Add IACA output for SOA

Signed-off-by: Rafael Ravedutti <rafaelravedutti@gmail.com>
This commit is contained in:
Rafael Ravedutti 2021-04-27 01:04:31 +02:00
parent d0d2bf8a0c
commit 5cb341ab1f

View File

@ -0,0 +1,104 @@
iwia021h@testfront1:~/MD-Bench/ICC$ iaca -arch SKX force.o
Intel(R) Architecture Code Analyzer Version - v3.0-28-g1ba2cbb build date: 2017-10-23;16:42:45
Analyzed File - force.o
Binary Format - 64Bit
Architecture - SKX
Analysis Type - Throughput
Throughput Analysis Report
--------------------------
Block Throughput: 36.70 Cycles Throughput Bottleneck: Backend
Loop Count: 23
Port Binding In Cycles Per Iteration:
--------------------------------------------------------------------------------------------------
| Port | 0 - DV | 1 | 2 - D | 3 - D | 4 | 5 | 6 | 7 |
--------------------------------------------------------------------------------------------------
| Cycles | 17.5 0.0 | 11.0 | 20.5 17.0 | 20.5 17.0 | 7.0 | 20.5 | 7.0 | 0.0 |
--------------------------------------------------------------------------------------------------
DV - Divider pipe (on port 0)
D - Data fetch pipe (on ports 2 and 3)
F - Macro Fusion with the previous instruction occurred
* - instruction micro-ops not bound to a port
^ - Micro Fusion occurred
# - ESP Tracking sync uop was issued
@ - SSE instruction followed an AVX256/AVX512 instruction, dozens of cycles penalty is expected
X - instruction not supported, was not accounted in Analysis
| Num Of | Ports pressure in cycles | |
| Uops | 0 - DV | 1 | 2 - D | 3 - D | 4 | 5 | 6 | 7 |
-----------------------------------------------------------------------------------------
| 1* | | | | | | | | | mov r13, r8
| 1 | | 1.0 | | | | | | | imul r13, rcx
| 1 | | | | | | 1.0 | | | vbroadcastsd zmm2, xmm6
| 1 | | | | | | 1.0 | | | vbroadcastsd zmm1, xmm7
| 1 | | | | | | 1.0 | | | vbroadcastsd zmm0, xmm12
| 1 | | | | | | | 1.0 | | movsxd rbx, r12d
| 1 | | | | | | | 1.0 | | add r13, r10
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x40], rax
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x38], r8
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x30], r10
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x28], rsi
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x20], rcx
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x50], r9
| 2^ | | | 0.5 | 0.5 | 1.0 | | | | mov qword ptr [rsp-0x48], rdx
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | vmovdqu ymm3, ymmword ptr [r13+rbx*4]
| 1 | | 1.0 | | | | | | | vpaddd ymm4, ymm3, ymm3
| 1 | | 1.0 | | | | | | | vpaddd ymm3, ymm3, ymm4
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov r10d, dword ptr [r13+rbx*4]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov r9d, dword ptr [r13+rbx*4+0x4]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov r8d, dword ptr [r13+rbx*4+0x8]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov esi, dword ptr [r13+rbx*4+0xc]
| 1 | | 1.0 | | | | | | | lea r10d, ptr [r10+r10*2]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov ecx, dword ptr [r13+rbx*4+0x10]
| 1 | | 1.0 | | | | | | | lea r9d, ptr [r9+r9*2]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov edx, dword ptr [r13+rbx*4+0x14]
| 1 | | 1.0 | | | | | | | lea r8d, ptr [r8+r8*2]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov eax, dword ptr [r13+rbx*4+0x18]
| 1 | | 1.0 | | | | | | | lea esi, ptr [rsi+rsi*2]
| 1 | | | 0.5 0.5 | 0.5 0.5 | | | | | mov r15d, dword ptr [r13+rbx*4+0x1c]
| 1 | | 1.0 | | | | | | | lea ecx, ptr [rcx+rcx*2]
| 1 | | 1.0 | | | | | | | lea edx, ptr [rdx+rdx*2]
| 1 | | 1.0 | | | | | | | lea eax, ptr [rax+rax*2]
| 1 | | 1.0 | | | | | | | lea r15d, ptr [r15+r15*2]
| 1 | | | | | | 1.0 | | | vpcmpeqb k1, xmm0, xmm0
| 1 | | | | | | 1.0 | | | vpcmpeqb k2, xmm0, xmm0
| 1 | | | | | | 1.0 | | | vpcmpeqb k3, xmm0, xmm0
| 1* | | | | | | | | | vpxord zmm4, zmm4, zmm4
| 1* | | | | | | | | | vpxord zmm17, zmm17, zmm17
| 1* | | | | | | | | | vpxord zmm18, zmm18, zmm18
| 5^ | 2.0 | | 4.0 4.0 | 4.0 4.0 | | | 1.0 | | vgatherdpd zmm4, k1, zmmword ptr [rdi+ymm3*8+0x10]
| 5^ | 1.5 | | 4.0 4.0 | 4.0 4.0 | | 0.5 | 1.0 | | vgatherdpd zmm17, k2, zmmword ptr [rdi+ymm3*8+0x8]
| 5^ | 1.0 | | 4.0 4.0 | 4.0 4.0 | | 1.0 | 1.0 | | vgatherdpd zmm18, k3, zmmword ptr [rdi+ymm3*8]
| 1 | | | | | | | 1.0 | | add r12d, 0x8
| 1 | | | | | | | 1.0 | | add rbx, 0x8
| 1 | 0.5 | | | | | 0.5 | | | vsubpd zmm26, zmm0, zmm4
| 1 | 0.5 | | | | | 0.5 | | | vsubpd zmm24, zmm1, zmm17
| 1 | 0.5 | | | | | 0.5 | | | vsubpd zmm23, zmm2, zmm18
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm3, zmm24, zmm24
| 1 | 0.5 | | | | | 0.5 | | | vfmadd231pd zmm3, zmm23, zmm23
| 1 | 0.5 | | | | | 0.5 | | | vfmadd231pd zmm3, zmm26, zmm26
| 3 | 2.0 | | | | | 1.0 | | | vrcp14pd zmm22, zmm3
| 1 | | | | | | 1.0 | | | vcmppd k2, zmm3, zmm14, 0x1
| 1 | | | | | | 1.0 | | | vfpclasspd k0, zmm22, 0x1e
| 2^ | 1.0 | | 0.5 0.5 | 0.5 0.5 | | | | | vfnmadd213pd zmm3, zmm22, qword ptr [rip]{1to8}
| 1 | 1.0 | | | | | | | | knotw k1, k0
| 1 | | | | | | 1.0 | | | vmulpd zmm4, zmm3, zmm3
| 1 | 0.5 | | | | | 0.5 | | | vfmadd213pd zmm22{k1}, zmm3, zmm22
| 1 | 0.5 | | | | | 0.5 | | | vfmadd213pd zmm22{k1}, zmm4, zmm22
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm17, zmm22, zmm13
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm19, zmm22, zmm10
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm20, zmm22, zmm17
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm18, zmm22, zmm20
| 1 | 0.5 | | | | | 0.5 | | | vfmsub213pd zmm22, zmm20, zmm5
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm21, zmm18, zmm19
| 1 | 0.5 | | | | | 0.5 | | | vmulpd zmm25, zmm21, zmm22
| 1 | 0.5 | | | | | 0.5 | | | vfmadd231pd zmm9{k2}, zmm25, zmm23
| 1 | 0.5 | | | | | 0.5 | | | vfmadd231pd zmm8{k2}, zmm25, zmm24
| 1 | 0.5 | | | | | 0.5 | | | vfmadd231pd zmm11{k2}, zmm25, zmm26
| 1* | | | | | | | | | cmp r12d, r14d
| 0*F | | | | | | | | | jb 0xfffffffffffffed3
Total Num Of Uops: 91
Analysis Notes:
Backend allocation was stalled due to unavailable allocation resources.
There were bubbles in the frontend.