OV - exec - Loop 2260

	Loop Id: 2260	Module: exec	Source: ams.c:3382-3401 [...]	Coverage: 0.03%

0x4c4fc4 MOVI	D31, #0

0x4c4fc8 CBNZ	X4, 4c5154

0x4c4fcc ADD	X12, X12, #1

0x4c4fd0 CMP	X26, X12

0x4c4fd4 B.EQ	4c5560

0x4c4fd8 LDR	X14, [X7, X12,LSL #3]

0x4c4fdc ORR	X3, XZR, X12

0x4c4fe0 LDR	X0, [X24, X12,LSL #3]

0x4c4fe4 STR	XZR, [X19, X12,LSL #3]

0x4c4fe8 CMP	X14, X0

0x4c4fec B.LE	4c4fc4

0x4c4ff0 ADD	X18, X21, X14,LSL #3

0x4c4ff4 UBFM	X14, X0, #61, #60

0x4c4ff8 MOVI	D31, #0

0x4c4ffc ADD	X0, X21, X0,LSL #3

0x4c5000 SUB	X16, X18, X0

0x4c5004 SUB	X11, X16, #8

0x4c5008 UBFM	X2, X11, #3, #63

0x4c500c ADD	X30, X2, #1

0x4c5010 ANDS	X17, X30, #0x7

0x4c5014 B.EQ	4c50c0

0x4c5018 CMP	X17, #1

0x4c501c B.EQ	4c50a8

0x4c5020 CMP	X17, #2

0x4c5024 B.EQ	4c5098

0x4c5028 CMP	X17, #3

0x4c502c B.EQ	4c5088

0x4c5030 CMP	X17, #4

0x4c5034 B.EQ	4c5078

0x4c5038 CMP	X17, #5

0x4c503c B.EQ	4c5068

0x4c5040 CMP	X17, #6

0x4c5044 B.EQ	4c5058

0x4c5048 LDR	D25, [X21, X14]

0x4c504c ADD	X0, X0, #8

0x4c5050 FABS	D31, D25

0x4c5054 STR	D31, [X19, X3,LSL #3]

0x4c5058 LDR	D26, [X0], #8

0x4c505c FABS	D27, D26

0x4c5060 FADD	D31, D31, D27

0x4c5064 STR	D31, [X19, X3,LSL #3]

0x4c5068 LDR	D6, [X0], #8

0x4c506c FABS	D0, D6

0x4c5070 FADD	D31, D31, D0

0x4c5074 STR	D31, [X19, X3,LSL #3]

0x4c5078 LDR	D1, [X0], #8

0x4c507c FABS	D2, D1

0x4c5080 FADD	D31, D31, D2

0x4c5084 STR	D31, [X19, X3,LSL #3]

0x4c5088 LDR	D3, [X0], #8

0x4c508c FABS	D4, D3

0x4c5090 FADD	D31, D31, D4

0x4c5094 STR	D31, [X19, X3,LSL #3]

0x4c5098 LDR	D5, [X0], #8

0x4c509c FABS	D7, D5

0x4c50a0 FADD	D31, D31, D7

0x4c50a4 STR	D31, [X19, X3,LSL #3]

0x4c50a8 LDR	D16, [X0], #8

0x4c50ac FABS	D17, D16

0x4c50b0 FADD	D31, D31, D17

0x4c50b4 STR	D31, [X19, X3,LSL #3]

0x4c50b8 CMP	X18, X0

0x4c50bc B.EQ	4c4fc8

  (2262) 0x4c50c0 ORR	X16, XZR, X0

  (2262) 0x4c50c4 ADD	X0, X0, #64

  (2262) 0x4c50c8 LDR	D17, [X16], #8

  (2262) 0x4c50cc FABS	D18, D17

  (2262) 0x4c50d0 FADD	D31, D31, D18

  (2262) 0x4c50d4 STR	D31, [X19, X3,LSL #3]

  (2262) 0x4c50d8 LDUR	D19, [X0, #456]

  (2262) 0x4c50dc FABS	D20, D19

  (2262) 0x4c50e0 FADD	D28, D31, D20

  (2262) 0x4c50e4 STR	D28, [X19, X3,LSL #3]

  (2262) 0x4c50e8 LDR	D29, [X16, #8]

  (2262) 0x4c50ec FABS	D24, D29

  (2262) 0x4c50f0 FADD	D21, D28, D24

  (2262) 0x4c50f4 STR	D21, [X19, X3,LSL #3]

  (2262) 0x4c50f8 LDUR	D22, [X0, #472]

  (2262) 0x4c50fc FABS	D30, D22

  (2262) 0x4c5100 FADD	D23, D21, D30

  (2262) 0x4c5104 STR	D23, [X19, X3,LSL #3]

  (2262) 0x4c5108 LDUR	D25, [X0, #480]

  (2262) 0x4c510c FABS	D26, D25

  (2262) 0x4c5110 FADD	D27, D23, D26

  (2262) 0x4c5114 STR	D27, [X19, X3,LSL #3]

  (2262) 0x4c5118 LDUR	D6, [X0, #488]

  (2262) 0x4c511c FABS	D0, D6

  (2262) 0x4c5120 FADD	D1, D27, D0

  (2262) 0x4c5124 STR	D1, [X19, X3,LSL #3]

  (2262) 0x4c5128 LDUR	D2, [X0, #496]

  (2262) 0x4c512c FABS	D3, D2

  (2262) 0x4c5130 FADD	D31, D1, D3

  (2262) 0x4c5134 STR	D31, [X19, X3,LSL #3]

  (2262) 0x4c5138 LDUR	D16, [X0, #504]

  (2262) 0x4c513c FABS	D17, D16

  (2262) 0x4c5140 FADD	D31, D31, D17

  (2262) 0x4c5144 STR	D31, [X19, X3,LSL #3]

  (2262) 0x4c5148 CMP	X18, X0

  (2262) 0x4c514c B.NE	4c50c0

0x4c5150 CBZ	X4, 4c4fcc

0x4c5154 LDR	X18, [X10, X3,LSL #3]

0x4c5158 LDR	X14, [X15, X3,LSL #3]

0x4c515c CMP	X14, X18

0x4c5160 B.LE	4c4fcc

0x4c5164 ADD	X16, X6, X14,LSL #3

0x4c5168 UBFM	X11, X18, #61, #60

0x4c516c ADD	X0, X6, X18,LSL #3

0x4c5170 SUB	X2, X16, X0

0x4c5174 SUB	X30, X2, #8

0x4c5178 UBFM	X17, X30, #3, #63

0x4c517c ADD	X18, X17, #1

0x4c5180 ANDS	X14, X18, #0x7

0x4c5184 B.EQ	4c5220

0x4c5188 CMP	X14, #1

0x4c518c B.EQ	4c5208

0x4c5190 CMP	X14, #2

0x4c5194 B.EQ	4c51f8

0x4c5198 CMP	X14, #3

0x4c519c B.EQ	4c51e8

0x4c51a0 CMP	X14, #4

0x4c51a4 B.EQ	4c51d8

0x4c51a8 CMP	X14, #5

0x4c51ac B.EQ	4c51c8

0x4c51b0 CMP	X14, #6

0x4c51b4 B.NE	4c83b0

0x4c51b8 LDR	D20, [X0], #8

0x4c51bc FABS	D28, D20

0x4c51c0 FADD	D31, D31, D28

0x4c51c4 STR	D31, [X19, X3,LSL #3]

0x4c51c8 LDR	D29, [X0], #8

0x4c51cc FABS	D24, D29

0x4c51d0 FADD	D31, D31, D24

0x4c51d4 STR	D31, [X19, X3,LSL #3]

0x4c51d8 LDR	D21, [X0], #8

0x4c51dc FABS	D22, D21

0x4c51e0 FADD	D31, D31, D22

0x4c51e4 STR	D31, [X19, X3,LSL #3]

0x4c51e8 LDR	D30, [X0], #8

0x4c51ec FABS	D23, D30

0x4c51f0 FADD	D31, D31, D23

0x4c51f4 STR	D31, [X19, X3,LSL #3]

0x4c51f8 LDR	D25, [X0], #8

0x4c51fc FABS	D26, D25

0x4c5200 FADD	D31, D31, D26

0x4c5204 STR	D31, [X19, X3,LSL #3]

0x4c5208 LDR	D27, [X0], #8

0x4c520c FABS	D6, D27

0x4c5210 FADD	D31, D31, D6

0x4c5214 STR	D31, [X19, X3,LSL #3]

0x4c5218 CMP	X0, X16

0x4c521c B.EQ	4c4fcc

  (2261) 0x4c5220 ORR	X2, XZR, X0

  (2261) 0x4c5224 ADD	X0, X0, #64

  (2261) 0x4c5228 LDR	D20, [X2], #8

  (2261) 0x4c522c FABS	D28, D20

  (2261) 0x4c5230 FADD	D29, D31, D28

  (2261) 0x4c5234 STR	D29, [X19, X3,LSL #3]

  (2261) 0x4c5238 LDUR	D24, [X0, #456]

  (2261) 0x4c523c FABS	D21, D24

  (2261) 0x4c5240 FADD	D22, D29, D21

  (2261) 0x4c5244 STR	D22, [X19, X3,LSL #3]

  (2261) 0x4c5248 LDR	D30, [X2, #8]

  (2261) 0x4c524c FABS	D23, D30

  (2261) 0x4c5250 FADD	D25, D22, D23

  (2261) 0x4c5254 STR	D25, [X19, X3,LSL #3]

  (2261) 0x4c5258 LDUR	D26, [X0, #472]

  (2261) 0x4c525c FABS	D27, D26

  (2261) 0x4c5260 FADD	D6, D25, D27

  (2261) 0x4c5264 STR	D6, [X19, X3,LSL #3]

  (2261) 0x4c5268 LDUR	D0, [X0, #480]

  (2261) 0x4c526c FABS	D1, D0

  (2261) 0x4c5270 FADD	D2, D6, D1

  (2261) 0x4c5274 STR	D2, [X19, X3,LSL #3]

  (2261) 0x4c5278 LDUR	D3, [X0, #488]

  (2261) 0x4c527c FABS	D4, D3

  (2261) 0x4c5280 FADD	D5, D2, D4

  (2261) 0x4c5284 STR	D5, [X19, X3,LSL #3]

  (2261) 0x4c5288 LDUR	D7, [X0, #496]

  (2261) 0x4c528c FABS	D16, D7

  (2261) 0x4c5290 FADD	D31, D5, D16

  (2261) 0x4c5294 STR	D31, [X19, X3,LSL #3]

  (2261) 0x4c5298 LDUR	D27, [X0, #504]

  (2261) 0x4c529c FABS	D6, D27

  (2261) 0x4c52a0 FADD	D31, D31, D6

  (2261) 0x4c52a4 STR	D31, [X19, X3,LSL #3]

  (2261) 0x4c52a8 CMP	X0, X16

  (2261) 0x4c52ac B.NE	4c5220

0x4c52b0 B	4c4fcc

0x4c83b0 LDR	D18, [X6, X11]

0x4c83b4 ADD	X0, X0, #8

0x4c83b8 FABS	D19, D18

0x4c83bc FADD	D31, D31, D19

0x4c83c0 STR	D31, [X19, X3,LSL #3]

0x4c83c4 B	4c51b8

/home/eoseret/qaas/qaas_runs/178-188-3659/intel/AMG/build/AMG/AMG/parcsr_ls/ams.c: 3382 - 3401

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

3382:          for (i = ns; i < ne; i++)

3383:          {

3384:             l1_norm[i] = 0.0;

3385:             if (cf_marker == NULL)

3386:             {

3387:                /* Add the l1 norm of the diag part of the ith row */

3388:                for (j = A_diag_I[i]; j < A_diag_I[i+1]; j++)

3389:                   l1_norm[i] += fabs(A_diag_data[j]);

3390:                /* Add the l1 norm of the offd part of the ith row */

3391:                if (num_cols_offd)

3392:                {

3393:                   for (j = A_offd_I[i]; j < A_offd_I[i+1]; j++)

3394:                      l1_norm[i] += fabs(A_offd_data[j]);

[...]

3401:                for (j = A_diag_I[i]; j < A_diag_I[i+1]; j++)

Coverage (%)	Name	Module
►100.00+	omp_fulfill_event	libgomp.so.1.0.0
○	start_thread	libc.so.6
○	thread_start	libc.so.6

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 /

Metric	Value
CQA speedup if no scalar integer	1.63
CQA speedup if FP arith vectorized	1.46
CQA speedup if fully vectorized	4.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.24
Bottlenecks	micro-operation queue,
Function	hypre_ParCSRComputeL1NormsThreads._omp_fn.0
Source	ams.c:3382-3384,ams.c:3388-3394,ams.c:3401-3401
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	15.25
CQA cycles if no scalar integer	9.33
CQA cycles if FP arith vectorized	10.47
CQA cycles if fully vectorized	3.81
Front-end cycles	15.25
P0 cycles	11.50
P1 cycles	11.50
P2 cycles	12.25
P3 cycles	12.25
P4 cycles	12.25
P5 cycles	12.25
P6 cycles	10.75
P7 cycles	10.75
P8 cycles	10.75
P9 cycles	10.75
P10 cycles	11.17
P11 cycles	10.83
P12 cycles	11.00
P13 cycles	0.50
P14 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	122.00
Nb uops	122.00
Nb loads	NA
Nb stores	15.00
Nb stack references	0.00
FLOP/cycle	0.85
Nb FLOP add-sub	13.00
Nb FLOP mul	0.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	0.00
Bytes prefetched	0.00
Bytes loaded	0.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	0.00
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	25.00
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	25.00

Metric	Value
CQA speedup if no scalar integer	1.63
CQA speedup if FP arith vectorized	1.46
CQA speedup if fully vectorized	4.00
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.24
Bottlenecks	micro-operation queue,
Function	hypre_ParCSRComputeL1NormsThreads._omp_fn.0
Source	ams.c:3382-3384,ams.c:3388-3394,ams.c:3401-3401
Source loop unroll info	NA
Source loop unroll confidence level	NA
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	15.25
CQA cycles if no scalar integer	9.33
CQA cycles if FP arith vectorized	10.47
CQA cycles if fully vectorized	3.81
Front-end cycles	15.25
P0 cycles	11.50
P1 cycles	11.50
P2 cycles	12.25
P3 cycles	12.25
P4 cycles	12.25
P5 cycles	12.25
P6 cycles	10.75
P7 cycles	10.75
P8 cycles	10.75
P9 cycles	10.75
P10 cycles	11.17
P11 cycles	10.83
P12 cycles	11.00
P13 cycles	0.50
P14 cycles	0.50
DIV/SQRT cycles	0.00
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	122.00
Nb uops	122.00
Nb loads	NA
Nb stores	15.00
Nb stack references	0.00
FLOP/cycle	0.85
Nb FLOP add-sub	13.00
Nb FLOP mul	0.00
Nb FLOP fma	0.00
Nb FLOP div	0.00
Nb FLOP rcp	0.00
Nb FLOP sqrt	0.00
Nb FLOP rsqrt	0.00
Bytes/cycle	0.00
Bytes prefetched	0.00
Bytes loaded	0.00
Bytes stored	0.00
Stride 0	NA
Stride 1	NA
Stride n	NA
Stride unknown	NA
Stride indirect	NA
Vectorization ratio all	0.00
Vectorization ratio load	0.00
Vectorization ratio store	0.00
Vectorization ratio mul	NA
Vectorization ratio add_sub	0.00
Vectorization ratio fma	NA
Vectorization ratio div_sqrt	NA
Vectorization ratio other	0.00
Vector-efficiency ratio all	25.00
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	NA
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	NA
Vector-efficiency ratio div_sqrt	NA
Vector-efficiency ratio other	25.00

Path /

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

Function	hypre_ParCSRComputeL1NormsThreads._omp_fn.0
Source file and lines	ams.c:3382-3401
Module	exec

The loop is defined in /home/eoseret/qaas/qaas_runs/178-188-3659/intel/AMG/build/AMG/AMG/parcsr_ls/ams.c:3382-3394,3401.

Analyzed code is defined in /home/eoseret/qaas/qaas_runs/178-188-3659/intel/AMG/build/AMG/AMG/parcsr_ls/ams.c:3382-3384,3388-3394,3401.

Warnings:

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Too many paths. If you really need to analyze all of the 272 paths individually, rerun with max-paths=272
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

Try to simplify control and/or increase the maximum number of paths per function/loop through the 'max-paths-nb' option.

This loop has 272 execution paths.

The presence of multiple execution paths is typically the main/first bottleneck.
Try to simplify control inside loop: ideally, try to remove all conditional expressions, for example by (if applicable):

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = (x<0 ? 0 : x) (or MAX(0,x) after defining the corresponding macro)

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 15.25 to 9.33 cycles (1.63x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)

Vectorization

Your loop is not vectorized. 4 data elements could be processed at once in vector registers. By vectorizing your loop, you can lower the cost of an iteration from 15.25 to 3.81 cycles (4.00x speedup).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop 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)

No data for this section

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

The binary loop is composed of 13 FP arithmetical operations:

13: addition or subtraction

The binary loop does not load or store any data.

General properties

nb instructions	122
nb uops	122
loop length	488
used w registers	0
used x registers	20
used b registers	0
used h registers	0
used s registers	0
used d registers	24
used q registers	0
used v registers	0
used z registers	0
nb stack references	0

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	15.25 cycles
front end	15.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	11.50	11.50	12.25	12.25	12.25	12.25	10.75	10.75	10.75	10.75	11.17	10.83	11.00	0.50	0.50
cycles	11.50	11.50	12.25	12.25	12.25	12.25	10.75	10.75	10.75	10.75	11.17	10.83	11.00	0.50	0.50

Execution ports to units layout:

P0: BRU
P1: BRU
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P7 (128 bits): VPU, ALU, FP store data
P8 (128 bits): VPU, ALU, DIV/SQRT
P9 (128 bits): ALU, VPU
P10 (256 bits): store address, load
P11 (256 bits): store address, load
P12 (256 bits): load
P13 (64 bits): store data
P14 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	15.25
Dispatch	12.25
Overall L1	15.25

Vectorization ratios

INT

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

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

INT+FP

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

Vector efficiency ratios

INT

all	25%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
other	25%

all	25%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	25%

INT+FP

all	25%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	25%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 15.25 cycles.

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 15.25 to 12.25 cycles (1.24x speedup).

ASM code

In the binary file, the address of the loop is: 4c4fe8

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	N/A
CBNZ X4, 4c5154 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1294>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X12, X12, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP X26, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5560 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x16a0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X14, [X7, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
ORR X3, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LDR X0, [X24, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
STR XZR, [X19, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (25.0%)
CMP X14, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.LE 4c4fc4 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1104>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X18, X21, X14,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X14, X0, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	N/A
ADD X0, X21, X0,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB X16, X18, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X11, X16, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
UBFM X2, X11, #3, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X30, X2, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ANDS X17, X30, #0x7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c50c0 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1200>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c50a8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11e8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5098 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5088 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5078 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5068 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11a8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5058 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1198>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D25, [X21, X14]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
ADD X0, X0, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
FABS D31, D25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D26, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D27, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D6, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D0, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D1, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D2, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D3, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D4, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D5, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D7, D5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D16, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D17, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
CMP X18, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c4fc8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1108>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CBZ X4, 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X18, [X10, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X14, [X15, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
CMP X14, X18	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.LE 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X16, X6, X14,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X11, X18, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X0, X6, X18,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB X2, X16, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X30, X2, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X17, X30, #3, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X18, X17, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ANDS X14, X18, #0x7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	N/A
B.EQ 4c5220 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5208 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1348>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51f8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1338>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51e8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1328>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51d8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1318>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51c8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1308>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.NE 4c83b0 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x44f0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D20, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D28, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D29, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D24, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D21, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D22, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D22	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D30, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D23, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D25, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D26, D25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D27, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D6, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
CMP X0, X16	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
B 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D18, [X6, X11]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
ADD X0, X0, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
FABS D19, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
B 4c51b8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x12f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	hypre_ParCSRComputeL1NormsThreads._omp_fn.0
Source file and lines	ams.c:3382-3401
Module	exec

Non-innermost loop: analyzing only self part (ignoring child loops).
Ignoring paths for analysis
Too many paths. If you really need to analyze all of the 272 paths individually, rerun with max-paths=272
RecMII not computed since number of paths is unknown or > max_paths
Streams not analyzed since number of paths is unknown or > max_paths

hoisting them (moving them outside the loop)
turning them into conditional moves, MIN or MAX

Ex: if (x<0) x=0 => x = (x<0 ? 0 : x) (or MAX(0,x) after defining the corresponding macro)

gain
potential
hint
expert

Code clean check

Detected a slowdown caused by scalar integer instructions (typically used for address computation). By removing them, you can lower the cost of an iteration from 15.25 to 9.33 cycles (1.63x speedup).

Workaround

Try to reorganize arrays of structures to structures of arrays
Consider to permute loops (see vectorization gain report)

Vectorization

Details

All VPU instructions are used in scalar version (process only one data element in vector registers). Since your execution units are vector units, only a vectorized loop can use their full power.

Workaround

Try another compiler or update/tune your current one
Remove inter-iterations dependences from your loop and make it unit-stride:
- If your arrays have 2 or more dimensions, check whether elements are accessed contiguously and, otherwise, try to permute loops accordingly: C storage order is row-major: for(i) for(j) a[j][i] = b[j][i]; (slow, non stride 1) => for(i) for(j) a[i][j] = b[i][j]; (fast, stride 1)
- If your loop 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)

No data for this section

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

The binary loop is composed of 13 FP arithmetical operations:

13: addition or subtraction

The binary loop does not load or store any data.

General properties

nb instructions	122
nb uops	122
loop length	488
used w registers	0
used x registers	20
used b registers	0
used h registers	0
used s registers	0
used d registers	24
used q registers	0
used v registers	0
used z registers	0
nb stack references	0

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	15.25 cycles
front end	15.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	11.50	11.50	12.25	12.25	12.25	12.25	10.75	10.75	10.75	10.75	11.17	10.83	11.00	0.50	0.50
cycles	11.50	11.50	12.25	12.25	12.25	12.25	10.75	10.75	10.75	10.75	11.17	10.83	11.00	0.50	0.50

Execution ports to units layout:

P0: BRU
P1: BRU
P2: ALU
P3: ALU
P4: ALU
P5: ALU
P6 (128 bits): VPU, FP store data, ALU, DIV/SQRT
P7 (128 bits): VPU, ALU, FP store data
P8 (128 bits): VPU, ALU, DIV/SQRT
P9 (128 bits): ALU, VPU
P10 (256 bits): store address, load
P11 (256 bits): store address, load
P12 (256 bits): load
P13 (64 bits): store data
P14 (64 bits): store data

Cycles executing div or sqrt instructions

Cycles summary

Front-end	15.25
Dispatch	12.25
Overall L1	15.25

Vectorization ratios

INT

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

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

INT+FP

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

Vector efficiency ratios

INT

all	25%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
other	25%

all	25%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	25%

INT+FP

all	25%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	25%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 15.25 cycles.

Front-end bottlenecks

ASM code

In the binary file, the address of the loop is: 4c4fe8

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	N/A
CBNZ X4, 4c5154 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1294>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X12, X12, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
CMP X26, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5560 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x16a0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X14, [X7, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
ORR X3, XZR, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
LDR X0, [X24, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
STR XZR, [X19, X12,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (25.0%)
CMP X14, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.LE 4c4fc4 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1104>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X18, X21, X14,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X14, X0, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	N/A
ADD X0, X21, X0,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB X16, X18, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X11, X16, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
UBFM X2, X11, #3, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X30, X2, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ANDS X17, X30, #0x7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c50c0 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1200>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c50a8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11e8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5098 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11d8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5088 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11c8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5078 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5068 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x11a8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X17, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5058 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1198>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D25, [X21, X14]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
ADD X0, X0, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
FABS D31, D25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D26, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D27, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D6, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D0, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D1, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D2, D1	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D2	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D3, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D4, D3	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D4	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D5, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D7, D5	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D7	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D16, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D17, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
CMP X18, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c4fc8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1108>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CBZ X4, 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR X18, [X10, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDR X14, [X15, X3,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	N/A
CMP X14, X18	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.LE 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD X16, X6, X14,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X11, X18, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
ADD X0, X6, X18,LSL #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
SUB X2, X16, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
SUB X30, X2, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
UBFM X17, X30, #3, #63	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ADD X18, X17, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
ANDS X14, X18, #0x7	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	N/A
B.EQ 4c5220 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1360>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c5208 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1348>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51f8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1338>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51e8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1328>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51d8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1318>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c51c8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x1308>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP X14, #6	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.NE 4c83b0 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x44f0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D20, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D28, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D29, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D24, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D21, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D22, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D22	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D30, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D23, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D25, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D26, D25	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
LDR D27, [X0], #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
FABS D6, D27	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D6	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
CMP X0, X16	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.EQ 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
B 4c4fcc <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x110c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D18, [X6, X11]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33	scal (25.0%)
ADD X0, X0, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25	N/A
FABS D19, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	2	0.25	scal (25.0%)
STR D31, [X19, X3,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50	scal (25.0%)
B 4c51b8 <hypre_ParCSRComputeL1NormsThreads._omp_fn.0+0x12f8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code