On Tue, Aug 7, 2012 at 5:21 AM, Richard Earnshaw <rearn...@arm.com> wrote: > > Wouldn't it be more useful on Linux to check the task's affinity > settings? Then when a task is locked to a limited set of cores it won't > overload those cores with threads.
This suggestion was implemented by Shenghou Ma, as attached. Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu. Committed to mainline. Ian
diff -r 8fc45a01251d libgo/runtime/getncpu-linux.c --- a/libgo/runtime/getncpu-linux.c Mon Aug 06 21:42:17 2012 -0700 +++ b/libgo/runtime/getncpu-linux.c Thu Aug 09 23:04:31 2012 -0700 @@ -2,46 +2,35 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -#include <string.h> -#include <sys/types.h> -#include <fcntl.h> -#include <unistd.h> +#include <features.h> +#include <sched.h> + +// CPU_COUNT is only provided by glibc 2.6 or higher +#if !defined(__GLIBC_PREREQ) || !__GLIBC_PREREQ(2, 6) +#define CPU_COUNT(set) _CPU_COUNT((unsigned int *)(set), sizeof(*(set))/sizeof(unsigned int)) +static int _CPU_COUNT(unsigned int *set, size_t len) { + int cnt; + + cnt = 0; + while (len--) + cnt += __builtin_popcount(*set++); + return cnt; +} +#endif #include "runtime.h" #include "defs.h" -#ifndef O_CLOEXEC -#define O_CLOEXEC 0 -#endif - int32 getproccount(void) { - int32 fd, rd, cnt, cpustrlen; - const char *cpustr; - const byte *pos; - byte *bufpos; - byte buf[256]; + cpu_set_t set; + int32 r, cnt; - fd = open("/proc/stat", O_RDONLY|O_CLOEXEC, 0); - if(fd == -1) - return 1; cnt = 0; - bufpos = buf; - cpustr = "\ncpu"; - cpustrlen = strlen(cpustr); - for(;;) { - rd = read(fd, bufpos, sizeof(buf)-cpustrlen); - if(rd == -1) - break; - bufpos[rd] = 0; - for(pos=buf; (pos=(const byte*)strstr((const char*)pos, cpustr)) != nil; cnt++, pos++) { - } - if(rd < cpustrlen) - break; - memmove(buf, bufpos+rd-cpustrlen+1, cpustrlen-1); - bufpos = buf+cpustrlen-1; - } - close(fd); + r = sched_getaffinity(0, sizeof(set), &set); + if(r == 0) + cnt += CPU_COUNT(&set); + return cnt ? cnt : 1; }