4 ms·

parsed = http_parser_execute(parser, &settings_null, buf, strlen(buf)); assert(parsed == strlen(buf)); He's calling strlen(buf) twice here. There's no guarant

by halayli 12y ago

parsed = http_parser_execute(parser, &settings_null, buf, strlen(buf)); assert(parsed == strlen(buf)); He's calling strlen(buf) twice here. There's no guarantee that this is optimized at compile time. Usually len of buf is determined from read/recv system calls and for a fair comparison strlen should be removed here.

Ono-Sendai 12y ago
assert only runs in debug mode anyway.
- haberman 12y ago
  It runs unless NDEBUG is defined, which the author did not do according to his methodology. Anyway I looked at the assembly output and strlen() is optimized away.
  - webreac 12y ago
    I really do not like when bad habits are encouraged by the optimizer.
- halayli 12y ago
  no, that's not correct. http://pubs.opengroup.org/onlinepubs/009695399/functions/assert.html http://pubs.opengroup.org/onlinepubs/009695399/functions/ass...
  - aidenn0 12y ago
    "Forcing a definition of the name NDEBUG, either from the compiler command line or with the preprocessor control statement #define NDEBUG ahead of the #include <assert.h> statement, shall stop assertions from being compiled into the program"
fukamachi 12y ago
No difference even though I ran the benchmark again after removing the line.
- halayli 12y ago
  Interesting. I was expecting at least cpu branch improvements given the lack of assert.
- halayli 12y ago
  Also use gettimeofday() instead of clock(), because you don't know how far into the second you are at the time it's called.
jeffreyrogers 12y ago
Any good compiler will optimize that out. (Actually, even a bad compiler will optimize that out... it's a really easy optimization).