OV - exec - Function qmcplusplus::TimerType<qmcplusplus::CPUClock>::start()

	Function: qmcplusplus::TimerType<qmcplusplus::CPUClock>::start()	Module: exec	Source: NewTimer.cpp:38-81 [...]	Coverage: 0.04%

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp: 38 - 81

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

38: {

39:   if (active)

[...]

53:     for (int level = omp_get_level(); level > 0; level--)

54:       if (omp_get_ancestor_thread_num(level) != 0)

55:         is_true_master = false;

56:     if (is_true_master)

57:     {

58:       if (manager)

59:       {

60:         // compute current_stack_key

61:         TimerType* parent = manager->current_timer();

62:         if (parent)

63:         {

64:           current_stack_key = parent->get_stack_key();

65:           current_stack_key.add_id(timer_id);

66:         }

67:         else

68:         {

69:           current_stack_key = StackKey();

70:           current_stack_key.add_id(timer_id);

71:         }

72:

73:         manager->push_timer(this);

74:       }

75:       start_time = CLOCK()();

[...]

81: }

/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h: 919 - 919

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

919:       { return size_type(this->_M_impl._M_finish - this->_M_impl._M_start); }

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h: 92 - 93

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

92:     if (CurrentTimerStack.size() > 0)

93:       current = CurrentTimerStack.back();

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/Clock.h: 33 - 33

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

33:     return omp_get_wtime();

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h: 90 - 97

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

90:     short_buckets[level] = c1;

91:     if (level >= max_level - 1)

92:     {

93:       timer_max_level_exceeded = true;

94:     }

95:     else

96:     {

97:       level++;

0x44ec60 STP	X29, X30, [SP, #976]!

0x44ec64 STR	X21, [SP, #16]

0x44ec68 STP	X20, X19, [SP, #32]

0x44ec6c ADD	X29, SP, #0

0x44ec70 LDRB	W8, [X0, #56]

0x44ec74 CBZ	W8, 44ed44

0x44ec78 ORR	X19, XZR, X0

0x44ec7c BL	403ca0

0x44ec80 CMP	W0, #1

0x44ec84 B.LT	44ecc0

0x44ec88 ORR	W20, WZR, W0

0x44ec8c MOVZ	W21, #1

0x44ec90 HINT	#0

0x44ec94 HINT	#0

0x44ec98 HINT	#0

0x44ec9c HINT	#0

(1371) 0x44eca0 ORR	W0, WZR, W20

(1371) 0x44eca4 BL	403cb0

(1371) 0x44eca8 CMP	W0, #0

(1371) 0x44ecac CSINC	W8, WZR, WZR, #1

(1371) 0x44ecb0 SUBS	W20, W20, #1

(1371) 0x44ecb4 AND	W21, W8, W21

(1371) 0x44ecb8 B.HI	44eca0

0x44ecbc TBZ	W21, #0, 44ed44

0x44ecc0 LDR	X9, [X19, #72]

0x44ecc4 CBZ	X9, 44ed3c

0x44ecc8 LDP	X9, X8, [X9, #72]

0x44eccc CMP	X8, X9

0x44ecd0 B.EQ	44ed14

0x44ecd4 LDUR	X8, [X8, #504]

0x44ecd8 CBZ	X8, 44ed14

0x44ecdc LDR	Q0, [X8, #80]

0x44ece0 LDRSW	X10, [X8, #96]

0x44ece4 LDRB	W8, [X19, #64]

0x44ece8 ADD	X9, X19, #80

0x44ecec STR	W10, [X19, #96]

0x44ecf0 STR	Q0, [X19, #80]

0x44ecf4 STRB	W8, [X9, X10]

0x44ecf8 LDR	W8, [X19, #96]

0x44ecfc CMP	W8, #15

0x44ed00 B.LT	44ed28

0x44ed04 ADRP	X8,

0x44ed08 MOVZ	W9, #1

0x44ed0c STRB	W9, [X8, #1340]

0x44ed10 B	44ed30

0x44ed14 LDRB	W8, [X19, #64]

0x44ed18 STP	XZR, XZR, [X19, #80]

0x44ed1c STRB	W8, [X19, #80]

0x44ed20 MOVZ	W8, #1

0x44ed24 B	44ed2c

0x44ed28 ADD	W8, W8, #1

0x44ed2c STR	W8, [X19, #96]

0x44ed30 LDR	X0, [X19, #72]

0x44ed34 ORR	X1, XZR, X19

0x44ed38 BL	454490

0x44ed3c BL	403a80

0x44ed40 STR	D0, [X19]

0x44ed44 LDP	X20, X19, [SP, #32]

0x44ed48 LDR	X21, [SP, #16]

0x44ed4c LDP	X29, X30, [SP], #48

0x44ed50 RET

0x44ed54 HINT	#0

0x44ed58 HINT	#0

0x44ed5c HINT	#0

Coverage (%)	Name	Source Location	Module
►25.00+	miniqmcreference::DiracDetermi[...]	DiracDeterminantRef.cpp:102	exec
○	qmcplusplus::WaveFunction::rat[...]	stl_vector.h:919	exec
○	.omp_outlined..62	miniqmc.cpp:437	exec
○	__kmp_invoke_microtask		libomp.so
►25.00+	.omp_outlined..62	miniqmc.cpp:444	exec
○	__kmp_invoke_microtask		libomp.so
►25.00+	qmcplusplus::WaveFunction::rat[...]	stl_vector.h:1046	exec
○	.omp_outlined..62	miniqmc.cpp:485	exec
○	__kmp_invoke_microtask		libomp.so
►25.00+	qmcplusplus::WaveFunction::rat[...]	stl_vector.h:919	exec
○	.omp_outlined..62	miniqmc.cpp:485	exec
○	__kmp_invoke_microtask		libomp.so

Path /

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

Source file and lines	NewTimer.cpp:38-81
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:38-39,53-75,81
/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h:919
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h:92-93
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/Clock.h:33
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:90-97

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 8.17 to 2.33 cycles (3.50x speedup).

Workaround

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

Vectorization

Your function is poorly vectorized. Only 25% of vector register length is used (average across all VPU instructions). By fully vectorizing your function, you can lower the cost of an iteration from 8.17 to 1.31 cycles (6.22x speedup).

Details

20% of VPU instructions are used in vector version (process two or more data elements in vector registers):

50% of VPU stores are used in vector version.
0% of VPU addition or subtraction instructions are used in vector version.
0% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully 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 8.17 to 1.50 cycles (5.44x 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 107 bytes. The binary function is storing 91 bytes.

General properties

nb instructions	57
loop length	228
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	6.50	6.50	4.25	4.25	4.25	4.25	1.00	1.00	0.00	0.00	8.17	7.83	8.00	4.50	4.50
cycles	6.50	6.50	4.25	4.25	4.25	4.25	1.00	1.00	0.00	0.00	8.17	7.83	8.00	4.50	4.50

Cycles executing div or sqrt instructions

Cycles summary

Front-end	6.25
Overall L1	8.17

Vectorization ratios

all	20%
load	100%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 8.17 cycles. At this rate:

13% of peak load performance is reached (13.10 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
23% of peak store performance is reached (11.14 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 44ec60

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #976]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X21, [SP, #16]	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, #32]	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, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDRB W8, [X0, #56]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ W8, 44ed44 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xe4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ORR X19, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 403ca0 <@plt_start@+0x370>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
CMP W0, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 44ecc0 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0x60>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ORR W20, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W21, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
HINT #0
HINT #0
HINT #0
HINT #0
TBZ W21, #0, 44ed44 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xe4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X9, [X19, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X9, 44ed3c <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xdc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X9, X8, [X9, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
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
B.EQ 44ed14 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDUR X8, [X8, #504]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X8, 44ed14 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR Q0, [X8, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDRSW X10, [X8, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDRB W8, [X19, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD X9, X19, #80	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STR W10, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR Q0, [X19, #80]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
STRB W8, [X9, X10]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR W8, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W8, #15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 44ed28 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xc8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADRP X8, <48ad04>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ 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
STRB W9, [X8, #1340]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
B 44ed30 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xd0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDRB W8, [X19, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
STP XZR, XZR, [X19, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STRB W8, [X19, #80]	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
B 44ed2c <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xcc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD W8, 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
STR W8, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR X0, [X19, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ORR X1, XZR, X19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 454490 <_ZN11qmcplusplus12TimerManagerINS_9TimerTypeINS_8CPUClockEEEE10push_timerEPS3_>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
BL 403a80 <@plt_start@+0x150>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
STR D0, [X19]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDP X20, X19, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDR X21, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDP X29, X30, [SP], #48	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
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
HINT #0

Source file and lines	NewTimer.cpp:38-81
Module	exec

The function is defined in:

/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.cpp:38-39,53-75,81
/usr/lib/gcc/aarch64-linux-gnu/11/../../../../include/c++/11/bits/stl_vector.h:919
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/TimerManager.h:92-93
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/Clock.h:33
/home/hbollore/qaas/qaas-runs/169-816-8620/intel/miniqmc/build/miniqmc/src/Utilities/NewTimer.h:90-97

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 8.17 to 2.33 cycles (3.50x speedup).

Workaround

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

Vectorization

Details

20% of VPU instructions are used in vector version (process two or more data elements in vector registers):

50% of VPU stores are used in vector version.
0% of VPU addition or subtraction instructions are used in vector version.
0% of VPU instructions that are not load, store, addition, subtraction nor multiply instructions are used in vector version.

Since your execution units are vector units, only a fully 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 8.17 to 1.50 cycles (5.44x 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 107 bytes. The binary function is storing 91 bytes.

General properties

nb instructions	57
loop length	228
nb stack references	0

Front-end

FIT IN UOP CACHE

front end

6.25 cycles

Back-end

	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14
uops	6.50	6.50	4.25	4.25	4.25	4.25	1.00	1.00	0.00	0.00	8.17	7.83	8.00	4.50	4.50
cycles	6.50	6.50	4.25	4.25	4.25	4.25	1.00	1.00	0.00	0.00	8.17	7.83	8.00	4.50	4.50

Cycles executing div or sqrt instructions

Cycles summary

Front-end	6.25
Overall L1	8.17

Vectorization ratios

all	20%
load	100%
store	50%
mul	NA (no mul vectorizable/vectorized instructions)
add-sub	0%
fma	NA (no fma vectorizable/vectorized instructions)
div/sqrt	NA (no div/sqrt vectorizable/vectorized instructions)
other	0%

Cycles and memory resources usage

Assuming all data fit into the L1 cache, each call to the function takes 8.17 cycles. At this rate:

13% of peak load performance is reached (13.10 out of 96.00 bytes loaded per cycle (GB/s @ 1GHz))
23% of peak store performance is reached (11.14 out of 48.00 bytes stored per cycle (GB/s @ 1GHz))

ASM code

In the binary file, the address of the function is: 44ec60

Instruction	Nb FU	P0	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	Latency	Recip. throughput
STP X29, X30, [SP, #976]!	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR X21, [SP, #16]	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, #32]	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, #0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
LDRB W8, [X0, #56]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ W8, 44ed44 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xe4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ORR X19, XZR, X0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 403ca0 <@plt_start@+0x370>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
CMP W0, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 44ecc0 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0x60>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ORR W20, WZR, W0	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ W21, #1	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
HINT #0
HINT #0
HINT #0
HINT #0
TBZ W21, #0, 44ed44 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xe4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR X9, [X19, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X9, 44ed3c <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xdc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDP X9, X8, [X9, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
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
B.EQ 44ed14 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDUR X8, [X8, #504]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CBZ X8, 44ed14 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xb4>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDR Q0, [X8, #80]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	6	0.33
LDRSW X10, [X8, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDRB W8, [X19, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ADD X9, X19, #80	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
STR W10, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STR Q0, [X19, #80]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
STRB W8, [X9, X10]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR W8, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
CMP W8, #15	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.33
B.LT 44ed28 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xc8>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADRP X8, <48ad04>	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
MOVZ 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
STRB W9, [X8, #1340]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
B 44ed30 <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xd0>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
LDRB W8, [X19, #64]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
STP XZR, XZR, [X19, #80]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
STRB W8, [X19, #80]	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
B 44ed2c <_ZN11qmcplusplus9TimerTypeINS_8CPUClockEE5startEv+0xcc>	1	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0.50
ADD W8, 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
STR W8, [X19, #96]	1	0	0	0	0	0	0	0	0	0	0	0.50	0.50	0	0.50	0.50	1	0.50
LDR X0, [X19, #72]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
ORR X1, XZR, X19	1	0	0	0.25	0.25	0.25	0.25	0	0	0	0	0	0	0	0	0	1	0.25
BL 454490 <_ZN11qmcplusplus12TimerManagerINS_9TimerTypeINS_8CPUClockEEEE10push_timerEPS3_>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
BL 403a80 <@plt_start@+0x150>	1	0.50	0.50	0.50	0.50	0	0	0	0	0	0	0	0	0	0	0	1	0.50
STR D0, [X19]	1	0	0	0	0	0	0	0.50	0.50	0	0	0.50	0.50	0	0	0	2	0.50
LDP X20, X19, [SP, #32]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
LDR X21, [SP, #16]	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	0.33
LDP X29, X30, [SP], #48	1	0	0	0	0	0	0	0	0	0	0	0.33	0.33	0.33	0	0	4	1
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
HINT #0

Name	Coverage (%)	Time (s)
▼qmcplusplus::TimerType::start()–	0.04	0.02
○Loop 1371 - NewTimer.cpp:53-54 - 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