smp-safe vmalloc (was Re: [Patch] Polling on more than 16000 file descriptors)

From: Tigran Aivazian (tigran@sco.COM)
Date: Tue Jan 18 2000 - 10:54:42 EST

On 18 Jan 2000, Andi Kleen wrote:
> tigran@sco.COM (Tigran Aivazian) writes:
> > And, btw, since now the number of users of vmalloc() in
> > performance-critical paths is growing, perhaps it is worth to come back to
> > the patch I submitted for review ages ago that adds vmlist_lock rw
> > spinlock and makes vmalloc() SMP-safe?
> I think vmalloc in every poll is totally out of question.

how is that related to my suggestion to revisit the vmlist_lock spinlock
patch? I agree that vmalloc in every poll is totally out of the question
(and never suggested that). But making vmalloc()/vfree()/ioremap()
SMP-safe is, imho, of benefit regardless. The more users of of those
functions are there, the more sense it makes to have vmlist spinlock. It
is because there is no such spinlock /proc/kcore cannot 100% guarantee
reliable results when walking vmlist (see XXX in fs/proc/kcore.c). We
could take big kernel lock, I suppose but, imo, it is best avoided.


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

This archive was generated by hypermail 2b29 : Sun Jan 23 2000 - 21:00:18 EST