Re: test_and_set_bit() not atomic forever? [cli/sti in char/vt.c [patch]]

ralf@uni-koblenz.de
Sun, 31 May 1998 15:52:08 +0200


On Sun, May 31, 1998 at 11:39:07AM +0100, Philip Blundell wrote:

> >After reading Pavel sentence I looked the MIPS implementation and it didn'
> >t seem so atomic for SMP (but I didn' t looked also asm-mips/smp.h...), so
>
> SMP isn't supported on MIPS so this is no big deal.

The MIPS implementation is atomic with respect to interrupts for all CPUs
and atomic with respect to SMP for R6000 and above. The same applies to
the other primitives like xchg or atomic_add.

Even though we don't support SMP on MIPS yet these operations have been
implemented in a SMP safe manor because that also avoids playing with
interrupts and make smaller and faster code.

Ralf

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu