This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: FOR REVIEW: New x86-64 vsyscall vgetcpu()
- From: Andi Kleen <ak at suse dot de>
- To: Zoltan Menyhart <Zoltan dot Menyhart at bull dot net>
- Cc: Jes Sorensen <jes at sgi dot com>, Tony Luck <tony dot luck at intel dot com>, discuss at x86-64 dot org, linux-kernel at vger dot kernel dot org, libc-alpha at sourceware dot org, vojtech at suse dot cz, linux-ia64 at vger dot kernel dot org
- Date: Fri, 16 Jun 2006 16:56:40 +0200
- Subject: Re: FOR REVIEW: New x86-64 vsyscall vgetcpu()
- References: <200606140942.31150.ak@suse.de> <44929CE6.4@sgi.com> <4492A5E4.9050702@bull.net>
On Friday 16 June 2006 14:36, Zoltan Menyhart wrote:
> Just to make sure I understand it correctly...
> Assuming I have allocated per CPU data (numa control, etc.) pointed at by:
>
> void *per_cpu[MAXCPUS];
That is not how user space TLS works. It usually has a base a register.
>
> Assuming a per CPU variable has got an "offset" in each per CPU data area.
> Accessing this variable can be done as follows:
>
> err = vgetcpu(&my_cpu, ...);
> if (err)
> goto ....
> pointer = (typeof pointer) (per_cpu[my_cpu] + offset);
> // use "pointer"...
>
> It is hundred times more long than "__get_per_cpu(var)++".
14 cycles is not a 100 times longer.
> My idea is to map the current task structure at an arch. dependent
> virtual address into the user space (obviously in RO).
>
> #define current ((struct task_struct *) 0x...)
This means it cannot be cache colored (because you would need a static
offset) and you couldn't share task_structs on a page.
Also you would make task_struct part of the userland ABI which
seems like a very very bad idea to me. It means we couldn't change
it anymore.
-Andi