75 lines
6.3 KiB
Plaintext
75 lines
6.3 KiB
Plaintext
iwia021h@testfront1:~/MD-Bench/asm$ iaca -arch SKX force_soa_geq1200_markers.o
|
|
Intel(R) Architecture Code Analyzer Version - v3.0-28-g1ba2cbb build date: 2017-10-23;16:42:45
|
|
Analyzed File - force_soa_geq1200_markers.o
|
|
Binary Format - 64Bit
|
|
Architecture - SKX
|
|
Analysis Type - Throughput
|
|
|
|
Throughput Analysis Report
|
|
--------------------------
|
|
Block Throughput: 31.47 Cycles Throughput Bottleneck: Backend
|
|
Loop Count: 22
|
|
Port Binding In Cycles Per Iteration:
|
|
--------------------------------------------------------------------------------------------------
|
|
| Port | 0 - DV | 1 | 2 - D | 3 - D | 4 | 5 | 6 | 7 |
|
|
--------------------------------------------------------------------------------------------------
|
|
| Cycles | 18.0 0.0 | 3.0 | 13.0 13.0 | 13.0 13.0 | 0.0 | 18.0 | 3.0 | 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 | | | | | | 1.0 | | | vpcmpgtd k5, ymm3, ymm4
|
|
| 1 | | 1.0 | | | | | | | vpaddd ymm4, ymm4, ymm18
|
|
| 2 | | 1.0 | 1.0 1.0 | | | | | | vmovdqu32 ymm20{k5}{z}, ymmword ptr [rcx+r15*4]
|
|
| 1* | | | | | | | | | vmovaps zmm22, zmm19
|
|
| 1 | | 1.0 | | | | | | | add r15, 0x8
|
|
| 1 | 1.0 | | | | | | | | kmovw k2, k5
|
|
| 1* | | | | | | | | | vmovaps zmm21, zmm19
|
|
| 1 | 1.0 | | | | | | | | kmovw k1, k5
|
|
| 1* | | | | | | | | | vmovaps zmm23, zmm19
|
|
| 1 | 1.0 | | | | | | | | kmovw k3, k5
|
|
| 5^ | 1.0 | | 4.0 4.0 | 4.0 4.0 | | 1.0 | 1.0 | | vgatherdpd zmm23, k3, zmmword ptr [rsi+ymm20*8]
|
|
| 5^ | 1.0 | | 4.0 4.0 | 4.0 4.0 | | 1.0 | 1.0 | | vgatherdpd zmm22, k2, zmmword ptr [rax+ymm20*8]
|
|
| 5^ | 1.0 | | 4.0 4.0 | 4.0 4.0 | | 1.0 | 1.0 | | vgatherdpd zmm21, k1, zmmword ptr [rdx+ymm20*8]
|
|
| 1 | | | | | | 1.0 | | | vsubpd zmm0, zmm5, zmm22
|
|
| 1 | | | | | | 1.0 | | | vsubpd zmm1, zmm2, zmm21
|
|
| 1 | | | | | | 1.0 | | | vsubpd zmm21, zmm6, zmm23
|
|
| 1 | 1.0 | | | | | | | | vmulpd zmm20, zmm0, zmm0
|
|
| 1 | | | | | | 1.0 | | | vfmadd231pd zmm20, zmm1, zmm1
|
|
| 1 | 1.0 | | | | | | | | vfmadd231pd zmm20, zmm21, zmm21
|
|
| 3 | 2.0 | | | | | 1.0 | | | vrcp14pd zmm31, zmm20
|
|
| 1 | | | | | | 1.0 | | | vcmppd k6{k5}, zmm20, zmm16, 0x1
|
|
| 1 | | | | | | 1.0 | | | vfpclasspd k0, zmm31, 0x1e
|
|
| 1* | | | | | | | | | vmovaps zmm24, zmm20
|
|
| 2^ | 1.0 | | | 1.0 1.0 | | | | | vfnmadd213pd zmm24, zmm31, qword ptr [rip]{1to8}
|
|
| 1 | 1.0 | | | | | | | | knotw k4, k0
|
|
| 1 | | | | | | 1.0 | | | vmulpd zmm25, zmm24, zmm24
|
|
| 1 | | | | | | 1.0 | | | vfmadd213pd zmm31{k4}, zmm24, zmm31
|
|
| 1 | 1.0 | | | | | | | | vfmadd213pd zmm31{k4}, zmm25, zmm31
|
|
| 1 | | | | | | 1.0 | | | vmulpd zmm26, zmm31, zmm15
|
|
| 1 | 1.0 | | | | | | | | vmulpd zmm28, zmm31, zmm14
|
|
| 1 | | | | | | 1.0 | | | vmulpd zmm29, zmm31, zmm26
|
|
| 1 | 1.0 | | | | | | | | vmulpd zmm27, zmm31, zmm29
|
|
| 1 | | | | | | 1.0 | | | vfmsub213pd zmm31, zmm29, zmm7
|
|
| 1 | 1.0 | | | | | | | | vmulpd zmm30, zmm27, zmm28
|
|
| 1 | | | | | | 1.0 | | | vmulpd zmm24, zmm30, zmm31
|
|
| 1 | 1.0 | | | | | | | | vfmadd231pd zmm13{k6}, zmm24, zmm1
|
|
| 1 | | | | | | 1.0 | | | vfmadd231pd zmm12{k6}, zmm24, zmm0
|
|
| 1 | 1.0 | | | | | | | | vfmadd231pd zmm11{k6}, zmm24, zmm21
|
|
| 1* | | | | | | | | | cmp r15, r14
|
|
| 0*F | | | | | | | | | jb 0xffffffffffffff19
|
|
Total Num Of Uops: 55
|
|
Analysis Notes:
|
|
Backend allocation was stalled due to unavailable allocation resources.
|
|
There were bubbles in the frontend.
|