Re: [PATCH] mm: fix boundary checking in free_bootmem_core

["Yinghai Lu" <yhlu.kernel@gmail.com>]

On Wed, Mar 12, 2008 at 4:33 PM, Andrew Morton
<akpm@linux-foundation.org> wrote:
> On Wed, 12 Mar 2008 16:21:22 -0700
>
>
> "Yinghai Lu" <yhlu.kernel@gmail.com> wrote:
>
>  > On Tue, Mar 11, 2008 at 6:01 PM, Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>  > > [PATCH] mm: fix boundary checking in free_bootmem_core
>  > >
>  > >  so call it when numa is enabled, we don't know which node have that range.
>  > >  and make it more robust.
>  > >
>  > >  try to trim it to get valid sidx, and eidx.
>  > >
>  > >  Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
>  > >
>  >
>  > Ingo,
>  >
>  > this one need to be applied before
>  > [PATCH] x86_64: reserve dma32 early for gart
>  >
>
>  So perhaps we should merge
>  mm-fix-boundary-checking-in-free_bootmem_core.patch into 2.6.25?
also we can remove some NODE_DATA(0) assumpation on free_bootmem path.
>
>  <looks at it>
>
>  Sorry, but I find the changelog very hard to amke sense of.  I presently
>  have:
>
>
>   So call it when numa is enabled, we don't know which node have that
>   range.  and make it more robust.
>
>   Try to trim it to get valid sidx, and eidx.
>
>  Could you please expand on this?

please check following...

---
with numa enabled, some caller could have range on one node but try to
free that on other node.
that could cause some pages is freed wrongly.
For example: when we try to allocate 128g boot ram early for
gart/swiotlb, and free that range later so gart/swiotlb could get some
range afterwards. then with this patch, we don't need to care which
node is the range, just loop to call free_bootmem_node for all online
nodes.

this patch make free_bootmem_core more robust by trim the sidx and
eidx according the ram range that
node have.

----