Re: sysctlbyname( "hw.availcpu" ) fails
Re: sysctlbyname( "hw.availcpu" ) fails
- Subject: Re: sysctlbyname( "hw.availcpu" ) fails
- From: Terry Lambert <email@hidden>
- Date: Wed, 27 Jun 2007 13:00:41 -0700
On Jun 27, 2007, at 12:01 PM, Steve Checkoway wrote:
On Jun 25, 2007, at 4:50 PM, Terry Lambert wrote:
activecpu is the correct name; the manifest constant name was not
changed because people have existing code that has to compile vs.
the MIB version of the call.
Okay, I'll use that then.
The preferred method for all sysctls going forward is to use
sysctlbyname(), so that's what you should use in new code.
I did wonder about that. All of the examples I've seen on this list
have been of the sysctlbyname() type.
Is there a list of the sysctls that are guaranteed to exist on all
machines (or by architecture).
No.
Also, a description of them would be nice. For example, on this
computer sysctl -a|grep cpu gives me (among others):
hw.ncpu = 1
hw.availcpu = 1
hw.ncpu: 1
hw.activecpu: 1
hw.logicalcpu_max: 1
hw.logicalcpu: 1
hw.physicalcpu_max: 1
hw.physicalcpu: 1
It isn't clear to me how ncpu and activecpu relate to the others. My
guess is that ncpu is the same as logicalcpu_max and activecpu is
the same as logicalcpu. I'm not at my G5 so I can't check my
assumptions at the moment. The comments in the header do not mention
physical/logical for either ncpu or activecpu.
This can all change in the future, but currently:
hw.ncpu is a wart; consider it to be deprecated.
hw.physicalcpu is the number of physical CPUs
hw.logicalcpu is the number of logical CPUs; this is for SMT, which
we don't support (maybe T1s?)
hw.availcpu are the number logical CPUs currently online
These interfaces are evolving, however, you are unlikely to get a
description of these. They are intended for internal ibrary use, and
not for use by applications, since applications should use the library
abstractions rather than trying to use this information directly
themselves.
Specifically, this information is not terrifically useful unless you
know exactly how the scheduler works, and this can change from release
to release, and even as a result of a software update. As a result,
the only way you can safely use the information for anything is to
have your code sent out in lockstep with software updates. This
happens for system libraries, when something that matters to the
library changes, but rarely happens anywhere else.
-- Terry
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden