User Peter Cordes - Stack Overflow

2143 votes

Accepted

Why does C++ code for testing the Collatz conjecture run faster than hand-written assembly?

answered Nov 1, 2016 at 7:04

433 votes

Accepted

Deoptimizing a program for the pipeline in Intel Sandybridge-family CPUs

answered May 21, 2016 at 11:17

382 votes

How much of ‘What Every Programmer Should Know About Memory’ is still valid?

answered Dec 8, 2017 at 12:32

334 votes

Accepted

What is the best way to set a register to zero in x86 assembly: xor, mov or and?

answered Nov 12, 2015 at 9:37

326 votes

Accepted

Does the C++ standard allow for an uninitialized bool to crash a program?

answered Jan 10, 2019 at 9:42

272 votes

Accepted

Is incrementing an int effectively atomic in specific cases?

answered Sep 8, 2016 at 17:30

185 votes

Why does glibc's strlen need to be so complicated to run quickly?

answered Aug 27, 2019 at 13:43

176 votes

Accepted

Bubble sort slower with -O3 than -O2 with GCC

answered Oct 9, 2021 at 3:09

161 votes

Accepted

Fastest way to do horizontal SSE vector sum (or other reduction)

answered Feb 8, 2016 at 12:46

156 votes

Accepted

How to remove "noise" from GCC/clang assembly output?

answered Jul 24, 2016 at 13:25

129 votes

Accepted

Why does this delay-loop start to run faster after several iterations with no sleep?

answered Jul 11, 2016 at 5:57

112 votes

Accepted

Where is the x86-64 System V ABI documented?

answered Oct 31, 2016 at 18:11

104 votes

Accepted

Possible GCC bug when returning struct from a function

answered Jan 23, 2020 at 5:00

93 votes

Accepted

INC instruction vs ADD 1: Does it matter?

answered Apr 8, 2016 at 23:24

82 votes

Accepted

Why is integer assignment on a naturally aligned variable atomic on x86?

answered Apr 18, 2016 at 3:34

81 votes

Accepted

What happens if you use the 32-bit int 0x80 Linux ABI in 64-bit code?

answered Sep 7, 2017 at 4:20

79 votes

Where is the lock for a std::atomic?

answered May 11, 2018 at 19:13

79 votes

Accepted

Is there a 128 bit integer in gcc?

answered Feb 21, 2019 at 19:43

79 votes

Accepted

Argument order to std::min changes compiler output for floating-point

answered Sep 26, 2020 at 22:29

78 votes

How to get the CPU cycle count in x86_64 from C++?

answered Aug 18, 2018 at 10:29

78 votes

Floating point division vs floating point multiplication

answered Aug 26, 2017 at 20:00

75 votes

Accepted

Why are loops always compiled into "do...while" style (tail jump)?

answered Dec 13, 2017 at 10:28

73 votes

Accepted

Boolean values as 8 bit in compilers. Are operations on them inefficient?

answered Nov 12, 2017 at 2:55

73 votes

Accepted

gcc optimization flag -O3 makes code slower than -O2

answered May 12, 2017 at 15:51

73 votes

32-bit absolute addresses no longer allowed in x86-64 Linux?

answered Sep 29, 2017 at 16:45

73 votes

Accepted

Can x86's MOV really be "free"? Why can't I reproduce this at all?

answered May 26, 2017 at 4:43

73 votes

Accepted

Why do none of the major compilers optimize this conditional store that checks if the value is already set?

answered May 16, 2023 at 9:48

70 votes

Why is this seemingly slower C loop actually twice as fast as the other way?

answered Sep 22, 2022 at 22:06

67 votes

Accepted

What is callq instruction?

answered Oct 15, 2017 at 9:19

67 votes

Why does Windows64 use a different calling convention from all other OSes on x86-64?

answered Feb 25, 2016 at 6:00

Collectives™ on Stack Overflow

6,680 Answers

Why does C++ code for testing the Collatz conjecture run faster than hand-written assembly?

Deoptimizing a program for the pipeline in Intel Sandybridge-family CPUs

How much of ‘What Every Programmer Should Know About Memory’ is still valid?

What is the best way to set a register to zero in x86 assembly: xor, mov or and?

Does the C++ standard allow for an uninitialized bool to crash a program?

Is incrementing an int effectively atomic in specific cases?

Why does glibc's strlen need to be so complicated to run quickly?

Bubble sort slower with -O3 than -O2 with GCC

Fastest way to do horizontal SSE vector sum (or other reduction)

How to remove "noise" from GCC/clang assembly output?

Why does this delay-loop start to run faster after several iterations with no sleep?

Where is the x86-64 System V ABI documented?

Possible GCC bug when returning struct from a function

INC instruction vs ADD 1: Does it matter?

Why is integer assignment on a naturally aligned variable atomic on x86?

What happens if you use the 32-bit int 0x80 Linux ABI in 64-bit code?

Where is the lock for a std::atomic?

Is there a 128 bit integer in gcc?

Argument order to std::min changes compiler output for floating-point

How to get the CPU cycle count in x86_64 from C++?

Floating point division vs floating point multiplication

Why are loops always compiled into "do...while" style (tail jump)?

Boolean values as 8 bit in compilers. Are operations on them inefficient?

gcc optimization flag -O3 makes code slower than -O2

32-bit absolute addresses no longer allowed in x86-64 Linux?

Can x86's MOV really be "free"? Why can't I reproduce this at all?

Why do none of the major compilers optimize this conditional store that checks if the value is already set?

Why is this seemingly slower C loop actually twice as fast as the other way?

What is callq instruction?

Why does Windows64 use a different calling convention from all other OSes on x86-64?