Re: sysctlbyname( "hw.availcpu" ) fails
Re: sysctlbyname( "hw.availcpu" ) fails
- Subject: Re: sysctlbyname( "hw.availcpu" ) fails
- From: Terry Lambert <email@hidden>
- Date: Fri, 29 Jun 2007 15:58:04 -0700
On Jun 29, 2007, at 2:39 PM, Eric Schlegel wrote:
On Jun 29, 2007, at 2:34 PM, Steve Checkoway wrote:
Then there are other indications like <http://developer.apple.com/samplecode/CPUGestalt/index.html
> which has an interesting warning.
I think that's mostly because we never recommend that you use the
specific CPU type to make decisions about code paths. It doesn't
matter whether you're getting the cputype from Gestalt or
sysctlbyname, it's still better to use CPU capabilities rather than
CPU types.
Specifically, certain CPU capabilities can be disabled at runtime, so
your software should not assume those capabilities are present merely
because you've been able to identify the CPU.
My standard example of this is that it's possible to disable Altivec
on PPC systems, and in so doing, test that the code will/won't run on
a G3 system, to make sure you didn't make a mistake in code
generation. It's also possible on Intel for specific features to be
enabled/disabled on various models and steppings as a result of EFI or
Platform Expert updates.
-- 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