Date: Wed, 8 Dec 2010 23:40:41 +0000 (UTC) From: Jung-uk Kim <jkim@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r216312 - in head/sys: amd64/amd64 i386/i386 pc98/pc98 Message-ID: <201012082340.oB8Nefff018113@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: jkim Date: Wed Dec 8 23:40:41 2010 New Revision: 216312 URL: http://svn.freebsd.org/changeset/base/216312 Log: Do not subtract 0.5% from estimated frequency if DELAY(9) is driven by TSC. Remove a confusing comment about converting to MHz as we never did. Modified: head/sys/amd64/amd64/machdep.c head/sys/i386/i386/machdep.c head/sys/pc98/pc98/machdep.c Modified: head/sys/amd64/amd64/machdep.c ============================================================================== --- head/sys/amd64/amd64/machdep.c Wed Dec 8 22:54:18 2010 (r216311) +++ head/sys/amd64/amd64/machdep.c Wed Dec 8 23:40:41 2010 (r216312) @@ -563,13 +563,7 @@ cpu_est_clockrate(int cpu_id, uint64_t * thread_unlock(curthread); #endif - /* - * Calculate the difference in readings, convert to Mhz, and - * subtract 0.5% of the total. Empirical testing has shown that - * overhead in DELAY() works out to approximately this value. - */ - tsc2 -= tsc1; - *rate = tsc2 * 1000 - tsc2 * 5; + *rate = (tsc2 - tsc1) * 1000; return (0); } Modified: head/sys/i386/i386/machdep.c ============================================================================== --- head/sys/i386/i386/machdep.c Wed Dec 8 22:54:18 2010 (r216311) +++ head/sys/i386/i386/machdep.c Wed Dec 8 23:40:41 2010 (r216312) @@ -1157,12 +1157,16 @@ cpu_est_clockrate(int cpu_id, uint64_t * thread_unlock(curthread); #endif + tsc2 -= tsc1; + if (tsc_freq != 0 && !tsc_is_broken) { + *rate = tsc2 * 1000; + return (0); + } + /* - * Calculate the difference in readings, convert to Mhz, and - * subtract 0.5% of the total. Empirical testing has shown that + * Subtract 0.5% of the total. Empirical testing has shown that * overhead in DELAY() works out to approximately this value. */ - tsc2 -= tsc1; *rate = tsc2 * 1000 - tsc2 * 5; return (0); } Modified: head/sys/pc98/pc98/machdep.c ============================================================================== --- head/sys/pc98/pc98/machdep.c Wed Dec 8 22:54:18 2010 (r216311) +++ head/sys/pc98/pc98/machdep.c Wed Dec 8 23:40:41 2010 (r216312) @@ -1092,12 +1092,16 @@ cpu_est_clockrate(int cpu_id, uint64_t * thread_unlock(curthread); #endif + tsc2 -= tsc1; + if (tsc_freq != 0 && !tsc_is_broken) { + *rate = tsc2 * 1000; + return (0); + } + /* - * Calculate the difference in readings, convert to Mhz, and - * subtract 0.5% of the total. Empirical testing has shown that + * Subtract 0.5% of the total. Empirical testing has shown that * overhead in DELAY() works out to approximately this value. */ - tsc2 -= tsc1; *rate = tsc2 * 1000 - tsc2 * 5; return (0); }
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201012082340.oB8Nefff018113>