The problem is that there are systems where cpu_possible_mask
has gaps between set CPUs, for example SPARC. In this scenario
addr_to_vb_xa() hash function can return an index which accesses
to not-possible and not setup CPU area using per_cpu() macro.
A per-cpu vmap_block_queue is also used as hash table, incorrectly
assuming the cpu_possible_mask has no gaps. Fix it by adjusting an
index to a next possible CPU.
v1 -> v2:
- update a commit message.
Fixes: 062eacf57ad9 ("mm: vmalloc: remove a global vmap_blocks xarray")
Reported-by: Nick Bowler <nbowler@xxxxxxxxxx>
Closes: https://lore.kernel.org/linux-kernel/ZntjIE6msJbF8zTa@MiWiFi-R3L-srv/T/
Signed-off-by: Uladzislau Rezki (Sony) <urezki@xxxxxxxxx>
---
mm/vmalloc.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
index b4c42da9f3901..6b783baf12a14 100644
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -2544,7 +2544,15 @@ static DEFINE_PER_CPU(struct vmap_block_queue, vmap_block_queue);
static struct xarray *
addr_to_vb_xa(unsigned long addr)
{
- int index = (addr / VMAP_BLOCK_SIZE) % num_possible_cpus();
+ int index = (addr / VMAP_BLOCK_SIZE) % nr_cpu_ids;
+
+ /*
+ * Please note, nr_cpu_ids points on a highest set
+ * possible bit, i.e. we never invoke cpumask_next()
+ * if an index points on it which is nr_cpu_ids - 1.
+ */
+ if (!cpu_possible(index))
+ index = cpumask_next(index, cpu_possible_mask);
return &per_cpu(vmap_block_queue, index).vmap_blocks;
}
--
2.39.2
