Merge tag 'bitmap-6.1-rc1' of https://github.com/norov/linux

Pull bitmap updates from Yury Norov:

 - Fix unsigned comparison to -1 in CPUMAP_FILE_MAX_BYTES (Phil Auld)

 - cleanup nr_cpu_ids vs nr_cpumask_bits mess (me)

   This series cleans that mess and adds new config FORCE_NR_CPUS that
   allows to optimize cpumask subsystem if the number of CPUs is known
   at compile-time.

 - optimize find_bit() functions (me)

   Reworks find_bit() functions based on new FIND_{FIRST,NEXT}_BIT()
   macros.

 - add find_nth_bit() (me)

   Adds find_nth_bit(), which is ~70 times faster than bitcounting with
   for_each() loop:

	for_each_set_bit(bit, mask, size)
		if (n-- == 0)
			return bit;

   Also adds bitmap_weight_and() to let people replace this pattern:

	tmp = bitmap_alloc(nbits);
	bitmap_and(tmp, map1, map2, nbits);
	weight = bitmap_weight(tmp, nbits);
	bitmap_free(tmp);

   with a single bitmap_weight_and() call.

 - repair cpumask_check() (me)

   After switching cpumask to use nr_cpu_ids, cpumask_check() started
   generating many false-positive warnings. This series fixes it.

 - Add for_each_cpu_andnot() and for_each_cpu_andnot() (Valentin
   Schneider)

   Extends the API with one more function and applies it in sched/core.

* tag 'bitmap-6.1-rc1' of https://github.com/norov/linux: (28 commits)
  sched/core: Merge cpumask_andnot()+for_each_cpu() into for_each_cpu_andnot()
  lib/test_cpumask: Add for_each_cpu_and(not) tests
  cpumask: Introduce for_each_cpu_andnot()
  lib/find_bit: Introduce find_next_andnot_bit()
  cpumask: fix checking valid cpu range
  lib/bitmap: add tests for for_each() loops
  lib/find: optimize for_each() macros
  lib/bitmap: introduce for_each_set_bit_wrap() macro
  lib/find_bit: add find_next{,_and}_bit_wrap
  cpumask: switch for_each_cpu{,_not} to use for_each_bit()
  net: fix cpu_max_bits_warn() usage in netif_attrmask_next{,_and}
  cpumask: add cpumask_nth_{,and,andnot}
  lib/bitmap: remove bitmap_ord_to_pos
  lib/bitmap: add tests for find_nth_bit()
  lib: add find_nth{,_and,_andnot}_bit()
  lib/bitmap: add bitmap_weight_and()
  lib/bitmap: don't call __bitmap_weight() in kernel code
  tools: sync find_bit() implementation
  lib/find_bit: optimize find_next_bit() functions
  lib/find_bit: create find_first_zero_bit_le()
  ...

arch/loongarch/kernel/setup.c

@@ -336,7 +336,7 @@ static void __init prefill_possible_map(void)
 	for (; i < NR_CPUS; i++)
 		set_cpu_possible(i, false);
 
-	nr_cpu_ids = possible;
+	set_nr_cpu_ids(possible);
 }
 #endif
 

arch/mips/kernel/setup.c

@@ -751,7 +751,7 @@ static void __init prefill_possible_map(void)
 	for (; i < NR_CPUS; i++)
 		set_cpu_possible(i, false);
 
-	nr_cpu_ids = possible;
+	set_nr_cpu_ids(possible);
 }
 #else
 static inline void prefill_possible_map(void) {}

arch/powerpc/kernel/head_64.S

@@ -393,8 +393,12 @@ generic_secondary_common_init:
 #else
 	LOAD_REG_ADDR(r8, paca_ptrs)	/* Load paca_ptrs pointe	 */
 	ld	r8,0(r8)		/* Get base vaddr of array	 */
+#if (NR_CPUS == 1) || defined(CONFIG_FORCE_NR_CPUS)
+	LOAD_REG_IMMEDIATE(r7, NR_CPUS)
+#else
 	LOAD_REG_ADDR(r7, nr_cpu_ids)	/* Load nr_cpu_ids address       */
 	lwz	r7,0(r7)		/* also the max paca allocated 	 */
+#endif
 	li	r5,0			/* logical cpu id                */
 1:
 	sldi	r9,r5,3			/* get paca_ptrs[] index from cpu id */

arch/x86/kernel/smpboot.c

@@ -1316,7 +1316,7 @@ static void __init smp_sanity_check(void)
 			nr++;
 		}
 
-		nr_cpu_ids = 8;
+		set_nr_cpu_ids(8);
 	}
 #endif
 
@@ -1569,7 +1569,7 @@ __init void prefill_possible_map(void)
 		possible = i;
 	}
 
-	nr_cpu_ids = possible;
+	set_nr_cpu_ids(possible);
 
 	pr_info("Allowing %d CPUs, %d hotplug CPUs\n",
 		possible, max_t(int, possible - num_processors, 0));

arch/x86/xen/smp_pv.c

@@ -179,7 +179,7 @@ static void __init _get_smp_config(unsigned int early)
 	 * hypercall to expand the max number of VCPUs an already
 	 * running guest has. So cap it up to X. */
 	if (subtract)
-		nr_cpu_ids = nr_cpu_ids - subtract;
+		set_nr_cpu_ids(nr_cpu_ids - subtract);
 #endif
 
 }

fs/ntfs3/bitmap.c

@@ -560,7 +560,7 @@ static int wnd_rescan(struct wnd_bitmap *wnd)
 
 		buf = (ulong *)bh->b_data;
 
-		used = __bitmap_weight(buf, wbits);
+		used = bitmap_weight(buf, wbits);
 		if (used < wbits) {
 			frb = wbits - used;
 			wnd->free_bits[iw] = frb;
@@ -1364,7 +1364,7 @@ int wnd_extend(struct wnd_bitmap *wnd, size_t new_bits)
 		buf = (ulong *)bh->b_data;
 
 		__bitmap_clear(buf, b0, blocksize * 8 - b0);
-		frb = wbits - __bitmap_weight(buf, wbits);
+		frb = wbits - bitmap_weight(buf, wbits);
 		wnd->total_zeroes += frb - wnd->free_bits[iw];
 		wnd->free_bits[iw] = frb;
 

include/linux/bitmap.h

@@ -51,6 +51,7 @@ struct device;
  *  bitmap_empty(src, nbits)                    Are all bits zero in *src?
  *  bitmap_full(src, nbits)                     Are all bits set in *src?
  *  bitmap_weight(src, nbits)                   Hamming Weight: number set bits
+ *  bitmap_weight_and(src1, src2, nbits)        Hamming Weight of and'ed bitmap
  *  bitmap_set(dst, pos, nbits)                 Set specified bit area
  *  bitmap_clear(dst, pos, nbits)               Clear specified bit area
  *  bitmap_find_next_zero_area(buf, len, pos, n, mask)  Find bit free area
@@ -164,6 +165,8 @@ bool __bitmap_intersects(const unsigned long *bitmap1,
 bool __bitmap_subset(const unsigned long *bitmap1,
 		     const unsigned long *bitmap2, unsigned int nbits);
 unsigned int __bitmap_weight(const unsigned long *bitmap, unsigned int nbits);
+unsigned int __bitmap_weight_and(const unsigned long *bitmap1,
+				 const unsigned long *bitmap2, unsigned int nbits);
 void __bitmap_set(unsigned long *map, unsigned int start, int len);
 void __bitmap_clear(unsigned long *map, unsigned int start, int len);
 
@@ -222,7 +225,6 @@ void bitmap_copy_le(unsigned long *dst, const unsigned long *src, unsigned int n
 #else
 #define bitmap_copy_le bitmap_copy
 #endif
-unsigned int bitmap_ord_to_pos(const unsigned long *bitmap, unsigned int ord, unsigned int nbits);
 int bitmap_print_to_pagebuf(bool list, char *buf,
 				   const unsigned long *maskp, int nmaskbits);
 
@@ -439,6 +441,15 @@ unsigned int bitmap_weight(const unsigned long *src, unsigned int nbits)
 	return __bitmap_weight(src, nbits);
 }
 
+static __always_inline
+unsigned long bitmap_weight_and(const unsigned long *src1,
+				const unsigned long *src2, unsigned int nbits)
+{
+	if (small_const_nbits(nbits))
+		return hweight_long(*src1 & *src2 & BITMAP_LAST_WORD_MASK(nbits));
+	return __bitmap_weight_and(src1, src2, nbits);
+}
+
 static __always_inline void bitmap_set(unsigned long *map, unsigned int start,
 		unsigned int nbits)
 {

include/linux/bitops.h

@@ -247,6 +247,25 @@ static inline unsigned long __ffs64(u64 word)
 	return __ffs((unsigned long)word);
 }
 
+/**
+ * fns - find N'th set bit in a word
+ * @word: The word to search
+ * @n: Bit to find
+ */
+static inline unsigned long fns(unsigned long word, unsigned int n)
+{
+	unsigned int bit;
+
+	while (word) {
+		bit = __ffs(word);
+		if (n-- == 0)
+			return bit;
+		__clear_bit(bit, &word);
+	}
+
+	return BITS_PER_LONG;
+}
+
 /**
  * assign_bit - Assign value to a bit in memory
  * @nr: the bit to set

include/linux/cpumask.h

@@ -35,19 +35,23 @@ typedef struct cpumask { DECLARE_BITMAP(bits, NR_CPUS); } cpumask_t;
  */
 #define cpumask_pr_args(maskp)		nr_cpu_ids, cpumask_bits(maskp)
 
-#if NR_CPUS == 1
-#define nr_cpu_ids		1U
+#if (NR_CPUS == 1) || defined(CONFIG_FORCE_NR_CPUS)
+#define nr_cpu_ids ((unsigned int)NR_CPUS)
 #else
 extern unsigned int nr_cpu_ids;
 #endif
 
-#ifdef CONFIG_CPUMASK_OFFSTACK
-/* Assuming NR_CPUS is huge, a runtime limit is more efficient.  Also,
- * not all bits may be allocated. */
-#define nr_cpumask_bits	nr_cpu_ids
+static inline void set_nr_cpu_ids(unsigned int nr)
+{
+#if (NR_CPUS == 1) || defined(CONFIG_FORCE_NR_CPUS)
+	WARN_ON(nr != nr_cpu_ids);
 #else
-#define nr_cpumask_bits	((unsigned int)NR_CPUS)
+	nr_cpu_ids = nr;
 #endif
+}
+
+/* Deprecated. Always use nr_cpu_ids. */
+#define nr_cpumask_bits	nr_cpu_ids
 
 /*
  * The following particular system cpumasks and operations manage
@@ -67,10 +71,6 @@ extern unsigned int nr_cpu_ids;
  *  cpu_online_mask is the dynamic subset of cpu_present_mask,
  *  indicating those CPUs available for scheduling.
  *
- *  If HOTPLUG is enabled, then cpu_possible_mask is forced to have
- *  all NR_CPUS bits set, otherwise it is just the set of CPUs that
- *  ACPI reports present at boot.
- *
  *  If HOTPLUG is enabled, then cpu_present_mask varies dynamically,
  *  depending on what ACPI reports as currently plugged in, otherwise
  *  cpu_present_mask is just a copy of cpu_possible_mask.
@@ -174,9 +174,8 @@ static inline unsigned int cpumask_last(const struct cpumask *srcp)
 static inline
 unsigned int cpumask_next(int n, const struct cpumask *srcp)
 {
-	/* -1 is a legal arg here. */
-	if (n != -1)
-		cpumask_check(n);
+	/* n is a prior cpu */
+	cpumask_check(n + 1);
 	return find_next_bit(cpumask_bits(srcp), nr_cpumask_bits, n + 1);
 }
 
@@ -189,9 +188,8 @@ unsigned int cpumask_next(int n, const struct cpumask *srcp)
  */
 static inline unsigned int cpumask_next_zero(int n, const struct cpumask *srcp)
 {
-	/* -1 is a legal arg here. */
-	if (n != -1)
-		cpumask_check(n);
+	/* n is a prior cpu */
+	cpumask_check(n + 1);
 	return find_next_zero_bit(cpumask_bits(srcp), nr_cpumask_bits, n+1);
 }
 
@@ -231,9 +229,8 @@ static inline
 unsigned int cpumask_next_and(int n, const struct cpumask *src1p,
 		     const struct cpumask *src2p)
 {
-	/* -1 is a legal arg here. */
-	if (n != -1)
-		cpumask_check(n);
+	/* n is a prior cpu */
+	cpumask_check(n + 1);
 	return find_next_and_bit(cpumask_bits(src1p), cpumask_bits(src2p),
 		nr_cpumask_bits, n + 1);
 }
@@ -246,9 +243,7 @@ unsigned int cpumask_next_and(int n, const struct cpumask *src1p,
  * After the loop, cpu is >= nr_cpu_ids.
  */
 #define for_each_cpu(cpu, mask)				\
-	for ((cpu) = -1;				\
-		(cpu) = cpumask_next((cpu), (mask)),	\
-		(cpu) < nr_cpu_ids;)
+	for_each_set_bit(cpu, cpumask_bits(mask), nr_cpumask_bits)
 
 /**
  * for_each_cpu_not - iterate over every cpu in a complemented mask
@@ -258,17 +253,15 @@ unsigned int cpumask_next_and(int n, const struct cpumask *src1p,
  * After the loop, cpu is >= nr_cpu_ids.
  */
 #define for_each_cpu_not(cpu, mask)				\
-	for ((cpu) = -1;					\
-		(cpu) = cpumask_next_zero((cpu), (mask)),	\
-		(cpu) < nr_cpu_ids;)
+	for_each_clear_bit(cpu, cpumask_bits(mask), nr_cpumask_bits)
 
 #if NR_CPUS == 1
 static inline
 unsigned int cpumask_next_wrap(int n, const struct cpumask *mask, int start, bool wrap)
 {
 	cpumask_check(start);
-	if (n != -1)
-		cpumask_check(n);
+	/* n is a prior cpu */
+	cpumask_check(n + 1);
 
 	/*
 	 * Return the first available CPU when wrapping, or when starting before cpu0,
@@ -293,10 +286,8 @@ unsigned int __pure cpumask_next_wrap(int n, const struct cpumask *mask, int sta
  *
  * After the loop, cpu is >= nr_cpu_ids.
  */
-#define for_each_cpu_wrap(cpu, mask, start)					\
-	for ((cpu) = cpumask_next_wrap((start)-1, (mask), (start), false);	\
-	     (cpu) < nr_cpumask_bits;						\
-	     (cpu) = cpumask_next_wrap((cpu), (mask), (start), true))
+#define for_each_cpu_wrap(cpu, mask, start)				\
+	for_each_set_bit_wrap(cpu, cpumask_bits(mask), nr_cpumask_bits, start)
 
 /**
  * for_each_cpu_and - iterate over every cpu in both masks
@@ -313,9 +304,25 @@ unsigned int __pure cpumask_next_wrap(int n, const struct cpumask *mask, int sta
  * After the loop, cpu is >= nr_cpu_ids.
  */
 #define for_each_cpu_and(cpu, mask1, mask2)				\
-	for ((cpu) = -1;						\
-		(cpu) = cpumask_next_and((cpu), (mask1), (mask2)),	\
-		(cpu) < nr_cpu_ids;)
+	for_each_and_bit(cpu, cpumask_bits(mask1), cpumask_bits(mask2), nr_cpumask_bits)
+
+/**
+ * for_each_cpu_andnot - iterate over every cpu present in one mask, excluding
+ *			 those present in another.
+ * @cpu: the (optionally unsigned) integer iterator
+ * @mask1: the first cpumask pointer
+ * @mask2: the second cpumask pointer
+ *
+ * This saves a temporary CPU mask in many places.  It is equivalent to:
+ *	struct cpumask tmp;
+ *	cpumask_andnot(&tmp, &mask1, &mask2);
+ *	for_each_cpu(cpu, &tmp)
+ *		...
+ *
+ * After the loop, cpu is >= nr_cpu_ids.
+ */
+#define for_each_cpu_andnot(cpu, mask1, mask2)				\
+	for_each_andnot_bit(cpu, cpumask_bits(mask1), cpumask_bits(mask2), nr_cpumask_bits)
 
 /**
  * cpumask_any_but - return a "random" in a cpumask, but not this one.
@@ -337,6 +344,50 @@ unsigned int cpumask_any_but(const struct cpumask *mask, unsigned int cpu)
 	return i;
 }
 
+/**
+ * cpumask_nth - get the first cpu in a cpumask
+ * @srcp: the cpumask pointer
+ * @cpu: the N'th cpu to find, starting from 0
+ *
+ * Returns >= nr_cpu_ids if such cpu doesn't exist.
+ */
+static inline unsigned int cpumask_nth(unsigned int cpu, const struct cpumask *srcp)
+{
+	return find_nth_bit(cpumask_bits(srcp), nr_cpumask_bits, cpumask_check(cpu));
+}
+
+/**
+ * cpumask_nth_and - get the first cpu in 2 cpumasks
+ * @srcp1: the cpumask pointer
+ * @srcp2: the cpumask pointer
+ * @cpu: the N'th cpu to find, starting from 0
+ *
+ * Returns >= nr_cpu_ids if such cpu doesn't exist.
+ */
+static inline
+unsigned int cpumask_nth_and(unsigned int cpu, const struct cpumask *srcp1,
+							const struct cpumask *srcp2)
+{
+	return find_nth_and_bit(cpumask_bits(srcp1), cpumask_bits(srcp2),
+				nr_cpumask_bits, cpumask_check(cpu));
+}
+
+/**
+ * cpumask_nth_andnot - get the first cpu set in 1st cpumask, and clear in 2nd.
+ * @srcp1: the cpumask pointer
+ * @srcp2: the cpumask pointer
+ * @cpu: the N'th cpu to find, starting from 0
+ *
+ * Returns >= nr_cpu_ids if such cpu doesn't exist.
+ */
+static inline
+unsigned int cpumask_nth_andnot(unsigned int cpu, const struct cpumask *srcp1,
+							const struct cpumask *srcp2)
+{
+	return find_nth_andnot_bit(cpumask_bits(srcp1), cpumask_bits(srcp2),
+				nr_cpumask_bits, cpumask_check(cpu));
+}
+
 #define CPU_BITS_NONE						\
 {								\
 	[0 ... BITS_TO_LONGS(NR_CPUS)-1] = 0UL			\
@@ -586,6 +637,17 @@ static inline unsigned int cpumask_weight(const struct cpumask *srcp)
 	return bitmap_weight(cpumask_bits(srcp), nr_cpumask_bits);
 }
 
+/**
+ * cpumask_weight_and - Count of bits in (*srcp1 & *srcp2)
+ * @srcp1: the cpumask to count bits (< nr_cpu_ids) in.
+ * @srcp2: the cpumask to count bits (< nr_cpu_ids) in.
+ */
+static inline unsigned int cpumask_weight_and(const struct cpumask *srcp1,
+						const struct cpumask *srcp2)
+{
+	return bitmap_weight_and(cpumask_bits(srcp1), cpumask_bits(srcp2), nr_cpumask_bits);
+}
+
 /**
  * cpumask_shift_right - *dstp = *srcp >> n
  * @dstp: the cpumask result