On October 5, 2018 6:25:38 AM GMT+03:00, Benjamin Herrenschmidt <benh@xxxxxxxxxxxxxxxxxxx> wrote:
On Fri, 2018-10-05 at 00:07 +0300, Mike Rapoport wrote:
When a memblock allocation APIs are called with align = 0, the
alignment is
implicitly set to SMP_CACHE_BYTES.
Replace all such uses of memblock APIs with the 'align' parameter
explicitly
set to SMP_CACHE_BYTES and stop implicit alignment assignment in the
memblock internal allocation functions.
For the case when memblock APIs are used via helper functions, e.g.
like
iommu_arena_new_node() in Alpha, the helper functions were detected
with
Coccinelle's help and then manually examined and updated where
appropriate.
The direct memblock APIs users were updated using the semantic patch
below:
What is the purpose of this ? It sounds rather counter-intuitive...
Why?
I think it actually more intuitive to explicitly set alignment to SMP_CACHE_BYTES rather than use align = 0 because deeply inside allocator it will be implicitly reset to SMP_CACHE_BYTES...
Ben.
--
Sincerely yours,
Mike.
