OV - exec - Function advec_mom_kernel(int, int, int, int, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer2D<double>&, clover::Buffer1D<double>&, clover::Buffer1D<double>&, int, int, int) [clone ._omp

	Function: advec_mom_kernel(int, int, int, int, clover::Buffer2D<double>&, clover::Buffer2D<double>&, ...	Module: exec	Source: advec_mom.cpp:85-88 [...]	Coverage (incl. loops): 1.06% \| (excl. loops): 0.00%

/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h: 69 - 69

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

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

/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp: 85 - 88

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

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

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

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

88:           node_flux(i, j) = 0.25 * (mass_flux_x(i + 0, j - 1) + mass_flux_x(i, j) + mass_flux_x(i + 1, j - 1) + mass_flux_x(i + 1, j + 0));

0x42b36c STP	X29, X30, [SP, #944]!

0x42b370 ADD	X29, SP, #0

0x42b374 STP	X19, X20, [SP, #16]

0x42b378 STP	X21, X22, [SP, #32]

0x42b37c STP	X25, X26, [SP, #64]

0x42b380 LDP	W25, W19, [X0, #24]

0x42b384 LDP	W20, W22, [X0, #16]

0x42b388 ADD	W25, W25, #1

0x42b38c ADD	W19, W19, #3

0x42b390 CMP	W25, W19

0x42b394 B.GE	42b4e4

0x42b398 SUB	W20, W20, #1

0x42b39c ADD	W22, W22, #4

0x42b3a0 SUB	W26, W19, W25

0x42b3a4 CMP	W20, W22

0x42b3a8 B.GE	42b4e4

0x42b3ac SUB	W21, W22, W20

0x42b3b0 STP	X23, X24, [SP, #48]

0x42b3b4 ORR	X24, XZR, X0

0x42b3b8 MUL	W26, W26, W21

0x42b3bc BL	410670

0x42b3c0 ORR	W23, WZR, W0

0x42b3c4 BL	410550

0x42b3c8 UDIV	W1, W26, W23

0x42b3cc ORR	W2, WZR, W0

0x42b3d0 MSUB	W3, W1, W23, W26

0x42b3d4 CMP	W0, W3

0x42b3d8 B.CC	42b4f8

0x42b3dc MADD	W2, W1, W2, W3

0x42b3e0 ADD	W11, W1, W2

0x42b3e4 CMP	W2, W11

0x42b3e8 B.CS	42b4e0

0x42b3ec UDIV	W0, W2, W21

0x42b3f0 LDP	X10, X9, [X24]

0x42b3f4 CNTD	X7, ALL

0x42b3f8 FDUP	Z31.D, #80

0x42b3fc MSUB	W5, W0, W21, W2

0x42b400 ADD	W0, W0, W25

0x42b404 SBFM	X8, X0, #0, #31

0x42b408 ADD	W5, W5, W20

0x42b40c SUB	W0, W22, W5

0x42b410 CMP	W1, W0

0x42b414 CSEL	X1, X1, X0, #9

0x42b418 ADD	W12, W2, W1

0x42b41c CMP	W2, W12

0x42b420 B.CS	42b4d0

  (195) 0x42b424 LDR	X4, [X10]

  (195) 0x42b428 SUB	W3, W8, #1

  (195) 0x42b42c SBFM	X5, X5, #0, #31

  (195) 0x42b430 SBFM	X3, X3, #0, #31

  (195) 0x42b434 MOVZ	X0, #0

  (195) 0x42b438 WHILELO	P7.D, XZR, X1

  (195) 0x42b43c LDR	X2, [X9]

  (195) 0x42b440 LDR	X13, [X10, #16]

  (195) 0x42b444 MADD	X3, X3, X4, X5

  (195) 0x42b448 LDR	X14, [X9, #16]

  (195) 0x42b44c MADD	X4, X4, X8, X5

  (195) 0x42b450 MADD	X2, X8, X2, X5

  (195) 0x42b454 ADD	X6, X13, X3,LSL #3

  (195) 0x42b458 ADD	X5, X13, X4,LSL #3

  (195) 0x42b45c ADD	X2, X14, X2,LSL #3

  (195) 0x42b460 ADD	X6, X6, #8

  (195) 0x42b464 ADD	X3, X13, X3,LSL #3

  (195) 0x42b468 ADD	X5, X5, #8

  (195) 0x42b46c ADD	X4, X13, X4,LSL #3

    (194) 0x42b470 LD1D	{Z30.D}, P7/Z, [X3, X0,LSL #3]

    (194) 0x42b474 LD1D	{Z1.D}, P7/Z, [X4, X0,LSL #3]

    (194) 0x42b478 LD1D	{Z0.D}, P7/Z, [X6, X0,LSL #3]

    (194) 0x42b47c LD1D	{Z29.D}, P7/Z, [X5, X0,LSL #3]

    (194) 0x42b480 FADD	Z30.D, P7/M, Z30.D, Z1.D

    (194) 0x42b484 FADD	Z30.D, P7/M, Z30.D, Z0.D

    (194) 0x42b488 FADD	Z30.D, P7/M, Z30.D, Z29.D

    (194) 0x42b48c FMUL	Z30.D, P7/M, Z30.D, Z31.D

    (194) 0x42b490 ST1D	{Z30.D}, P7, [X2, X0,LSL #3]

    (194) 0x42b494 ADD	X0, X0, X7

    (194) 0x42b498 WHILELO	P7.D, X0, X1

    (194) 0x42b49c B.NE	42b470

  (195) 0x42b4a0 ADD	X8, X8, #1

  (195) 0x42b4a4 CMP	W19, W8

  (195) 0x42b4a8 B.LE	42b4e0

  (195) 0x42b4ac SUB	W1, W11, W12

  (195) 0x42b4b0 ORR	W0, WZR, W21

  (195) 0x42b4b4 ORR	W2, WZR, W12

  (195) 0x42b4b8 CMP	W1, W0

  (195) 0x42b4bc CSEL	X1, X1, X0, #9

  (195) 0x42b4c0 ADD	W12, W2, W1

  (195) 0x42b4c4 ORR	W5, WZR, W20

  (195) 0x42b4c8 CMP	W2, W12

  (195) 0x42b4cc B.CC	42b424

(196) 0x42b4d0 ADD	X8, X8, #1

(196) 0x42b4d4 ORR	W12, WZR, W2

(196) 0x42b4d8 CMP	W19, W8

(196) 0x42b4dc B.GT	42b4ac

0x42b4e0 LDP	X23, X24, [SP, #48]

0x42b4e4 LDP	X19, X20, [SP, #16]

0x42b4e8 LDP	X21, X22, [SP, #32]

0x42b4ec LDP	X25, X26, [SP, #64]

0x42b4f0 LDP	X29, X30, [SP], #80

0x42b4f4 RET

0x42b4f8 ADD	W1, W1, #1

0x42b4fc MOVZ	W3, #0

0x42b500 B	42b3dc

Coverage (%)	Name	Module
►98.45+	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 /

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 gcc_0

Source file and lines	advec_mom.cpp:85-88
Module	exec

The function is defined in:

/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:85-88

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 25.00 to 12.50 cycles (2.00x 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 25% of vector register length is used (average across all VPU instructions).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers).

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 25.00 to 8.25 cycles (3.03x speedup).

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.

unknown: 2 occurrences►
- /home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:86
- /home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:86

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 does not load or store any data.

General properties

nb instructions	55
nb uops	55
loop length	220
used w registers	15
used x registers	17
used b registers	0
used h registers	0
used s registers	1
used d registers	0
used q registers	0
used v registers	0
used z registers	1
nb stack references	10

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	6.88 cycles
front end	6.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	4.50	4.50	8.75	8.75	8.75	8.75	0.50	0.50	0.00	0.00	4.50	4.17	4.33	2.50	2.50
cycles	4.50	4.50	8.75	8.75	8.75	8.75	0.50	0.50	0.00	0.00	4.50	4.17	4.33	2.50	2.50

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

10.00-25.00

Cycles summary

Front-end	6.88
Dispatch	8.75
DIV/SQRT	10.00-25.00
Overall L1	10.00-25.00

Vectorization ratios

INT

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

Vector efficiency ratios

INT

all	26%
load	42%
store	50%
mul	12%
add-sub	13%
fma	12%
other	21%

all	3%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	3%

INT+FP

all	25%
load	42%
store	50%
mul	12%
add-sub	13%
fma	12%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	20%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 25.00 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the function is: 42b36c

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
STP X29, X30, [SP, #944]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
STP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
STP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
STP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
LDP W25, W19, [X0, #24]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDP W20, W22, [X0, #16]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
ADD W25, W25, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
ADD W19, W19, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W25, W19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 42b4e4 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x178>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB W20, W20, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
ADD W22, W22, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
SUB W26, W19, W25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W20, W22	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 42b4e4 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x178>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB W21, W22, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
STP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
ORR X24, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MUL W26, W26, W21	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50	scal (12.5%)
BL 410670 <@plt_start@+0x650>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ORR W23, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
BL 410550 <@plt_start@+0x530>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UDIV W1, W26, W23	1	0	0	0	0	1	0	0	0	0	0	0	0	0	5-12	5-12.50	N/A
ORR W2, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
MSUB W3, W1, W23, W26	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
CMP W0, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CC 42b4f8 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x18c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MADD W2, W1, W2, W3	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W11, W1, W2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W2, W11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CS 42b4e0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x174>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UDIV W0, W2, W21	1	0	0	0	0	1	0	0	0	0	0	0	0	0	5-12	5-12.50	N/A
LDP X10, X9, [X24]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
CNTD X7, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
FDUP Z31.D, #80	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (3.1%)
MSUB W5, W0, W21, W2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W0, W0, W25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X8, X0, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (100.0%)
ADD W5, W5, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
SUB W0, W22, W5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W1, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
CSEL X1, X1, X0, #9	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ADD W12, W2, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W2, W12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CS 42b4d0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x164>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X29, X30, [SP], #80	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD W1, W1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MOVZ W3, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
B 42b3dc <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x70>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

The code analyzed by CQA in that panel excludes loops and represents 0.00% of application time for run gcc_0

Source file and lines	advec_mom.cpp:85-88
Module	exec

The function is defined in:

/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:85-88

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

Your function is not vectorized. Only 25% of vector register length is used (average across all VPU instructions).

Details

All VPU instructions are used in scalar version (process only one data element in vector registers).

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 25.00 to 8.25 cycles (3.03x speedup).

CALL instructions

Detected function call instructions.

Details

unknown: 2 occurrences►
- /home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:86
- /home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_mom.cpp:86

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 does not load or store any data.

General properties

nb instructions	55
nb uops	55
loop length	220
used w registers	15
used x registers	17
used b registers	0
used h registers	0
used s registers	1
used d registers	0
used q registers	0
used v registers	0
used z registers	1
nb stack references	10

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	6.88 cycles
front end	6.88 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	4.50	4.50	8.75	8.75	8.75	8.75	0.50	0.50	0.00	0.00	4.50	4.17	4.33	2.50	2.50
cycles	4.50	4.50	8.75	8.75	8.75	8.75	0.50	0.50	0.00	0.00	4.50	4.17	4.33	2.50	2.50

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

10.00-25.00

Cycles summary

Front-end	6.88
Dispatch	8.75
DIV/SQRT	10.00-25.00
Overall L1	10.00-25.00

Vectorization ratios

INT

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

Vector efficiency ratios

INT

all	26%
load	42%
store	50%
mul	12%
add-sub	13%
fma	12%
other	21%

all	3%
load	NA (no load vectorizable/vectorized instructions)
store	NA (no store vectorizable/vectorized instructions)
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	NA (no add-sub vectorizable/vectorized instructions)
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	3%

INT+FP

all	25%
load	42%
store	50%
mul	12%
add-sub	13%
fma	12%
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	20%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 25.00 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the function is: 42b36c

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P10	P11	P12	P13	P14	Latency	Recip. throughput	Vectorization
STP X29, X30, [SP, #944]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
ADD X29, SP, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
STP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
STP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
STP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
LDP W25, W19, [X0, #24]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
LDP W20, W22, [X0, #16]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33	scal (25.0%)
ADD W25, W25, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
ADD W19, W19, #3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W25, W19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 42b4e4 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x178>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB W20, W20, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
ADD W22, W22, #4	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
SUB W26, W19, W25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W20, W22	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.GE 42b4e4 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x178>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB W21, W22, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
STP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50	scal (50.0%)
ORR X24, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MUL W26, W26, W21	1	0	0	0	0	0.50	0.50	0	0	0	0	0	0	0	2	0.50	scal (12.5%)
BL 410670 <@plt_start@+0x650>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ORR W23, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
BL 410550 <@plt_start@+0x530>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UDIV W1, W26, W23	1	0	0	0	0	1	0	0	0	0	0	0	0	0	5-12	5-12.50	N/A
ORR W2, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
MSUB W3, W1, W23, W26	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
CMP W0, W3	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CC 42b4f8 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x18c>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MADD W2, W1, W2, W3	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W11, W1, W2	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W2, W11	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CS 42b4e0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x174>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
UDIV W0, W2, W21	1	0	0	0	0	1	0	0	0	0	0	0	0	0	5-12	5-12.50	N/A
LDP X10, X9, [X24]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
CNTD X7, ALL	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (25.0%)
FDUP Z31.D, #80	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (3.1%)
MSUB W5, W0, W21, W2	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	scal (12.5%)
ADD W0, W0, W25	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X8, X0, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (100.0%)
ADD W5, W5, W20	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
SUB W0, W22, W5	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP W1, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
CSEL X1, X1, X0, #9	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ADD W12, W2, W1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
CMP W2, W12	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (12.5%)
B.CS 42b4d0 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x164>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDP X23, X24, [SP, #48]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	N/A
LDP X19, X20, [SP, #16]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X21, X22, [SP, #32]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X25, X26, [SP, #64]	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
LDP X29, X30, [SP], #80	1	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1	scal (50.0%)
RET	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ADD W1, W1, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MOVZ W3, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (12.5%)
B 42b3dc <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.4+0x70>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Name	Coverage (%)	Time (s)
▼advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clo...–	1.06	1.45
▼Loop 196 - advec_mom.cpp:87-88 - exec–	0.00	0.00
▼Loop 195 - advec_mom.cpp:87-88 - exec–	0.00	0.02
○Loop 194 - advec_mom.cpp:88-88 - exec	1.06	1.40

Report Configuration

Code clean check

Workaround

Vectorization

Details

Expensive FP math instructions/calls

CALL instructions

Details

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

Code clean check

Workaround

Vectorization

Details

Expensive FP math instructions/calls

CALL instructions

Details

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