On 1/23/06, Hugh Dickins <[email protected]> wrote:
> On Sun, 22 Jan 2006, Don Dupuis wrote:
> > On 1/21/06, Nick Piggin <[email protected]> wrote:
> > > Andrew Morton wrote:
> > > > Don Dupuis <[email protected]> wrote:
> > > >>I have an app that mlocks hugepages. The same app works just fine in 2.6.14.
> > > > That being said, we shouldn't have broken your application.
> > > Don, an strace log of the failing sequence of syscalls could be helpful.
> >
> > sducstart:
> > open("/pivot3/mem/sduc", O_RDWR|O_CREAT, 0666) = 3
> > mmap2(NULL, 1761607680, PROT_READ|PROT_WRITE, MAP_SHARED|MAP_LOCKED,
> > 3, 0) = 0x4e000000
> >
> > This is the strace output of sductest that is a test program to access
> > the shared memory that was setup by sducstart:
> > open("/pivot3/mem/sduc", O_RDWR) = 3
> > mmap2(NULL, 4194304, PROT_READ|PROT_WRITE, MAP_SHARED|MAP_LOCKED, 3,
> > 0) = -1 ENOMEM (Cannot allocate memory)
>
> Thanks a lot for the strace, that indeed helped to track it down.
>
> This has nothing to do with mlock or MAP_LOCKED - which by the way do
> make more sense in 2.6.15, since they provide a way of prefaulting the
> hugepage area like in earlier releases (now hugepages are being faulted
> in on demand, though never paged out, as Andrew said).
>
> Please try the patch below, and let us know if it works for you - thanks.
> Looks like we'll need this in 2.6.16-rc-git and 2.6.15-stable.
>
>
> 2.6.15's hugepage faulting introduced huge_pages_needed accounting into
> hugetlbfs: to count how many pages are already in cache, for spot check
> on how far a new mapping may be allowed to extend the file. But it's
> muddled: each hugepage found covers HPAGE_SIZE, not PAGE_SIZE. Once
> pages were already in cache, it would overshoot, wrap its hugepages
> count backwards, and so fail a harmless repeat mapping with -ENOMEM.
> Fixes the problem found by Don Dupuis.
>
> Signed-off-by: Hugh Dickins <[email protected]>
> ---
>
> fs/hugetlbfs/inode.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> --- 2.6.15/fs/hugetlbfs/inode.c 2006-01-03 03:21:10.000000000 +0000
> +++ linux/fs/hugetlbfs/inode.c 2006-01-23 18:39:47.000000000 +0000
> @@ -71,8 +71,8 @@ huge_pages_needed(struct address_space *
> unsigned long start = vma->vm_start;
> unsigned long end = vma->vm_end;
> unsigned long hugepages = (end - start) >> HPAGE_SHIFT;
> - pgoff_t next = vma->vm_pgoff;
> - pgoff_t endpg = next + ((end - start) >> PAGE_SHIFT);
> + pgoff_t next = vma->vm_pgoff >> (HPAGE_SHIFT - PAGE_SHIFT);
> + pgoff_t endpg = next + hugepages;
>
> pagevec_init(&pvec, 0);
> while (next < endpg) {
>
This patch fixed my problem.
Thanks very much
Don
-
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]
