Ack: [PATCH 0/2] [SRU] [SAUCE] [MNO/ti-omap4 and L/fsl-imx51] large mmap() support

[Seth Forshee]

On Fri, Dec 02, 2011 at 05:30:32PM +0100, Paolo Pisati wrote:
> BugLink: http://bugs.launchpad.net/bugs/861296
> 
> SRU Justification:
> 
> Impact: 
> 
> Mmap() fails to allocate large chunks of contiguos memory due to limitations in the allocation strategy.
> 
> Testcase:
> 
> See mmap-test.c and mmap-test-slices.c attached to the LP bug.
> 
> Fix:
> 
> The patches are not mainline yet but entered rmk patch system (and thus should be in the 3.3 merge window).
> (http://www.arm.linux.org.uk/developer/patches/viewpatch.php?id=7169/1)

The fact that they're in the patch system doesn't necessarily mean that
they'll be merged, since afaict Russell hasn't yet weighed in on them.
Still, they have good acks and have been tested in the Linaro tree for a
couple of weeks, so they seem likely to make it.

The first one looks perfectly reasonable to me. I don't feel fully
qualified to comment on the correctness of the second patch, but I don't
see any deal-breakers. The backports also look reasonable, except that
for lucid patch 2 should precede patch 3 to ensure that every commit can
build. Assuming that is done:

Acked-by: Seth Forshee <seth.forshee at canonical.com>

> 
> Natty and Oneiric applied as-is, while Maverick needed some mechanical modifications.
> Lucid required some backporting (no address randomization there) and some helper functions.
> 
> Kernels with the aforementioned patches were tested and confirmed to fix the problem.
> 
> Rob Herring (2):
>   use cache type functions for arch_get_unmapped_area
>   topdown mmap support
> 
>  arch/arm/include/asm/pgtable.h   |    1 +
>  arch/arm/include/asm/processor.h |    2 +
>  arch/arm/mm/mmap.c               |  196 +++++++++++++++++++++++++++++++++----
>  3 files changed, 177 insertions(+), 22 deletions(-)
> 
> -- 
> 1.7.5.4
> 
> 
> -- 
> kernel-team mailing list
> kernel-team at lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team