Subject: Re: [patch] generic nonlinear mappings, 2.5.44-mm2-D0 - msg#00333

List: linux.kernel.mm

Date: Prev Next Index Thread: Prev Next Index

On Tue, 22 Oct 2002, Andrew Morton wrote:

> We seem to have lost a pte_page_unlock() from fremap.c:zap_pte()? I
> fixed up the ifdef tangle in there within the shpte-ng patch and then
> put the pte_page_unlock() back.

ok. I too fixed up the shpte #ifdef tangle in there as well, and it was
complex for no good reason, so i suspected that it was missing a line or
two.

> I also added a page_cache_release() to the error path in
> filemap_populate(), if install_page() failed.

hm, i somehow missed to add this, this was reported once already.

> The 2TB file size limit for mmap on non-PAE is a little worrisome. [...]

the limit is only there for 32-bit ptes on 32-bit platforms. 64-bit ptes
(both true 64-bit architectures and x86-PAE) has a ~64 zetabyte filesize
limit. I do not realistically believe that any 32-bit x86 box that is
connected to a larger than 2 TB disk array cannot possibly run a PAE
kernel. Just like you need PAE for more than 4 GB physical RAM. I find it
a bit worrisome that 32-bit x86 ptes can only support up to 4 GB of
physical RAM, but such is life :-)

Ingo

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/

Was this page helpful?

Thread at a glance:

Previous Message by Date: click to view message preview

Re: [PATCH 2.5.43-mm2] New shared page table patch

On Tue, 2002-10-22 at 20:03, Martin J. Bligh wrote: > > Can we delete the specialty syscalls now? > > I was lead to believe that Linus designed them, so he may be emotionally > attatched > to them, but I think there would be few others that would cry over the loss > ... You mean like the wonderfully pointless sys_readahead. The sooner these calls go the better. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/

Next Message by Date: click to view message preview

Re: [patch] generic nonlinear mappings, 2.5.44-mm2-D0

On Tue, 2002-10-22 at 21:27, Ingo Molnar wrote: > limit. I do not realistically believe that any 32-bit x86 box that is > connected to a larger than 2 TB disk array cannot possibly run a PAE > kernel. Just like you need PAE for more than 4 GB physical RAM. I find it > a bit worrisome that 32-bit x86 ptes can only support up to 4 GB of > physical RAM, but such is life :-) Actually I know a few. 2Tb is cheap - its one pci controller and eight ide disks. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/

Previous Message by Thread: click to view message preview

Re: [patch] generic nonlinear mappings, 2.5.44-mm2-D0

Christoph Hellwig wrote: > > On Tue, Oct 22, 2002 at 07:57:00PM +0200, Ingo Molnar wrote: > > the attached patch (ontop of 2.5.44-mm2) implements generic (swappable!) > > nonlinear mappings and sys_remap_file_pages() support. Ie. no more > > MAP_LOCKED restrictions and strange pagefault semantics. > > > > to implement this i added a new pte concept: "file pte's". This means that > > upon swapout, shared-named mappings do not get cleared but get converted > > into file pte's, which can then be decoded by the pagefault path and can > > be looked up in the pagecache. > > > > the normal linear pagefault path from now on does not assume linearity and > > decodes the offset in the pte. This also tests pte encoding/decoding in > > the pagecache case, and the ->populate functions. > > Ingo, > > what is the reason for that interface? It looks like a gross performance > hack for misdesigned applications to me, kindof windowsish.. > So that evicted pages in non-linear mappings can be reestablished at fault time by the kernel, rather than by delegation to userspace via SIGBUS. We seem to have lost a pte_page_unlock() from fremap.c:zap_pte()? I fixed up the ifdef tangle in there within the shpte-ng patch and then put the pte_page_unlock() back. I also added a page_cache_release() to the error path in filemap_populate(), if install_page() failed. The 2TB file size limit for mmap on non-PAE is a little worrisome. I wonder if we can only instantiate the pte_file() bit if the mapping is using MAP_POPULATE? Seems hard to do. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/

Next Message by Thread: click to view message preview

Loading Comments...