crush: remove mutable part of CRUSH map

Then add it to the working state. It would be very nice if we didn't
have to take a lock to calculate a crush placement. By moving the
permutation array into the working data, we can treat the CRUSH map as
immutable.

Reflects ceph.git commit cbcd039651c0569551cb90d26ce27e1432671f2a.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

3 files changed, 38 insertions, 8 deletions

diff --git a/include/linux/ceph/osdmap.h b/include/linux/ceph/osdmap.h
index 412906609954..cef1cab789b9 100644
--- a/include/linux/ceph/osdmap.h
+++ b/include/linux/ceph/osdmap.h
@@ -175,6 +175,7 @@ struct ceph_osdmap {
 
 	struct mutex crush_scratch_mutex;
 	int crush_scratch_ary[CEPH_PG_MAX_SIZE * 3];
+	void *crush_workspace;
 };
 
 static inline bool ceph_osd_exists(struct ceph_osdmap *map, int osd)
diff --git a/include/linux/crush/crush.h b/include/linux/crush/crush.h
index be8f12b8f195..fbecbd089d75 100644
--- a/include/linux/crush/crush.h
+++ b/include/linux/crush/crush.h
@@ -135,13 +135,6 @@ struct crush_bucket {
 	__u32 size;      /* num items */
 	__s32 *items;
 
-	/*
-	 * cached random permutation: used for uniform bucket and for
-	 * the linear search fallback for the other bucket types.
-	 */
-	__u32 perm_x;  /* @x for which *perm is defined */
-	__u32 perm_n;  /* num elements of *perm that are permuted/defined */
-	__u32 *perm;
 };
 
 struct crush_bucket_uniform {
@@ -211,6 +204,21 @@ struct crush_map {
 	 * device fails. */
 	__u8 chooseleaf_stable;
 
+	/*
+	 * This value is calculated after decode or construction by
+	 * the builder. It is exposed here (rather than having a
+	 * 'build CRUSH working space' function) so that callers can
+	 * reserve a static buffer, allocate space on the stack, or
+	 * otherwise avoid calling into the heap allocator if they
+	 * want to. The size of the working space depends on the map,
+	 * while the size of the scratch vector passed to the mapper
+	 * depends on the size of the desired result set.
+	 *
+	 * Nothing stops the caller from allocating both in one swell
+	 * foop and passing in two points, though.
+	 */
+	size_t working_size;
+
 #ifndef __KERNEL__
 	/*
 	 * version 0 (original) of straw_calc has various flaws.  version 1
@@ -248,4 +256,23 @@ static inline int crush_calc_tree_node(int i)
 	return ((i+1) << 1)-1;
 }
 
+/*
+ * These data structures are private to the CRUSH implementation. They
+ * are exposed in this header file because builder needs their
+ * definitions to calculate the total working size.
+ *
+ * Moving this out of the crush map allow us to treat the CRUSH map as
+ * immutable within the mapper and removes the requirement for a CRUSH
+ * map lock.
+ */
+struct crush_work_bucket {
+	__u32 perm_x; /* @x for which *perm is defined */
+	__u32 perm_n; /* num elements of *perm that are permuted/defined */
+	__u32 *perm;  /* Permutation of the bucket's items */
+};
+
+struct crush_work {
+	struct crush_work_bucket **work; /* Per-bucket working store */
+};
+
 #endif
diff --git a/include/linux/crush/mapper.h b/include/linux/crush/mapper.h
index 5dfd5b1125d2..3303c7fd8a31 100644
--- a/include/linux/crush/mapper.h
+++ b/include/linux/crush/mapper.h
@@ -15,6 +15,8 @@ extern int crush_do_rule(const struct crush_map *map,
 			 int ruleno,
 			 int x, int *result, int result_max,
 			 const __u32 *weights, int weight_max,
-			 int *scratch);
+			 void *cwin, int *scratch);
+
+void crush_init_workspace(const struct crush_map *map, void *v);
 
 #endif