From 6398268d2bc454735f11e08705e858f9fdf5c750 Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@redhat.com>
Date: Tue, 5 Jun 2012 13:23:44 +0200
Subject: x86/apic: Factor out default cpu_mask_to_apicid() operations

Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120605112340.GA11454@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/apic/apic_noop.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index a6e4c6e06c0..7c3dd4fe068 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -159,8 +159,8 @@ struct apic apic_noop = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0x0F << 24,
 
-	.cpu_mask_to_apicid		= default_cpu_mask_to_apicid,
-	.cpu_mask_to_apicid_and		= default_cpu_mask_to_apicid_and,
+	.cpu_mask_to_apicid		= flat_cpu_mask_to_apicid,
+	.cpu_mask_to_apicid_and		= flat_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= noop_send_IPI_mask,
 	.send_IPI_mask_allbutself	= noop_send_IPI_mask_allbutself,
-- 
cgit v1.2.3-70-g09d2


From 9d8e10667624ea6411f04495aef1fa4a8a778ee8 Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@redhat.com>
Date: Thu, 7 Jun 2012 15:14:49 +0200
Subject: x86/apic: Factor out default vector_allocation_domain() operation

Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120607131449.GC4759@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/include/asm/apic.h          | 21 +++++++++++++++++++++
 arch/x86/kernel/apic/apic_flat_64.c  | 22 +---------------------
 arch/x86/kernel/apic/apic_noop.c     |  3 +--
 arch/x86/kernel/apic/apic_numachip.c |  8 +-------
 arch/x86/kernel/apic/bigsmp_32.c     |  8 +-------
 arch/x86/kernel/apic/es7000_32.c     | 19 ++-----------------
 arch/x86/kernel/apic/numaq_32.c      | 16 +---------------
 arch/x86/kernel/apic/probe_32.c      | 17 +----------------
 arch/x86/kernel/apic/summit_32.c     | 16 +---------------
 arch/x86/kernel/apic/x2apic_phys.c   | 11 +----------
 arch/x86/kernel/apic/x2apic_uv_x.c   |  8 +-------
 11 files changed, 32 insertions(+), 117 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
index bef571769e6..feb2dbdae9e 100644
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -615,6 +615,27 @@ extern unsigned int
 default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			       const struct cpumask *andmask);
 
+static inline void
+flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
+{
+	/* Careful. Some cpus do not strictly honor the set of cpus
+	 * specified in the interrupt destination when using lowest
+	 * priority interrupt delivery mode.
+	 *
+	 * In particular there was a hyperthreading cpu observed to
+	 * deliver interrupts to the wrong hyperthread when only one
+	 * hyperthread was specified in the interrupt desitination.
+	 */
+	cpumask_clear(retmask);
+	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
+}
+
+static inline void
+default_vector_allocation_domain(int cpu, struct cpumask *retmask)
+{
+	cpumask_copy(retmask, cpumask_of(cpu));
+}
+
 static inline unsigned long default_check_apicid_used(physid_mask_t *map, int apicid)
 {
 	return physid_isset(apicid, *map);
diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
index 55b97ce4fa1..bddc92566d0 100644
--- a/arch/x86/kernel/apic/apic_flat_64.c
+++ b/arch/x86/kernel/apic/apic_flat_64.c
@@ -36,20 +36,6 @@ static int flat_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
 	return 1;
 }
 
-static void flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	/* Careful. Some cpus do not strictly honor the set of cpus
-	 * specified in the interrupt destination when using lowest
-	 * priority interrupt delivery mode.
-	 *
-	 * In particular there was a hyperthreading cpu observed to
-	 * deliver interrupts to the wrong hyperthread when only one
-	 * hyperthread was specified in the interrupt desitination.
-	 */
-	cpumask_clear(retmask);
-	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-}
-
 /*
  * Set up the logical destination ID.
  *
@@ -257,12 +243,6 @@ static int physflat_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
 	return 0;
 }
 
-static void physflat_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
-}
-
 static void physflat_send_IPI_mask(const struct cpumask *cpumask, int vector)
 {
 	default_send_IPI_mask_sequence_phys(cpumask, vector);
@@ -309,7 +289,7 @@ static struct apic apic_physflat =  {
 	.check_apicid_used		= NULL,
 	.check_apicid_present		= NULL,
 
-	.vector_allocation_domain	= physflat_vector_allocation_domain,
+	.vector_allocation_domain	= default_vector_allocation_domain,
 	/* not needed, but shouldn't hurt: */
 	.init_apic_ldr			= flat_init_apic_ldr,
 
diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index 7c3dd4fe068..3e43cf52893 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -104,8 +104,7 @@ static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	if (cpu != 0)
 		pr_warning("APIC: Vector allocated for non-BSP cpu\n");
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
+	cpumask_copy(retmask, cpumask_of(cpu));
 }
 
 static u32 noop_apic_read(u32 reg)
diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c
index dba4bf2ed56..c028132ad35 100644
--- a/arch/x86/kernel/apic/apic_numachip.c
+++ b/arch/x86/kernel/apic/apic_numachip.c
@@ -72,12 +72,6 @@ static int numachip_phys_pkg_id(int initial_apic_id, int index_msb)
 	return initial_apic_id >> index_msb;
 }
 
-static void numachip_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
-}
-
 static int __cpuinit numachip_wakeup_secondary(int phys_apicid, unsigned long start_rip)
 {
 	union numachip_csr_g3_ext_irq_gen int_gen;
@@ -222,7 +216,7 @@ static struct apic apic_numachip __refconst = {
 	.check_apicid_used		= NULL,
 	.check_apicid_present		= NULL,
 
-	.vector_allocation_domain	= numachip_vector_allocation_domain,
+	.vector_allocation_domain	= default_vector_allocation_domain,
 	.init_apic_ldr			= flat_init_apic_ldr,
 
 	.ioapic_phys_id_map		= NULL,
diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c
index 907aa3d112a..df342fe4d6a 100644
--- a/arch/x86/kernel/apic/bigsmp_32.c
+++ b/arch/x86/kernel/apic/bigsmp_32.c
@@ -142,12 +142,6 @@ static const struct dmi_system_id bigsmp_dmi_table[] = {
 	{ } /* NULL entry stops DMI scanning */
 };
 
-static void bigsmp_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
-}
-
 static int probe_bigsmp(void)
 {
 	if (def_to_bigsmp)
@@ -176,7 +170,7 @@ static struct apic apic_bigsmp = {
 	.check_apicid_used		= bigsmp_check_apicid_used,
 	.check_apicid_present		= bigsmp_check_apicid_present,
 
-	.vector_allocation_domain	= bigsmp_vector_allocation_domain,
+	.vector_allocation_domain	= default_vector_allocation_domain,
 	.init_apic_ldr			= bigsmp_init_apic_ldr,
 
 	.ioapic_phys_id_map		= bigsmp_ioapic_phys_id_map,
diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c
index db4ab1be3c7..3c42865757e 100644
--- a/arch/x86/kernel/apic/es7000_32.c
+++ b/arch/x86/kernel/apic/es7000_32.c
@@ -394,21 +394,6 @@ static void es7000_enable_apic_mode(void)
 		WARN(1, "Command failed, status = %x\n", mip_status);
 }
 
-static void es7000_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	/* Careful. Some cpus do not strictly honor the set of cpus
-	 * specified in the interrupt destination when using lowest
-	 * priority interrupt delivery mode.
-	 *
-	 * In particular there was a hyperthreading cpu observed to
-	 * deliver interrupts to the wrong hyperthread when only one
-	 * hyperthread was specified in the interrupt desitination.
-	 */
-	cpumask_clear(retmask);
-	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-}
-
-
 static void es7000_wait_for_init_deassert(atomic_t *deassert)
 {
 	while (!atomic_read(deassert))
@@ -638,7 +623,7 @@ static struct apic __refdata apic_es7000_cluster = {
 	.check_apicid_used		= es7000_check_apicid_used,
 	.check_apicid_present		= es7000_check_apicid_present,
 
-	.vector_allocation_domain	= es7000_vector_allocation_domain,
+	.vector_allocation_domain	= flat_vector_allocation_domain,
 	.init_apic_ldr			= es7000_init_apic_ldr_cluster,
 
 	.ioapic_phys_id_map		= es7000_ioapic_phys_id_map,
@@ -705,7 +690,7 @@ static struct apic __refdata apic_es7000 = {
 	.check_apicid_used		= es7000_check_apicid_used,
 	.check_apicid_present		= es7000_check_apicid_present,
 
-	.vector_allocation_domain	= es7000_vector_allocation_domain,
+	.vector_allocation_domain	= flat_vector_allocation_domain,
 	.init_apic_ldr			= es7000_init_apic_ldr,
 
 	.ioapic_phys_id_map		= es7000_ioapic_phys_id_map,
diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c
index f00a68cca37..eb2d466fd81 100644
--- a/arch/x86/kernel/apic/numaq_32.c
+++ b/arch/x86/kernel/apic/numaq_32.c
@@ -441,20 +441,6 @@ static int probe_numaq(void)
 	return found_numaq;
 }
 
-static void numaq_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	/* Careful. Some cpus do not strictly honor the set of cpus
-	 * specified in the interrupt destination when using lowest
-	 * priority interrupt delivery mode.
-	 *
-	 * In particular there was a hyperthreading cpu observed to
-	 * deliver interrupts to the wrong hyperthread when only one
-	 * hyperthread was specified in the interrupt desitination.
-	 */
-	cpumask_clear(retmask);
-	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-}
-
 static void numaq_setup_portio_remap(void)
 {
 	int num_quads = num_online_nodes();
@@ -491,7 +477,7 @@ static struct apic __refdata apic_numaq = {
 	.check_apicid_used		= numaq_check_apicid_used,
 	.check_apicid_present		= numaq_check_apicid_present,
 
-	.vector_allocation_domain	= numaq_vector_allocation_domain,
+	.vector_allocation_domain	= flat_vector_allocation_domain,
 	.init_apic_ldr			= numaq_init_apic_ldr,
 
 	.ioapic_phys_id_map		= numaq_ioapic_phys_id_map,
diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c
index 71b6ac48ab2..2c6f003b2e4 100644
--- a/arch/x86/kernel/apic/probe_32.c
+++ b/arch/x86/kernel/apic/probe_32.c
@@ -66,21 +66,6 @@ static void setup_apic_flat_routing(void)
 #endif
 }
 
-static void default_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	/*
-	 * Careful. Some cpus do not strictly honor the set of cpus
-	 * specified in the interrupt destination when using lowest
-	 * priority interrupt delivery mode.
-	 *
-	 * In particular there was a hyperthreading cpu observed to
-	 * deliver interrupts to the wrong hyperthread when only one
-	 * hyperthread was specified in the interrupt desitination.
-	 */
-	cpumask_clear(retmask);
-	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-}
-
 /* should be called last. */
 static int probe_default(void)
 {
@@ -105,7 +90,7 @@ static struct apic apic_default = {
 	.check_apicid_used		= default_check_apicid_used,
 	.check_apicid_present		= default_check_apicid_present,
 
-	.vector_allocation_domain	= default_vector_allocation_domain,
+	.vector_allocation_domain	= flat_vector_allocation_domain,
 	.init_apic_ldr			= default_init_apic_ldr,
 
 	.ioapic_phys_id_map		= default_ioapic_phys_id_map,
diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c
index 659897c0075..35d254c1fec 100644
--- a/arch/x86/kernel/apic/summit_32.c
+++ b/arch/x86/kernel/apic/summit_32.c
@@ -320,20 +320,6 @@ static int probe_summit(void)
 	return 0;
 }
 
-static void summit_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	/* Careful. Some cpus do not strictly honor the set of cpus
-	 * specified in the interrupt destination when using lowest
-	 * priority interrupt delivery mode.
-	 *
-	 * In particular there was a hyperthreading cpu observed to
-	 * deliver interrupts to the wrong hyperthread when only one
-	 * hyperthread was specified in the interrupt desitination.
-	 */
-	cpumask_clear(retmask);
-	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-}
-
 #ifdef CONFIG_X86_SUMMIT_NUMA
 static struct rio_table_hdr *rio_table_hdr;
 static struct scal_detail   *scal_devs[MAX_NUMNODES];
@@ -509,7 +495,7 @@ static struct apic apic_summit = {
 	.check_apicid_used		= summit_check_apicid_used,
 	.check_apicid_present		= summit_check_apicid_present,
 
-	.vector_allocation_domain	= summit_vector_allocation_domain,
+	.vector_allocation_domain	= flat_vector_allocation_domain,
 	.init_apic_ldr			= summit_init_apic_ldr,
 
 	.ioapic_phys_id_map		= summit_ioapic_phys_id_map,
diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c
index f730269edef..f109388a0e8 100644
--- a/arch/x86/kernel/apic/x2apic_phys.c
+++ b/arch/x86/kernel/apic/x2apic_phys.c
@@ -88,15 +88,6 @@ static int x2apic_phys_probe(void)
 	return apic == &apic_x2apic_phys;
 }
 
-/*
- * Each logical cpu is in its own vector allocation domain.
- */
-static void x2apic_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
-}
-
 static struct apic apic_x2apic_phys = {
 
 	.name				= "physical x2apic",
@@ -114,7 +105,7 @@ static struct apic apic_x2apic_phys = {
 	.check_apicid_used		= NULL,
 	.check_apicid_present		= NULL,
 
-	.vector_allocation_domain	= x2apic_vector_allocation_domain,
+	.vector_allocation_domain	= default_vector_allocation_domain,
 	.init_apic_ldr			= init_x2apic_ldr,
 
 	.ioapic_phys_id_map		= NULL,
diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 16efb92bfea..df89a7d7874 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -185,12 +185,6 @@ EXPORT_SYMBOL_GPL(uv_possible_blades);
 unsigned long sn_rtc_cycles_per_second;
 EXPORT_SYMBOL(sn_rtc_cycles_per_second);
 
-static void uv_vector_allocation_domain(int cpu, struct cpumask *retmask)
-{
-	cpumask_clear(retmask);
-	cpumask_set_cpu(cpu, retmask);
-}
-
 static int __cpuinit uv_wakeup_secondary(int phys_apicid, unsigned long start_rip)
 {
 #ifdef CONFIG_SMP
@@ -363,7 +357,7 @@ static struct apic __refdata apic_x2apic_uv_x = {
 	.check_apicid_used		= NULL,
 	.check_apicid_present		= NULL,
 
-	.vector_allocation_domain	= uv_vector_allocation_domain,
+	.vector_allocation_domain	= default_vector_allocation_domain,
 	.init_apic_ldr			= uv_init_apic_ldr,
 
 	.ioapic_phys_id_map		= NULL,
-- 
cgit v1.2.3-70-g09d2


From 8637e38aff14d048b649075114023023a2e80fba Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@redhat.com>
Date: Thu, 7 Jun 2012 15:15:44 +0200
Subject: x86/apic: Avoid useless scanning thru a cpumask in
 assign_irq_vector()

In case of static vector allocation domains (i.e. flat) if all
vector numbers are exhausted, an attempt to assign a new vector
will lead to useless scans through all CPUs in the cpumask, even
though it is known that each new pass would fail. Make this
corner case less painful by letting report whether the vector
allocation domain depends on passed arguments or not and stop
scanning early.

The same could have been achived by introducing a static flag to
the apic operations. But let's allow vector_allocation_domain()
have more intelligence here and decide dynamically, in case we
would need it in the future.

Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120607131542.GE4759@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/include/asm/apic.h      |  8 +++++---
 arch/x86/kernel/apic/apic_noop.c |  3 ++-
 arch/x86/kernel/apic/io_apic.c   | 12 +++++++++---
 3 files changed, 16 insertions(+), 7 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
index feb2dbdae9e..e3fecd50d5c 100644
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -306,7 +306,7 @@ struct apic {
 	unsigned long (*check_apicid_used)(physid_mask_t *map, int apicid);
 	unsigned long (*check_apicid_present)(int apicid);
 
-	void (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
+	bool (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
 	void (*init_apic_ldr)(void);
 
 	void (*ioapic_phys_id_map)(physid_mask_t *phys_map, physid_mask_t *retmap);
@@ -615,7 +615,7 @@ extern unsigned int
 default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			       const struct cpumask *andmask);
 
-static inline void
+static inline bool
 flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	/* Careful. Some cpus do not strictly honor the set of cpus
@@ -628,12 +628,14 @@ flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
 	 */
 	cpumask_clear(retmask);
 	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
+	return false;
 }
 
-static inline void
+static inline bool
 default_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	cpumask_copy(retmask, cpumask_of(cpu));
+	return true;
 }
 
 static inline unsigned long default_check_apicid_used(physid_mask_t *map, int apicid)
diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index 3e43cf52893..ac9edf247b1 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -100,11 +100,12 @@ static unsigned long noop_check_apicid_present(int bit)
 	return physid_isset(bit, phys_cpu_present_map);
 }
 
-static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static bool noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	if (cpu != 0)
 		pr_warning("APIC: Vector allocated for non-BSP cpu\n");
 	cpumask_copy(retmask, cpumask_of(cpu));
+	return true;
 }
 
 static u32 noop_apic_read(u32 reg)
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 05af3d341aa..4061a7dee5c 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -1137,8 +1137,9 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask)
 	for_each_cpu_and(cpu, mask, cpu_online_mask) {
 		int new_cpu;
 		int vector, offset;
+		bool more_domains;
 
-		apic->vector_allocation_domain(cpu, tmp_mask);
+		more_domains = apic->vector_allocation_domain(cpu, tmp_mask);
 
 		if (cpumask_subset(tmp_mask, cfg->domain)) {
 			free_cpumask_var(tmp_mask);
@@ -1153,8 +1154,13 @@ next:
 			offset = (offset + 1) % 16;
 			vector = FIRST_EXTERNAL_VECTOR + offset;
 		}
-		if (unlikely(current_vector == vector))
-			continue;
+
+		if (unlikely(current_vector == vector)) {
+			if (more_domains)
+				continue;
+			else
+				break;
+		}
 
 		if (test_bit(vector, used_vectors))
 			goto next;
-- 
cgit v1.2.3-70-g09d2


From a5a391561bc25898ba1a702a0c4b028aa5b11ce9 Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@redhat.com>
Date: Thu, 14 Jun 2012 09:49:35 +0200
Subject: x86/apic: Eliminate cpu_mask_to_apicid() operation

Since there are only two locations where cpu_mask_to_apicid() is
called from, remove the operation and use only
cpu_mask_to_apicid_and() instead.

Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
Suggested-and-acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20120614074935.GE3383@dhcp-26-207.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/include/asm/apic.h           | 33 ++++++++-------------------------
 arch/x86/kernel/apic/apic.c           | 24 ++++++------------------
 arch/x86/kernel/apic/apic_flat_64.c   |  2 --
 arch/x86/kernel/apic/apic_noop.c      |  1 -
 arch/x86/kernel/apic/apic_numachip.c  |  1 -
 arch/x86/kernel/apic/bigsmp_32.c      |  1 -
 arch/x86/kernel/apic/es7000_32.c      |  4 +---
 arch/x86/kernel/apic/io_apic.c        |  3 ++-
 arch/x86/kernel/apic/numaq_32.c       |  8 --------
 arch/x86/kernel/apic/probe_32.c       |  1 -
 arch/x86/kernel/apic/summit_32.c      |  3 +--
 arch/x86/kernel/apic/x2apic_cluster.c | 17 -----------------
 arch/x86/kernel/apic/x2apic_phys.c    |  1 -
 arch/x86/kernel/apic/x2apic_uv_x.c    | 29 ++++++-----------------------
 arch/x86/platform/uv/uv_irq.c         |  2 +-
 15 files changed, 25 insertions(+), 105 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
index 1ed3eead203..eec240e1209 100644
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -331,8 +331,6 @@ struct apic {
 	unsigned long (*set_apic_id)(unsigned int id);
 	unsigned long apic_id_mask;
 
-	int (*cpu_mask_to_apicid)(const struct cpumask *cpumask,
-				  unsigned int *apicid);
 	int (*cpu_mask_to_apicid_and)(const struct cpumask *cpumask,
 				      const struct cpumask *andmask,
 				      unsigned int *apicid);
@@ -594,9 +592,15 @@ static inline int default_phys_pkg_id(int cpuid_apic, int index_msb)
 #endif
 
 static inline int
-__flat_cpu_mask_to_apicid(unsigned long cpu_mask, unsigned int *apicid)
+flat_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
+			    const struct cpumask *andmask,
+			    unsigned int *apicid)
 {
-	cpu_mask = cpu_mask & APIC_ALL_CPUS & cpumask_bits(cpu_online_mask)[0];
+	unsigned long cpu_mask = cpumask_bits(cpumask)[0] &
+				 cpumask_bits(andmask)[0] &
+				 cpumask_bits(cpu_online_mask)[0] &
+				 APIC_ALL_CPUS;
+
 	if (likely(cpu_mask)) {
 		*apicid = (unsigned int)cpu_mask;
 		return 0;
@@ -605,27 +609,6 @@ __flat_cpu_mask_to_apicid(unsigned long cpu_mask, unsigned int *apicid)
 	}
 }
 
-static inline int
-flat_cpu_mask_to_apicid(const struct cpumask *cpumask,
-			unsigned int *apicid)
-{
-	return __flat_cpu_mask_to_apicid(cpumask_bits(cpumask)[0], apicid);
-}
-
-static inline int
-flat_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
-			    const struct cpumask *andmask,
-			    unsigned int *apicid)
-{
-	unsigned long mask1 = cpumask_bits(cpumask)[0];
-	unsigned long mask2 = cpumask_bits(andmask)[0];
-	return __flat_cpu_mask_to_apicid(mask1 & mask2, apicid);
-}
-
-extern int
-default_cpu_mask_to_apicid(const struct cpumask *cpumask,
-			   unsigned int *apicid);
-
 extern int
 default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			       const struct cpumask *andmask,
diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index 7e9bbe73bc5..048a4f806d4 100644
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -2123,23 +2123,6 @@ void default_init_apic_ldr(void)
 	apic_write(APIC_LDR, val);
 }
 
-static inline int __default_cpu_to_apicid(int cpu, unsigned int *apicid)
-{
-	if (likely((unsigned int)cpu < nr_cpu_ids)) {
-		*apicid = per_cpu(x86_cpu_to_apicid, cpu);
-		return 0;
-	} else {
-		return -EINVAL;
-	}
-}
-
-int default_cpu_mask_to_apicid(const struct cpumask *cpumask,
-			       unsigned int *apicid)
-{
-	int cpu = cpumask_first_and(cpumask, cpu_online_mask);
-	return __default_cpu_to_apicid(cpu, apicid);
-}
-
 int default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 				   const struct cpumask *andmask,
 				   unsigned int *apicid)
@@ -2151,7 +2134,12 @@ int default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			break;
 	}
 
-	return __default_cpu_to_apicid(cpu, apicid);
+	if (likely((unsigned int)cpu < nr_cpu_ids)) {
+		*apicid = per_cpu(x86_cpu_to_apicid, cpu);
+		return 0;
+	} else {
+		return -EINVAL;
+	}
 }
 
 /*
diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
index bddc92566d0..00c77cf78e9 100644
--- a/arch/x86/kernel/apic/apic_flat_64.c
+++ b/arch/x86/kernel/apic/apic_flat_64.c
@@ -191,7 +191,6 @@ static struct apic apic_flat =  {
 	.set_apic_id			= set_apic_id,
 	.apic_id_mask			= 0xFFu << 24,
 
-	.cpu_mask_to_apicid		= flat_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= flat_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= flat_send_IPI_mask,
@@ -308,7 +307,6 @@ static struct apic apic_physflat =  {
 	.set_apic_id			= set_apic_id,
 	.apic_id_mask			= 0xFFu << 24,
 
-	.cpu_mask_to_apicid		= default_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= default_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= physflat_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index ac9edf247b1..65c07fc630a 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -159,7 +159,6 @@ struct apic apic_noop = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0x0F << 24,
 
-	.cpu_mask_to_apicid		= flat_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= flat_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= noop_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c
index c028132ad35..bc552cff257 100644
--- a/arch/x86/kernel/apic/apic_numachip.c
+++ b/arch/x86/kernel/apic/apic_numachip.c
@@ -234,7 +234,6 @@ static struct apic apic_numachip __refconst = {
 	.set_apic_id			= set_apic_id,
 	.apic_id_mask			= 0xffU << 24,
 
-	.cpu_mask_to_apicid		= default_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= default_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= numachip_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c
index df342fe4d6a..d50e3640d5a 100644
--- a/arch/x86/kernel/apic/bigsmp_32.c
+++ b/arch/x86/kernel/apic/bigsmp_32.c
@@ -188,7 +188,6 @@ static struct apic apic_bigsmp = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0xFF << 24,
 
-	.cpu_mask_to_apicid		= default_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= default_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= bigsmp_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c
index b35cfb9b696..2c5317ea1b8 100644
--- a/arch/x86/kernel/apic/es7000_32.c
+++ b/arch/x86/kernel/apic/es7000_32.c
@@ -525,7 +525,7 @@ static int es7000_check_phys_apicid_present(int cpu_physical_apicid)
 	return 1;
 }
 
-static int
+static inline int
 es7000_cpu_mask_to_apicid(const struct cpumask *cpumask, unsigned int *dest_id)
 {
 	unsigned int round = 0;
@@ -643,7 +643,6 @@ static struct apic __refdata apic_es7000_cluster = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0xFF << 24,
 
-	.cpu_mask_to_apicid		= es7000_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= es7000_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= es7000_send_IPI_mask,
@@ -710,7 +709,6 @@ static struct apic __refdata apic_es7000 = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0xFF << 24,
 
-	.cpu_mask_to_apicid		= es7000_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= es7000_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= es7000_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 0deb773404e..0540f083f45 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -1492,7 +1492,8 @@ static void __init setup_timer_IRQ0_pin(unsigned int ioapic_idx,
 	 * We use logical delivery to get the timer IRQ
 	 * to the first CPU.
 	 */
-	if (unlikely(apic->cpu_mask_to_apicid(apic->target_cpus(), &dest)))
+	if (unlikely(apic->cpu_mask_to_apicid_and(apic->target_cpus(),
+						  apic->target_cpus(), &dest)))
 		dest = BAD_APICID;
 
 	entry.dest_mode = apic->irq_dest_mode;
diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c
index 2b55514c328..d661ee95cab 100644
--- a/arch/x86/kernel/apic/numaq_32.c
+++ b/arch/x86/kernel/apic/numaq_32.c
@@ -406,13 +406,6 @@ static inline int numaq_check_phys_apicid_present(int phys_apicid)
  * We use physical apicids here, not logical, so just return the default
  * physical broadcast to stop people from breaking us
  */
-static int
-numaq_cpu_mask_to_apicid(const struct cpumask *cpumask, unsigned int *apicid)
-{
-	*apicid = 0x0F;
-	return 0;
-}
-
 static int
 numaq_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			     const struct cpumask *andmask,
@@ -499,7 +492,6 @@ static struct apic __refdata apic_numaq = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0x0F << 24,
 
-	.cpu_mask_to_apicid		= numaq_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= numaq_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= numaq_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c
index 2c6f003b2e4..eef6bcd1bf1 100644
--- a/arch/x86/kernel/apic/probe_32.c
+++ b/arch/x86/kernel/apic/probe_32.c
@@ -108,7 +108,6 @@ static struct apic apic_default = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0x0F << 24,
 
-	.cpu_mask_to_apicid		= flat_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= flat_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= default_send_IPI_mask_logical,
diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c
index 79d360f6729..bbad180f289 100644
--- a/arch/x86/kernel/apic/summit_32.c
+++ b/arch/x86/kernel/apic/summit_32.c
@@ -263,7 +263,7 @@ static int summit_check_phys_apicid_present(int physical_apicid)
 	return 1;
 }
 
-static int
+static inline int
 summit_cpu_mask_to_apicid(const struct cpumask *cpumask, unsigned int *dest_id)
 {
 	unsigned int round = 0;
@@ -516,7 +516,6 @@ static struct apic apic_summit = {
 	.set_apic_id			= NULL,
 	.apic_id_mask			= 0xFF << 24,
 
-	.cpu_mask_to_apicid		= summit_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= summit_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= summit_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c
index 1885a73b7f3..943d03fc6fc 100644
--- a/arch/x86/kernel/apic/x2apic_cluster.c
+++ b/arch/x86/kernel/apic/x2apic_cluster.c
@@ -96,22 +96,6 @@ static void x2apic_send_IPI_all(int vector)
 	__x2apic_send_IPI_mask(cpu_online_mask, vector, APIC_DEST_ALLINC);
 }
 
-static int
-x2apic_cpu_mask_to_apicid(const struct cpumask *cpumask, unsigned int *apicid)
-{
-	int cpu = cpumask_first_and(cpumask, cpu_online_mask);
-	int i;
-
-	if (cpu >= nr_cpu_ids)
-		return -EINVAL;
-
-	*apicid = 0;
-	for_each_cpu_and(i, cpumask, per_cpu(cpus_in_cluster, cpu))
-		*apicid |= per_cpu(x86_cpu_to_logical_apicid, i);
-
-	return 0;
-}
-
 static int
 x2apic_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			      const struct cpumask *andmask,
@@ -270,7 +254,6 @@ static struct apic apic_x2apic_cluster = {
 	.set_apic_id			= x2apic_set_apic_id,
 	.apic_id_mask			= 0xFFFFFFFFu,
 
-	.cpu_mask_to_apicid		= x2apic_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= x2apic_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= x2apic_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c
index f109388a0e8..e03a1e180e8 100644
--- a/arch/x86/kernel/apic/x2apic_phys.c
+++ b/arch/x86/kernel/apic/x2apic_phys.c
@@ -123,7 +123,6 @@ static struct apic apic_x2apic_phys = {
 	.set_apic_id			= x2apic_set_apic_id,
 	.apic_id_mask			= 0xFFFFFFFFu,
 
-	.cpu_mask_to_apicid		= default_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= default_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= x2apic_send_IPI_mask,
diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 307aa076bd6..026de0114d1 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -269,27 +269,6 @@ static void uv_init_apic_ldr(void)
 {
 }
 
-static inline int __uv_cpu_to_apicid(int cpu, unsigned int *apicid)
-{
-	if (likely((unsigned int)cpu < nr_cpu_ids)) {
-		*apicid = per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits;
-		return 0;
-	} else {
-		return -EINVAL;
-	}
-}
-
-static int
-uv_cpu_mask_to_apicid(const struct cpumask *cpumask, unsigned int *apicid)
-{
-	/*
-	 * We're using fixed IRQ delivery, can only return one phys APIC ID.
-	 * May as well be the first.
-	 */
-	int cpu = cpumask_first_and(cpumask, cpu_online_mask);
-	return __uv_cpu_to_apicid(cpu, apicid);
-}
-
 static int
 uv_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			  const struct cpumask *andmask,
@@ -306,7 +285,12 @@ uv_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			break;
 	}
 
-	return __uv_cpu_to_apicid(cpu, apicid);
+	if (likely((unsigned int)cpu < nr_cpu_ids)) {
+		*apicid = per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits;
+		return 0;
+	} else {
+		return -EINVAL;
+	}
 }
 
 static unsigned int x2apic_get_apic_id(unsigned long x)
@@ -384,7 +368,6 @@ static struct apic __refdata apic_x2apic_uv_x = {
 	.set_apic_id			= set_apic_id,
 	.apic_id_mask			= 0xFFFFFFFFu,
 
-	.cpu_mask_to_apicid		= uv_cpu_mask_to_apicid,
 	.cpu_mask_to_apicid_and		= uv_cpu_mask_to_apicid_and,
 
 	.send_IPI_mask			= uv_send_IPI_mask,
diff --git a/arch/x86/platform/uv/uv_irq.c b/arch/x86/platform/uv/uv_irq.c
index dd1ff39a464..a67c7a6bac7 100644
--- a/arch/x86/platform/uv/uv_irq.c
+++ b/arch/x86/platform/uv/uv_irq.c
@@ -144,7 +144,7 @@ arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade,
 	if (err != 0)
 		return err;
 
-	err = apic->cpu_mask_to_apicid(eligible_cpu, &dest);
+	err = apic->cpu_mask_to_apicid_and(eligible_cpu, eligible_cpu, &dest);
 	if (err != 0)
 		return err;
 
-- 
cgit v1.2.3-70-g09d2


From b39f25a849d7677a7dbf183f2483fd41c201a5ce Mon Sep 17 00:00:00 2001
From: Suresh Siddha <suresh.b.siddha@intel.com>
Date: Mon, 25 Jun 2012 13:38:27 -0700
Subject: x86/apic: Optimize cpu traversal in __assign_irq_vector() using
 domain membership

Currently __assign_irq_vector() goes through each cpu in the
specified mask until it finds a free vector in all the cpu's
that are part of the same interrupt domain. We visit all the
interrupt domain sibling cpus to reserve the free vector. So,
when we fail to find a free vector in an interrupt domain, it is
safe to continue our search with a cpu belonging to a new
interrupt domain. No need to go through each cpu, if the domain
containing that cpu is already visited.

Use the irq_cfg's old_domain to track the visited domains and
optimize the cpu traversal while finding a free vector in the
given cpumask.

NOTE: We can also optimize the search by using for_each_cpu() and
skip the current cpu, if it is not the first cpu in the mask
returned by the vector_allocation_domain(). But re-using the
cfg->old_domain to track the visited domains will be slightly
faster.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Alexander Gordeev <agordeev@redhat.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Link: http://lkml.kernel.org/r/1340656709-11423-2-git-send-email-suresh.b.siddha@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/include/asm/apic.h           |  8 +++-----
 arch/x86/kernel/apic/apic_noop.c      |  3 +--
 arch/x86/kernel/apic/io_apic.c        | 15 ++++++++-------
 arch/x86/kernel/apic/x2apic_cluster.c |  3 +--
 arch/x86/kernel/vsmp_64.c             |  3 +--
 5 files changed, 14 insertions(+), 18 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
index eec240e1209..8bebeb8952f 100644
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -306,7 +306,7 @@ struct apic {
 	unsigned long (*check_apicid_used)(physid_mask_t *map, int apicid);
 	unsigned long (*check_apicid_present)(int apicid);
 
-	bool (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
+	void (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
 	void (*init_apic_ldr)(void);
 
 	void (*ioapic_phys_id_map)(physid_mask_t *phys_map, physid_mask_t *retmap);
@@ -614,7 +614,7 @@ default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			       const struct cpumask *andmask,
 			       unsigned int *apicid);
 
-static inline bool
+static inline void
 flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	/* Careful. Some cpus do not strictly honor the set of cpus
@@ -627,14 +627,12 @@ flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
 	 */
 	cpumask_clear(retmask);
 	cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
-	return false;
 }
 
-static inline bool
+static inline void
 default_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	cpumask_copy(retmask, cpumask_of(cpu));
-	return true;
 }
 
 static inline unsigned long default_check_apicid_used(physid_mask_t *map, int apicid)
diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index 65c07fc630a..08c337bc49f 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -100,12 +100,11 @@ static unsigned long noop_check_apicid_present(int bit)
 	return physid_isset(bit, phys_cpu_present_map);
 }
 
-static bool noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	if (cpu != 0)
 		pr_warning("APIC: Vector allocated for non-BSP cpu\n");
 	cpumask_copy(retmask, cpumask_of(cpu));
-	return true;
 }
 
 static u32 noop_apic_read(u32 reg)
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index a951ef7decb..8a08f09aa50 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -1134,12 +1134,13 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask)
 
 	/* Only try and allocate irqs on cpus that are present */
 	err = -ENOSPC;
-	for_each_cpu_and(cpu, mask, cpu_online_mask) {
+	cpumask_clear(cfg->old_domain);
+	cpu = cpumask_first_and(mask, cpu_online_mask);
+	while (cpu < nr_cpu_ids) {
 		int new_cpu;
 		int vector, offset;
-		bool more_domains;
 
-		more_domains = apic->vector_allocation_domain(cpu, tmp_mask);
+		apic->vector_allocation_domain(cpu, tmp_mask);
 
 		if (cpumask_subset(tmp_mask, cfg->domain)) {
 			free_cpumask_var(tmp_mask);
@@ -1156,10 +1157,10 @@ next:
 		}
 
 		if (unlikely(current_vector == vector)) {
-			if (more_domains)
-				continue;
-			else
-				break;
+			cpumask_or(cfg->old_domain, cfg->old_domain, tmp_mask);
+			cpumask_andnot(tmp_mask, mask, cfg->old_domain);
+			cpu = cpumask_first_and(tmp_mask, cpu_online_mask);
+			continue;
 		}
 
 		if (test_bit(vector, used_vectors))
diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c
index 943d03fc6fc..b5d889b5659 100644
--- a/arch/x86/kernel/apic/x2apic_cluster.c
+++ b/arch/x86/kernel/apic/x2apic_cluster.c
@@ -212,11 +212,10 @@ static int x2apic_cluster_probe(void)
 /*
  * Each x2apic cluster is an allocation domain.
  */
-static bool cluster_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void cluster_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	cpumask_clear(retmask);
 	cpumask_copy(retmask, per_cpu(cpus_in_cluster, cpu));
-	return true;
 }
 
 static struct apic apic_x2apic_cluster = {
diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c
index fa5adb7c228..3f0285ac00f 100644
--- a/arch/x86/kernel/vsmp_64.c
+++ b/arch/x86/kernel/vsmp_64.c
@@ -208,10 +208,9 @@ static int apicid_phys_pkg_id(int initial_apic_id, int index_msb)
  * In vSMP, all cpus should be capable of handling interrupts, regardless of
  * the APIC used.
  */
-static bool fill_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void fill_vector_allocation_domain(int cpu, struct cpumask *retmask)
 {
 	cpumask_setall(retmask);
-	return false;
 }
 
 static void vsmp_apic_post_init(void)
-- 
cgit v1.2.3-70-g09d2


From 1ac322d0b169c95ce34d55b3ed6d40ce1a5f3a02 Mon Sep 17 00:00:00 2001
From: Suresh Siddha <suresh.b.siddha@intel.com>
Date: Mon, 25 Jun 2012 13:38:28 -0700
Subject: x86/apic/x2apic: Limit the vector reservation to the user specified
 mask

For the x2apic cluster mode, vector for an interrupt is
currently reserved on all the cpu's that are part of the x2apic
cluster. But the interrupts will be routed only to the cluster
(derived from the first cpu in the mask) members specified in
the mask. So there is no need to reserve the vector in the
unused cluster members.

Modify __assign_irq_vector() to reserve the vectors based on the
user specified irq destination mask. If the new mask is a proper
subset of the currently used mask, cleanup the vector allocation
on the unused cpu members.

Also, allow the apic driver to tune the vector domain based on
the affinity mask (which in most cases is the user-specified
mask).

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Alexander Gordeev <agordeev@redhat.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Link: http://lkml.kernel.org/r/1340656709-11423-3-git-send-email-suresh.b.siddha@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/include/asm/apic.h           |  9 ++++++---
 arch/x86/kernel/apic/apic_noop.c      |  3 ++-
 arch/x86/kernel/apic/io_apic.c        | 31 +++++++++++++++----------------
 arch/x86/kernel/apic/x2apic_cluster.c |  6 +++---
 arch/x86/kernel/vsmp_64.c             |  3 ++-
 5 files changed, 28 insertions(+), 24 deletions(-)

(limited to 'arch/x86/kernel/apic/apic_noop.c')

diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
index 8bebeb8952f..88093c1d44f 100644
--- a/arch/x86/include/asm/apic.h
+++ b/arch/x86/include/asm/apic.h
@@ -306,7 +306,8 @@ struct apic {
 	unsigned long (*check_apicid_used)(physid_mask_t *map, int apicid);
 	unsigned long (*check_apicid_present)(int apicid);
 
-	void (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
+	void (*vector_allocation_domain)(int cpu, struct cpumask *retmask,
+					 const struct cpumask *mask);
 	void (*init_apic_ldr)(void);
 
 	void (*ioapic_phys_id_map)(physid_mask_t *phys_map, physid_mask_t *retmap);
@@ -615,7 +616,8 @@ default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
 			       unsigned int *apicid);
 
 static inline void
-flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
+flat_vector_allocation_domain(int cpu, struct cpumask *retmask,
+			      const struct cpumask *mask)
 {
 	/* Careful. Some cpus do not strictly honor the set of cpus
 	 * specified in the interrupt destination when using lowest
@@ -630,7 +632,8 @@ flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
 }
 
 static inline void
-default_vector_allocation_domain(int cpu, struct cpumask *retmask)
+default_vector_allocation_domain(int cpu, struct cpumask *retmask,
+				 const struct cpumask *mask)
 {
 	cpumask_copy(retmask, cpumask_of(cpu));
 }
diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
index 08c337bc49f..e145f28b409 100644
--- a/arch/x86/kernel/apic/apic_noop.c
+++ b/arch/x86/kernel/apic/apic_noop.c
@@ -100,7 +100,8 @@ static unsigned long noop_check_apicid_present(int bit)
 	return physid_isset(bit, phys_cpu_present_map);
 }
 
-static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask,
+					  const struct cpumask *mask)
 {
 	if (cpu != 0)
 		pr_warning("APIC: Vector allocated for non-BSP cpu\n");
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 8a08f09aa50..9684f963bef 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -1113,7 +1113,6 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask)
 	 */
 	static int current_vector = FIRST_EXTERNAL_VECTOR + VECTOR_OFFSET_START;
 	static int current_offset = VECTOR_OFFSET_START % 16;
-	unsigned int old_vector;
 	int cpu, err;
 	cpumask_var_t tmp_mask;
 
@@ -1123,28 +1122,28 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask)
 	if (!alloc_cpumask_var(&tmp_mask, GFP_ATOMIC))
 		return -ENOMEM;
 
-	old_vector = cfg->vector;
-	if (old_vector) {
-		cpumask_and(tmp_mask, mask, cpu_online_mask);
-		if (cpumask_subset(tmp_mask, cfg->domain)) {
-			free_cpumask_var(tmp_mask);
-			return 0;
-		}
-	}
-
 	/* Only try and allocate irqs on cpus that are present */
 	err = -ENOSPC;
 	cpumask_clear(cfg->old_domain);
 	cpu = cpumask_first_and(mask, cpu_online_mask);
 	while (cpu < nr_cpu_ids) {
-		int new_cpu;
-		int vector, offset;
+		int new_cpu, vector, offset;
 
-		apic->vector_allocation_domain(cpu, tmp_mask);
+		apic->vector_allocation_domain(cpu, tmp_mask, mask);
 
 		if (cpumask_subset(tmp_mask, cfg->domain)) {
-			free_cpumask_var(tmp_mask);
-			return 0;
+			err = 0;
+			if (cpumask_equal(tmp_mask, cfg->domain))
+				break;
+			/*
+			 * New cpumask using the vector is a proper subset of
+			 * the current in use mask. So cleanup the vector
+			 * allocation for the members that are not used anymore.
+			 */
+			cpumask_andnot(cfg->old_domain, cfg->domain, tmp_mask);
+			cfg->move_in_progress = 1;
+			cpumask_and(cfg->domain, cfg->domain, tmp_mask);
+			break;
 		}
 
 		vector = current_vector;
@@ -1172,7 +1171,7 @@ next:
 		/* Found one! */
 		current_vector = vector;
 		current_offset = offset;
-		if (old_vector) {
+		if (cfg->vector) {
 			cfg->move_in_progress = 1;
 			cpumask_copy(cfg->old_domain, cfg->domain);
 		}
diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c
index b5d889b5659..bde78d0098a 100644
--- a/arch/x86/kernel/apic/x2apic_cluster.c
+++ b/arch/x86/kernel/apic/x2apic_cluster.c
@@ -212,10 +212,10 @@ static int x2apic_cluster_probe(void)
 /*
  * Each x2apic cluster is an allocation domain.
  */
-static void cluster_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void cluster_vector_allocation_domain(int cpu, struct cpumask *retmask,
+					     const struct cpumask *mask)
 {
-	cpumask_clear(retmask);
-	cpumask_copy(retmask, per_cpu(cpus_in_cluster, cpu));
+	cpumask_and(retmask, mask, per_cpu(cpus_in_cluster, cpu));
 }
 
 static struct apic apic_x2apic_cluster = {
diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c
index 3f0285ac00f..992f890283e 100644
--- a/arch/x86/kernel/vsmp_64.c
+++ b/arch/x86/kernel/vsmp_64.c
@@ -208,7 +208,8 @@ static int apicid_phys_pkg_id(int initial_apic_id, int index_msb)
  * In vSMP, all cpus should be capable of handling interrupts, regardless of
  * the APIC used.
  */
-static void fill_vector_allocation_domain(int cpu, struct cpumask *retmask)
+static void fill_vector_allocation_domain(int cpu, struct cpumask *retmask,
+					  const struct cpumask *mask)
 {
 	cpumask_setall(retmask);
 }
-- 
cgit v1.2.3-70-g09d2