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()
  ...

5 files changed, 9 insertions, 5 deletions

diff --git a/arch/loongarch/kernel/setup.c b/arch/loongarch/kernel/setup.c
index 6ca5a0dcbb1c..5b49c78c23f4 100644
--- a/arch/loongarch/kernel/setup.c
+++ b/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
 
diff --git a/arch/mips/kernel/setup.c b/arch/mips/kernel/setup.c
index 39c79f67c7a3..f1c88f8a1dc5 100644
--- a/arch/mips/kernel/setup.c
+++ b/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) {}
diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S
index cfc09a96a420..dedcc6fe2263 100644
--- a/arch/powerpc/kernel/head_64.S
+++ b/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 */
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index f24227bc3220..3f3ea0287f69 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/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));
diff --git a/arch/x86/xen/smp_pv.c b/arch/x86/xen/smp_pv.c
index ba7af2eca755..480be82e9b7b 100644
--- a/arch/x86/xen/smp_pv.c
+++ b/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
 
 }