Messages in this thread | | | From | Mark Rustad <> | Subject | Re: 2.6.16 hugetlbfs problem | Date | Wed, 22 Mar 2006 17:32:57 -0600 |
| |
On Mar 22, 2006, at 4:13 PM, David S. Miller wrote:
> From: Mark Rustad <mrustad@mac.com> > Date: Wed, 22 Mar 2006 16:10:33 -0600 > >> I seem to be having trouble using hugetlbfs with kernel 2.6.16. I >> have a small test program that worked with 2.6.16-rc5, but fails with >> 2.6.16-rc6 or the release. The program is below. Given a path to a >> file on a hugetlbfs, it opens/creates the file, mmaps it and tries to >> access the first word. On 2.6.16-rc5, it works. On 2.6.16, it hangs >> page-faulting until it is killed. > > On what platform? Things like hugetlb and address space layout > (you're requesting a specific mmap() address I noticed) are very > platform specific.
This is on a Xeon, without PAE with the 1GB no-highmem memory map, in all three cases. This is a 32-bit kernel running on a Nacona CPU. I also had an unmap call over the range to be mmap-ed, but the failure/ success cases were the same, so I removed it to reduce the test program further.
-- Mark Rustad, MRustad@mac.com
- 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/
| |