In-Reply-To: <[email protected]>
On Tue, 20 Jun 2006 22:57:16 +0200, Andi Kleen wrote:
> On Tuesday 20 June 2006 22:25, Chuck Ebbert wrote:
> > Use the limit field of a GDT entry to store the current CPU
> > number for fast userspace access. This still leaves 12 bits
> > free for other information.
>
> Nice trick. Maybe I'll even add that to the x86-64 implementation
> if it's fast enough. Do you have numbers?
>
I got ~13 clocks on x86_64 and 21 on PII. Test program below.
> But it needs to be encapsulated in a wrapper I think. Just exposing
> it to user space is the wrong way to do this.
Well there's no real way to hide it, but something more is definitely
needed. A new vdso entry point is easy but how do you tell userspace
it's there? Does glibc look at the kernel version in the note?
/* test how fast lsl, test for success + mask result runs
* leave DO_TEST undefined to measure overhead
*/
#define _GNU_SOURCE
#include <stdio.h>
#include <stdlib.h>
#define rdtscll(t) asm volatile ("rdtsc" : "=A" (t))
#ifndef ITERS
#define ITERS 1000000
#endif
int main(int argc, char * const argv[])
{
unsigned long long tsc1, tsc2;
int count, cpu, junk;
rdtscll(tsc1);
asm (
" pushl %%ds \n"
" popl %2 \n"
"1: \n"
#ifdef DO_TEST
" lsl %2,%0 \n"
" jnz 2f \n"
" and $0xff,%0 \n"
#endif
" dec %1 \n"
" jnz 1b \n"
"2: \n"
: "=&r" (cpu), "=&r" (count), "=&r" (junk)
: "1" (ITERS), "0" (-1)
);
rdtscll(tsc2);
if (count == 0)
printf("loops: %d, avg: %llu clocks\n",
ITERS, (tsc2 - tsc1) / ITERS);
return 0;
}
--
Chuck
"You can't read a newspaper if you can't read." --George W. Bush
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
