Re: [PATCH 10/14] s390/sclp_cmd: Replace page allocator calls with kzalloc()

From: Heiko Carstens

Date: Mon Nov 16 2026 - 12:24:39 EST


On Sun, Nov 08, 2026 at 10:58:41AM +0200, Mike Rapoport (Microsoft) wrote:
> _sclp_get_core_info(), do_chp_configure() and sclp_chp_read_info()
> allocate the SCCBs used to read the CPU information, to configure a
> channel path and to read the channel path information.
>
> These buffers can be allocated with kmalloc() as there's nothing special
> about them to go directly to the page allocator.
>
> kmalloc() provides a better API that does not require ugly casts and
> kfree() does not need to know the size of the freed object.
>
> Performance difference between kmalloc() and __get_free_pages() is not
> measurable as both allocators take an object/page from a per-CPU list for
> fast path allocations.
>
> For the slow path the performance is anyway determined by the amount of
> reclaim involved rather than by what allocator is used.
>
> While on it, allocate only the amount of memory the SCCB actually
> requires in _sclp_get_core_info() rather than rounding it up to the next
> power of two. An SCCB must be page aligned and kmalloc() guarantees that
> the alignment of an allocation is at least its largest power of two
> divisor, so the extended SCCB is still page aligned.
>
> Replace use of __get_free_pages() and get_zeroed_page() with kzalloc()
> and free_page[s]() with kfree().
>
> Link: https://lore.kernel.org/all/635405e4-9423-4a25-a6e7-e03c8ea0bcbe@xxxxxxxxxx
> Signed-off-by: Mike Rapoport (Microsoft) <rppt@xxxxxxxxxx>
> ---
> drivers/s390/char/sclp_cmd.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/s390/char/sclp_cmd.c b/drivers/s390/char/sclp_cmd.c
> index 1dc93d945be6f..6aaf1f6a3633f 100644
> --- a/drivers/s390/char/sclp_cmd.c
> +++ b/drivers/s390/char/sclp_cmd.c
> @@ -102,7 +102,7 @@ int _sclp_get_core_info(struct sclp_core_info *info)
> return -EOPNOTSUPP;
>
> length = test_facility(140) ? EXT_SCCB_READ_CPU : PAGE_SIZE;
> - sccb = (void *)__get_free_pages(GFP_KERNEL | GFP_DMA | __GFP_ZERO, get_order(length));
> + sccb = kzalloc(length, GFP_KERNEL | GFP_DMA);

It is a bit subtle that this really works. EXT_SCCB_READ_CPU is 3 * PAGE_SIZE,
so we end up with a guaranteed 4k alignment from the slab cache, which is
indeed sufficient here. But this is really not obvious, and not as stable as
before, IMHO.

I'd prefer to have this changed to:

sccb = kzalloc(PAGE_SIZE << get_order(length), GFP_KERNEL | GFP_DMA);

which comes with the same alignment guarantees as before.