| From | Sent On | Attachments |
|---|
| Subject: | Re: New hw.cpuhz sysctl as per PR i386/27627 | |
|---|---|---|
| From: | Bruce Evans (bd...@zeta.org.au) | |
| Date: | Jul 7, 2001 4:19:53 am | |
| List: | org.freebsd.freebsd-arch | |
Refine Search
| From | Sent On | Attachments |
|---|---|---|
| Peter Pentchev | Jul 6, 2001 2:38 am | |
| Arun Sharma | Jul 6, 2001 11:10 am | |
| Mike Smith | Jul 6, 2001 2:00 pm | |
| Mike Smith | Jul 6, 2001 2:35 pm | |
| Lars Eggert | Jul 6, 2001 3:09 pm | |
| Mike Smith | Jul 6, 2001 3:26 pm | |
| Bruce Evans | Jul 7, 2001 4:19 am |
| Subject: | Re: New hw.cpuhz sysctl as per PR i386/27627 | |
|---|---|---|
| From: | Bruce Evans (bd...@zeta.org.au) | |
| Date: | Jul 7, 2001 4:19:53 am | |
| List: | org.freebsd.freebsd-arch | |
On Fri, 6 Jul 2001, Mike Smith wrote:
Mike Smith wrote:
CPU speeds aren't constants; this is probably not the right place to put it. It also assumes there's only one CPU, which isn't acceptable.
Being the author of the original patch, I agree that its a hack - but it's better than nothing :-) Plus, it doesn't necessarily assume only one CPU - it does assume they all have the same fixed speed though.
It doesn't seem to serve much useful purpose; the information is already available to the user in the boot-time message buffer, and it's not useful as a number to a running process.
The same information is already available to the user via the machdep.tsc_freq sysctl if there is a tsc, modulo bugs in the timecounter code (this sysctl is actually only available if there is a tsc AND the tsc timecounter was initialized at boot time). If there is no tsc, then machdep.tsc_freq fails properly but hw.cpuhz bogusly says that the frequency is 0.
Bruce
To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message

