Messages in this thread | | | Date | Sat, 19 Jun 2004 14:34:33 -0700 | From | William Lee Irwin III <> | Subject | Re: [patch] flexible-mmap-2.6.7-D5 |
| |
* William Lee Irwin III <wli@holomorphy.com> wrote: >> Also, I suspect some more graceful fallback would make sense >> particularly for the case of RLIM_INFINITY, which would leave users >> that run with, say, all rlimits at RLIM_INFINITY in the interest of >> having full access to system resources with a mere 512MB of >> virtualspace for the heap, which IIRC glibc is intelligent enough to >> circumvent for malloc(), but not for mmap(NULL, ...). [...]
On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > well, the 5/6=stack 1/6=malloc rule in the RLIM_INFINITY can be changed. > What would make the most sense - 1/2 for both?
I had in mind fallback as opposed to a changed base, but a particular choice of the base may cover enough cases. The bugreport below seems to say there's no need for a change.
* William Lee Irwin III <wli@holomorphy.com> wrote: >> If it's been in production that long, I find it hard to believe that's >> never been tripped over. [...]
On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > it's been tripped over and the 5/6 rule was a fix for such a bugreport. > What happens more in practice frequently is that someone needs a big > stack and sets the stack ulimit to RLIM_INFINITY.
This sounds like nothing is needed, then.
* William Lee Irwin III <wli@holomorphy.com> wrote: >> [...] (also, that 128MB is currently wasted); [...]
On Sat, Jun 19, 2004 at 01:38:36PM +0200, Ingo Molnar wrote: > the 128MB is 'wasted' to give some flexibility to the stack rlimits > changing runtime. But in practice it's far more important to have the > mmap()/malloc() space maximized and flexible than to give the stack > automatic flexibility.
Fishing around down there to utilize it for mmap() placement should happen anyway if things fall back far enough in the top-down scheme.
Answers like "I've thought about it" or "I've seen this and dealt with it" are good enough for me. There isn't much of a normative aspect to user virtualspace layout.
Thanks.
-- wli - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |