@@ -103,6 +103,7 @@ acpi_parse_x2apic(struct acpi_subtable_header *header, const unsigned long end)
processor->lapic_flags & ACPI_MADT_ENABLED
? KERN_WARNING "WARNING: " : KERN_INFO,
processor->local_apic_id, processor->uid);
+ unaccounted_cpus = true;
/*
* Must not return an error here, to prevent
* acpi_table_parse_entries() from terminating early.
@@ -61,11 +61,14 @@ unsigned int __read_mostly boot_cpu_physical_apicid = BAD_APICID;
/* Internal processor count */
static unsigned int num_processors;
-static unsigned int __initdata disabled_cpus;
+unsigned int __read_mostly disabled_cpus;
/* Bitmask of physically existing CPUs */
physid_mask_t phys_cpu_present_map;
+/* Record whether CPUs haven't been added due to overflows. */
+bool __read_mostly unaccounted_cpus;
+
void __init set_nr_cpu_ids(unsigned int max_cpus)
{
unsigned int tot_cpus = num_processors + disabled_cpus;
@@ -160,6 +163,7 @@ static int MP_processor_info_x(struct mpc_config_processor *m,
printk_once(XENLOG_WARNING
"WARNING: NR_CPUS limit of %u reached - ignoring further processors\n",
nr_cpu_ids);
+ unaccounted_cpus = true;
return -ENOSPC;
}
@@ -167,6 +171,7 @@ static int MP_processor_info_x(struct mpc_config_processor *m,
&& genapic.name == apic_default.name) {
printk_once(XENLOG_WARNING
"WARNING: CPUs limit of 8 reached - ignoring futher processors\n");
+ unaccounted_cpus = true;
return -ENOSPC;
}
@@ -8,6 +8,7 @@
* later.
*/
+#include <xen/cpu.h>
#include <xen/irq.h>
#include <xen/sched.h>
#include <xen/delay.h>
@@ -64,7 +65,39 @@ static void send_IPI_shortcut(unsigned int shortcut, int vector,
void send_IPI_mask(const cpumask_t *mask, int vector)
{
- alternative_vcall(genapic.send_IPI_mask, mask, vector);
+ bool cpus_locked = false;
+ cpumask_t *scratch = this_cpu(scratch_cpumask);
+
+ /*
+ * This can only be safely used when no CPU hotplug or unplug operations
+ * are taking place, there are no offline CPUs (unless those have been
+ * onlined and parked), there are no disabled CPUs and all possible CPUs in
+ * the system have been accounted for.
+ */
+ if ( system_state > SYS_STATE_smp_boot &&
+ !unaccounted_cpus && !disabled_cpus &&
+ /* NB: get_cpu_maps lock requires enabled interrupts. */
+ local_irq_is_enabled() && (cpus_locked = get_cpu_maps()) &&
+ (park_offline_cpus ||
+ cpumask_equal(&cpu_online_map, &cpu_present_map)) )
+ cpumask_or(scratch, mask, cpumask_of(smp_processor_id()));
+ else
+ {
+ if ( cpus_locked )
+ {
+ put_cpu_maps();
+ cpus_locked = false;
+ }
+ cpumask_clear(scratch);
+ }
+
+ if ( cpumask_equal(scratch, &cpu_online_map) )
+ send_IPI_shortcut(APIC_DEST_ALLBUT, vector, APIC_DEST_PHYSICAL);
+ else
+ alternative_vcall(genapic.send_IPI_mask, mask, vector);
+
+ if ( cpus_locked )
+ put_cpu_maps();
}
void send_IPI_self(int vector)
@@ -84,6 +84,9 @@ extern cpumask_t **socket_cpumask;
#define get_cpu_current(cpu) \
(get_cpu_info_from_stack((unsigned long)stack_base[cpu])->current_vcpu)
+extern unsigned int disabled_cpus;
+extern bool unaccounted_cpus;
+
#endif /* !__ASSEMBLY__ */
#endif
If the IPI destination mask matches the mask of online CPUs use the APIC ALLBUT destination shorthand in order to send an IPI to all CPUs on the system except the current one. This can only be safely used when no CPU hotplug or unplug operations are taking place, no offline CPUs or those have been onlined and parked, all CPUs in the system have been accounted for (ie: the number of CPUs doesn't exceed NR_CPUS and APIC IDs are below MAX_APICS) and there's no possibility of CPU hotplug (ie: no disabled CPUs have been reported by the firmware tables). This is specially beneficial when using the PV shim, since using the shorthand avoids performing an APIC register write (or multiple ones if using xAPIC mode) for each destination when doing a global TLB flush. The lock time of flush_lock on a 32 vCPU guest using the shim in x2APIC mode without the shorthand is: Global lock flush_lock: addr=ffff82d0804b21c0, lockval=f602f602, not locked lock:228455938(79406065573135), block:205908580(556416605761539) Average lock time: 347577ns While the same guest using the shorthand: Global lock flush_lock: addr=ffff82d0804b41c0, lockval=d9c4d9bc, cpu=12 lock:1890775(416719148054), block:1663958(2500161282949) Average lock time: 220395ns Approximately a 1/3 improvement in the lock time. Note that this requires locking the CPU maps (get_cpu_maps) which uses a trylock. This is currently safe as all users of cpu_add_remove_lock do a trylock, but will need reevaluating if non-trylock users appear. Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> --- Changes since v2: - Place the code moment in a pre-patch. - Rename cpu_overflow to unaccounted_cpus (leaving it as a bool). - Make disabled_cpus non-init and non-static (and mark it read_mostly). - Prevent using the shorthand if there are disabled CPU slots. - Cope correctly with send_IPI_mask getting passed an empty mask. - Reword comment about shorthand usage in send_IPI_mask. Changes since v1: - Move the shorthand logic to send_IPI_mask. - Check interrupts are enabled before trying to get the cpu maps lock. - Move __prepare_ICR and __default_send_IPI_shortcut. --- xen/arch/x86/acpi/boot.c | 1 + xen/arch/x86/mpparse.c | 7 ++++++- xen/arch/x86/smp.c | 35 ++++++++++++++++++++++++++++++++++- xen/include/asm-x86/smp.h | 3 +++ 4 files changed, 44 insertions(+), 2 deletions(-)