summaryrefslogtreecommitdiffstats
path: root/arch/x86/kernel/apic
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kernel/apic')
-rw-r--r--arch/x86/kernel/apic/apic.c112
-rw-r--r--arch/x86/kernel/apic/es7000_32.c5
-rw-r--r--arch/x86/kernel/apic/io_apic.c375
-rw-r--r--arch/x86/kernel/apic/ipi.c5
-rw-r--r--arch/x86/kernel/apic/nmi.c20
-rw-r--r--arch/x86/kernel/apic/numaq_32.c3
-rw-r--r--arch/x86/kernel/apic/probe_32.c11
-rw-r--r--arch/x86/kernel/apic/probe_64.c16
-rw-r--r--arch/x86/kernel/apic/summit_32.c1
-rw-r--r--arch/x86/kernel/apic/x2apic_cluster.c10
-rw-r--r--arch/x86/kernel/apic/x2apic_phys.c10
-rw-r--r--arch/x86/kernel/apic/x2apic_uv_x.c42
12 files changed, 318 insertions, 292 deletions
diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index 8c7c042ecad..159740decc4 100644
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -49,6 +49,7 @@
#include <asm/mtrr.h>
#include <asm/smp.h>
#include <asm/mce.h>
+#include <asm/kvm_para.h>
unsigned int num_processors;
@@ -140,7 +141,6 @@ int x2apic_mode;
#ifdef CONFIG_X86_X2APIC
/* x2apic enabled before OS handover */
static int x2apic_preenabled;
-static int disable_x2apic;
static __init int setup_nox2apic(char *str)
{
if (x2apic_enabled()) {
@@ -149,7 +149,6 @@ static __init int setup_nox2apic(char *str)
return 0;
}
- disable_x2apic = 1;
setup_clear_cpu_cap(X86_FEATURE_X2APIC);
return 0;
}
@@ -1363,52 +1362,80 @@ void enable_x2apic(void)
}
#endif /* CONFIG_X86_X2APIC */
-void __init enable_IR_x2apic(void)
+int __init enable_IR(void)
{
#ifdef CONFIG_INTR_REMAP
- int ret;
- unsigned long flags;
- struct IO_APIC_route_entry **ioapic_entries = NULL;
-
- ret = dmar_table_init();
- if (ret) {
- pr_debug("dmar_table_init() failed with %d:\n", ret);
- goto ir_failed;
- }
-
if (!intr_remapping_supported()) {
pr_debug("intr-remapping not supported\n");
- goto ir_failed;
+ return 0;
}
-
if (!x2apic_preenabled && skip_ioapic_setup) {
pr_info("Skipped enabling intr-remap because of skipping "
"io-apic setup\n");
- return;
+ return 0;
}
+ if (enable_intr_remapping(x2apic_supported()))
+ return 0;
+
+ pr_info("Enabled Interrupt-remapping\n");
+
+ return 1;
+
+#endif
+ return 0;
+}
+
+void __init enable_IR_x2apic(void)
+{
+ unsigned long flags;
+ struct IO_APIC_route_entry **ioapic_entries = NULL;
+ int ret, x2apic_enabled = 0;
+ int dmar_table_init_ret = 0;
+
+#ifdef CONFIG_INTR_REMAP
+ dmar_table_init_ret = dmar_table_init();
+ if (dmar_table_init_ret)
+ pr_debug("dmar_table_init() failed with %d:\n",
+ dmar_table_init_ret);
+#endif
+
ioapic_entries = alloc_ioapic_entries();
if (!ioapic_entries) {
- pr_info("Allocate ioapic_entries failed: %d\n", ret);
- goto end;
+ pr_err("Allocate ioapic_entries failed\n");
+ goto out;
}
ret = save_IO_APIC_setup(ioapic_entries);
if (ret) {
pr_info("Saving IO-APIC state failed: %d\n", ret);
- goto end;
+ goto out;
}
local_irq_save(flags);
- mask_IO_APIC_setup(ioapic_entries);
mask_8259A();
+ mask_IO_APIC_setup(ioapic_entries);
- ret = enable_intr_remapping(x2apic_supported());
- if (ret)
- goto end_restore;
+ if (dmar_table_init_ret)
+ ret = 0;
+ else
+ ret = enable_IR();
- pr_info("Enabled Interrupt-remapping\n");
+ if (!ret) {
+ /* IR is required if there is APIC ID > 255 even when running
+ * under KVM
+ */
+ if (max_physical_apicid > 255 || !kvm_para_available())
+ goto nox2apic;
+ /*
+ * without IR all CPUs can be addressed by IOAPIC/MSI
+ * only in physical mode
+ */
+ x2apic_force_phys();
+ }
+
+ x2apic_enabled = 1;
if (x2apic_supported() && !x2apic_mode) {
x2apic_mode = 1;
@@ -1416,41 +1443,25 @@ void __init enable_IR_x2apic(void)
pr_info("Enabled x2apic\n");
}
-end_restore:
- if (ret)
- /*
- * IR enabling failed
- */
+nox2apic:
+ if (!ret) /* IR enabling failed */
restore_IO_APIC_setup(ioapic_entries);
-
unmask_8259A();
local_irq_restore(flags);
-end:
+out:
if (ioapic_entries)
free_ioapic_entries(ioapic_entries);
- if (!ret)
+ if (x2apic_enabled)
return;
-ir_failed:
if (x2apic_preenabled)
- panic("x2apic enabled by bios. But IR enabling failed");
+ panic("x2apic: enabled by BIOS but kernel init failed.");
else if (cpu_has_x2apic)
- pr_info("Not enabling x2apic,Intr-remapping\n");
-#else
- if (!cpu_has_x2apic)
- return;
-
- if (x2apic_preenabled)
- panic("x2apic enabled prior OS handover,"
- " enable CONFIG_X86_X2APIC, CONFIG_INTR_REMAP");
-#endif
-
- return;
+ pr_info("Not enabling x2apic, Intr-remapping init failed.\n");
}
-
#ifdef CONFIG_X86_64
/*
* Detect and enable local APICs on non-SMP boards.
@@ -1551,8 +1562,6 @@ no_apic:
#ifdef CONFIG_X86_64
void __init early_init_lapic_mapping(void)
{
- unsigned long phys_addr;
-
/*
* If no local APIC can be found then go out
* : it means there is no mpatable and MADT
@@ -1560,11 +1569,9 @@ void __init early_init_lapic_mapping(void)
if (!smp_found_config)
return;
- phys_addr = mp_lapic_addr;
-
- set_fixmap_nocache(FIX_APIC_BASE, phys_addr);
+ set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr);
apic_printk(APIC_VERBOSE, "mapped APIC to %16lx (%16lx)\n",
- APIC_BASE, phys_addr);
+ APIC_BASE, mp_lapic_addr);
/*
* Fetch the APIC ID of the BSP in case we have a
@@ -1653,7 +1660,6 @@ int __init APIC_init_uniprocessor(void)
APIC_INTEGRATED(apic_version[boot_cpu_physical_apicid])) {
pr_err("BIOS bug, local APIC 0x%x not detected!...\n",
boot_cpu_physical_apicid);
- clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC);
return -1;
}
#endif
diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c
index 69328ac8de9..89174f847b4 100644
--- a/arch/x86/kernel/apic/es7000_32.c
+++ b/arch/x86/kernel/apic/es7000_32.c
@@ -167,7 +167,7 @@ static int es7000_apic_is_cluster(void)
{
/* MPENTIUMIII */
if (boot_cpu_data.x86 == 6 &&
- (boot_cpu_data.x86_model >= 7 || boot_cpu_data.x86_model <= 11))
+ (boot_cpu_data.x86_model >= 7 && boot_cpu_data.x86_model <= 11))
return 1;
return 0;
@@ -652,7 +652,8 @@ static int es7000_mps_oem_check_cluster(struct mpc_table *mpc, char *oem,
return ret && es7000_apic_is_cluster();
}
-struct apic apic_es7000_cluster = {
+/* We've been warned by a false positive warning.Use __refdata to keep calm. */
+struct apic __refdata apic_es7000_cluster = {
.name = "es7000",
.probe = probe_es7000,
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 29d0752d951..3c8f9e75d03 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -66,6 +66,8 @@
#include <asm/apic.h>
#define __apicdebuginit(type) static type __init
+#define for_each_irq_pin(entry, head) \
+ for (entry = head; entry; entry = entry->next)
/*
* Is the SiS APIC rmw bug present ?
@@ -85,6 +87,9 @@ int nr_ioapic_registers[MAX_IO_APICS];
struct mpc_ioapic mp_ioapics[MAX_IO_APICS];
int nr_ioapics;
+/* IO APIC gsi routing info */
+struct mp_ioapic_gsi mp_gsi_routing[MAX_IO_APICS];
+
/* MP IRQ source entries */
struct mpc_intsrc mp_irqs[MAX_IRQ_SOURCES];
@@ -116,15 +121,6 @@ static int __init parse_noapic(char *str)
}
early_param("noapic", parse_noapic);
-struct irq_pin_list;
-
-/*
- * This is performance-critical, we want to do it O(1)
- *
- * the indexing order of this array favors 1:1 mappings
- * between pins and IRQs.
- */
-
struct irq_pin_list {
int apic, pin;
struct irq_pin_list *next;
@@ -139,6 +135,11 @@ static struct irq_pin_list *get_one_free_irq_2_pin(int node)
return pin;
}
+/*
+ * This is performance-critical, we want to do it O(1)
+ *
+ * Most irqs are mapped 1:1 with pins.
+ */
struct irq_cfg {
struct irq_pin_list *irq_2_pin;
cpumask_var_t domain;
@@ -414,13 +415,10 @@ static bool io_apic_level_ack_pending(struct irq_cfg *cfg)
unsigned long flags;
spin_lock_irqsave(&ioapic_lock, flags);
- entry = cfg->irq_2_pin;
- for (;;) {
+ for_each_irq_pin(entry, cfg->irq_2_pin) {
unsigned int reg;
int pin;
- if (!entry)
- break;
pin = entry->pin;
reg = io_apic_read(entry->apic, 0x10 + pin*2);
/* Is the remote IRR bit set? */
@@ -428,9 +426,6 @@ static bool io_apic_level_ack_pending(struct irq_cfg *cfg)
spin_unlock_irqrestore(&ioapic_lock, flags);
return true;
}
- if (!entry->next)
- break;
- entry = entry->next;
}
spin_unlock_irqrestore(&ioapic_lock, flags);
@@ -498,72 +493,68 @@ static void ioapic_mask_entry(int apic, int pin)
* shared ISA-space IRQs, so we have to support them. We are super
* fast in the common case, and fast for shared ISA-space IRQs.
*/
-static void add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin)
+static int
+add_pin_to_irq_node_nopanic(struct irq_cfg *cfg, int node, int apic, int pin)
{
- struct irq_pin_list *entry;
+ struct irq_pin_list **last, *entry;
- entry = cfg->irq_2_pin;
- if (!entry) {
- entry = get_one_free_irq_2_pin(node);
- if (!entry) {
- printk(KERN_ERR "can not alloc irq_2_pin to add %d - %d\n",
- apic, pin);
- return;
- }
- cfg->irq_2_pin = entry;
- entry->apic = apic;
- entry->pin = pin;
- return;
- }
-
- while (entry->next) {
- /* not again, please */
+ /* don't allow duplicates */
+ last = &cfg->irq_2_pin;
+ for_each_irq_pin(entry, cfg->irq_2_pin) {
if (entry->apic == apic && entry->pin == pin)
- return;
-
- entry = entry->next;
+ return 0;
+ last = &entry->next;
}
- entry->next = get_one_free_irq_2_pin(node);
- entry = entry->next;
+ entry = get_one_free_irq_2_pin(node);
+ if (!entry) {
+ printk(KERN_ERR "can not alloc irq_pin_list (%d,%d,%d)\n",
+ node, apic, pin);
+ return -ENOMEM;
+ }
entry->apic = apic;
entry->pin = pin;
+
+ *last = entry;
+ return 0;
+}
+
+static void add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin)
+{
+ if (add_pin_to_irq_node_nopanic(cfg, node, apic, pin))
+ panic("IO-APIC: failed to add irq-pin. Can not proceed\n");
}
/*
* Reroute an IRQ to a different pin.
*/
static void __init replace_pin_at_irq_node(struct irq_cfg *cfg, int node,
- int oldapic, int oldpin,
- int newapic, int newpin)
+ int oldapic, int oldpin,
+ int newapic, int newpin)
{
- struct irq_pin_list *entry = cfg->irq_2_pin;
- int replaced = 0;
+ struct irq_pin_list *entry;
- while (entry) {
+ for_each_irq_pin(entry, cfg->irq_2_pin) {
if (entry->apic == oldapic && entry->pin == oldpin) {
entry->apic = newapic;
entry->pin = newpin;
- replaced = 1;
/* every one is different, right? */
- break;
+ return;
}
- entry = entry->next;
}
- /* why? call replace before add? */
- if (!replaced)
- add_pin_to_irq_node(cfg, node, newapic, newpin);
+ /* old apic/pin didn't exist, so just add new ones */
+ add_pin_to_irq_node(cfg, node, newapic, newpin);
}
-static inline void io_apic_modify_irq(struct irq_cfg *cfg,
- int mask_and, int mask_or,
- void (*final)(struct irq_pin_list *entry))
+static void io_apic_modify_irq(struct irq_cfg *cfg,
+ int mask_and, int mask_or,
+ void (*final)(struct irq_pin_list *entry))
{
int pin;
struct irq_pin_list *entry;
- for (entry = cfg->irq_2_pin; entry != NULL; entry = entry->next) {
+ for_each_irq_pin(entry, cfg->irq_2_pin) {
unsigned int reg;
pin = entry->pin;
reg = io_apic_read(entry->apic, 0x10 + pin * 2);
@@ -580,7 +571,6 @@ static void __unmask_IO_APIC_irq(struct irq_cfg *cfg)
io_apic_modify_irq(cfg, ~IO_APIC_REDIR_MASKED, 0, NULL);
}
-#ifdef CONFIG_X86_64
static void io_apic_sync(struct irq_pin_list *entry)
{
/*
@@ -596,11 +586,6 @@ static void __mask_IO_APIC_irq(struct irq_cfg *cfg)
{
io_apic_modify_irq(cfg, ~0, IO_APIC_REDIR_MASKED, &io_apic_sync);
}
-#else /* CONFIG_X86_32 */
-static void __mask_IO_APIC_irq(struct irq_cfg *cfg)
-{
- io_apic_modify_irq(cfg, ~0, IO_APIC_REDIR_MASKED, NULL);
-}
static void __mask_and_edge_IO_APIC_irq(struct irq_cfg *cfg)
{
@@ -613,7 +598,6 @@ static void __unmask_and_level_IO_APIC_irq(struct irq_cfg *cfg)
io_apic_modify_irq(cfg, ~IO_APIC_REDIR_MASKED,
IO_APIC_REDIR_LEVEL_TRIGGER, NULL);
}
-#endif /* CONFIG_X86_32 */
static void mask_IO_APIC_irq_desc(struct irq_desc *desc)
{
@@ -1414,6 +1398,9 @@ int setup_ioapic_entry(int apic_id, int irq,
irte.vector = vector;
irte.dest_id = IRTE_DEST(destination);
+ /* Set source-id of interrupt request */
+ set_ioapic_sid(&irte, apic_id);
+
modify_irte(irq, &irte);
ir_entry->index2 = (index >> 15) & 0x1;
@@ -1699,12 +1686,8 @@ __apicdebuginit(void) print_IO_APIC(void)
if (!entry)
continue;
printk(KERN_DEBUG "IRQ%d ", irq);
- for (;;) {
+ for_each_irq_pin(entry, cfg->irq_2_pin)
printk("-> %d:%d", entry->apic, entry->pin);
- if (!entry->next)
- break;
- entry = entry->next;
- }
printk("\n");
}
@@ -1713,25 +1696,19 @@ __apicdebuginit(void) print_IO_APIC(void)
return;
}
-__apicdebuginit(void) print_APIC_bitfield(int base)
+__apicdebuginit(void) print_APIC_field(int base)
{
- unsigned int v;
- int i, j;
+ int i;
if (apic_verbosity == APIC_QUIET)
return;
- printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG);
- for (i = 0; i < 8; i++) {
- v = apic_read(base + i*0x10);
- for (j = 0; j < 32; j++) {
- if (v & (1<<j))
- printk("1");
- else
- printk("0");
- }
- printk("\n");
- }
+ printk(KERN_DEBUG);
+
+ for (i = 0; i < 8; i++)
+ printk(KERN_CONT "%08x", apic_read(base + i*0x10));
+
+ printk(KERN_CONT "\n");
}
__apicdebuginit(void) print_local_APIC(void *dummy)
@@ -1742,7 +1719,7 @@ __apicdebuginit(void) print_local_APIC(void *dummy)
if (apic_verbosity == APIC_QUIET)
return;
- printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
+ printk(KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
smp_processor_id(), hard_smp_processor_id());
v = apic_read(APIC_ID);
printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, read_apic_id());
@@ -1783,11 +1760,11 @@ __apicdebuginit(void) print_local_APIC(void *dummy)
printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
printk(KERN_DEBUG "... APIC ISR field:\n");
- print_APIC_bitfield(APIC_ISR);
+ print_APIC_field(APIC_ISR);
printk(KERN_DEBUG "... APIC TMR field:\n");
- print_APIC_bitfield(APIC_TMR);
+ print_APIC_field(APIC_TMR);
printk(KERN_DEBUG "... APIC IRR field:\n");
- print_APIC_bitfield(APIC_IRR);
+ print_APIC_field(APIC_IRR);
if (APIC_INTEGRATED(ver)) { /* !82489DX */
if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */
@@ -2004,7 +1981,9 @@ void disable_IO_APIC(void)
/*
* Use virtual wire A mode when interrupt remapping is enabled.
*/
- disconnect_bsp_APIC(!intr_remapping_enabled && ioapic_i8259.pin != -1);
+ if (cpu_has_apic)
+ disconnect_bsp_APIC(!intr_remapping_enabled &&
+ ioapic_i8259.pin != -1);
}
#ifdef CONFIG_X86_32
@@ -2212,7 +2191,6 @@ static unsigned int startup_ioapic_irq(unsigned int irq)
return was_pending;
}
-#ifdef CONFIG_X86_64
static int ioapic_retrigger_irq(unsigned int irq)
{
@@ -2225,14 +2203,6 @@ static int ioapic_retrigger_irq(unsigned int irq)
return 1;
}
-#else
-static int ioapic_retrigger_irq(unsigned int irq)
-{
- apic->send_IPI_self(irq_cfg(irq)->vector);
-
- return 1;
-}
-#endif
/*
* Level and edge triggered IO-APIC interrupts need different handling,
@@ -2270,13 +2240,9 @@ static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq
struct irq_pin_list *entry;
u8 vector = cfg->vector;
- entry = cfg->irq_2_pin;
- for (;;) {
+ for_each_irq_pin(entry, cfg->irq_2_pin) {
unsigned int reg;
- if (!entry)
- break;
-
apic = entry->apic;
pin = entry->pin;
/*
@@ -2289,9 +2255,6 @@ static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq
reg &= ~IO_APIC_REDIR_VECTOR_MASK;
reg |= vector;
io_apic_modify(apic, 0x10 + pin*2, reg);
- if (!entry->next)
- break;
- entry = entry->next;
}
}
@@ -2516,11 +2479,8 @@ atomic_t irq_mis_count;
static void ack_apic_level(unsigned int irq)
{
struct irq_desc *desc = irq_to_desc(irq);
-
-#ifdef CONFIG_X86_32
unsigned long v;
int i;
-#endif
struct irq_cfg *cfg;
int do_unmask_irq = 0;
@@ -2533,31 +2493,28 @@ static void ack_apic_level(unsigned int irq)
}
#endif
-#ifdef CONFIG_X86_32
/*
- * It appears there is an erratum which affects at least version 0x11
- * of I/O APIC (that's the 82093AA and cores integrated into various
- * chipsets). Under certain conditions a level-triggered interrupt is
- * erroneously delivered as edge-triggered one but the respective IRR
- * bit gets set nevertheless. As a result the I/O unit expects an EOI
- * message but it will never arrive and further interrupts are blocked
- * from the source. The exact reason is so far unknown, but the
- * phenomenon was observed when two consecutive interrupt requests
- * from a given source get delivered to the same CPU and the source is
- * temporarily disabled in between.
- *
- * A workaround is to simulate an EOI message manually. We achieve it
- * by setting the trigger mode to edge and then to level when the edge
- * trigger mode gets detected in the TMR of a local APIC for a
- * level-triggered interrupt. We mask the source for the time of the
- * operation to prevent an edge-triggered interrupt escaping meanwhile.
- * The idea is from Manfred Spraul. --macro
- */
+ * It appears there is an erratum which affects at least version 0x11
+ * of I/O APIC (that's the 82093AA and cores integrated into various
+ * chipsets). Under certain conditions a level-triggered interrupt is
+ * erroneously delivered as edge-triggered one but the respective IRR
+ * bit gets set nevertheless. As a result the I/O unit expects an EOI
+ * message but it will never arrive and further interrupts are blocked
+ * from the source. The exact reason is so far unknown, but the
+ * phenomenon was observed when two consecutive interrupt requests
+ * from a given source get delivered to the same CPU and the source is
+ * temporarily disabled in between.
+ *
+ * A workaround is to simulate an EOI message manually. We achieve it
+ * by setting the trigger mode to edge and then to level when the edge
+ * trigger mode gets detected in the TMR of a local APIC for a
+ * level-triggered interrupt. We mask the source for the time of the
+ * operation to prevent an edge-triggered interrupt escaping meanwhile.
+ * The idea is from Manfred Spraul. --macro
+ */
cfg = desc->chip_data;
i = cfg->vector;
-
v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-#endif
/*
* We must acknowledge the irq before we move it or the acknowledge will
@@ -2599,7 +2556,7 @@ static void ack_apic_level(unsigned int irq)
unmask_IO_APIC_irq_desc(desc);
}
-#ifdef CONFIG_X86_32
+ /* Tail end of version 0x11 I/O APIC bug workaround */
if (!(v & (1 << (i & 0x1f)))) {
atomic_inc(&irq_mis_count);
spin_lock(&ioapic_lock);
@@ -2607,26 +2564,15 @@ static void ack_apic_level(unsigned int irq)
__unmask_and_level_IO_APIC_irq(cfg);
spin_unlock(&ioapic_lock);
}
-#endif
}
#ifdef CONFIG_INTR_REMAP
static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg)
{
- int apic, pin;
struct irq_pin_list *entry;
- entry = cfg->irq_2_pin;
- for (;;) {
-
- if (!entry)
- break;
-
- apic = entry->apic;
- pin = entry->pin;
- io_apic_eoi(apic, pin);
- entry = entry->next;
- }
+ for_each_irq_pin(entry, cfg->irq_2_pin)
+ io_apic_eoi(entry->apic, entry->pin);
}
static void
@@ -3242,8 +3188,7 @@ void destroy_irq(unsigned int irq)
cfg = desc->chip_data;
dynamic_irq_cleanup(irq);
/* connect back irq_cfg */
- if (desc)
- desc->chip_data = cfg;
+ desc->chip_data = cfg;
free_irte(irq);
spin_lock_irqsave(&vector_lock, flags);
@@ -3288,6 +3233,9 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms
irte.vector = cfg->vector;
irte.dest_id = IRTE_DEST(dest);
+ /* Set source-id of interrupt request */
+ set_msi_sid(&irte, pdev);
+
modify_irte(irq, &irte);
msg->address_hi = MSI_ADDR_BASE_HI;
@@ -3791,6 +3739,9 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade,
mmr_pnode = uv_blade_to_pnode(mmr_blade);
uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value);
+ if (cfg->move_in_progress)
+ send_cleanup_vector(cfg);
+
return irq;
}
@@ -3907,7 +3858,11 @@ static int __io_apic_set_pci_routing(struct device *dev, int irq,
*/
if (irq >= NR_IRQS_LEGACY) {
cfg = desc->chip_data;
- add_pin_to_irq_node(cfg, node, ioapic, pin);
+ if (add_pin_to_irq_node_nopanic(cfg, node, ioapic, pin)) {
+ printk(KERN_INFO "can not add pin %d for irq %d\n",
+ pin, irq);
+ return 0;
+ }
}
setup_IO_APIC_irq(ioapic, pin, irq, desc, trigger, polarity);
@@ -3936,11 +3891,28 @@ int io_apic_set_pci_routing(struct device *dev, int irq,
return __io_apic_set_pci_routing(dev, irq, irq_attr);
}
-/* --------------------------------------------------------------------------
- ACPI-based IOAPIC Configuration
- -------------------------------------------------------------------------- */
+u8 __init io_apic_unique_id(u8 id)
+{
+#ifdef CONFIG_X86_32
+ if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) &&
+ !APIC_XAPIC(apic_version[boot_cpu_physical_apicid]))
+ return io_apic_get_unique_id(nr_ioapics, id);
+ else
+ return id;
+#else
+ int i;
+ DECLARE_BITMAP(used, 256);
-#ifdef CONFIG_ACPI
+ bitmap_zero(used, 256);
+ for (i = 0; i < nr_ioapics; i++) {
+ struct mpc_ioapic *ia = &mp_ioapics[i];
+ __set_bit(ia->apicid, used);
+ }
+ if (!test_bit(id, used))
+ return id;
+ return find_first_zero_bit(used, 256);
+#endif
+}
#ifdef CONFIG_X86_32
int __init io_apic_get_unique_id(int ioapic, int apic_id)
@@ -4049,8 +4021,6 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
return 0;
}
-#endif /* CONFIG_ACPI */
-
/*
* This function currently is only a helper for the i386 smp boot process where
* we need to reprogram the ioredtbls to cater for the cpus which have come online
@@ -4104,7 +4074,7 @@ void __init setup_ioapic_dest(void)
static struct resource *ioapic_resources;
-static struct resource * __init ioapic_setup_resources(void)
+static struct resource * __init ioapic_setup_resources(int nr_ioapics)
{
unsigned long n;
struct resource *res;
@@ -4120,15 +4090,13 @@ static struct resource * __init ioapic_setup_resources(void)
mem = alloc_bootmem(n);
res = (void *)mem;
- if (mem != NULL) {
- mem += sizeof(struct resource) * nr_ioapics;
+ mem += sizeof(struct resource) * nr_ioapics;
- for (i = 0; i < nr_ioapics; i++) {
- res[i].name = mem;
- res[i].flags = IORESOURCE_MEM | IORESOURCE_BUSY;
- sprintf(mem, "IOAPIC %u", i);
- mem += IOAPIC_RESOURCE_NAME_SIZE;
- }
+ for (i = 0; i < nr_ioapics; i++) {
+ res[i].name = mem;
+ res[i].flags = IORESOURCE_MEM | IORESOURCE_BUSY;
+ sprintf(mem, "IOAPIC %u", i);
+ mem += IOAPIC_RESOURCE_NAME_SIZE;
}
ioapic_resources = res;
@@ -4142,7 +4110,7 @@ void __init ioapic_init_mappings(void)
struct resource *ioapic_res;
int i;
- ioapic_res = ioapic_setup_resources();
+ ioapic_res = ioapic_setup_resources(nr_ioapics);
for (i = 0; i < nr_ioapics; i++) {
if (smp_found_config) {
ioapic_phys = mp_ioapics[i].apicaddr;
@@ -4171,36 +4139,99 @@ fake_ioapic_page:
__fix_to_virt(idx), ioapic_phys);
idx++;
- if (ioapic_res != NULL) {
- ioapic_res->start = ioapic_phys;
- ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
- ioapic_res++;
- }
+ ioapic_res->start = ioapic_phys;
+ ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
+ ioapic_res++;
}
}
-static int __init ioapic_insert_resources(void)
+void __init ioapic_insert_resources(void)
{
int i;
struct resource *r = ioapic_resources;
if (!r) {
- if (nr_ioapics > 0) {
+ if (nr_ioapics > 0)
printk(KERN_ERR
"IO APIC resources couldn't be allocated.\n");
- return -1;
- }
- return 0;
+ return;
}
for (i = 0; i < nr_ioapics; i++) {
insert_resource(&iomem_resource, r);
r++;
}
+}
+
+int mp_find_ioapic(int gsi)
+{
+ int i = 0;
+
+ /* Find the IOAPIC that manages this GSI. */
+ for (i = 0; i < nr_ioapics; i++) {
+ if ((gsi >= mp_gsi_routing[i].gsi_base)
+ && (gsi <= mp_gsi_routing[i].gsi_end))
+ return i;
+ }
+
+ printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi);
+ return -1;
+}
+
+int mp_find_ioapic_pin(int ioapic, int gsi)
+{
+ if (WARN_ON(ioapic == -1))
+ return -1;
+ if (WARN_ON(gsi > mp_gsi_routing[ioapic].gsi_end))
+ return -1;
+
+ return gsi - mp_gsi_routing[ioapic].gsi_base;
+}
+static int bad_ioapic(unsigned long address)
+{
+ if (nr_ioapics >= MAX_IO_APICS) {
+ printk(KERN_WARNING "WARING: Max # of I/O APICs (%d) exceeded "
+ "(found %d), skipping\n", MAX_IO_APICS, nr_ioapics);
+ return 1;
+ }
+ if (!address) {
+ printk(KERN_WARNING "WARNING: Bogus (zero) I/O APIC address"
+ " found in table, skipping!\n");
+ return 1;
+ }
return 0;
}
-/* Insert the IO APIC resources after PCI initialization has occured to handle
- * IO APICS that are mapped in on a BAR in PCI space. */
-late_initcall(ioapic_insert_resources);
+void __init mp_register_ioapic(int id, u32 address, u32 gsi_base)
+{
+ int idx = 0;
+
+ if (bad_ioapic(address))
+ return;
+
+ idx = nr_ioapics;
+
+ mp_ioapics[idx].type = MP_IOAPIC;
+ mp_ioapics[idx].flags = MPC_APIC_USABLE;
+ mp_ioapics[idx].apicaddr = address;
+
+ set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address);
+ mp_ioapics[idx].apicid = io_apic_unique_id(id);
+ mp_ioapics[idx].apicver = io_apic_get_version(idx);
+
+ /*
+ * Build basic GSI lookup table to facilitate gsi->io_apic lookups
+ * and to prevent reprogramming of IOAPIC pins (PCI GSIs).
+ */
+ mp_gsi_routing[idx].gsi_base = gsi_base;
+ mp_gsi_routing[idx].gsi_end = gsi_base +
+ io_apic_get_redir_entries(idx);
+
+ printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, "
+ "GSI %d-%d\n", idx, mp_ioapics[idx].apicid,
+ mp_ioapics[idx].apicver, mp_ioapics[idx].apicaddr,
+ mp_gsi_routing[idx].gsi_base, mp_gsi_routing[idx].gsi_end);
+
+ nr_ioapics++;
+}
diff --git a/arch/x86/kernel/apic/ipi.c b/arch/x86/kernel/apic/ipi.c
index dbf5445727a..08385e090a6 100644
--- a/arch/x86/kernel/apic/ipi.c
+++ b/arch/x86/kernel/apic/ipi.c
@@ -106,6 +106,9 @@ void default_send_IPI_mask_logical(const struct cpumask *cpumask, int vector)
unsigned long mask = cpumask_bits(cpumask)[0];
unsigned long flags;
+ if (WARN_ONCE(!mask, "empty IPI mask"))
+ return;
+
local_irq_save(flags);
WARN_ON(mask & ~cpumask_bits(cpu_online_mask)[0]);
__default_send_IPI_dest_field(mask, vector, apic->dest_logical);
@@ -150,7 +153,7 @@ int safe_smp_processor_id(void)
{
int apicid, cpuid;
- if (!boot_cpu_has(X86_FEATURE_APIC))
+ if (!cpu_has_apic)
return 0;
apicid = hard_smp_processor_id();
diff --git a/arch/x86/kernel/apic/nmi.c b/arch/x86/kernel/apic/nmi.c
index f4227289caf..cb66a22d98a 100644
--- a/arch/x86/kernel/apic/nmi.c
+++ b/arch/x86/kernel/apic/nmi.c
@@ -39,7 +39,7 @@
int unknown_nmi_panic;
int nmi_watchdog_enabled;
-static cpumask_var_t backtrace_mask;
+static cpumask_t backtrace_mask __read_mostly;
/* nmi_active:
* >0: the lapic NMI watchdog is active, but can be disabled
@@ -138,7 +138,6 @@ int __init check_nmi_watchdog(void)
if (!prev_nmi_count)
goto error;
- alloc_cpumask_var(&backtrace_mask, GFP_KERNEL|__GFP_ZERO);
printk(KERN_INFO "Testing NMI watchdog ... ");
#ifdef CONFIG_SMP
@@ -415,14 +414,17 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason)
}
/* We can be called before check_nmi_watchdog, hence NULL check. */
- if (backtrace_mask != NULL && cpumask_test_cpu(cpu, backtrace_mask)) {
+ if (cpumask_test_cpu(cpu, &backtrace_mask)) {
static DEFINE_SPINLOCK(lock); /* Serialise the printks */
spin_lock(&lock);
printk(KERN_WARNING "NMI backtrace for cpu %d\n", cpu);
+ show_regs(regs);
dump_stack();
spin_unlock(&lock);
- cpumask_clear_cpu(cpu, backtrace_mask);
+ cpumask_clear_cpu(cpu, &backtrace_mask);
+
+ rc = 1;
}
/* Could check oops_in_progress here too, but it's safer not to */
@@ -552,14 +554,18 @@ int do_nmi_callback(struct pt_regs *regs, int cpu)
return 0;
}
-void __trigger_all_cpu_backtrace(void)
+void arch_trigger_all_cpu_backtrace(void)
{
int i;
- cpumask_copy(backtrace_mask, cpu_online_mask);
+ cpumask_copy(&backtrace_mask, cpu_online_mask);
+
+ printk(KERN_INFO "sending NMI to all CPUs:\n");
+ apic->send_IPI_all(NMI_VECTOR);
+
/* Wait for up to 10 seconds for all CPUs to do the backtrace */
for (i = 0; i < 10 * 1000; i++) {
- if (cpumask_empty(backtrace_mask))
+ if (cpumask_empty(&backtrace_mask))
break;
mdelay(1);
}
diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c
index 533e59c6fc8..ca96e68f0d2 100644
--- a/arch/x86/kernel/apic/numaq_32.c
+++ b/arch/x86/kernel/apic/numaq_32.c
@@ -493,7 +493,8 @@ static void numaq_setup_portio_remap(void)
(u_long) xquad_portio, (u_long) num_quads*XQUAD_PORTIO_QUAD);
}
-struct apic apic_numaq = {
+/* Use __refdata to keep false positive warning calm. */
+struct apic __refdata apic_numaq = {
.name = "NUMAQ",
.probe = probe_numaq,
diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c
index 440a8bccd91..0c0182cc947 100644
--- a/arch/x86/kernel/apic/probe_32.c
+++ b/arch/x86/kernel/apic/probe_32.c
@@ -20,23 +20,12 @@
#include <asm/apic.h>
#include <asm/setup.h>
-#include <linux/threads.h>
-#include <linux/cpumask.h>
-#include <asm/mpspec.h>
-#include <asm/fixmap.h>
-#include <asm/apicdef.h>
-#include <linux/kernel.h>
-#include <linux/string.h>
#include <linux/smp.h>
-#include <linux/init.h>
#include <asm/ipi.h>
-#include <linux/smp.h>
-#include <linux/init.h>
#include <linux/interrupt.h>
#include <asm/acpi.h>
#include <asm/e820.h>
-#include <asm/setup.h>
#ifdef CONFIG_HOTPLUG_CPU
#define DEFAULT_SEND_IPI (1)
diff --git a/arch/x86/kernel/apic/probe_64.c b/arch/x86/kernel/apic/probe_64.c
index bc3e880f9b8..65edc180fc8 100644
--- a/arch/x86/kernel/apic/probe_64.c
+++ b/arch/x86/kernel/apic/probe_64.c
@@ -44,17 +44,22 @@ static struct apic *apic_probe[] __initdata = {
NULL,
};
+static int apicid_phys_pkg_id(int initial_apic_id, int index_msb)
+{
+ return hard_smp_processor_id() >> index_msb;
+}
+
/*
* Check the APIC IDs in bios_cpu_apicid and choose the APIC mode.
*/
void __init default_setup_apic_routing(void)
{
#ifdef CONFIG_X86_X2APIC
- if (x2apic_mode && (apic != &apic_x2apic_phys &&
+ if (x2apic_mode
#ifdef CONFIG_X86_UV
- apic != &apic_x2apic_uv_x &&
+ && apic != &apic_x2apic_uv_x
#endif
- apic != &apic_x2apic_cluster)) {
+ ) {
if (x2apic_phys)
apic = &apic_x2apic_phys;
else
@@ -69,6 +74,11 @@ void __init default_setup_apic_routing(void)
printk(KERN_INFO "Setting APIC routing to %s\n", apic->name);
}
+ if (is_vsmp_box()) {
+ /* need to update phys_pkg_id */
+ apic->phys_pkg_id = apicid_phys_pkg_id;
+ }
+
/*
* Now that apic routing model is selected, configure the
* fault handling for intr remapping.
diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c
index 344eee4ac0a..eafdfbd1ea9 100644
--- a/arch/x86/kernel/apic/summit_32.c
+++ b/arch/x86/kernel/apic/summit_32.c
@@ -44,7 +44,6 @@
#include <asm/ipi.h>
#include <linux/kernel.h>
#include <linux/string.h>
-#include <linux/init.h>
#include <linux/gfp.h>
#include <linux/smp.h>
diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c
index 8e4cbb255c3..a5371ec3677 100644
--- a/arch/x86/kernel/apic/x2apic_cluster.c
+++ b/arch/x86/kernel/apic/x2apic_cluster.c
@@ -17,11 +17,13 @@ static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
return x2apic_enabled();
}
-/* Start with all IRQs pointing to boot CPU. IRQ balancing will shift them. */
-
+/*
+ * need to use more than cpu 0, because we need more vectors when
+ * MSI-X are used.
+ */
static const struct cpumask *x2apic_target_cpus(void)
{
- return cpumask_of(0);
+ return cpu_online_mask;
}
/*
@@ -170,7 +172,7 @@ static unsigned long set_apic_id(unsigned int id)
static int x2apic_cluster_phys_pkg_id(int initial_apicid, int index_msb)
{
- return current_cpu_data.initial_apicid >> index_msb;
+ return initial_apicid >> index_msb;
}
static void x2apic_send_IPI_self(int vector)
diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c
index a284359627e..a8989aadc99 100644
--- a/arch/x86/kernel/apic/x2apic_phys.c
+++ b/arch/x86/kernel/apic/x2apic_phys.c
@@ -27,11 +27,13 @@ static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
return 0;
}
-/* Start with all IRQs pointing to boot CPU. IRQ balancing will shift them. */
-
+/*
+ * need to use more than cpu 0, because we need more vectors when
+ * MSI-X are used.
+ */
static const struct cpumask *x2apic_target_cpus(void)
{
- return cpumask_of(0);
+ return cpu_online_mask;
}
static void x2apic_vector_allocation_domain(int cpu, struct cpumask *retmask)
@@ -162,7 +164,7 @@ static unsigned long set_apic_id(unsigned int id)
static int x2apic_phys_pkg_id(int initial_apicid, int index_msb)
{
- return current_cpu_data.initial_apicid >> index_msb;
+ return initial_apicid >> index_msb;
}
static void x2apic_send_IPI_self(int vector)
diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 096d19aea2f..601159374e8 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -46,7 +46,7 @@ static int early_get_nodeid(void)
return node_id.s.node_id;
}
-static int uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
+static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
{
if (!strcmp(oem_id, "SGI")) {
if (!strcmp(oem_table_id, "UVL"))
@@ -253,7 +253,7 @@ static void uv_send_IPI_self(int vector)
apic_write(APIC_SELF_IPI, vector);
}
-struct apic apic_x2apic_uv_x = {
+struct apic __refdata apic_x2apic_uv_x = {
.name = "UV large system",
.probe = NULL,
@@ -261,7 +261,7 @@ struct apic apic_x2apic_uv_x = {
.apic_id_registered = uv_apic_id_registered,
.irq_delivery_mode = dest_Fixed,
- .irq_dest_mode = 1, /* logical */
+ .irq_dest_mode = 0, /* physical */
.target_cpus = uv_target_cpus,
.disable_esr = 0,
@@ -362,12 +362,6 @@ static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size)
BUG();
}
-static __init void map_low_mmrs(void)
-{
- init_extra_mapping_uc(UV_GLOBAL_MMR32_BASE, UV_GLOBAL_MMR32_SIZE);
- init_extra_mapping_uc(UV_LOCAL_MMR_BASE, UV_LOCAL_MMR_SIZE);
-}
-
enum map_type {map_wb, map_uc};
static __init void map_high(char *id, unsigned long base, int shift,
@@ -395,26 +389,6 @@ static __init void map_gru_high(int max_pnode)
map_high("GRU", gru.s.base, shift, max_pnode, map_wb);
}
-static __init void map_config_high(int max_pnode)
-{
- union uvh_rh_gam_cfg_overlay_config_mmr_u cfg;
- int shift = UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_SHFT;
-
- cfg.v = uv_read_local_mmr(UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR);
- if (cfg.s.enable)
- map_high("CONFIG", cfg.s.base, shift, max_pnode, map_uc);
-}
-
-static __init void map_mmr_high(int max_pnode)
-{
- union uvh_rh_gam_mmr_overlay_config_mmr_u mmr;
- int shift = UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT;
-
- mmr.v = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR);
- if (mmr.s.enable)
- map_high("MMR", mmr.s.base, shift, max_pnode, map_uc);
-}
-
static __init void map_mmioh_high(int max_pnode)
{
union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh;
@@ -566,8 +540,6 @@ void __init uv_system_init(void)
unsigned long mmr_base, present, paddr;
unsigned short pnode_mask;
- map_low_mmrs();
-
m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG);
m_val = m_n_config.s.m_skt;
n_val = m_n_config.s.n_skt;
@@ -591,6 +563,8 @@ void __init uv_system_init(void)
bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades();
uv_blade_info = kmalloc(bytes, GFP_KERNEL);
BUG_ON(!uv_blade_info);
+ for (blade = 0; blade < uv_num_possible_blades(); blade++)
+ uv_blade_info[blade].memory_nid = -1;
get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size);
@@ -629,6 +603,9 @@ void __init uv_system_init(void)
lcpu = uv_blade_info[blade].nr_possible_cpus;
uv_blade_info[blade].nr_possible_cpus++;
+ /* Any node on the blade, else will contain -1. */
+ uv_blade_info[blade].memory_nid = nid;
+
uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base;
uv_cpu_hub_info(cpu)->lowmem_remap_top = lowmem_redir_size;
uv_cpu_hub_info(cpu)->m_val = m_val;
@@ -662,11 +639,10 @@ void __init uv_system_init(void)
pnode = (paddr >> m_val) & pnode_mask;
blade = boot_pnode_to_blade(pnode);
uv_node_to_blade[nid] = blade;
+ max_pnode = max(pnode, max_pnode);
}
map_gru_high(max_pnode);
- map_mmr_high(max_pnode);
- map_config_high(max_pnode);
map_mmioh_high(max_pnode);
uv_cpu_init();