Backed out previous commit. The seconds part of microtime() is
not the necessarily the same as the seconds part of getmicrotime() yet, and anyway, we should have used `time_second' if we only wanted a sloppy value for the seconds part. There is no point in making ibcs2's time(2) more efficient than FreeBSD's time(3).
This commit is contained in:
parent
2d2f8ae7ad
commit
084a3966cc
|
@ -45,7 +45,7 @@
|
|||
*
|
||||
* @(#)sun_misc.c 8.1 (Berkeley) 6/18/93
|
||||
*
|
||||
* $Id: ibcs2_misc.c,v 1.26 1998/04/06 08:25:58 phk Exp $
|
||||
* $Id: ibcs2_misc.c,v 1.27 1998/04/06 11:37:17 phk Exp $
|
||||
*/
|
||||
|
||||
/*
|
||||
|
@ -695,7 +695,7 @@ ibcs2_time(p, uap)
|
|||
{
|
||||
struct timeval tv;
|
||||
|
||||
getmicrotime(&tv);
|
||||
microtime(&tv);
|
||||
p->p_retval[0] = tv.tv_sec;
|
||||
if (SCARG(uap, tp))
|
||||
return copyout((caddr_t)&tv.tv_sec, (caddr_t)SCARG(uap, tp),
|
||||
|
|
Loading…
Reference in New Issue