Merge 6.3-rc5 into driver-core-next

We need the fixes in here for testing, as well as the driver core
changes for documentation updates to build on.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

8 files changed, 104 insertions, 49 deletions

diff --git a/lib/dhry_run.c b/lib/dhry_run.c
index f9d33efa6d09..f15ac666e9d3 100644
--- a/lib/dhry_run.c
+++ b/lib/dhry_run.c
@@ -31,6 +31,7 @@ MODULE_PARM_DESC(iterations,
 
 static void dhry_benchmark(void)
 {
+	unsigned int cpu = get_cpu();
 	int i, n;
 
 	if (iterations > 0) {
@@ -45,9 +46,10 @@ static void dhry_benchmark(void)
 	}
 
 report:
+	put_cpu();
 	if (n >= 0)
-		pr_info("CPU%u: Dhrystones per Second: %d (%d DMIPS)\n",
-			smp_processor_id(), n, n / DHRY_VAX);
+		pr_info("CPU%u: Dhrystones per Second: %d (%d DMIPS)\n", cpu,
+			n, n / DHRY_VAX);
 	else if (n == -EAGAIN)
 		pr_err("Please increase the number of iterations\n");
 	else
diff --git a/lib/find_bit.c b/lib/find_bit.c
index c10920e66788..32f99e9a670e 100644
--- a/lib/find_bit.c
+++ b/lib/find_bit.c
@@ -182,6 +182,15 @@ unsigned long _find_next_andnot_bit(const unsigned long *addr1, const unsigned l
 EXPORT_SYMBOL(_find_next_andnot_bit);
 #endif
 
+#ifndef find_next_or_bit
+unsigned long _find_next_or_bit(const unsigned long *addr1, const unsigned long *addr2,
+					unsigned long nbits, unsigned long start)
+{
+	return FIND_NEXT_BIT(addr1[idx] | addr2[idx], /* nop */, nbits, start);
+}
+EXPORT_SYMBOL(_find_next_or_bit);
+#endif
+
 #ifndef find_next_zero_bit
 unsigned long _find_next_zero_bit(const unsigned long *addr, unsigned long nbits,
 					 unsigned long start)
diff --git a/lib/maple_tree.c b/lib/maple_tree.c
index 646297cae5d1..9e2735cbc2b4 100644
--- a/lib/maple_tree.c
+++ b/lib/maple_tree.c
@@ -5099,35 +5099,21 @@ static inline bool mas_rewind_node(struct ma_state *mas)
  */
 static inline bool mas_skip_node(struct ma_state *mas)
 {
-	unsigned char slot, slot_count;
-	unsigned long *pivots;
-	enum maple_type mt;
+	if (mas_is_err(mas))
+		return false;
 
-	mt = mte_node_type(mas->node);
-	slot_count = mt_slots[mt] - 1;
 	do {
 		if (mte_is_root(mas->node)) {
-			slot = mas->offset;
-			if (slot > slot_count) {
+			if (mas->offset >= mas_data_end(mas)) {
 				mas_set_err(mas, -EBUSY);
 				return false;
 			}
 		} else {
 			mas_ascend(mas);
-			slot = mas->offset;
-			mt = mte_node_type(mas->node);
-			slot_count = mt_slots[mt] - 1;
 		}
-	} while (slot > slot_count);
-
-	mas->offset = ++slot;
-	pivots = ma_pivots(mas_mn(mas), mt);
-	if (slot > 0)
-		mas->min = pivots[slot - 1] + 1;
-
-	if (slot <= slot_count)
-		mas->max = pivots[slot];
+	} while (mas->offset >= mas_data_end(mas));
 
+	mas->offset++;
 	return true;
 }
 
diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c
index dba56c5c1837..5004463c4f9f 100644
--- a/lib/percpu_counter.c
+++ b/lib/percpu_counter.c
@@ -122,8 +122,19 @@ void percpu_counter_sync(struct percpu_counter *fbc)
 }
 EXPORT_SYMBOL(percpu_counter_sync);
 
-static s64 __percpu_counter_sum_mask(struct percpu_counter *fbc,
-			      const struct cpumask *cpu_mask)
+/*
+ * Add up all the per-cpu counts, return the result.  This is a more accurate
+ * but much slower version of percpu_counter_read_positive().
+ *
+ * We use the cpu mask of (cpu_online_mask | cpu_dying_mask) to capture sums
+ * from CPUs that are in the process of being taken offline. Dying cpus have
+ * been removed from the online mask, but may not have had the hotplug dead
+ * notifier called to fold the percpu count back into the global counter sum.
+ * By including dying CPUs in the iteration mask, we avoid this race condition
+ * so __percpu_counter_sum() just does the right thing when CPUs are being taken
+ * offline.
+ */
+s64 __percpu_counter_sum(struct percpu_counter *fbc)
 {
 	s64 ret;
 	int cpu;
@@ -131,35 +142,15 @@ static s64 __percpu_counter_sum_mask(struct percpu_counter *fbc,
 
 	raw_spin_lock_irqsave(&fbc->lock, flags);
 	ret = fbc->count;
-	for_each_cpu(cpu, cpu_mask) {
+	for_each_cpu_or(cpu, cpu_online_mask, cpu_dying_mask) {
 		s32 *pcount = per_cpu_ptr(fbc->counters, cpu);
 		ret += *pcount;
 	}
 	raw_spin_unlock_irqrestore(&fbc->lock, flags);
 	return ret;
 }
-
-/*
- * Add up all the per-cpu counts, return the result.  This is a more accurate
- * but much slower version of percpu_counter_read_positive()
- */
-s64 __percpu_counter_sum(struct percpu_counter *fbc)
-{
-	return __percpu_counter_sum_mask(fbc, cpu_online_mask);
-}
 EXPORT_SYMBOL(__percpu_counter_sum);
 
-/*
- * This is slower version of percpu_counter_sum as it traverses all possible
- * cpus. Use this only in the cases where accurate data is needed in the
- * presense of CPUs getting offlined.
- */
-s64 percpu_counter_sum_all(struct percpu_counter *fbc)
-{
-	return __percpu_counter_sum_mask(fbc, cpu_possible_mask);
-}
-EXPORT_SYMBOL(percpu_counter_sum_all);
-
 int __percpu_counter_init(struct percpu_counter *fbc, s64 amount, gfp_t gfp,
 			  struct lock_class_key *key)
 {
diff --git a/lib/test_maple_tree.c b/lib/test_maple_tree.c
index 3d19b1f78d71..f1db333270e9 100644
--- a/lib/test_maple_tree.c
+++ b/lib/test_maple_tree.c
@@ -2670,6 +2670,49 @@ static noinline void check_empty_area_window(struct maple_tree *mt)
 	rcu_read_unlock();
 }
 
+static noinline void check_empty_area_fill(struct maple_tree *mt)
+{
+	const unsigned long max = 0x25D78000;
+	unsigned long size;
+	int loop, shift;
+	MA_STATE(mas, mt, 0, 0);
+
+	mt_set_non_kernel(99999);
+	for (shift = 12; shift <= 16; shift++) {
+		loop = 5000;
+		size = 1 << shift;
+		while (loop--) {
+			mas_set(&mas, 0);
+			mas_lock(&mas);
+			MT_BUG_ON(mt, mas_empty_area(&mas, 0, max, size) != 0);
+			MT_BUG_ON(mt, mas.last != mas.index + size - 1);
+			mas_store_gfp(&mas, (void *)size, GFP_KERNEL);
+			mas_unlock(&mas);
+			mas_reset(&mas);
+		}
+	}
+
+	/* No space left. */
+	size = 0x1000;
+	rcu_read_lock();
+	MT_BUG_ON(mt, mas_empty_area(&mas, 0, max, size) != -EBUSY);
+	rcu_read_unlock();
+
+	/* Fill a depth 3 node to the maximum */
+	for (unsigned long i = 629440511; i <= 629440800; i += 6)
+		mtree_store_range(mt, i, i + 5, (void *)i, GFP_KERNEL);
+	/* Make space in the second-last depth 4 node */
+	mtree_erase(mt, 631668735);
+	/* Make space in the last depth 4 node */
+	mtree_erase(mt, 629506047);
+	mas_reset(&mas);
+	/* Search from just after the gap in the second-last depth 4 */
+	rcu_read_lock();
+	MT_BUG_ON(mt, mas_empty_area(&mas, 629506048, 690000000, 0x5000) != 0);
+	rcu_read_unlock();
+	mt_set_non_kernel(0);
+}
+
 static DEFINE_MTREE(tree);
 static int maple_tree_seed(void)
 {
@@ -2926,6 +2969,11 @@ static int maple_tree_seed(void)
 	check_empty_area_window(&tree);
 	mtree_destroy(&tree);
 
+	mt_init_flags(&tree, MT_FLAGS_ALLOC_RANGE);
+	check_empty_area_fill(&tree);
+	mtree_destroy(&tree);
+
+
 #if defined(BENCH)
 skip:
 #endif
diff --git a/lib/zstd/common/zstd_deps.h b/lib/zstd/common/zstd_deps.h
index 7a5bf44839c9..f06df065dec0 100644
--- a/lib/zstd/common/zstd_deps.h
+++ b/lib/zstd/common/zstd_deps.h
@@ -84,7 +84,7 @@ static uint64_t ZSTD_div64(uint64_t dividend, uint32_t divisor) {
 
 #include <linux/kernel.h>
 
-#define assert(x) WARN_ON((x))
+#define assert(x) WARN_ON(!(x))
 
 #endif /* ZSTD_DEPS_ASSERT */
 #endif /* ZSTD_DEPS_NEED_ASSERT */
diff --git a/lib/zstd/decompress/huf_decompress.c b/lib/zstd/decompress/huf_decompress.c
index 89b269a641c7..60958afebc41 100644
--- a/lib/zstd/decompress/huf_decompress.c
+++ b/lib/zstd/decompress/huf_decompress.c
@@ -985,7 +985,7 @@ static void HUF_fillDTableX2Level2(HUF_DEltX2* DTable, U32 targetLog, const U32
 
 static void HUF_fillDTableX2(HUF_DEltX2* DTable, const U32 targetLog,
                            const sortedSymbol_t* sortedList,
-                           const U32* rankStart, rankVal_t rankValOrigin, const U32 maxWeight,
+                           const U32* rankStart, rankValCol_t *rankValOrigin, const U32 maxWeight,
                            const U32 nbBitsBaseline)
 {
     U32* const rankVal = rankValOrigin[0];
diff --git a/lib/zstd/decompress/zstd_decompress.c b/lib/zstd/decompress/zstd_decompress.c
index b9b935a9f5c0..6b3177c94711 100644
--- a/lib/zstd/decompress/zstd_decompress.c
+++ b/lib/zstd/decompress/zstd_decompress.c
@@ -798,7 +798,7 @@ static size_t ZSTD_copyRawBlock(void* dst, size_t dstCapacity,
         if (srcSize == 0) return 0;
         RETURN_ERROR(dstBuffer_null, "");
     }
-    ZSTD_memcpy(dst, src, srcSize);
+    ZSTD_memmove(dst, src, srcSize);
     return srcSize;
 }
 
@@ -858,6 +858,7 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
 
     /* Loop on each block */
     while (1) {
+        BYTE* oBlockEnd = oend;
         size_t decodedSize;
         blockProperties_t blockProperties;
         size_t const cBlockSize = ZSTD_getcBlockSize(ip, remainingSrcSize, &blockProperties);
@@ -867,16 +868,34 @@ static size_t ZSTD_decompressFrame(ZSTD_DCtx* dctx,
         remainingSrcSize -= ZSTD_blockHeaderSize;
         RETURN_ERROR_IF(cBlockSize > remainingSrcSize, srcSize_wrong, "");
 
+        if (ip >= op && ip < oBlockEnd) {
+            /* We are decompressing in-place. Limit the output pointer so that we
+             * don't overwrite the block that we are currently reading. This will
+             * fail decompression if the input & output pointers aren't spaced
+             * far enough apart.
+             *
+             * This is important to set, even when the pointers are far enough
+             * apart, because ZSTD_decompressBlock_internal() can decide to store
+             * literals in the output buffer, after the block it is decompressing.
+             * Since we don't want anything to overwrite our input, we have to tell
+             * ZSTD_decompressBlock_internal to never write past ip.
+             *
+             * See ZSTD_allocateLiteralsBuffer() for reference.
+             */
+            oBlockEnd = op + (ip - op);
+        }
+
         switch(blockProperties.blockType)
         {
         case bt_compressed:
-            decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oend-op), ip, cBlockSize, /* frame */ 1, not_streaming);
+            decodedSize = ZSTD_decompressBlock_internal(dctx, op, (size_t)(oBlockEnd-op), ip, cBlockSize, /* frame */ 1, not_streaming);
             break;
         case bt_raw :
+            /* Use oend instead of oBlockEnd because this function is safe to overlap. It uses memmove. */
             decodedSize = ZSTD_copyRawBlock(op, (size_t)(oend-op), ip, cBlockSize);
             break;
         case bt_rle :
-            decodedSize = ZSTD_setRleBlock(op, (size_t)(oend-op), *ip, blockProperties.origSize);
+            decodedSize = ZSTD_setRleBlock(op, (size_t)(oBlockEnd-op), *ip, blockProperties.origSize);
             break;
         case bt_reserved :
         default: