciprocal calcs when fast-math allows it. This is very similar to the rsqrt optimization enabled in D5658 ( http://reviews.llvm.org/rL220570 ). For now, be conservative and only enable this for AMD btver2 where performance improves significantly both in terms of latency and throughput. We may never enable this codegen for Intel Core* chips because the divider circuits are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21 cycle critical path for the rcp + mul + sub + mul + add estimate. Follow-on patches may allow configuration of the number of Newton-Raphson refinement steps, add AVX512 support, and enable the optimization for more chips. More background here: http://llvm.org/bugs/show_bug.cgi?id=21385 Differential Revision: http://reviews.llvm.org/D6175 llvm-svn: 221706
2014-11-11 21:51:00 +01:00
}
attributes #0 = { "unsafe-fp-math"="true" }
Reference in New Issue Copy Permalink