Date: Wed, 07 Jan 2009 00:31:13 +0000 From: Craig Butler <craig001@lerwick.hopto.org> To: freebsd-sparc64@FreeBSD.org Subject: Re: sparc64/119240 top has WCPU over 100% on UP system Message-ID: <1231288273.83664.4.camel@main.lerwick.hopto.org> In-Reply-To: <200901062120.n06LK6cq079473@freefall.freebsd.org> References: <200901062120.n06LK6cq079473@freefall.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi All I think its safe to shelve this PR for now, it does not appear to be a fault and I have seen the same result on other architectures, namely x86 and x64. I think its due to the way that top works the weighted average out, if its polling it over a period it could be possible to get above 100% result for WCPU. The -C flag appears to display the raw CPU usage at that point of time and does not go over 100% Thanks Craig Butler On Tue, 2009-01-06 at 21:20 +0000, Craig Butler wrote: > The following reply was made to PR sparc64/119240; it has been noted by GNATS. > > From: Craig Butler <craig001@lerwick.hopto.org> > To: bug-followup <bug-followup@freebsd.org> > Cc: > Subject: sparc64/119240 top has WCPU over 100% on UP system > Date: Tue, 06 Jan 2009 21:16:36 +0000 > > Hello > > I have also seen the same result on sparc64 and amd64 kit, I have > looked into it a bit on google and the general feeling is that the WCPU > (or weighted CPU) count is not all that accurate. > > A more accurate representation can be obtained by passing the -C flag > when starting top. This is for raw CPU rather than a weighted average, > on some quick tests on a V120 whilst running buildworld I have not seen > a reading above 100%. > > Kind Regards > > Craig Butler > > _______________________________________________ > freebsd-sparc64@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-sparc64 > To unsubscribe, send any mail to "freebsd-sparc64-unsubscribe@freebsd.org"
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1231288273.83664.4.camel>