Re: [patch 046/181] mm: remove sparsemem allocation details from thebootmem allocator

[Yinghai Lu]

On Sat, Jun 23, 2012 at 2:17 AM, Johannes Weiner <hannes@xxxxxxxxxxx> wrote:
> On Fri, Jun 22, 2012 at 07:05:45PM -0700, Yinghai Lu wrote:
>> On Fri, Jun 22, 2012 at 6:11 PM, Yinghai Lu <yinghai@xxxxxxxxxx> wrote:
>> > On Tue, May 29, 2012 at 3:06 PM,  <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
>> >> From: Johannes Weiner <hannes@xxxxxxxxxxx>
>> >> Subject: mm: remove sparsemem allocation details from the bootmem allocator
>> >>
>> >> alloc_bootmem_section() derives allocation area constraints from the
>> >> specified sparsemem section.  This is a bit specific for a generic memory
>> >> allocator like bootmem, though, so move it over to sparsemem.
>> >>
>> >> As __alloc_bootmem_node_nopanic() already retries failed allocations with
>> >> relaxed area constraints, the fallback code in sparsemem.c can be removed
>> >> and the code becomes a bit more compact overall.
>> >>
>> >> [akpm@xxxxxxxxxxxxxxxxxxxx: fix build]
>> >> Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
>> >> Acked-by: Tejun Heo <tj@xxxxxxxxxx>
>> >> Acked-by: David S. Miller <davem@xxxxxxxxxxxxx>
>> >> Cc: Yinghai Lu <yinghai@xxxxxxxxxx>
>> >> Cc: Gavin Shan <shangw@xxxxxxxxxxxxxxxxxx>
>> >> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
>> >
>> > hi, this one cause regression, will put usemap to last node's memory
>> > instead of each node.
>>
>> attached patch fixes the problem.
>
> Sorry for the trouble and thanks for the patch!  The number of bugs in
> these three lines is too damn high...
>
yeah, i should run Andrew -mm early.

Andrew, Do you have git tree for your -mm?
So I could merge it to my local tree.
now i only have tip, pci, scsi, net, driver-core, usb, tty in my local tree.

...
>>        */
>> -     goal = __pa(pgdat) & PAGE_SECTION_MASK;
>> +     goal = ((__pa(pgdat) >> PAGE_SHIFT) & PAGE_SECTION_MASK) << PAGE_SHIFT;
>
> How about
>
>        goal = __pa(pgdat) & (PAGE_SECTION_MASK << PAGE_SHIFT);
>
> instead?

yeah.

After closely looking, looks like we need to revert the commit or
apply -v2 patch.

old sequence is : try same section range at first.
this commit will try [start_of_section, end_of_same_node_range], so
could have chance to get

Yinghai

Attachment: fix_usemap_goal_v2.patch
Description: Binary data