OV - exec - Loop 178

	Loop Id: 178	Module: exec	Source: advec_cell.cpp:163-202 [...]	Coverage: 3.67%

0x42a550 ADD	X8, X8, #1

0x42a554 LDR	D23, [X18, X0,LSL #3]    [7]

0x42a558 FCMPE	D23, #0

0x42a55c B.GT	42a6f0

0x42a560 CMP	W24, W25

0x42a564 CSEL	W1, W24, W25, #13

0x42a568 SBFM	X1, X1, #0, #31

0x42a56c ORR	X5, XZR, X1

0x42a570 ORR	X6, XZR, X15

0x42a574 ORR	X4, XZR, X14

0x42a578 MADD	X3, X22, X4, X0

0x42a57c LDR	D31, [X20, X1,LSL #3]    [3]

0x42a580 FABS	D30, D23

0x42a584 MADD	X1, X12, X4, X0

0x42a588 LDR	D22, [X23]    [4]

0x42a58c MADD	X30, X12, X6, X0

0x42a590 UBFM	X3, X3, #61, #60

0x42a594 MADD	X27, X12, X5, X0

0x42a598 UBFM	X1, X1, #61, #60

0x42a59c LDR	D20, [X16, X3]    [5]

0x42a5a0 LDR	D21, [X9, X1]    [1]

0x42a5a4 FDIV	D22, D22, D31

0x42a5a8 LDR	D31, [X9, X30,LSL #3]    [1]

0x42a5ac LDR	D29, [X9, X27,LSL #3]    [1]

0x42a5b0 FDIV	D30, D30, D20

0x42a5b4 FSUB	D31, D31, S21

0x42a5b8 FSUB	D29, D21, S29

0x42a5bc FCMPE	D31, #0

0x42a5c0 FMUL	D20, D31, D29

0x42a5c4 FCSEL	D18, D27, D28, #9

0x42a5c8 FCMPE	D20, #0

0x42a5cc FADD	D20, D30, D28

0x42a5d0 FSUB	D19, D25, S30

0x42a5d4 FMUL	D22, D22, D20

0x42a5d8 B.GT	42a684

0x42a5dc MOVI	D31, #0

0x42a5e0 FADD	D31, D31, D21

0x42a5e4 MADD	X4, X11, X4, X0

0x42a5e8 MADD	X6, X11, X6, X0

0x42a5ec MADD	X5, X11, X5, X0

0x42a5f0 FMUL	D31, D31, D23

0x42a5f4 STR	D31, [X17, X0,LSL #3]    [6]

0x42a5f8 LDR	D29, [X10, X4,LSL #3]    [2]

0x42a5fc LDR	D30, [X10, X6,LSL #3]    [2]

0x42a600 LDR	D23, [X10, X5,LSL #3]    [2]

0x42a604 FSUB	D30, D30, S29

0x42a608 FSUB	D23, D29, S23

0x42a60c FCMPE	D30, #0

0x42a610 FMUL	D21, D30, D23

0x42a614 FCSEL	D18, D27, D28, #9

0x42a618 FCMPE	D21, #0

0x42a61c B.GT	42a6b0

0x42a620 FADD	D29, D29, D24

0x42a624 FMUL	D29, D29, D31

0x42a628 STR	D29, [X19, X0,LSL #3]    [8]

0x42a62c ORR	X0, XZR, X8

0x42a630 CMP	X13, X8

0x42a634 B.NE	42a550

0x42a684 FABS	D31, D31

0x42a688 FABS	D29, D29

0x42a68c FMINNM	D20, D31, D29

0x42a690 FSUB	D30, D28, S30

0x42a694 FMUL	D31, D31, D19

0x42a698 FMUL	D30, D30, D18

0x42a69c FMADD	D31, D29, D22, D31

0x42a6a0 FMUL	D31, D31, D26

0x42a6a4 FMINNM	D31, D31, D20

0x42a6a8 FMUL	D31, D31, D30

0x42a6ac B	42a5e0

0x42a6b0 LDR	D20, [X16, X3]    [5]

0x42a6b4 FABS	D30, D30

0x42a6b8 FABS	D23, D23

0x42a6bc FMINNM	D17, D30, D23

0x42a6c0 FABS	D21, D31

0x42a6c4 LDR	D16, [X9, X1]    [1]

0x42a6c8 FMUL	D30, D30, D19

0x42a6cc FMUL	D20, D20, D16

0x42a6d0 FMADD	D30, D23, D22, D30

0x42a6d4 FDIV	D23, D21, D20

0x42a6d8 FMUL	D30, D30, D26

0x42a6dc FMINNM	D30, D30, D17

0x42a6e0 FSUB	D23, D28, S23

0x42a6e4 FMUL	D23, D23, D18

0x42a6e8 FMADD	D29, D30, D23, D29

0x42a6ec B	42a624

0x42a6f0 ORR	X5, XZR, X21

0x42a6f4 ORR	X6, XZR, X14

0x42a6f8 ORR	X1, XZR, X15

0x42a6fc ORR	X4, XZR, X15

0x42a700 B	42a578

/usr/include/c++/14/bits/stl_algobase.h: 238 - 238

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

238:       if (__b < __a)

/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp: 163 - 202

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

163:           if (vol_flux_y(i, j) > 0.0) {

[...]

174:           sigmat = std::fabs(vol_flux_y(i, j)) / pre_vol(i, donor);

175:           sigma3 = (1.0 + sigmat) * (vertexdy[j] / vertexdy[dif]);

176:           sigma4 = 2.0 - sigmat;

177:           sigmav = sigmat;

178:           diffuw = density1(i, donor) - density1(i, upwind);

179:           diffdw = density1(i, downwind) - density1(i, donor);

180:           wind = 1.0;

181:           if (diffdw <= 0.0) wind = -1.0;

182:           if (diffuw * diffdw > 0.0) {

183:             limiter = (1.0 - sigmav) * wind *

184:                       std::fmin(std::fmin(std::fabs(diffuw), std::fabs(diffdw)),

185:                                 one_by_six * (sigma3 * std::fabs(diffuw) + sigma4 * std::fabs(diffdw)));

186:           } else {

187:             limiter = 0.0;

188:           }

189:           mass_flux_y(i, j) = vol_flux_y(i, j) * (density1(i, donor) + limiter);

190:           sigmam = std::fabs(mass_flux_y(i, j)) / (density1(i, donor) * pre_vol(i, donor));

191:           diffuw = energy1(i, donor) - energy1(i, upwind);

192:           diffdw = energy1(i, downwind) - energy1(i, donor);

193:           wind = 1.0;

194:           if (diffdw <= 0.0) wind = -1.0;

195:           if (diffuw * diffdw > 0.0) {

196:             limiter = (1.0 - sigmam) * wind *

197:                       std::fmin(std::fmin(std::fabs(diffuw), std::fabs(diffdw)),

198:                                 one_by_six * (sigma3 * std::fabs(diffuw) + sigma4 * std::fabs(diffdw)));

199:           } else {

200:             limiter = 0.0;

201:           }

202:           ener_flux(i, j) = mass_flux_y(i, j) * (energy1(i, donor) + limiter);

/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]; }

Coverage (%)	Name	Module
►98.42+	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	2.62
CQA speedup if FP arith vectorized	1.85
CQA speedup if fully vectorized	4.00 - 2.34
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.09
Bottlenecks	P6, P7, P8, P9,
Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone ._omp_fn.6]
Source	stl_algobase.h:238-238,advec_cell.cpp:163-163,advec_cell.cpp:174-202,context.h:69-69
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	12.25
CQA cycles if no scalar integer	4.67
CQA cycles if FP arith vectorized	6.63
CQA cycles if fully vectorized	3.06 - 5.24
Front-end cycles	11.25
P0 cycles	3.50
P1 cycles	3.50
P2 cycles	5.00
P3 cycles	5.00
P4 cycles	7.00
P5 cycles	5.00
P6 cycles	12.25
P7 cycles	12.25
P8 cycles	12.25
P9 cycles	12.25
P10 cycles	4.67
P11 cycles	4.67
P12 cycles	4.67
P13 cycles	0.00
P14 cycles	0.00
DIV/SQRT cycles	5.25 - 10.49
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	90.00
Nb uops	90.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	2.61
Nb FLOP add-sub	10.00
Nb FLOP mul	13.00
Nb FLOP fma	3.00
Nb FLOP div	3.00
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	0.00
Vectorization ratio add_sub	0.00
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	0.00
Vectorization ratio other	0.00
Vector-efficiency ratio all	24.80
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	25.00
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	25.00
Vector-efficiency ratio div_sqrt	25.00
Vector-efficiency ratio other	24.38

Metric	Value
CQA speedup if no scalar integer	2.62
CQA speedup if FP arith vectorized	1.85
CQA speedup if fully vectorized	4.00 - 2.34
CQA speedup if no inter-iteration dependency	NA
CQA speedup if next bottleneck killed	1.09
Bottlenecks	P6, P7, P8, P9,
Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone ._omp_fn.6]
Source	stl_algobase.h:238-238,advec_cell.cpp:163-163,advec_cell.cpp:174-202,context.h:69-69
Source loop unroll info	not unrolled or unrolled with no peel/tail loop
Source loop unroll confidence level	max
Unroll/vectorization loop type	NA
Unroll factor	NA
CQA cycles	12.25
CQA cycles if no scalar integer	4.67
CQA cycles if FP arith vectorized	6.63
CQA cycles if fully vectorized	3.06 - 5.24
Front-end cycles	11.25
P0 cycles	3.50
P1 cycles	3.50
P2 cycles	5.00
P3 cycles	5.00
P4 cycles	7.00
P5 cycles	5.00
P6 cycles	12.25
P7 cycles	12.25
P8 cycles	12.25
P9 cycles	12.25
P10 cycles	4.67
P11 cycles	4.67
P12 cycles	4.67
P13 cycles	0.00
P14 cycles	0.00
DIV/SQRT cycles	5.25 - 10.49
Inter-iter dependencies cycles	NA
FE+BE cycles (UFS)	NA
Stall cycles (UFS)	NA
Nb insns	90.00
Nb uops	90.00
Nb loads	NA
Nb stores	2.00
Nb stack references	0.00
FLOP/cycle	2.61
Nb FLOP add-sub	10.00
Nb FLOP mul	13.00
Nb FLOP fma	3.00
Nb FLOP div	3.00
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	0.00
Vectorization ratio add_sub	0.00
Vectorization ratio fma	0.00
Vectorization ratio div_sqrt	0.00
Vectorization ratio other	0.00
Vector-efficiency ratio all	24.80
Vector-efficiency ratio load	25.00
Vector-efficiency ratio store	25.00
Vector-efficiency ratio mul	25.00
Vector-efficiency ratio add_sub	25.00
Vector-efficiency ratio fma	25.00
Vector-efficiency ratio div_sqrt	25.00
Vector-efficiency ratio other	24.38

Path /

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

Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone ._omp_fn.6]
Source file and lines	advec_cell.cpp:163-202
Module	exec

The loop is defined in:

/usr/include/c++/14/bits/stl_algobase.h:238
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp:163,174-202
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69

The related source loop is not unrolled or unrolled with no peel/tail loop.
Warnings:

Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
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 8 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 12.25 to 4.67 cycles (2.63x 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. Only 24% of vector register length is used (average across all VPU instructions). By vectorizing your loop, you can lower the cost of an iteration from 12.25 to 5.24 cycles (2.34x 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:
- recompile with fassociative-math (included in Ofast or ffast-math) to extend loop vectorization to FP reductions.
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)

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 12.25 to 11.50 cycles (1.07x speedup).

FMA

Detected 3 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

Try to change order in which elements are evaluated (using parentheses) in arithmetic expressions containing both ADD/SUB and MUL operations to enable your compiler to generate FMA instructions wherever possible. For instance a + b*c is a valid FMA (MUL then ADD). However (a+b)* c cannot be translated into an FMA (ADD then MUL).

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

The binary loop is composed of 32 FP arithmetical operations:

13: addition or subtraction (3 inside FMA instructions)
16: multiply (3 inside FMA instructions)
3: divide

The binary loop does not load or store any data.

General properties

nb instructions	90
nb uops	90
loop length	360
used w registers	3
used x registers	25
used b registers	0
used h registers	0
used s registers	4
used d registers	16
used q registers	0
used v registers	0
used z registers	0
nb stack references	0
ADD-SUB / MUL ratio	0.77

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	11.25 cycles
front end	11.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	5.00	5.00	7.00	5.00	12.25	12.25	12.25	12.25	4.67	4.67	4.67	0.00	0.00
cycles	3.50	3.50	5.00	5.00	7.00	5.00	12.25	12.25	12.25	12.25	4.67	4.67	4.67	0.00	0.00

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

5.25-10.49

Cycles summary

Front-end	11.25
Dispatch	12.25
DIV/SQRT	5.25-10.49
Overall L1	12.25

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	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

all	0%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

Vector efficiency ratios

INT

all	24%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
other	20%

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

INT+FP

all	24%
load	25%
store	25%
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	24%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 12.25 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 42a554

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput	Vectorization
ADD X8, X8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
LDR D23, [X18, X0,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FCMPE D23, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 42a6f0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x330>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP W24, W25	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 W1, W24, W25, #13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X1, X1, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X5, XZR, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X4, XZR, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X3, X22, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D31, [X20, X1,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FABS D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
MADD X1, X12, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D22, [X23]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X30, X12, X6, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
UBFM X3, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X27, X12, X5, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
UBFM X1, X1, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D20, [X16, X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D21, [X9, X1]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D22, D22, D31	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
LDR D31, [X9, X30,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D29, [X9, X27,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D30, D30, D20	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D31, D31, S21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D29, D21, S29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D31, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D20, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D18, D27, D28, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D20, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FADD D20, D30, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D19, D25, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D22, D22, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B.GT 42a684 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x2c4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
MADD X4, X11, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X6, X11, X6, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X5, X11, X5, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
FMUL D31, D31, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D31, [X17, X0,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
LDR D29, [X10, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D30, [X10, X6,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D23, [X10, X5,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FSUB D30, D30, S29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D23, D29, S23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D30, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D21, D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D18, D27, D28, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D21, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 42a6b0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x2f0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FADD D29, D29, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D29, D29, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D29, [X19, X0,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ORR X0, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP X13, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.NE 42a550 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x190>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FABS D31, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D29, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D20, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D30, D28, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D31, D31, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMUL D30, D30, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D31, D29, D22, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D31, D31, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D31, D31, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D31, D31, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B 42a5e0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D20, [X16, X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FABS D30, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D23, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D17, D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D21, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D16, [X9, X1]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D30, D30, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMUL D20, D20, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D30, D23, D22, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FDIV D23, D21, D20	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FMUL D30, D30, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D30, D30, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D23, D28, S23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D23, D23, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D29, D30, D23, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
B 42a624 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x264>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ORR X5, XZR, X21	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X1, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X4, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
B 42a578 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x1b8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A

Function	advec_cell_kernel(int, int, int, int, int, int, clover::Buffer1D&, clover::Buffer1D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&) [clone ._omp_fn.6]
Source file and lines	advec_cell.cpp:163-202
Module	exec

The loop is defined in:

/usr/include/c++/14/bits/stl_algobase.h:238
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/advec_cell.cpp:163,174-202
/home/eoseret/qaas/qaas_runs/178-219-7589/intel/CloverLeaf2.0-CXX/build/CloverLeaf2.0-CXX/src/omp/context.h:69

The related source loop is not unrolled or unrolled with no peel/tail loop.
Warnings:

Ignoring paths for analysis
Too many paths. Rerun with max-paths=8
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 12.25 to 4.67 cycles (2.63x 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:
- recompile with fassociative-math (included in Ofast or ffast-math) to extend loop vectorization to FP reductions.
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)

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 12.25 to 11.50 cycles (1.07x speedup).

FMA

Detected 3 FMA (fused multiply-add) operations. Presence of both ADD/SUB and MUL operations.

Workaround

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

The binary loop is composed of 32 FP arithmetical operations:

13: addition or subtraction (3 inside FMA instructions)
16: multiply (3 inside FMA instructions)
3: divide

The binary loop does not load or store any data.

General properties

nb instructions	90
nb uops	90
loop length	360
used w registers	3
used x registers	25
used b registers	0
used h registers	0
used s registers	4
used d registers	16
used q registers	0
used v registers	0
used z registers	0
nb stack references	0
ADD-SUB / MUL ratio	0.77

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	11.25 cycles
front end	11.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	3.50	3.50	5.00	5.00	7.00	5.00	12.25	12.25	12.25	12.25	4.67	4.67	4.67	0.00	0.00
cycles	3.50	3.50	5.00	5.00	7.00	5.00	12.25	12.25	12.25	12.25	4.67	4.67	4.67	0.00	0.00

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

5.25-10.49

Cycles summary

Front-end	11.25
Dispatch	12.25
DIV/SQRT	5.25-10.49
Overall L1	12.25

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	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

INT+FP

all	0%
load	0%
store	0%
mul	0%
add-sub	0%
fma	0%
div/sqrt	0%
other	0%

Vector efficiency ratios

INT

all	24%
load	25%
store	25%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	25%
fma	NA (no fma vectorizable/vectorized instructions)
other	20%

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

INT+FP

all	24%
load	25%
store	25%
mul	25%
add-sub	25%
fma	25%
div/sqrt	25%
other	24%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each iteration of the binary loop takes 12.25 cycles.

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the loop is: 42a554

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	Latency	Recip. throughput	Vectorization
ADD X8, X8, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	scal (25.0%)
LDR D23, [X18, X0,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FCMPE D23, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 42a6f0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x330>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
CMP W24, W25	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 W1, W24, W25, #13	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
SBFM X1, X1, #0, #31	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X5, XZR, X1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X4, XZR, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X3, X22, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D31, [X20, X1,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FABS D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
MADD X1, X12, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
LDR D22, [X23]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
MADD X30, X12, X6, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
UBFM X3, X3, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
MADD X27, X12, X5, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
UBFM X1, X1, #61, #60	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
LDR D20, [X16, X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D21, [X9, X1]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D22, D22, D31	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
LDR D31, [X9, X30,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D29, [X9, X27,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FDIV D30, D30, D20	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FSUB D31, D31, S21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D29, D21, S29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D31, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D20, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D18, D27, D28, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D20, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FADD D20, D30, D28	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D19, D25, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D22, D22, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B.GT 42a684 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x2c4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
MOVI D31, #0	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FADD D31, D31, D21	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
MADD X4, X11, X4, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X6, X11, X6, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
MADD X5, X11, X5, X0	1	0	0	0	0	1	0	0	0	0	0	0	0	0	2	1	N/A
FMUL D31, D31, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D31, [X17, X0,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
LDR D29, [X10, X4,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D30, [X10, X6,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
LDR D23, [X10, X5,LSL #3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FSUB D30, D30, S29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D23, D29, S23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FCMPE D30, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
FMUL D21, D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FCSEL D18, D27, D28, #9	1	0	0	0	0	0	0	0.50	0.50	0	0	0	0	0	2	0.50	scal (25.0%)
FCMPE D21, #0	1	0	0	0	0	0	0	1	0	0	0	0	0	0	2	1	scal (25.0%)
B.GT 42a6b0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x2f0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FADD D29, D29, D24	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D29, D29, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
STR D29, [X19, X0,LSL #3]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	2	0.50	scal (25.0%)
ORR X0, XZR, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
CMP X13, X8	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.33	scal (25.0%)
B.NE 42a550 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x190>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
FABS D31, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D29, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D20, D31, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D30, D28, S30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D31, D31, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMUL D30, D30, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D31, D29, D22, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FMUL D31, D31, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D31, D31, D20	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D31, D31, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
B 42a5e0 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x220>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
LDR D20, [X16, X3]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FABS D30, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D23, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMINNM D17, D30, D23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FABS D21, D31	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
LDR D16, [X9, X1]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	6	0.33	scal (25.0%)
FMUL D30, D30, D19	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMUL D20, D20, D16	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D30, D23, D22, D30	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
FDIV D23, D21, D20	1	0	0	0	0	0	0	0.50	0	0.50	0	0	0	0	7-15	1.75-3.50	scal (25.0%)
FMUL D30, D30, D26	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMINNM D30, D30, D17	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FSUB D23, D28, S23	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	2	0.25	scal (25.0%)
FMUL D23, D23, D18	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	3	0.25	scal (25.0%)
FMADD D29, D30, D23, D29	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	0	0	0	4	0.25	scal (25.0%)
B 42a624 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x264>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
ORR X5, XZR, X21	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X6, XZR, X14	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X1, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
ORR X4, XZR, X15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	1	0.25	N/A
B 42a578 <_Z17advec_cell_kerneliiiiiiRN6clover8Buffer1DIdEES2_RNS_8Buffer2DIdEES5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_S5_._omp_fn.6+0x1b8>	1	0.50	0.50	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

Expensive FP math instructions/calls

FMA

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

Expensive FP math instructions/calls

FMA

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