On 6/13/2017 5:35 AM, Pietro Cerutti wrote: > Author: gahr (ports committer) > Date: Tue Jun 13 12:35:01 2017 > New Revision: 319897 > URL: https://svnweb.freebsd.org/changeset/base/319897 > > Log: > Improve yes' throughput > > On my system, this brings up the throughput from ~20 to ~600 MiB/s. > > Inspired by: > https://www.reddit.com/r/unix/comments/6gxduc/how_is_gnu_yes_so_fast/ > > Reviewed by: cognet > Approved by: cognet > > Modified: > head/usr.bin/yes/yes.c > > Modified: head/usr.bin/yes/yes.c > ============================================================================== > --- head/usr.bin/yes/yes.c Tue Jun 13 12:07:18 2017 (r319896) > +++ head/usr.bin/yes/yes.c Tue Jun 13 12:35:01 2017 (r319897) > @@ -44,20 +44,42 @@ static const char rcsid[] = "$FreeBSD$"; > #include <capsicum_helpers.h> > #include <err.h> > #include <stdio.h> > +#include <string.h> > +#include <unistd.h> > > int > main(int argc, char **argv) > { > + char buf[8192];
8192 feels arbitrary but it has a purpose. Shouldn't it be based on PAGE_SIZE or something? > + char y[2] = { 'y', '\n' }; > + char * exp = y; > + size_t buflen = 0; > + size_t explen = sizeof(y); More style bugs here > > if (caph_limit_stdio() < 0 || (cap_enter() < 0 && errno != ENOSYS)) > err(1, "capsicum"); > > if (argc > 1) > - while (puts(argv[1]) != EOF) > - ; > - else > - while (puts("y") != EOF) > - ; > + { > + exp = argv[1]; > + explen = strlen(exp) + 1; > + exp[explen - 1] = '\n'; > + } > + > + if (explen <= sizeof(buf)) > + { > + while (buflen < sizeof(buf) - explen) > + { > + memcpy(buf + buflen, exp, explen); > + buflen += explen; > + } > + exp = buf; > + explen = buflen; > + } > + > + while (write(STDOUT_FILENO, exp, explen) > 0) > + ; > + > err(1, "stdout"); > /*NOTREACHED*/ > } > -- Regards, Bryan Drewery
signature.asc
Description: OpenPGP digital signature