OV - exec - Function hypre_CSRMatrixMatvecT.omp

	Function: hypre_CSRMatrixMatvecT.omp_outlined.24	Module: exec	Source: csr_matvec.c:554-579 [...]	Coverage (incl. loops): 1.23% \| (excl. loops): 0.00%

/home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c: 554 - 579

--------------------------------------------------------------------------------

554: #pragma omp parallel private(i,jj,j,my_thread_num,offset)

555: #endif

556:          {

557:             my_thread_num = hypre_GetThreadNum();

558:             offset =  y_size*my_thread_num;

559: #ifdef HYPRE_USING_OPENMP

560: #pragma omp for HYPRE_SMP_SCHEDULE

561: #endif

562:             for (i = 0; i < num_rows; i++)

563:             {

564:                for (jj = A_i[i]; jj < A_i[i+1]; jj++)

565:                {

566:                   j = A_j[jj];

567:                   y_data_expand[offset + j] += A_data[jj] * x_data[i];

[...]

573: #pragma omp for HYPRE_SMP_SCHEDULE

574: #endif

575:             for (i = 0; i < y_size; i++)

576:             {

577:                for (j = 0; j < num_threads; j++)

578:                {

579:                   y_data[i] += y_data_expand[j*y_size + i];

0x2fa300 PUSH	%RBP

0x2fa301 MOV	%RSP,%RBP

0x2fa304 PUSH	%R15

0x2fa306 PUSH	%R14

0x2fa308 PUSH	%R13

0x2fa30a PUSH	%R12

0x2fa30c PUSH	%RBX

0x2fa30d SUB	$0x58,%RSP

0x2fa311 MOV	%R9,-0x58(%RBP)

0x2fa315 MOV	%R8,-0x50(%RBP)

0x2fa319 MOV	%RCX,%RBX

0x2fa31c MOV	%RDX,%R14

0x2fa31f MOV	%RDI,%R12

0x2fa322 CALL	303460 <hypre_GetThreadNum>

0x2fa327 MOV	(%RBX),%R13

0x2fa32a MOV	(%R12),%EBX

0x2fa32e TEST	%R13,%R13

0x2fa331 JLE	2fa3b2

0x2fa333 MOV	%RAX,%R15

0x2fa336 IMUL	(%R14),%R15

0x2fa33a DEC	%R13

0x2fa33d MOVQ	$0,-0x48(%RBP)

0x2fa345 MOVQ	$0x1,-0x60(%RBP)

0x2fa34d MOVL	$0,-0x3c(%RBP)

0x2fa354 MOV	%R13,-0x38(%RBP)

0x2fa358 SUB	$0x8,%RSP

0x2fa35c LEA	-0x60(%RBP),%RAX

0x2fa360 LEA	0x11509(%RIP),%RDI

0x2fa367 LEA	-0x3c(%RBP),%RCX

0x2fa36b LEA	-0x48(%RBP),%R8

0x2fa36f LEA	-0x38(%RBP),%R9

0x2fa373 MOV	%EBX,%ESI

0x2fa375 MOV	$0x22,%EDX

0x2fa37a PUSH	$0x1

0x2fa37c PUSH	$0x1

0x2fa37e PUSH	%RAX

0x2fa37f CALL	305fa0 <@plt_start@+0x90>

0x2fa384 ADD	$0x20,%RSP

0x2fa388 MOV	-0x38(%RBP),%RAX

0x2fa38c CMP	%R13,%RAX

0x2fa38f CMOVL	%RAX,%R13

0x2fa393 MOV	-0x48(%RBP),%RAX

0x2fa397 MOV	%R13,-0x38(%RBP)

0x2fa39b CMP	%R13,%RAX

0x2fa39e JLE	2fa66a

0x2fa3a4 LEA	0x114dd(%RIP),%RDI

0x2fa3ab MOV	%EBX,%ESI

0x2fa3ad CALL	305fb0 <@plt_start@+0xa0>

0x2fa3b2 LEA	0x114e7(%RIP),%RDI

0x2fa3b9 MOV	%EBX,%ESI

0x2fa3bb CALL	305f90 <@plt_start@+0x80>

0x2fa3c0 MOV	(%R14),%R15

0x2fa3c3 TEST	%R15,%R15

0x2fa3c6 JLE	2fa64d

0x2fa3cc DEC	%R15

0x2fa3cf MOVQ	$0,-0x48(%RBP)

0x2fa3d7 MOVQ	$0x1,-0x60(%RBP)

0x2fa3df MOVL	$0,-0x3c(%RBP)

0x2fa3e6 MOV	%R15,-0x38(%RBP)

0x2fa3ea SUB	$0x8,%RSP

0x2fa3ee LEA	-0x60(%RBP),%RAX

0x2fa3f2 LEA	0x114bf(%RIP),%RDI

0x2fa3f9 LEA	-0x3c(%RBP),%RCX

0x2fa3fd LEA	-0x48(%RBP),%R8

0x2fa401 LEA	-0x38(%RBP),%R9

0x2fa405 MOV	%EBX,-0x64(%RBP)

0x2fa408 MOV	%EBX,%ESI

0x2fa40a MOV	$0x22,%EDX

0x2fa40f PUSH	$0x1

0x2fa411 PUSH	$0x1

0x2fa413 PUSH	%RAX

0x2fa414 CALL	305fa0 <@plt_start@+0x90>

0x2fa419 ADD	$0x20,%RSP

0x2fa41d MOV	-0x38(%RBP),%RAX

0x2fa421 CMP	%R15,%RAX

0x2fa424 CMOVL	%RAX,%R15

0x2fa428 MOV	-0x48(%RBP),%RAX

0x2fa42c MOV	%R15,-0x38(%RBP)

0x2fa430 CMP	%R15,%RAX

0x2fa433 JG	2fa639

0x2fa439 MOV	0x28(%RBP),%RCX

0x2fa43d MOV	(%RCX),%RCX

0x2fa440 TEST	%RCX,%RCX

0x2fa443 JLE	2fa639

0x2fa449 MOV	0x10(%RBP),%RDX

0x2fa44d MOV	0x30(%RBP),%RSI

0x2fa451 MOV	(%R14),%RDI

0x2fa454 LEA	(,%RAX,8),%R8

0x2fa45c CMP	$0x4,%RCX

0x2fa460 MOV	%RCX,%R14

0x2fa463 MOV	%RCX,%R12

0x2fa466 MOV	%R8,-0x50(%RBP)

0x2fa46a LEA	0x8(,%RAX,8),%R8

0x2fa472 MOV	%R8,-0x70(%RBP)

0x2fa476 MOV	(%RDX),%RDX

0x2fa479 MOV	(%RSI),%R9

0x2fa47c LEA	(%RAX,%RCX,1),%RSI

0x2fa480 MOV	%RDI,-0x58(%RBP)

0x2fa484 MOV	%RSI,-0x78(%RBP)

0x2fa488 SETAE	%SIL

0x2fa48c CMP	$0x1,%RDI

0x2fa490 LEA	(,%RDI,8),%RDI

0x2fa498 SETE	%R11B

0x2fa49c AND	$-0x20,%R14

0x2fa4a0 AND	$-0x4,%R12

0x2fa4a4 XOR	%R10D,%R10D

0x2fa4a7 AND	%SIL,%R11B

0x2fa4aa MOV	%R11B,-0x29(%RBP)

0x2fa4ae LEA	0xc0(%RDX,%RAX,8),%R13

0x2fa4b6 LEA	(%RDX,%RAX,8),%R8

0x2fa4ba JMP	2fa4db

0x2fa4bc NOPL	(%RAX)

(3651) 0x2fa4c0 LEA	0x1(%RAX),%RSI

(3651) 0x2fa4c4 INC	%R10

(3651) 0x2fa4c7 ADD	$0x8,%R13

(3651) 0x2fa4cb ADD	$0x8,%R8

(3651) 0x2fa4cf CMP	%R15,%RAX

(3651) 0x2fa4d2 MOV	%RSI,%RAX

(3651) 0x2fa4d5 JGE	2fa639

(3651) 0x2fa4db VMOVQ	(%R9,%RAX,8),%XMM0

(3651) 0x2fa4e1 TEST	%R11B,%R11B

(3651) 0x2fa4e4 JE	2fa51a

(3651) 0x2fa4e6 MOV	-0x78(%RBP),%RSI

(3651) 0x2fa4ea MOV	-0x50(%RBP),%R11

(3651) 0x2fa4ee ADD	%R10,%RSI

(3651) 0x2fa4f1 LEA	(%RDX,%RSI,8),%RBX

(3651) 0x2fa4f5 LEA	(%R11,%R10,8),%RSI

(3651) 0x2fa4f9 LEA	(%R9,%RSI,1),%R11

(3651) 0x2fa4fd CMP	%RBX,%R11

(3651) 0x2fa500 JAE	2fa550

(3651) 0x2fa502 MOV	-0x70(%RBP),%R11

(3651) 0x2fa506 ADD	%RDX,%RSI

(3651) 0x2fa509 LEA	(%R11,%R10,8),%R11

(3651) 0x2fa50d ADD	%R9,%R11

(3651) 0x2fa510 CMP	%R11,%RSI

(3651) 0x2fa513 JAE	2fa550

(3651) 0x2fa515 MOVZX	-0x29(%RBP),%R11D

(3651) 0x2fa51a XOR	%ESI,%ESI

(3651) 0x2fa51c MOV	%RCX,%RBX

(3651) 0x2fa51f SUB	%RSI,%RBX

(3651) 0x2fa522 IMUL	-0x58(%RBP),%RSI

(3651) 0x2fa527 ADD	%RAX,%RSI

(3651) 0x2fa52a LEA	(%RDX,%RSI,8),%RSI

(3651) 0x2fa52e XCHG	%AX,%AX

  (3650) 0x2fa530 VADDSD	(%RSI),%XMM0,%XMM0

  (3650) 0x2fa534 ADD	%RDI,%RSI

  (3650) 0x2fa537 DEC	%RBX

  (3650) 0x2fa53a VMOVSD	%XMM0,(%R9,%RAX,8)

  (3650) 0x2fa540 JNE	2fa530

(3651) 0x2fa542 JMP	2fa4c0

0x2fa547 NOPW	(%RAX,%RAX,1)

(3651) 0x2fa550 MOVZX	-0x29(%RBP),%R11D

(3651) 0x2fa555 CMP	$0x20,%RCX

(3651) 0x2fa559 JAE	2fa562

(3651) 0x2fa55b XOR	%EBX,%EBX

(3651) 0x2fa55d JMP	2fa5f9

(3651) 0x2fa562 VMOVQ	%XMM0,%XMM0

(3651) 0x2fa566 VXORPD	%XMM1,%XMM1,%XMM1

(3651) 0x2fa56a VXORPD	%XMM2,%XMM2,%XMM2

(3651) 0x2fa56e VXORPD	%XMM3,%XMM3,%XMM3

(3651) 0x2fa572 XOR	%EBX,%EBX

(3651) 0x2fa574 NOPW	%CS:(%RAX,%RAX,1)

  (3652) 0x2fa580 VADDPD	-0xc0(%R13,%RBX,8),%ZMM0,%ZMM0

  (3652) 0x2fa588 VADDPD	-0x80(%R13,%RBX,8),%ZMM1,%ZMM1

  (3652) 0x2fa590 VADDPD	-0x40(%R13,%RBX,8),%ZMM2,%ZMM2

  (3652) 0x2fa598 VADDPD	(%R13,%RBX,8),%ZMM3,%ZMM3

  (3652) 0x2fa5a0 ADD	$0x20,%RBX

  (3652) 0x2fa5a4 CMP	%RBX,%R14

  (3652) 0x2fa5a7 JNE	2fa580

(3651) 0x2fa5a9 VADDPD	%ZMM0,%ZMM1,%ZMM0

(3651) 0x2fa5af VADDPD	%ZMM2,%ZMM3,%ZMM2

(3651) 0x2fa5b5 VADDPD	%ZMM0,%ZMM2,%ZMM0

(3651) 0x2fa5bb VEXTRACTF64X4	$0x1,%ZMM0,%YMM1

(3651) 0x2fa5c2 VADDPD	%ZMM1,%ZMM0,%ZMM0

(3651) 0x2fa5c8 VEXTRACTF128	$0x1,%YMM0,%XMM1

(3651) 0x2fa5ce VADDPD	%XMM1,%XMM0,%XMM0

(3651) 0x2fa5d2 VSHUFPD	$0x1,%XMM0,%XMM0,%XMM1

(3651) 0x2fa5d7 VADDSD	%XMM1,%XMM0,%XMM0

(3651) 0x2fa5db VMOVSD	%XMM0,(%R9,%RAX,8)

(3651) 0x2fa5e1 CMP	%R14,%RCX

(3651) 0x2fa5e4 JE	2fa4c0

(3651) 0x2fa5ea MOV	%R14,%RBX

(3651) 0x2fa5ed MOV	%R14,%RSI

(3651) 0x2fa5f0 TEST	$0x1c,%CL

(3651) 0x2fa5f3 JE	2fa51c

(3651) 0x2fa5f9 VMOVQ	%XMM0,%XMM0

(3651) 0x2fa5fd NOPL	(%RAX)

  (3653) 0x2fa600 VADDPD	(%R8,%RBX,8),%YMM0,%YMM0

  (3653) 0x2fa606 ADD	$0x4,%RBX

  (3653) 0x2fa60a CMP	%RBX,%R12

  (3653) 0x2fa60d JNE	2fa600

(3651) 0x2fa60f VEXTRACTF128	$0x1,%YMM0,%XMM1

(3651) 0x2fa615 MOV	%R12,%RSI

(3651) 0x2fa618 VADDPD	%XMM1,%XMM0,%XMM0

(3651) 0x2fa61c VSHUFPD	$0x1,%XMM0,%XMM0,%XMM1

(3651) 0x2fa621 VADDSD	%XMM1,%XMM0,%XMM0

(3651) 0x2fa625 VMOVSD	%XMM0,(%R9,%RAX,8)

(3651) 0x2fa62b CMP	%R12,%RCX

(3651) 0x2fa62e JE	2fa4c0

(3651) 0x2fa634 JMP	2fa51c

0x2fa639 MOV	-0x64(%RBP),%EBX

0x2fa63c LEA	0x1128d(%RIP),%RDI

0x2fa643 MOV	%EBX,%ESI

0x2fa645 VZEROUPPER

0x2fa648 CALL	305fb0 <@plt_start@+0xa0>

0x2fa64d LEA	0x11294(%RIP),%RDI

0x2fa654 MOV	%EBX,%ESI

0x2fa656 CALL	305f90 <@plt_start@+0x80>

0x2fa65b ADD	$0x58,%RSP

0x2fa65f POP	%RBX

0x2fa660 POP	%R12

0x2fa662 POP	%R13

0x2fa664 POP	%R14

0x2fa666 POP	%R15

0x2fa668 POP	%RBP

0x2fa669 RET

0x2fa66a MOV	-0x50(%RBP),%RCX

0x2fa66e MOV	0x20(%RBP),%RDI

0x2fa672 MOV	0x18(%RBP),%RSI

0x2fa676 MOV	-0x58(%RBP),%RDX

0x2fa67a MOV	0x10(%RBP),%R8

0x2fa67e MOV	(%RCX),%RCX

0x2fa681 MOV	(%RDX),%RDX

0x2fa684 MOV	(%RSI),%RSI

0x2fa687 MOV	(%RDI),%RDI

0x2fa68a MOV	(%R8),%R8

0x2fa68d MOV	(%RCX,%RAX,8),%R9

0x2fa691 JMP	2fa6b0

0x2fa693 NOPW	%CS:(%RAX,%RAX,1)

(3654) 0x2fa6a0 LEA	0x1(%RAX),%R10

(3654) 0x2fa6a4 CMP	%R13,%RAX

(3654) 0x2fa6a7 MOV	%R10,%RAX

(3654) 0x2fa6aa JGE	2fa3a4

(3654) 0x2fa6b0 MOV	%R9,%R10

(3654) 0x2fa6b3 MOV	0x8(%RCX,%RAX,8),%R9

(3654) 0x2fa6b8 CMP	%R9,%R10

(3654) 0x2fa6bb JGE	2fa6a0

(3654) 0x2fa6bd NOPL	(%RAX)

  (3655) 0x2fa6c0 MOV	(%RDX,%R10,8),%R11

  (3655) 0x2fa6c4 VMOVSD	(%RSI,%R10,8),%XMM0

  (3655) 0x2fa6ca VMOVSD	(%RDI,%RAX,8),%XMM1

  (3655) 0x2fa6cf INC	%R10

  (3655) 0x2fa6d2 ADD	%R15,%R11

  (3655) 0x2fa6d5 VFMADD213SD	(%R8,%R11,8),%XMM0,%XMM1

  (3655) 0x2fa6db VMOVSD	%XMM1,(%R8,%R11,8)

  (3655) 0x2fa6e1 CMP	%R10,%R9

  (3655) 0x2fa6e4 JNE	2fa6c0

(3654) 0x2fa6e6 JMP	2fa6a0

0x2fa6e8 NOPL	(%RAX,%RAX,1)

Coverage (%)	Name	Source Location	Module

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

min	med	avg	max

Percentile Index	10	20	30	40	50	60	70	80	90	100
Value

Path /

Average path: Display a virtual path defined by average values of all real paths

The code analyzed by CQA in that panel excludes loops and represents 0.00% of application time for run aocc_2

Source file and lines	csr_matvec.c:554-579
Module	exec

The function is defined in /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:554-567,573-579.

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Vectorization

Your function is probably not vectorized. Only 11% of vector register length is used (average across all SSE/AVX instructions). By vectorizing your function, you can lower the cost of an iteration from 18.63 to 1.40 cycles (13.30x speedup).

Details

Store and arithmetical SSE/AVX instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a vectorized function can use their full power.

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

No data for this section

CALL instructions

Detected function call instructions.

Details

Calling (and then returning from) a function prevents many compiler optimizations (like vectorization), breaks control flow (which reduces pipeline performance) and executes extra instructions to save/restore the registers used inside it, which is very expensive (dozens of cycles). Consider to inline small functions.

hypre_GetThreadNum: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:557
unknown: 6 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

These instructions generate more than one micro-operation and only one of them can be decoded during a cycle and the extra micro-operations increase pressure on execution units.

CMOVL: 2 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:562
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:575
LEA: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573

Type of elements and instruction set

No instructions are processing arithmetic or math operations on FP elements. This function is probably writing/copying data or processing integer elements.

Matching between your function (in the source code) and the binary function

The binary function does not contain any FP arithmetical operations. The binary function is loading 208 bytes. The binary function is storing 125 bytes.

General properties

nb instructions	143
nb uops	149
loop length	568
used x86 registers	16
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	15

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	18.63 cycles
front end	18.63 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	10.67	10.67	10.67	10.67	10.67	10.67	13.25	13.25	13.25	13.25	0.00	0.00	0.00	0.00	0.00	0.00
cycles	10.67	10.67	10.67	10.67	10.67	10.67	13.25	13.25	13.25	13.25	0.00	0.00	0.00	0.00	0.00	0.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	18.63
Dispatch	13.25
Overall L1	18.63

Vectorization ratios

all	2%
load	0%
store	0%
mul	0%
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	7%

Vector efficiency ratios

all	11%
load	12%
store	9%
mul	12%
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	12%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 18.63 cycles. At this rate:

4% of peak load performance is reached (11.17 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (6.71 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Performance is limited by instruction throughput (loading/decoding program instructions to execution core) (front-end is a bottleneck). By removing all these bottlenecks, you can lower the cost of an iteration from 18.63 to 13.25 cycles (1.41x speedup).

ASM code

In the binary file, the address of the function is: 2fa300

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x58,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %R9,-0x58(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %R8,-0x50(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %RCX,%RBX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RDX,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RDI,%R12	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 303460 <hypre_GetThreadNum>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV (%RBX),%R13	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
MOV (%R12),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
TEST %R13,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa3b2 <hypre_CSRMatrixMatvecT.omp_outlined.24+0xb2>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%R15	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
IMUL (%R14),%R15	1	0.33	0.33	0.33	0	0	0	0.25	0.25	0.25	0.25	3	0.33	scal (12.5%)
DEC %R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOVQ $0,-0x48(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVQ $0x1,-0x60(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVL $0,-0x3c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %R13,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SUB $0x8,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
LEA -0x60(%RBP),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA 0x11509(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x3c(%RBP),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x48(%RBP),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x22,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 305fa0 <@plt_start@+0x90>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x20,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV -0x38(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMP %R13,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
CMOVL %RAX,%R13	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	scal (12.5%)
MOV -0x48(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R13,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R13,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa66a <hypre_CSRMatrixMatvecT.omp_outlined.24+0x36a>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
LEA 0x114dd(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305fb0 <@plt_start@+0xa0>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
LEA 0x114e7(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305f90 <@plt_start@+0x80>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV (%R14),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
TEST %R15,%R15	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa64d <hypre_CSRMatrixMatvecT.omp_outlined.24+0x34d>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
DEC %R15	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOVQ $0,-0x48(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVQ $0x1,-0x60(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVL $0,-0x3c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SUB $0x8,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
LEA -0x60(%RBP),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA 0x114bf(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x3c(%RBP),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x48(%RBP),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,-0x64(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x22,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 305fa0 <@plt_start@+0x90>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x20,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV -0x38(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMP %R15,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
CMOVL %RAX,%R15	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	scal (12.5%)
MOV -0x48(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R15,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JG 2fa639 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x339>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV 0x28(%RBP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
TEST %RCX,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa639 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x339>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV 0x10(%RBP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x30(%RBP),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%R14),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA (,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
CMP $0x4,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOV %RCX,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RCX,%R12	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %R8,-0x50(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
LEA 0x8(,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %R8,-0x70(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV (%RDX),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RSI),%R9	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
LEA (%RAX,%RCX,1),%RSI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %RDI,-0x58(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %RSI,-0x78(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SETAE %SIL	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
CMP $0x1,%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
LEA (,%RDI,8),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
SETE %R11B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
AND $-0x20,%R14	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
AND $-0x4,%R12	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
XOR %R10D,%R10D	1	0	0	0	0	0	0	0	0	0	0	0	0.25	scal (6.3%)
AND %SIL,%R11B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %R11B,-0x29(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	N/A
LEA 0xc0(%RDX,%RAX,8),%R13	2	0	0	0.50	0.50	0	0	0.25	0.25	0.25	0.25	2	0.50	N/A
LEA (%RDX,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
JMP 2fa4db <hypre_CSRMatrixMatvecT.omp_outlined.24+0x1db>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV -0x64(%RBP),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA 0x1128d(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
CALL 305fb0 <@plt_start@+0xa0>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
LEA 0x11294(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305f90 <@plt_start@+0x80>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x58,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50	N/A
MOV -0x50(%RBP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x20(%RBP),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x18(%RBP),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV -0x58(%RBP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x10(%RBP),%R8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RDX),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RSI),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RDI),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%R8),%R8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX,%RAX,8),%R9	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
JMP 2fa6b0 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x3b0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPL (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

The code analyzed by CQA in that panel excludes loops and represents 0.00% of application time for run aocc_2

Source file and lines	csr_matvec.c:554-579
Module	exec

The function is defined in /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:554-567,573-579.

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Vectorization

Details

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function streams arrays of structures (AoS), try to use structures of arrays instead (SoA): for(i) a[i].x = b[i].x; (slow, non stride 1) => for(i) a.x[i] = b.x[i]; (fast, stride 1)

Execution units bottlenecks

Found no such bottlenecks but see expert reports for more complex bottlenecks.

No data for this section

CALL instructions

Detected function call instructions.

Details

hypre_GetThreadNum: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:557
unknown: 6 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:560
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

These instructions generate more than one micro-operation and only one of them can be decoded during a cycle and the extra micro-operations increase pressure on execution units.

CMOVL: 2 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:562
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:575
LEA: 1 occurrences►
- /home/eoseret/qaas_runs_ZEN5/173-989-6379/intel/AMG/build/AMG/AMG/seq_mv/csr_matvec.c:573

Type of elements and instruction set

No instructions are processing arithmetic or math operations on FP elements. This function is probably writing/copying data or processing integer elements.

Matching between your function (in the source code) and the binary function

The binary function does not contain any FP arithmetical operations. The binary function is loading 208 bytes. The binary function is storing 125 bytes.

General properties

nb instructions	143
nb uops	149
loop length	568
used x86 registers	16
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	15

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	18.63 cycles
front end	18.63 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	10.67	10.67	10.67	10.67	10.67	10.67	13.25	13.25	13.25	13.25	0.00	0.00	0.00	0.00	0.00	0.00
cycles	10.67	10.67	10.67	10.67	10.67	10.67	13.25	13.25	13.25	13.25	0.00	0.00	0.00	0.00	0.00	0.00

Execution ports to units layout:

P0: ALU
P1: ALU
P2: ALU
P3: ALU, BRU
P4: ALU, BRU
P5: ALU, BRU
P6 (512 bits): store address, load
P7 (512 bits): store address, load
P8 (512 bits): store address, load
P9 (512 bits): store address, load
P10 (512 bits): VPU, DIV/SQRT
P11 (512 bits): VPU, DIV/SQRT
P12 (512 bits): VPU
P13 (512 bits): VPU
P14 (512 bits): FP store data
P15 (512 bits): FP store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	18.63
Dispatch	13.25
Overall L1	18.63

Vectorization ratios

all	2%
load	0%
store	0%
mul	0%
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	7%

Vector efficiency ratios

all	11%
load	12%
store	9%
mul	12%
add-sub	12%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	12%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 18.63 cycles. At this rate:

4% of peak load performance is reached (11.17 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
5% of peak store performance is reached (6.71 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

ASM code

In the binary file, the address of the function is: 2fa300

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	Latency	Recip. throughput	Vectorization
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOV %RSP,%RBP	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x58,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %R9,-0x58(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %R8,-0x50(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %RCX,%RBX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RDX,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RDI,%R12	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 303460 <hypre_GetThreadNum>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV (%RBX),%R13	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
MOV (%R12),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
TEST %R13,%R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa3b2 <hypre_CSRMatrixMatvecT.omp_outlined.24+0xb2>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV %RAX,%R15	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (12.5%)
IMUL (%R14),%R15	1	0.33	0.33	0.33	0	0	0	0.25	0.25	0.25	0.25	3	0.33	scal (12.5%)
DEC %R13	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOVQ $0,-0x48(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVQ $0x1,-0x60(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVL $0,-0x3c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %R13,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SUB $0x8,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
LEA -0x60(%RBP),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA 0x11509(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x3c(%RBP),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x48(%RBP),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x22,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 305fa0 <@plt_start@+0x90>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x20,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV -0x38(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMP %R13,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
CMOVL %RAX,%R13	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	scal (12.5%)
MOV -0x48(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R13,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R13,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa66a <hypre_CSRMatrixMatvecT.omp_outlined.24+0x36a>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
LEA 0x114dd(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305fb0 <@plt_start@+0xa0>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
LEA 0x114e7(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305f90 <@plt_start@+0x80>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV (%R14),%R15	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
TEST %R15,%R15	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa64d <hypre_CSRMatrixMatvecT.omp_outlined.24+0x34d>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
DEC %R15	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOVQ $0,-0x48(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVQ $0x1,-0x60(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOVL $0,-0x3c(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SUB $0x8,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
LEA -0x60(%RBP),%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA 0x114bf(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x3c(%RBP),%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x48(%RBP),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA -0x38(%RBP),%R9	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,-0x64(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV $0x22,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH $0x1	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RAX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CALL 305fa0 <@plt_start@+0x90>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x20,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV -0x38(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMP %R15,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
CMOVL %RAX,%R15	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	scal (12.5%)
MOV -0x48(%RBP),%RAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R15,-0x38(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R15,%RAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JG 2fa639 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x339>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV 0x28(%RBP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
TEST %RCX,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
JLE 2fa639 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x339>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV 0x10(%RBP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x30(%RBP),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%R14),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA (,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
CMP $0x4,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
MOV %RCX,%R14	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %RCX,%R12	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %R8,-0x50(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
LEA 0x8(,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %R8,-0x70(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV (%RDX),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RSI),%R9	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
LEA (%RAX,%RCX,1),%RSI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %RDI,-0x58(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV %RSI,-0x78(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
SETAE %SIL	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
CMP $0x1,%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (12.5%)
LEA (,%RDI,8),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
SETE %R11B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
AND $-0x20,%R14	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
AND $-0x4,%R12	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
XOR %R10D,%R10D	1	0	0	0	0	0	0	0	0	0	0	0	0.25	scal (6.3%)
AND %SIL,%R11B	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %R11B,-0x29(%RBP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	N/A
LEA 0xc0(%RDX,%RAX,8),%R13	2	0	0	0.50	0.50	0	0	0.25	0.25	0.25	0.25	2	0.50	N/A
LEA (%RDX,%RAX,8),%R8	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
JMP 2fa4db <hypre_CSRMatrixMatvecT.omp_outlined.24+0x1db>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPL (%RAX)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
MOV -0x64(%RBP),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA 0x1128d(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
VZEROUPPER	1	0	0	0	0	0	0	0	0	0	0	0	1	vect (25.0%)
CALL 305fb0 <@plt_start@+0xa0>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
LEA 0x11294(%RIP),%RDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EBX,%ESI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CALL 305f90 <@plt_start@+0x80>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
ADD $0x58,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50	N/A
MOV -0x50(%RBP),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x20(%RBP),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x18(%RBP),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV -0x58(%RBP),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x10(%RBP),%R8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX),%RCX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RDX),%RDX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RSI),%RSI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RDI),%RDI	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%R8),%R8	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV (%RCX,%RAX,8),%R9	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
JMP 2fa6b0 <hypre_CSRMatrixMatvecT.omp_outlined.24+0x3b0>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPW %CS:(%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
NOPL (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

Name	Coverage (%)	Time (s)
▼hypre_CSRMatrixMatvecT.omp_outlined.24–	1.23	0.65
▼Loop 3654 - csr_matvec.c:560-567 - exec–	0.35	0.18
○Loop 3655 - csr_matvec.c:564-567 - exec	0.81	0.40
▼Loop 3651 - csr_matvec.c:573-579 - exec–	0.00	0.01
○Loop 3650 - csr_matvec.c:577-579 - exec	0.07	0.03
○Loop 3653 - csr_matvec.c:577-579 - exec	0.00	0.00
○Loop 3652 - csr_matvec.c:577-579 - exec	0.00	0.00

Report Configuration

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex instructions

Details

Type of elements and instruction set

Matching between your function (in the source code) and the binary function

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Vectorization

Details

Workaround

Execution units bottlenecks

CALL instructions

Details

Complex instructions

Details

Type of elements and instruction set

Matching between your function (in the source code) and the binary function

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code