* Andrew Morton <[email protected]> wrote:
> On Tue, 24 Mar 2015 16:08:34 -0600 Toshi Kani <[email protected]> wrote:
>
> > This patchset enhances MTRR checks for the kernel huge I/O mapping,
> > which was enabled by the patchset below:
> > https://lkml.org/lkml/2015/3/3/589
> >
> > The following functional changes are made in patch 7/7.
> > - Allow pud_set_huge() and pmd_set_huge() to create a huge page
> > mapping to a range covered by a single MTRR entry of any memory
> > type.
> > - Log a pr_warn() message when a specified PMD map range spans more
> > than a single MTRR entry. Drivers should make a mapping request
> > aligned to a single MTRR entry when the range is covered by MTRRs.
> >
>
> OK, I grabbed these after barely looking at them, to get them a bit of
> runtime testing.
>
> I'll await guidance from the x86 maintainers regarding next steps?
Could you please send the current version of them over to us if your
testing didn't find any problems?
I'd like to take a final look and have them cook in the x86 tree as
well for a while and want to preserve your testing effort.
Thanks!
Ingo
On Fri, 2015-04-03 at 08:33 +0200, Ingo Molnar wrote:
> * Andrew Morton <[email protected]> wrote:
>
> > On Tue, 24 Mar 2015 16:08:34 -0600 Toshi Kani <[email protected]> wrote:
> >
> > > This patchset enhances MTRR checks for the kernel huge I/O mapping,
> > > which was enabled by the patchset below:
> > > https://lkml.org/lkml/2015/3/3/589
> > >
> > > The following functional changes are made in patch 7/7.
> > > - Allow pud_set_huge() and pmd_set_huge() to create a huge page
> > > mapping to a range covered by a single MTRR entry of any memory
> > > type.
> > > - Log a pr_warn() message when a specified PMD map range spans more
> > > than a single MTRR entry. Drivers should make a mapping request
> > > aligned to a single MTRR entry when the range is covered by MTRRs.
> > >
> >
> > OK, I grabbed these after barely looking at them, to get them a bit of
> > runtime testing.
> >
> > I'll await guidance from the x86 maintainers regarding next steps?
>
> Could you please send the current version of them over to us if your
> testing didn't find any problems?
>
> I'd like to take a final look and have them cook in the x86 tree as
> well for a while and want to preserve your testing effort.
This patchset is on top of the following patches in the -mm tree.
(Patches apply from the bottom to the top.)
2. Build error fixes and cleanups
http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-kva-mappings-on-x86-fix.patch
http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-vunmap-to-tear-down-huge-kva-mappings-fix.patch
http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-ioremap-to-set-up-huge-i-o-mappings-fix.patch
http://ozlabs.org/~akpm/mmotm/broken-out/lib-add-huge-i-o-map-capability-interfaces-fix.patch
1. Kernel huge I/O mapping support
http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-kva-mappings-on-x86.patch
http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-i-o-mapping-capability-i-f.patch
http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-vunmap-to-tear-down-huge-kva-mappings.patch
http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-ioremap-to-set-up-huge-i-o-mappings.patch
http://ozlabs.org/~akpm/mmotm/broken-out/lib-add-huge-i-o-map-capability-interfaces.patch
http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-__get_vm_area_node-to-use-fls_long.patch
Thanks,
-Toshi
On Fri, 2015-04-03 at 09:22 -0600, Toshi Kani wrote:
> On Fri, 2015-04-03 at 08:33 +0200, Ingo Molnar wrote:
> > * Andrew Morton <[email protected]> wrote:
> >
> > > On Tue, 24 Mar 2015 16:08:34 -0600 Toshi Kani <[email protected]> wrote:
> > >
> > > > This patchset enhances MTRR checks for the kernel huge I/O mapping,
> > > > which was enabled by the patchset below:
> > > > https://lkml.org/lkml/2015/3/3/589
> > > >
> > > > The following functional changes are made in patch 7/7.
> > > > - Allow pud_set_huge() and pmd_set_huge() to create a huge page
> > > > mapping to a range covered by a single MTRR entry of any memory
> > > > type.
> > > > - Log a pr_warn() message when a specified PMD map range spans more
> > > > than a single MTRR entry. Drivers should make a mapping request
> > > > aligned to a single MTRR entry when the range is covered by MTRRs.
> > > >
> > >
> > > OK, I grabbed these after barely looking at them, to get them a bit of
> > > runtime testing.
> > >
> > > I'll await guidance from the x86 maintainers regarding next steps?
> >
> > Could you please send the current version of them over to us if your
> > testing didn't find any problems?
> >
> > I'd like to take a final look and have them cook in the x86 tree as
> > well for a while and want to preserve your testing effort.
>
> This patchset is on top of the following patches in the -mm tree.
> (Patches apply from the bottom to the top.)
Ingo,
The following patches (2 got squashed to 1) went to 4.1-rc1, but this
patch-set is still sitting in the -mm tree. I confirmed that the
patch-set applies cleanly to 4.1-rc1. Please take a final look and let
me know if you have any comment.
Thanks,
-Toshi
> 2. Build error fixes and cleanups
> http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-kva-mappings-on-x86-fix.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-vunmap-to-tear-down-huge-kva-mappings-fix.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-ioremap-to-set-up-huge-i-o-mappings-fix.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/lib-add-huge-i-o-map-capability-interfaces-fix.patch
>
> 1. Kernel huge I/O mapping support
> http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-kva-mappings-on-x86.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/x86-mm-support-huge-i-o-mapping-capability-i-f.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-vunmap-to-tear-down-huge-kva-mappings.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-ioremap-to-set-up-huge-i-o-mappings.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/lib-add-huge-i-o-map-capability-interfaces.patch
> http://ozlabs.org/~akpm/mmotm/broken-out/mm-change-__get_vm_area_node-to-use-fls_long.patch
>
> Thanks,
> -Toshi
>