Re: [Patch 2/3] fast VMA recycling
- From: Benjamin LaHaise <bcrl@xxxxxxxxx>
- Date: Thu, 23 Feb 2006 11:37:24 -0500
On Thu, Feb 23, 2006 at 10:30:29AM +0100, Arjan van de Ven wrote:
One could argue that this should be a generic slab feature (the preloading) but
that only gives some of the gains and not all, and vma's are small creatures,
with a high "recycling" rate in typical cases.
I think this is wrong, and far too narrow in useful scope to be merged.
It's unnecessary bloat of task_struct and if this is really a problem, then
there is a [performance] bug in the slab allocator. Please at least provide
the slab statistics for what is going on, as the workload might well be
thrashing the creation and destruction of slabs, which would be a much better
thing to fix.
-ben
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
- References:
- [Patch 0/3] threaded mmap tweaks
- From: Arjan van de Ven
- [Patch 2/3] fast VMA recycling
- From: Arjan van de Ven
- [Patch 0/3] threaded mmap tweaks
- Prev by Date: Re: Patch to reorder functions in the vmlinux to a defined order
- Next by Date: Re: [OT] portable Makefiles (was: CD writing in future Linux (stirring up a hornets' nest))
- Previous by thread: Re: [Patch 2/3] fast VMA recycling
- Next by thread: [Patch 1/3] prefetch the mmap_sem in the fault path
- Index(es):
Relevant Pages
|
