From: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Since gfp & GFP_ATOMIC == GFP_ATOMIC is true for GFP_KERNEL | GFP_HIGH,
it will use kmalloc if user specifies that combination. Here the reason
why combining the __vmalloc_node() and kmalloc_node() is that the vmalloc
does not support all GFP flag, especially GFP_ATOMIC. So we should check
if gfp & (GFP_ATOMIC | GFP_KERNEL) != GFP_ATOMIC for vmalloc first. This
ensures caller can sleep.
And for the robustness, even if vmalloc fails, it should retry with
kmalloc to allocate it.
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Closes: https://lore.kernel.org/all/CAHk-=whO+vSH+XVRio8byJU8idAWES0SPGVZ7KAVdc4qrV0VUA@mail.gmail.com/
Fixes: aff1871bfc81 ("objpool: fix choosing allocation for percpu slots")
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
---
lib/objpool.c | 18 ++++++++++++------
1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/lib/objpool.c b/lib/objpool.c
index fd108fe0d095..b998b720c732 100644
--- a/lib/objpool.c
+++ b/lib/objpool.c
@@ -74,15 +74,21 @@ objpool_init_percpu_slots(struct objpool_head *pool, int nr_objs,
* warm caches and TLB hits. in default vmalloc is used to
* reduce the pressure of kernel slab system. as we know,
* mimimal size of vmalloc is one page since vmalloc would
- * always align the requested size to page size
+ * always align the requested size to page size.
+ * but if vmalloc fails or it is not available (e.g. GFP_ATOMIC)
+ * allocate percpu slot with kmalloc.
*/
- if ((pool->gfp & GFP_ATOMIC) == GFP_ATOMIC)
- slot = kmalloc_node(size, pool->gfp, cpu_to_node(i));
- else
+ slot = NULL;
+
+ if ((pool->gfp & (GFP_ATOMIC | GFP_KERNEL)) != GFP_ATOMIC)
slot = __vmalloc_node(size, sizeof(void *), pool->gfp,
cpu_to_node(i), __builtin_return_address(0));
- if (!slot)
- return -ENOMEM;
+
+ if (!slot) {
+ slot = kmalloc_node(size, pool->gfp, cpu_to_node(i));
+ if (!slot)
+ return -ENOMEM;
+ }
memset(slot, 0, size);
pool->cpu_slots[i] = slot;