12 files changed, 49 insertions, 19 deletions

diff --git a/lib/Kconfig b/lib/Kconfig
index a5ce0c7f6c30..54cf309a92a5 100644
--- a/lib/Kconfig
+++ b/lib/Kconfig
@@ -51,6 +51,9 @@ config PERCPU_RWSEM
 config ARCH_USE_CMPXCHG_LOCKREF
         bool
 
+config ARCH_HAS_FAST_MULTIPLIER
+        bool
+
 config CRC_CCITT
         tristate "CRC-CCITT functions"
         help
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 07c28323f88f..a28590083622 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -892,6 +892,10 @@ config DEBUG_WW_MUTEX_SLOWPATH
          the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this
          will test all possible w/w mutex interface abuse with the
          exception of simply not acquiring all the required locks.
+         Note that this feature can introduce significant overhead, so
+         it really should not be enabled in a production or distro kernel,
+         even a debug kernel.  If you are a driver writer, enable it.  If
+         you are a distro, do not.
 
 config DEBUG_LOCK_ALLOC
         bool "Lock debugging: detect incorrect freeing of live locks"
@@ -1032,8 +1036,13 @@ config TRACE_IRQFLAGS
           either tracing or lock debugging.
 
 config STACKTRACE
-        bool
+        bool "Stack backtrace support"
         depends on STACKTRACE_SUPPORT
+        help
+          This option causes the kernel to create a /proc/pid/stack for
+          every process, showing its current stack trace.
+          It is also used by various kernel debugging features that require
+          stack trace generation.
 
 config DEBUG_KOBJECT
         bool "kobject debugging"
diff --git a/lib/assoc_array.c b/lib/assoc_array.c
index c0b1007011e1..2404d03e251a 100644
--- a/lib/assoc_array.c
+++ b/lib/assoc_array.c
@@ -1723,11 +1723,13 @@ ascend_old_tree:
                 shortcut = assoc_array_ptr_to_shortcut(ptr);
                 slot = shortcut->parent_slot;
                 cursor = shortcut->back_pointer;
+                if (!cursor)
+                        goto gc_complete;
         } else {
                 slot = node->parent_slot;
                 cursor = ptr;
         }
-        BUG_ON(!ptr);
+        BUG_ON(!cursor);
         node = assoc_array_ptr_to_node(cursor);
         slot++;
         goto continue_node;
@@ -1735,7 +1737,7 @@ ascend_old_tree:
 gc_complete:
         edit->set[0].to = new_root;
         assoc_array_apply_edit(edit);
-        edit->array->nr_leaves_on_tree = nr_leaves_on_tree;
+        array->nr_leaves_on_tree = nr_leaves_on_tree;
         return 0;
 
 enomem:
diff --git a/lib/bitmap.c b/lib/bitmap.c
index 1e031f2c9aba..cd250a2e14cb 100644
--- a/lib/bitmap.c
+++ b/lib/bitmap.c
@@ -884,7 +884,7 @@ EXPORT_SYMBOL(bitmap_bitremap);
  * read it, you're overqualified for your current job.)
  *
  * In other words, @orig is mapped onto (surjectively) @dst,
- * using the the map { <n, m> | the n-th bit of @relmap is the
+ * using the map { <n, m> | the n-th bit of @relmap is the
  * m-th set bit of @relmap }.
  *
  * Any set bits in @orig above bit number W, where W is the
@@ -932,7 +932,7 @@ EXPORT_SYMBOL(bitmap_bitremap);
  *
  *  Further lets say we use the following code, invoking
  *  bitmap_fold() then bitmap_onto, as suggested above to
- *  avoid the possitility of an empty @dst result:
+ *  avoid the possibility of an empty @dst result:
  *
  *      unsigned long *tmp;     // a temporary bitmap's bits
  *
diff --git a/lib/dma-debug.c b/lib/dma-debug.c
index 98f2d7e91a91..add80cc02dbe 100644
--- a/lib/dma-debug.c
+++ b/lib/dma-debug.c
@@ -1149,7 +1149,7 @@ static void check_unmap(struct dma_debug_entry *ref)
 static void check_for_stack(struct device *dev, void *addr)
 {
         if (object_is_on_stack(addr))
-                err_printk(dev, NULL, "DMA-API: device driver maps memory from"
+                err_printk(dev, NULL, "DMA-API: device driver maps memory from "
                                 "stack [addr=%p]\n", addr);
 }
 
diff --git a/lib/genalloc.c b/lib/genalloc.c
index bdb9a456bcbb..38d2db82228c 100644
--- a/lib/genalloc.c
+++ b/lib/genalloc.c
@@ -588,6 +588,7 @@ struct gen_pool *of_get_named_gen_pool(struct device_node *np,
         if (!np_pool)
                 return NULL;
         pdev = of_find_device_by_node(np_pool);
+        of_node_put(np_pool);
         if (!pdev)
                 return NULL;
         return dev_get_gen_pool(&pdev->dev);
diff --git a/lib/hweight.c b/lib/hweight.c
index b7d81ba143d1..9a5c1f221558 100644
--- a/lib/hweight.c
+++ b/lib/hweight.c
@@ -11,7 +11,7 @@
 
 unsigned int __sw_hweight32(unsigned int w)
 {
-#ifdef ARCH_HAS_FAST_MULTIPLIER
+#ifdef CONFIG_ARCH_HAS_FAST_MULTIPLIER
         w -= (w >> 1) & 0x55555555;
         w =  (w & 0x33333333) + ((w >> 2) & 0x33333333);
         w =  (w + (w >> 4)) & 0x0f0f0f0f;
@@ -49,7 +49,7 @@ unsigned long __sw_hweight64(__u64 w)
         return __sw_hweight32((unsigned int)(w >> 32)) +
                __sw_hweight32((unsigned int)w);
 #elif BITS_PER_LONG == 64
-#ifdef ARCH_HAS_FAST_MULTIPLIER
+#ifdef CONFIG_ARCH_HAS_FAST_MULTIPLIER
         w -= (w >> 1) & 0x5555555555555555ul;
         w =  (w & 0x3333333333333333ul) + ((w >> 2) & 0x3333333333333333ul);
         w =  (w + (w >> 4)) & 0x0f0f0f0f0f0f0f0ful;
diff --git a/lib/idr.c b/lib/idr.c
index 50be3fa9b657..e654aebd5f80 100644
--- a/lib/idr.c
+++ b/lib/idr.c
@@ -626,7 +626,7 @@ static void __idr_remove_all(struct idr *idp)
  * idr_destroy().
  *
  * A typical clean-up sequence for objects stored in an idr tree will use
- * idr_for_each() to free all objects, if necessay, then idr_destroy() to
+ * idr_for_each() to free all objects, if necessary, then idr_destroy() to
  * free up the id mappings and cached idr_layers.
  */
 void idr_destroy(struct idr *idp)
diff --git a/lib/percpu-refcount.c b/lib/percpu-refcount.c
index fe5a3342e960..a89cf09a8268 100644
--- a/lib/percpu-refcount.c
+++ b/lib/percpu-refcount.c
@@ -184,3 +184,19 @@ void percpu_ref_kill_and_confirm(struct percpu_ref *ref,
         call_rcu_sched(&ref->rcu, percpu_ref_kill_rcu);
 }
 EXPORT_SYMBOL_GPL(percpu_ref_kill_and_confirm);
+
+/*
+ * XXX: Temporary kludge to work around SCSI blk-mq stall.  Used only by
+ * block/blk-mq.c::blk_mq_freeze_queue().  Will be removed during v3.18
+ * devel cycle.  Do not use anywhere else.
+ */
+void __percpu_ref_kill_expedited(struct percpu_ref *ref)
+{
+        WARN_ONCE(ref->pcpu_count_ptr & PCPU_REF_DEAD,
+                  "percpu_ref_kill() called more than once on %pf!",
+                  ref->release);
+
+        ref->pcpu_count_ptr |= PCPU_REF_DEAD;
+        synchronize_sched_expedited();
+        percpu_ref_kill_rcu(&ref->rcu);
+}
diff --git a/lib/rhashtable.c b/lib/rhashtable.c
index a2c78810ebc1..3d2b4733f6cb 100644
--- a/lib/rhashtable.c
+++ b/lib/rhashtable.c
@@ -23,7 +23,6 @@
 #include <linux/hash.h>
 #include <linux/random.h>
 #include <linux/rhashtable.h>
-#include <linux/log2.h>
 
 #define HASH_DEFAULT_SIZE       64UL
 #define HASH_MIN_SIZE           4UL
@@ -55,7 +54,7 @@ static u32 __hashfn(const struct rhashtable *ht, const void *key,
 
 /**
  * rhashtable_hashfn - compute hash for key of given length
- * @ht:         hash table to compuate for
+ * @ht:         hash table to compute for
  * @key:        pointer to key
  * @len:        length of key
  *
@@ -86,7 +85,7 @@ static u32 obj_hashfn(const struct rhashtable *ht, const void *ptr, u32 hsize)
 
 /**
  * rhashtable_obj_hashfn - compute hash for hashed object
- * @ht:         hash table to compuate for
+ * @ht:         hash table to compute for
  * @ptr:        pointer to hashed object
  *
  * Computes the hash value using the hash function `hashfn` respectively
@@ -589,13 +588,13 @@ EXPORT_SYMBOL_GPL(rhashtable_init);
  * rhashtable_destroy - destroy hash table
  * @ht:         the hash table to destroy
  *
- * Frees the bucket array.
+ * Frees the bucket array. This function is not rcu safe, therefore the caller
+ * has to make sure that no resizing may happen by unpublishing the hashtable
+ * and waiting for the quiescent cycle before releasing the bucket array.
  */
 void rhashtable_destroy(const struct rhashtable *ht)
 {
-        const struct bucket_table *tbl = rht_dereference(ht->tbl, ht);
-
-        bucket_table_free(tbl);
+        bucket_table_free(ht->tbl);
 }
 EXPORT_SYMBOL_GPL(rhashtable_destroy);
 
diff --git a/lib/string.c b/lib/string.c
index 992bf30af759..f3c6ff596414 100644
--- a/lib/string.c
+++ b/lib/string.c
@@ -807,9 +807,9 @@ void *memchr_inv(const void *start, int c, size_t bytes)
                 return check_bytes8(start, value, bytes);
 
         value64 = value;
-#if defined(ARCH_HAS_FAST_MULTIPLIER) && BITS_PER_LONG == 64
+#if defined(CONFIG_ARCH_HAS_FAST_MULTIPLIER) && BITS_PER_LONG == 64
         value64 *= 0x0101010101010101;
-#elif defined(ARCH_HAS_FAST_MULTIPLIER)
+#elif defined(CONFIG_ARCH_HAS_FAST_MULTIPLIER)
         value64 *= 0x01010101;
         value64 |= value64 << 32;
 #else
diff --git a/lib/vsprintf.c b/lib/vsprintf.c
index 6fe2c84eb055..ba3cd0a35640 100644
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -1937,7 +1937,7 @@ EXPORT_SYMBOL(sprintf);
  * @args: Arguments for the format string
  *
  * The format follows C99 vsnprintf, except %n is ignored, and its argument
- * is skiped.
+ * is skipped.
  *
  * The return value is the number of words(32bits) which would be generated for
  * the given input.