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: :0-0	Coverage (incl. loops): 2.52% \| (excl. loops): 0.00%

  ***  This Panel is Intentionally Left Blank.  ***

It is due to a lack of debug symbols in the given object
(loop or function).

0x426a60 PUSH	%R15

0x426a62 PUSH	%R14

0x426a64 PUSH	%R13

0x426a66 PUSH	%R12

0x426a68 PUSH	%RBP

0x426a69 PUSH	%RBX

0x426a6a SUB	$0x68,%RSP

0x426a6e MOV	0x28(%RDI),%EAX

0x426a71 MOV	0x20(%RDI),%EBX

0x426a74 LEA	-0x1(%RAX),%EBP

0x426a77 MOV	0x2c(%RDI),%EAX

0x426a7a LEA	-0x1(%RBX),%ECX

0x426a7d MOV	0x24(%RDI),%EBX

0x426a80 MOV	%ECX,0x3c(%RSP)

0x426a84 ADD	$0x4,%EAX

0x426a87 MOV	%EAX,0x38(%RSP)

0x426a8b CMP	%EAX,%EBP

0x426a8d JGE	426cbf

0x426a93 LEA	0x4(%RBX),%R14D

0x426a97 MOV	%EAX,%EBX

0x426a99 SUB	%EBP,%EBX

0x426a9b CMP	%R14D,%ECX

0x426a9e JGE	426cbf

0x426aa4 MOV	%R14D,%R15D

0x426aa7 MOV	%RDI,%R13

0x426aaa SUB	%ECX,%R15D

0x426aad MOV	%R15D,0x40(%RSP)

0x426ab2 CALL	404660 <omp_get_num_threads@plt>

0x426ab7 MOV	%EAX,%R12D

0x426aba CALL	404550 <omp_get_thread_num@plt>

0x426abf XOR	%EDX,%EDX

0x426ac1 MOV	%EAX,%ECX

0x426ac3 MOV	%R15D,%EAX

0x426ac6 IMUL	%EBX,%EAX

0x426ac9 DIV	%R12D

0x426acc MOV	%EAX,%EDI

0x426ace CMP	%EDX,%ECX

0x426ad0 JB	426cce

0x426ad6 IMUL	%EDI,%ECX

0x426ad9 LEA	(%RCX,%RDX,1),%ESI

0x426adc LEA	(%RDI,%RSI,1),%EAX

0x426adf MOV	%EAX,0x44(%RSP)

0x426ae3 CMP	%EAX,%ESI

0x426ae5 JAE	426cbf

0x426aeb XOR	%EDX,%EDX

0x426aed MOV	%ESI,%EAX

0x426aef MOV	0x3c(%RSP),%EBX

0x426af3 DIVL	0x40(%RSP)

0x426af7 LEA	(%RDX,%RBX,1),%R15D

0x426afb MOV	0x8(%R13),%RBX

0x426aff MOV	%R14D,%EDX

0x426b02 ADD	%EBP,%EAX

0x426b04 SUB	%R15D,%EDX

0x426b07 MOV	(%R13),%R14

0x426b0b MOVSXD	%EAX,%RCX

0x426b0e MOV	%RBX,0x48(%RSP)

0x426b13 CMP	%EDX,%EDI

0x426b15 MOV	0x18(%R13),%RBX

0x426b19 CMOVA	%EDX,%EDI

0x426b1c MOV	%RBX,0x50(%RSP)

0x426b21 MOV	0x10(%R13),%RBX

0x426b25 LEA	(%RSI,%RDI,1),%R12D

0x426b29 MOV	%RBX,0x58(%RSP)

0x426b2e CMP	%R12D,%ESI

0x426b31 JAE	426cb1

0x426b37 NOPW	(%RAX,%RAX,1)

  (221) 0x426b40 MOV	0x48(%RSP),%RAX

  (221) 0x426b45 LEA	0x1(%RCX),%R13

  (221) 0x426b49 MOV	0x10(%R14),%RBP

  (221) 0x426b4d MOV	%R13,%RBX

  (221) 0x426b50 MOV	0x10(%RAX),%R9

  (221) 0x426b54 MOV	(%RAX),%RSI

  (221) 0x426b57 MOV	0x50(%RSP),%RAX

  (221) 0x426b5c IMUL	%RCX,%RSI

  (221) 0x426b60 MOV	%R9,0x28(%RSP)

  (221) 0x426b65 MOV	0x10(%RAX),%R10

  (221) 0x426b69 MOV	(%RAX),%R8

  (221) 0x426b6c MOV	(%R14),%RAX

  (221) 0x426b6f IMUL	%RCX,%R8

  (221) 0x426b73 MOV	%RSI,0x30(%RSP)

  (221) 0x426b78 IMUL	%RAX,%RBX

  (221) 0x426b7c MOV	%R10,0x18(%RSP)

  (221) 0x426b81 MOV	%R8,0x20(%RSP)

  (221) 0x426b86 MOV	%RBX,%RDX

  (221) 0x426b89 SUB	%RAX,%RDX

  (221) 0x426b8c MOV	0x58(%RSP),%RAX

  (221) 0x426b91 MOV	%RDX,0x10(%RSP)

  (221) 0x426b96 MOV	0x10(%RAX),%R11

  (221) 0x426b9a IMUL	(%RAX),%RCX

  (221) 0x426b9e MOV	%R11,0x8(%RSP)

  (221) 0x426ba3 CMP	$0x1,%EDI

  (221) 0x426ba6 JE	426c38

  (221) 0x426bac MOVSXD	%R15D,%RAX

  (221) 0x426baf LEA	(%RSI,%RAX,1),%RDX

  (221) 0x426bb3 MOV	0x8(%RSP),%RSI

  (221) 0x426bb8 LEA	(%R9,%RDX,8),%R11

  (221) 0x426bbc LEA	(%R8,%RAX,1),%RDX

  (221) 0x426bc0 LEA	(%R10,%RDX,8),%R10

  (221) 0x426bc4 LEA	(%RBX,%RAX,1),%RDX

  (221) 0x426bc8 LEA	(%RBP,%RDX,8),%R9

  (221) 0x426bcd MOV	0x10(%RSP),%RDX

  (221) 0x426bd2 ADD	%RAX,%RDX

  (221) 0x426bd5 ADD	%RCX,%RAX

  (221) 0x426bd8 LEA	(%RBP,%RDX,8),%R8

  (221) 0x426bdd MOV	%EDI,%EDX

  (221) 0x426bdf LEA	(%RSI,%RAX,8),%RSI

  (221) 0x426be3 XOR	%EAX,%EAX

  (221) 0x426be5 SHR	$0x1,%EDX

  (221) 0x426be7 SAL	$0x4,%RDX

  (221) 0x426beb NOPW	%CS:(%RAX,%RAX,1)

  (221) 0x426bf6 NOPW	%CS:(%RAX,%RAX,1)

    (222) 0x426c00 MOVUPD	(%R11,%RAX,1),%XMM0

    (222) 0x426c06 MOVUPS	%XMM0,(%R10,%RAX,1)

    (222) 0x426c0b MOVUPD	(%R9,%RAX,1),%XMM1

    (222) 0x426c11 MOVUPD	(%R8,%RAX,1),%XMM2

    (222) 0x426c17 ADDPD	%XMM1,%XMM0

    (222) 0x426c1b SUBPD	%XMM2,%XMM0

    (222) 0x426c1f MOVUPS	%XMM0,(%RSI,%RAX,1)

    (222) 0x426c23 ADD	$0x10,%RAX

    (222) 0x426c27 CMP	%RAX,%RDX

    (222) 0x426c2a JNE	426c00

  (221) 0x426c2c TEST	$0x1,%DIL

  (221) 0x426c30 JE	426c83

  (221) 0x426c32 AND	$-0x2,%EDI

  (221) 0x426c35 ADD	%EDI,%R15D

  (221) 0x426c38 MOV	0x30(%RSP),%RAX

  (221) 0x426c3d MOVSXD	%R15D,%R15

  (221) 0x426c40 MOV	0x28(%RSP),%RDI

  (221) 0x426c45 ADD	%R15,%RBX

  (221) 0x426c48 ADD	%R15,%RCX

  (221) 0x426c4b ADD	%R15,%RAX

  (221) 0x426c4e MOVSD	(%RDI,%RAX,8),%XMM0

  (221) 0x426c53 MOV	0x20(%RSP),%RAX

  (221) 0x426c58 MOV	0x18(%RSP),%RDI

  (221) 0x426c5d ADD	%R15,%RAX

  (221) 0x426c60 MOVSD	%XMM0,(%RDI,%RAX,8)

  (221) 0x426c65 MOV	0x10(%RSP),%RAX

  (221) 0x426c6a ADDSD	(%RBP,%RBX,8),%XMM0

  (221) 0x426c70 ADD	%R15,%RAX

  (221) 0x426c73 SUBSD	(%RBP,%RAX,8),%XMM0

  (221) 0x426c79 MOV	0x8(%RSP),%RAX

  (221) 0x426c7e MOVSD	%XMM0,(%RAX,%RCX,8)

  (221) 0x426c83 MOV	%R12D,%ESI

  (221) 0x426c86 MOV	%R13,%RCX

  (221) 0x426c89 CMP	%R13D,0x38(%RSP)

  (221) 0x426c8e JLE	426cbf

  (221) 0x426c90 MOV	0x44(%RSP),%EDI

  (221) 0x426c94 MOV	0x40(%RSP),%EDX

  (221) 0x426c98 MOV	0x3c(%RSP),%R15D

  (221) 0x426c9d SUB	%ESI,%EDI

  (221) 0x426c9f CMP	%EDX,%EDI

  (221) 0x426ca1 CMOVA	%EDX,%EDI

  (221) 0x426ca4 LEA	(%RSI,%RDI,1),%R12D

  (221) 0x426ca8 CMP	%R12D,%ESI

  (221) 0x426cab JB	426b40

(223) 0x426cb1 LEA	0x1(%RCX),%R13

(223) 0x426cb5 MOV	%R13,%RCX

(223) 0x426cb8 CMP	%R13D,0x38(%RSP)

(223) 0x426cbd JG	426c90

0x426cbf ADD	$0x68,%RSP

0x426cc3 POP	%RBX

0x426cc4 POP	%RBP

0x426cc5 POP	%R12

0x426cc7 POP	%R13

0x426cc9 POP	%R14

0x426ccb POP	%R15

0x426ccd RET

0x426cce ADD	$0x1,%EDI

0x426cd1 XOR	%EDX,%EDX

0x426cd3 JMP	426ad6

0x426cd8 NOPL	(%RAX,%RAX,1)

Coverage (%)	Name	Source Location	Module
►96.56+	gomp_thread_start	team.c:130	libgomp.so.1.0.0
○	start_thread		libc.so.6
○	__GI___clone3		libc.so.6
►1.55+	GOMP_parallel	libgomp.h:982	libgomp.so.1.0.0
○	advec_mom_kernel(int, int, int[...]		exec
○	advec_mom_driver(global_variab[...]		exec
○	advection(global_variables&)		exec
○	hydro(global_variables&, paral[...]		exec
○	main		exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start		exec
►1.52+	GOMP_parallel	libgomp.h:982	libgomp.so.1.0.0
○	advec_mom_kernel(int, int, int[...]		exec
○	advec_mom_driver(global_variab[...]		exec
○	advection(global_variables&)		exec
○	hydro(global_variables&, paral[...]		exec
○	main		exec
○	__libc_start_call_main		libc.so.6
○	__libc_start_main		libc.so.6
○	_start		exec

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
Module	exec

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Vectorization

Your function is not vectorized. Only 7% of vector register length is used (average across all SSE/AVX instructions). By vectorizing your function, you can lower the cost of an iteration from 12.00 to 0.75 cycles (16.00x speedup).

Details

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

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function streams arrays of structures (AoS), try to use structures of arrays instead (SoA)

Execution units bottlenecks

Performance is limited by execution of divide and square root operations (the divide/square root unit is a bottleneck). By removing all these bottlenecks, you can lower the cost of an iteration from 12.00 to 10.13 cycles (1.19x speedup).

Workaround

Reduce the number of division or square root instructions:

If denominator is constant over iterations, use reciprocal (replace x/y with x*(1/y)). Check precision impact.

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 12.00 to 9.88 cycles (1.22x 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.

omp_get_num_threads@plt: 1 occurrences►
omp_get_thread_num@plt: 1 occurrences►

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

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

CMOVA: 1 occurrences►
DIV: 2 occurrences►

Type of elements and instruction set

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

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

The binary function does not contain any FP arithmetical operations. The binary function is loading 56 bytes. The binary function is storing 40 bytes.

General properties

nb instructions	78
nb uops	81
loop length	257
used x86 registers	12
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	7

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	10.13 cycles
front end	10.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	6.50	6.50	6.50	6.50	6.50	6.50	4.75	4.75	4.75	4.75	0.00	0.00	0.00	0.00	0.00	0.00
cycles	6.50	6.50	6.50	6.50	6.50	6.50	4.75	4.75	4.75	4.75	0.00	0.00	0.00	0.00	0.00	0.00

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

12.00

Cycles summary

Front-end	10.13
Dispatch	6.50
DIV/SQRT	12.00
Overall L1	12.00

Vectorization ratios

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	0%
other	0%

Vector efficiency ratios

all	7%
load	9%
store	8%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	6%
other	6%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 12.00 cycles. At this rate:

1% of peak load performance is reached (4.67 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
2% of peak store performance is reached (3.33 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the function is: 426a60

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	Latency	Recip. throughput	Vectorization
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x68,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV 0x28(%RDI),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x20(%RDI),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA -0x1(%RAX),%EBP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x2c(%RDI),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA -0x1(%RBX),%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x24(%RDI),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %ECX,0x3c(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
ADD $0x4,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %EAX,0x38(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CMP %EAX,%EBP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JGE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
LEA 0x4(%RBX),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EAX,%EBX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SUB %EBP,%EBX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
CMP %R14D,%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JGE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV %R14D,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
MOV %RDI,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SUB %ECX,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
MOV %R15D,0x40(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CALL 404660 <omp_get_num_threads@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV %EAX,%R12D	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
CALL 404550 <omp_get_thread_num@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV %EAX,%ECX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %R15D,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
IMUL %EBX,%EAX	1	0.33	0.33	0.33	0	0	0	0	0	0	0	3	0.33	N/A
DIV %R12D	2	0	0	0	2	0	0	0	0	0	0	10-17	6	scal (6.3%)
MOV %EAX,%EDI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CMP %EDX,%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JB 426cce <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x26e>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
IMUL %EDI,%ECX	1	0.33	0.33	0.33	0	0	0	0	0	0	0	3	0.33	N/A
LEA (%RCX,%RDX,1),%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA (%RDI,%RSI,1),%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EAX,0x44(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CMP %EAX,%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JAE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV %ESI,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x3c(%RSP),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
DIVL 0x40(%RSP)	2	0	0	0	2	0	0	0.25	0.25	0.25	0.25	10-17	6	scal (6.3%)
LEA (%RDX,%RBX,1),%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x8(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R14D,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
ADD %EBP,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
SUB %R15D,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV (%R13),%R14	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
MOVSXD %EAX,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %RBX,0x48(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %EDX,%EDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
MOV 0x18(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMOVA %EDX,%EDI	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	N/A
MOV %RBX,0x50(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV 0x10(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA (%RSI,%RDI,1),%R12D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %RBX,0x58(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R12D,%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JAE 426cb1 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x251>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
ADD $0x68,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50	N/A
ADD $0x1,%EDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
JMP 426ad6 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x76>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPL (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

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

Source file and lines
Module	exec

Warnings:
Ignoring paths for analysis

gain
potential
hint
expert

Vectorization

Details

Workaround

Try another compiler or update/tune your current one
Make array accesses unit-stride:
- If your function streams arrays of structures (AoS), try to use structures of arrays instead (SoA)

Execution units bottlenecks

Workaround

Reduce the number of division or square root instructions:

If denominator is constant over iterations, use reciprocal (replace x/y with x*(1/y)). Check precision impact.

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 12.00 to 9.88 cycles (1.22x speedup).

CALL instructions

Detected function call instructions.

Details

omp_get_num_threads@plt: 1 occurrences►
omp_get_thread_num@plt: 1 occurrences►

Complex instructions

Detected COMPLEX INSTRUCTIONS.

Details

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

CMOVA: 1 occurrences►
DIV: 2 occurrences►

Type of elements and instruction set

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

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

The binary function does not contain any FP arithmetical operations. The binary function is loading 56 bytes. The binary function is storing 40 bytes.

General properties

nb instructions	78
nb uops	81
loop length	257
used x86 registers	12
used mmx registers	0
used xmm registers	0
used ymm registers	0
used zmm registers	0
nb stack references	7

Front-end

MACRO FUSION NOT POSSIBLE FIT IN UOP CACHE

micro-operation queue	10.13 cycles
front end	10.13 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15
uops	6.50	6.50	6.50	6.50	6.50	6.50	4.75	4.75	4.75	4.75	0.00	0.00	0.00	0.00	0.00	0.00
cycles	6.50	6.50	6.50	6.50	6.50	6.50	4.75	4.75	4.75	4.75	0.00	0.00	0.00	0.00	0.00	0.00

Execution ports to units layout:

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

Cycles executing div or sqrt instructions

12.00

Cycles summary

Front-end	10.13
Dispatch	6.50
DIV/SQRT	12.00
Overall L1	12.00

Vectorization ratios

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	0%
other	0%

Vector efficiency ratios

all	7%
load	9%
store	8%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	6%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	6%
other	6%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 12.00 cycles. At this rate:

1% of peak load performance is reached (4.67 out of 256.00 bytes loaded per cycle (GB/s @ 1GHz))
2% of peak store performance is reached (3.33 out of 128.00 bytes stored per cycle (GB/s @ 1GHz))

Front-end bottlenecks

Found no such bottlenecks.

ASM code

In the binary file, the address of the function is: 426a60

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	Latency	Recip. throughput	Vectorization
PUSH %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
PUSH %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.50	N/A
SUB $0x68,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV 0x28(%RDI),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV 0x20(%RDI),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA -0x1(%RAX),%EBP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x2c(%RDI),%EAX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA -0x1(%RBX),%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x24(%RDI),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %ECX,0x3c(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
ADD $0x4,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %EAX,0x38(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CMP %EAX,%EBP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JGE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
LEA 0x4(%RBX),%R14D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EAX,%EBX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SUB %EBP,%EBX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
CMP %R14D,%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JGE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
MOV %R14D,%R15D	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
MOV %RDI,%R13	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
SUB %ECX,%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
MOV %R15D,0x40(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CALL 404660 <omp_get_num_threads@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
MOV %EAX,%R12D	1	0	0	0	0	0	0	0	0	0	0	0	0.17	scal (6.3%)
CALL 404550 <omp_get_thread_num@plt>	2	0	0	0	0.33	0.33	0.33	0.25	0.25	0.25	0.25	0	0.50	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV %EAX,%ECX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV %R15D,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
IMUL %EBX,%EAX	1	0.33	0.33	0.33	0	0	0	0	0	0	0	3	0.33	N/A
DIV %R12D	2	0	0	0	2	0	0	0	0	0	0	10-17	6	scal (6.3%)
MOV %EAX,%EDI	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
CMP %EDX,%ECX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JB 426cce <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x26e>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
IMUL %EDI,%ECX	1	0.33	0.33	0.33	0	0	0	0	0	0	0	3	0.33	N/A
LEA (%RCX,%RDX,1),%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
LEA (%RDI,%RSI,1),%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %EAX,0x44(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (6.3%)
CMP %EAX,%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JAE 426cbf <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x25f>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
MOV %ESI,%EAX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
MOV 0x3c(%RSP),%EBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
DIVL 0x40(%RSP)	2	0	0	0	2	0	0	0.25	0.25	0.25	0.25	10-17	6	scal (6.3%)
LEA (%RDX,%RBX,1),%R15D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV 0x8(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
MOV %R14D,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.17	N/A
ADD %EBP,%EAX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
SUB %R15D,%EDX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV (%R13),%R14	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	scal (12.5%)
MOVSXD %EAX,%RCX	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
MOV %RBX,0x48(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %EDX,%EDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
MOV 0x18(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
CMOVA %EDX,%EDI	2	0.33	0.33	0.33	0.33	0.33	0.33	0	0	0	0	1	0.33	N/A
MOV %RBX,0x50(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
MOV 0x10(%R13),%RBX	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	3	0.25	N/A
LEA (%RSI,%RDI,1),%R12D	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.25	N/A
MOV %RBX,0x58(%RSP)	1	0	0	0	0	0	0	0.25	0.25	0.25	0.25	4	0.50	scal (12.5%)
CMP %R12D,%ESI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	scal (6.3%)
JAE 426cb1 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x251>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50-1	N/A
NOPW (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A
ADD $0x68,%RSP	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
POP %RBX	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %RBP	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R12	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R13	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R14	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
POP %R15	1	0	0	0	0	0	0	0	0	0	0	1	0.33	N/A
RET	1	0	0	0	0.33	0.33	0.33	0	0	0	0	1	0.50	N/A
ADD $0x1,%EDI	1	0.17	0.17	0.17	0.17	0.17	0.17	0	0	0	0	1	0.17	N/A
XOR %EDX,%EDX	1	0	0	0	0	0	0	0	0	0	0	0	0.25	N/A
JMP 426ad6 <_Z16advec_mom_kerneliiiiRN6clover8Buffer2DIdEES2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_S2_RNS_8Buffer1DIdEES5_iii._omp_fn.2+0x76>	1	0	0	0	0.33	0.33	0.33	0	0	0	0	0	1	N/A
NOPL (%RAX,%RAX,1)	0	0	0	0	0	0	0	0	0	0	0	0	0.09	N/A

Name	Coverage (%)	Time (s)
▼advec_mom_kernel(int, int, int, int, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clover::Buffer2D&, clo...–	2.52	0.97
▼Loop 223 - - exec–	0.00	0.00
▼Loop 221 - - exec–	0.00	0.00
○Loop 222 - - exec	2.52	0.96

Report Configuration

Vectorization

Details

Workaround

Execution units bottlenecks

Workaround

Expensive FP math instructions/calls

CALL instructions

Details

Complex instructions

Details

Type of elements and instruction set

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code

Vectorization

Details

Workaround

Execution units bottlenecks

Workaround

Expensive FP math instructions/calls

CALL instructions

Details

Complex instructions

Details

Type of elements and instruction set

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

General properties

Front-end

Back-end

Cycles summary

Vectorization ratios

Vector efficiency ratios

Cycles and memory resources usage

Front-end bottlenecks

ASM code