The RT patch "mm-disable-slab-on-rt.patch" unconditionally
converts kmem_cache_node's list_lock into a raw lock. As
of mainline commit ca34956b804b7554fc4e88826773380d9d5122a8
("slab: Common definition for kmem_cache_node") the definition
is shared -- but slab.c still thinks the lock is non raw.
At the moment SLAB depends on !RT_FULL, however with the lock
being raw for !RT_FULL, we can't build the SLAB + !RT_FULL
combination because of the above. So only convert the lock if
SLAB is not enabled.
Signed-off-by: Paul Gortmaker <paul.gortmaker@xxxxxxxxxxxxx>
---
[Should be squished into mm-disable-slab-on-rt.patch ]
mm/slab.h | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/mm/slab.h b/mm/slab.h
index 2e6c8b7..fc3c097 100644
--- a/mm/slab.h
+++ b/mm/slab.h
@@ -247,7 +247,11 @@ static inline struct kmem_cache *cache_from_obj(struct kmem_cache *s, void *x)
* The slab lists for all objects.
*/
struct kmem_cache_node {
+#ifdef CONFIG_SLAB
+ spinlock_t list_lock;
+#else
raw_spinlock_t list_lock;
+#endif
#ifdef CONFIG_SLAB
struct list_head slabs_partial; /* partial list first, better asm code */
--
1.8.1.2
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
