Date: Tue, 04 Mar 2008 11:09:44 +0800 From: David Xu <davidxu@FreeBSD.org> To: Jeff Roberson <jroberson@chesapeake.net> Cc: cvs-src@FreeBSD.org, Jeff Roberson <jeff@FreeBSD.org>, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h Message-ID: <47CCBD78.5040708@freebsd.org> In-Reply-To: <20080303164227.S920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CCAF49.20903@freebsd.org> <20080303164227.S920@desktop>
next in thread | previous in thread | raw e-mail | index | archive | help
Jeff Roberson wrote:
>
> On Tue, 4 Mar 2008, David Xu wrote:
>
>> Jeff Roberson wrote:
>>
>>>> One question is how I can determine the size of cpuset the kernel is
>>>> using ?
>>>
>>> I wrote it to tolerate user masks that were much larger than the
>>> kernel mask. I set the default CPU_SETSIZE in userspace to 128 and
>>> in kernel it's MAXCPU. So in practice an application shouldn't have
>>> to redefine CPU_SETSIZE. If your set is too small the kernel will
>>> return ERANGE however. Unfortunately, if your set is larger than the
>>> kernel's CPU_MAXSIZE it'll also return ERANGE. Maybe I should use
>>> different errnos for those cases.
>>>
>>
>> From my point, userland has to write some urgly code to guess what
>> kernel code wants, it is rather frustrate.
>
> You can use sysctl kern.smp.maxcpus to get the precise size.
>
if kern.smp.maxcpus is a stable ABI, I may use it, can it be guaranteed?
I saw following code in kern_cpuset.c, obviously, maxcpus is not
respected.
if (uap->cpusetsize < CPU_SETSIZE || uap->cpusetsize > CPU_MAXSIZE)
return (ERANGE);
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?47CCBD78.5040708>