OV - exec - Function .omp

	Function: .omp_outlined..10	Module: exec	Source: advec_cell.cpp:146-150 [...]	Coverage: 1.11%

/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/advec_cell.cpp: 146 - 150

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

146: #pragma omp parallel for simd collapse(2)

147:       for (int j = (y_min - 2 + 1); j < (y_max + 2 + 2); j++) {

148:         for (int i = (x_min - 2 + 1); i < (x_max + 2 + 2); i++) {

149:           pre_vol(i, j) = volume(i, j) + vol_flux_y(i + 0, j + 1) - vol_flux_y(i, j);

150:           post_vol(i, j) = volume(i, j);

/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/context.h: 69 - 69

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

69:   T &operator()(size_t i, size_t j) const { return data[i + j * sizeX]; }

0x41b660 SUB	SP, SP, #160

0x41b664 STP	X29, X30, [SP, #64]

0x41b668 STP	X28, X27, [SP, #80]

0x41b66c STP	X26, X25, [SP, #96]

0x41b670 STP	X24, X23, [SP, #112]

0x41b674 STP	X22, X21, [SP, #128]

0x41b678 STP	X20, X19, [SP, #144]

0x41b67c ADD	X29, SP, #64

0x41b680 LDR	W8, [X3]

0x41b684 LDR	W20, [X2]

0x41b688 LDR	W10, [X5]

0x41b68c ADD	W8, W8, #4

0x41b690 LDR	W9, [X4]

0x41b694 ADD	W10, W10, #4

0x41b698 CMP	W8, W20

0x41b69c CCMP	W10, W9, #8, #10

0x41b6a0 B.LT	41b890

0x41b6a4 LDP	X12, X11, [X29, #96]

0x41b6a8 SUB	W19, W20, #1

0x41b6ac SUB	W21, W9, #1

0x41b6b0 LDR	X14, [X6]

0x41b6b4 LDR	X13, [X7]

0x41b6b8 MOVN	X9, #0

0x41b6bc SUB	X3, X29, #28

0x41b6c0 SUB	X4, X29, #8

0x41b6c4 LDR	W1, [X0]

0x41b6c8 SUB	W23, W10, W21

0x41b6cc SUB	W8, W8, W19

0x41b6d0 LDR	X22, [X6, #16]

0x41b6d4 LDR	X24, [X7, #16]

0x41b6d8 ADRP	X0,

0x41b6dc ADD	X0, X0, #3384

0x41b6e0 STUR	WZR, [X29, #484]

0x41b6e4 SUB	X5, X29, #16

0x41b6e8 LDR	X25, [X12]

0x41b6ec LDR	X26, [X12, #16]

0x41b6f0 LDR	X27, [X11]

0x41b6f4 LDR	X28, [X11, #16]

0x41b6f8 UMADDL	X8, W23, W8, X9

0x41b6fc SUB	X6, X29, #24

0x41b700 MOVZ	W2, #34

0x41b704 MOVZ	W7, #1

0x41b708 STP	X13, X14, [SP, #8]

0x41b70c STR	W1, [SP, #32]

0x41b710 STR	X8, [SP, #24]

0x41b714 STP	X8, XZR, [X29, #1008]

0x41b718 MOVZ	W8, #1

0x41b71c STUR	X8, [X29, #488]

0x41b720 STR	X8, [SP]

0x41b724 BL	402ee0

0x41b728 LDP	X8, X13, [X29, #1008]

0x41b72c LDR	X9, [SP, #24]

0x41b730 CMP	X8, X9

0x41b734 CSEL	X8, X8, X9, #11

0x41b738 CMP	X13, X8

0x41b73c B.GT	41b880

0x41b740 ADD	X8, X8, #1

0x41b744 CNTD	X10, ALL

0x41b748 ORR	X9, XZR, X13

0x41b74c SUB	X11, X8, X13

0x41b750 CMP	X11, X10

0x41b754 B.CC	41b818

0x41b758 UDIV	X9, X11, X10

0x41b75c PTRUE	P0.D, ALL

0x41b760 MADD	X12, X9, X10, XZR

0x41b764 INDEX	Z0.D, X13, #1

0x41b768 ADD	X9, X13, X12

0x41b76c LDR	X13, [SP, #8]

0x41b770 DUP	Z1.D, X10

0x41b774 SUB	X11, X11, X12

0x41b778 DUP	Z2.D, X23

0x41b77c DUP	Z3.D, X19

0x41b780 DUP	Z4.D, X21

0x41b784 DUP	Z6.D, X20

0x41b788 DUP	Z7.D, X25

0x41b78c DUP	Z17.D, X27

0x41b790 DUP	Z5.D, X13

0x41b794 LDR	X13, [SP, #16]

0x41b798 DUP	Z16.D, X13

0x41b79c HINT	#0

(292) 0x41b7a0 MOVPRFX	Z18, Z0

(292) 0x41b7a4 SDIV	Z18.D, P0/M, Z18.D, Z2.D

(292) 0x41b7a8 ADD	Z19.D, Z3.D, Z18.D

(292) 0x41b7ac MOVPRFX	Z20, Z0

(292) 0x41b7b0 MLS	Z20.D, P0/M, Z18.D, Z2.D

(292) 0x41b7b4 ADD	Z20.D, Z4.D, Z20.D

(292) 0x41b7b8 SUBS	X12, X12, X10

(292) 0x41b7bc ADD	Z18.D, Z6.D, Z18.D

(292) 0x41b7c0 ADD	Z0.D, Z0.D, Z1.D

(292) 0x41b7c4 SXTW	Z20.D, P0/M, Z20.D

(292) 0x41b7c8 SXTW	Z18.D, P0/M, Z18.D

(292) 0x41b7cc SXTW	Z19.D, P0/M, Z19.D

(292) 0x41b7d0 MOVPRFX	Z21, Z20

(292) 0x41b7d4 MLA	Z21.D, P0/M, Z5.D, Z19.D

(292) 0x41b7d8 LD1D	{Z22.D}, P0/Z, [X24, Z21.D,LSL #3]

(292) 0x41b7dc MOVPRFX	Z23, Z20

(292) 0x41b7e0 MLA	Z23.D, P0/M, Z7.D, Z19.D

(292) 0x41b7e4 MAD	Z18.D, P0/M, Z7.D, Z20.D

(292) 0x41b7e8 LD1D	{Z18.D}, P0/Z, [X26, Z18.D,LSL #3]

(292) 0x41b7ec LD1D	{Z23.D}, P0/Z, [X26, Z23.D,LSL #3]

(292) 0x41b7f0 FADD	Z18.D, Z22.D, Z18.D

(292) 0x41b7f4 MOVPRFX	Z22, Z20

(292) 0x41b7f8 MLA	Z22.D, P0/M, Z16.D, Z19.D

(292) 0x41b7fc FSUB	Z18.D, Z18.D, Z23.D

(292) 0x41b800 MAD	Z19.D, P0/M, Z17.D, Z20.D

(292) 0x41b804 ST1D	{Z18.D}, P0, [X22, Z22.D,LSL #3]

(292) 0x41b808 LD1D	{Z18.D}, P0/Z, [X24, Z21.D,LSL #3]

(292) 0x41b80c ST1D	{Z18.D}, P0, [X28, Z19.D,LSL #3]

(292) 0x41b810 B.NE	41b7a0

0x41b814 CBZ	X11, 41b880

0x41b818 LDP	X16, X15, [SP, #8]

0x41b81c SUB	W10, WZR, W23

(291) 0x41b820 SDIV	X11, X9, X23

(291) 0x41b824 ADD	W13, W21, W9

(291) 0x41b828 ADD	X9, X9, #1

(291) 0x41b82c CMP	X8, X9

(291) 0x41b830 ADD	W12, W19, W11

(291) 0x41b834 MADD	W13, W10, W11, W13

(291) 0x41b838 ADD	W11, W20, W11

(291) 0x41b83c SBFM	X13, X13, #0, #31

(291) 0x41b840 SBFM	X12, X12, #0, #31

(291) 0x41b844 SBFM	X11, X11, #0, #31

(291) 0x41b848 MADD	X14, X16, X12, X13

(291) 0x41b84c MADD	X11, X25, X11, X13

(291) 0x41b850 LDR	D0, [X24, X14,LSL #3]

(291) 0x41b854 LDR	D1, [X26, X11,LSL #3]

(291) 0x41b858 MADD	X11, X25, X12, X13

(291) 0x41b85c LDR	D2, [X26, X11,LSL #3]

(291) 0x41b860 MADD	X11, X15, X12, X13

(291) 0x41b864 FADD	D0, D0, D1

(291) 0x41b868 FSUB	D0, D0, S2

(291) 0x41b86c STR	D0, [X22, X11,LSL #3]

(291) 0x41b870 MADD	X11, X27, X12, X13

(291) 0x41b874 LDR	D0, [X24, X14,LSL #3]

(291) 0x41b878 STR	D0, [X28, X11,LSL #3]

(291) 0x41b87c B.NE	41b820

0x41b880 LDR	W1, [SP, #32]

0x41b884 ADRP	X0,

0x41b888 ADD	X0, X0, #3408

0x41b88c BL	402d80

0x41b890 LDP	X20, X19, [SP, #144]

0x41b894 LDP	X22, X21, [SP, #128]

0x41b898 LDP	X24, X23, [SP, #112]

0x41b89c LDP	X26, X25, [SP, #96]

0x41b8a0 LDP	X28, X27, [SP, #80]

0x41b8a4 LDP	X29, X30, [SP, #64]

0x41b8a8 ADD	SP, SP, #160

0x41b8ac RET

0x42afe8 HINT	#0

0x42afec HINT	#0

Coverage (%)	Name	Source Location	Module
○100.00	__kmp_invoke_microtask		libomp.so

Path /

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

Source file and lines	advec_cell.cpp:146-150
Module	exec

The function is defined in:

/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/advec_cell.cpp:146-150
/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/context.h:69

Warnings:
Ignoring paths for analysis

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 14.00 to 10.00 cycles (1.40x speedup).

Workaround

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

Vectorization

Your function is not vectorized. Only 23% of vector register length is used (average across all VPU instructions). By vectorizing your function, you can lower the cost of an iteration from 14.00 to 13.17 cycles (1.06x 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 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)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 14.00 to 1.00 cycles (14.00x speedup).

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 256 bytes. The binary function is storing 160 bytes.

General properties

nb instructions	97
loop length	388
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

11.75 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	11.75	11.75	14.00	11.50	1.00	0.00	0.00	0.00	13.17	12.83	13.00	6.50	6.50
cycles	3.50	3.50	11.75	11.75	14.00	11.50	1.00	0.00	0.00	0.00	13.17	12.83	13.00	6.50	6.50

Cycles executing div or sqrt instructions

1.00-0.50

Cycles summary

Front-end	11.75
Overall L1	14.00

Vectorization ratios

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	0%
div/sqrt	0%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 14.00 cycles. At this rate:

19% of peak load performance is reached (18.29 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
23% of peak store performance is reached (11.43 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 41b660

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
SUB SP, SP, #160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X29, X30, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X28, X27, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X26, X25, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X24, X23, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X22, X21, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X20, X19, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #64	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W8, [X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR W20, [X2]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR W10, [X5]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD W8, W8, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W9, [X4]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD W10, W10, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP W8, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
CCMP W10, W9, #8, #10	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
B.LT 41b890 <.omp_outlined..10+0x230>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X12, X11, [X29, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
SUB W19, W20, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB W21, W9, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X14, [X6]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X13, [X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVN X9, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X3, X29, #28	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X4, X29, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W1, [X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
SUB W23, W10, W21	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB W8, W8, W19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X22, [X6, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X24, [X7, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADRP X0, <4626d8>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X0, X0, #3384	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STUR WZR, [X29, #484]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
SUB X5, X29, #16	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X25, [X12]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X26, [X12, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X27, [X11]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X28, [X11, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
UMADDL X8, W23, W8, X9	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
SUB X6, X29, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W2, #34	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W7, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X13, X14, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR W1, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X8, [SP, #24]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X8, XZR, [X29, #1008]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
MOVZ W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STUR X8, [X29, #488]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X8, [SP]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
BL 402ee0 <@plt_start@+0x190>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X8, X13, [X29, #1008]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDR X9, [SP, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP X8, X9	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
CSEL X8, X8, X9, #11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP X13, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.GT 41b880 <.omp_outlined..10+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X8, X8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CNTD X10, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
ORR X9, XZR, X13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X11, X8, X13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP X11, X10	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.CC 41b818 <.omp_outlined..10+0x1b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
UDIV X9, X11, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	5-20	1-0.50
PTRUE P0.D, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
MADD X12, X9, X10, XZR	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
INDEX Z0.D, X13, #1	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5	2
ADD X9, X13, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X13, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z1.D, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
SUB X11, X11, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
DUP Z2.D, X23	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z3.D, X19	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z4.D, X21	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z6.D, X20	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z7.D, X25	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z17.D, X27	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z5.D, X13	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
LDR X13, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z16.D, X13	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
HINT #0
CBZ X11, 41b880 <.omp_outlined..10+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X16, X15, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
SUB W10, WZR, W23	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W1, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADRP X0, <462884>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X0, X0, #3408	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 402d80 <@plt_start@+0x30>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X20, X19, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X22, X21, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X24, X23, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X26, X25, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X28, X27, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
ADD SP, SP, #160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0

Source file and lines	advec_cell.cpp:146-150
Module	exec

The function is defined in:

/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/advec_cell.cpp:146-150
/home/hbollore/qaas-runs/170-290-5445/intel/CloverLeafCXX/build/CloverLeafCXX/src/omp/context.h:69

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Code clean check

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 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)

Expensive FP math instructions/calls

Detected performance impact from expensive FP math instructions/calls. By removing/reexpressing them, you can lower the cost of an iteration from 14.00 to 1.00 cycles (14.00x speedup).

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 256 bytes. The binary function is storing 160 bytes.

General properties

nb instructions	97
loop length	388
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

11.75 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	11.75	11.75	14.00	11.50	1.00	0.00	0.00	0.00	13.17	12.83	13.00	6.50	6.50
cycles	3.50	3.50	11.75	11.75	14.00	11.50	1.00	0.00	0.00	0.00	13.17	12.83	13.00	6.50	6.50

Cycles executing div or sqrt instructions

1.00-0.50

Cycles summary

Front-end	11.75
Overall L1	14.00

Vectorization ratios

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	0%
div/sqrt	0%
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 14.00 cycles. At this rate:

19% of peak load performance is reached (18.29 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
23% of peak store performance is reached (11.43 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 41b660

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
SUB SP, SP, #160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X29, X30, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X28, X27, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X26, X25, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X24, X23, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X22, X21, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X20, X19, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
ADD X29, SP, #64	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W8, [X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR W20, [X2]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR W10, [X5]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD W8, W8, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W9, [X4]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD W10, W10, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP W8, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
CCMP W10, W9, #8, #10	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
B.LT 41b890 <.omp_outlined..10+0x230>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X12, X11, [X29, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
SUB W19, W20, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB W21, W9, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X14, [X6]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X13, [X7]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
MOVN X9, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X3, X29, #28	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X4, X29, #8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W1, [X0]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
SUB W23, W10, W21	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB W8, W8, W19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X22, [X6, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X24, [X7, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADRP X0, <4626d8>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X0, X0, #3384	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STUR WZR, [X29, #484]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
SUB X5, X29, #16	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X25, [X12]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X26, [X12, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X27, [X11]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDR X28, [X11, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
UMADDL X8, W23, W8, X9	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
SUB X6, X29, #24	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W2, #34	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W7, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STP X13, X14, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR W1, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X8, [SP, #24]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STP X8, XZR, [X29, #1008]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
MOVZ W8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STUR X8, [X29, #488]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X8, [SP]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
BL 402ee0 <@plt_start@+0x190>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X8, X13, [X29, #1008]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDR X9, [SP, #24]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP X8, X9	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
CSEL X8, X8, X9, #11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP X13, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.GT 41b880 <.omp_outlined..10+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD X8, X8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CNTD X10, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
ORR X9, XZR, X13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
SUB X11, X8, X13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
CMP X11, X10	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.CC 41b818 <.omp_outlined..10+0x1b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
UDIV X9, X11, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	5-20	1-0.50
PTRUE P0.D, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
MADD X12, X9, X10, XZR	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	2	1
INDEX Z0.D, X13, #1	1	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	5	2
ADD X9, X13, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR X13, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z1.D, X10	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
SUB X11, X11, X12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
DUP Z2.D, X23	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z3.D, X19	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z4.D, X21	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z6.D, X20	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z7.D, X25	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z17.D, X27	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
DUP Z5.D, X13	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
LDR X13, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
DUP Z16.D, X13	1	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	3	1
HINT #0
CBZ X11, 41b880 <.omp_outlined..10+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X16, X15, [SP, #8]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
SUB W10, WZR, W23	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDR W1, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADRP X0, <462884>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
ADD X0, X0, #3408	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 402d80 <@plt_start@+0x30>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X20, X19, [SP, #144]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X22, X21, [SP, #128]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X24, X23, [SP, #112]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X26, X25, [SP, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X28, X27, [SP, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDP X29, X30, [SP, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
ADD SP, SP, #160	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
HINT #0
HINT #0

Name	Coverage (%)	Time (s)
▼.omp_outlined..10–	1.11	1.48
○Loop 292 - advec_cell.cpp:147-150 - exec	1.11	1.47
○Loop 291 - advec_cell.cpp:147-150 - exec	0	0

Report Configuration

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Cycles and memory resources usage

ASM code

Code clean check

Workaround

Vectorization

Details

Workaround

Expensive FP math instructions/calls

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Cycles and memory resources usage

ASM code