KVM: arm64: Avoid reading ID_AA64DFR0_EL1 for debug save/restore

Similar to other per-CPU profiling/debug features we handle, store the
number of breakpoints/watchpoints in kvm_host_data to avoid reading the
ID register 4 times on every guest entry/exit. And if you're in the
nested virt business that's quite a few avoidable exits to the L0
hypervisor.

Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20241219224116.3941496-18-oliver.upton@linux.dev
Signed-off-by: Marc Zyngier <maz@kernel.org>

Author: oupton

arch/arm64/include/asm/kvm_host.h

@@ -661,6 +661,10 @@ struct kvm_host_data {
 
 	/* Number of programmable event counters (PMCR_EL0.N) for this CPU */
 	unsigned int nr_event_counters;
+
+	/* Number of debug breakpoints/watchpoints for this CPU (minus 1) */
+	unsigned int debug_brps;
+	unsigned int debug_wrps;
 };
 
 struct kvm_host_psci_config {

arch/arm64/kvm/debug.c

@@ -71,6 +71,9 @@ void kvm_init_host_debug_data(void)
 		*host_data_ptr(nr_event_counters) = FIELD_GET(ARMV8_PMU_PMCR_N,
 							      read_sysreg(pmcr_el0));
 
+	*host_data_ptr(debug_brps) = SYS_FIELD_GET(ID_AA64DFR0_EL1, BRPs, dfr0);
+	*host_data_ptr(debug_wrps) = SYS_FIELD_GET(ID_AA64DFR0_EL1, WRPs, dfr0);
+
 	if (has_vhe())
 		return;
 

arch/arm64/kvm/hyp/include/hyp/debug-sr.h

@@ -106,12 +106,8 @@ static struct kvm_guest_debug_arch *__vcpu_debug_regs(struct kvm_vcpu *vcpu)
 static void __debug_save_state(struct kvm_guest_debug_arch *dbg,
 			       struct kvm_cpu_context *ctxt)
 {
-	u64 aa64dfr0;
-	int brps, wrps;
-
-	aa64dfr0 = read_sysreg(id_aa64dfr0_el1);
-	brps = (aa64dfr0 >> 12) & 0xf;
-	wrps = (aa64dfr0 >> 20) & 0xf;
+	int brps = *host_data_ptr(debug_brps);
+	int wrps = *host_data_ptr(debug_wrps);
 
 	save_debug(dbg->dbg_bcr, dbgbcr, brps);
 	save_debug(dbg->dbg_bvr, dbgbvr, brps);
@@ -124,13 +120,8 @@ static void __debug_save_state(struct kvm_guest_debug_arch *dbg,
 static void __debug_restore_state(struct kvm_guest_debug_arch *dbg,
 				  struct kvm_cpu_context *ctxt)
 {
-	u64 aa64dfr0;
-	int brps, wrps;
-
-	aa64dfr0 = read_sysreg(id_aa64dfr0_el1);
-
-	brps = (aa64dfr0 >> 12) & 0xf;
-	wrps = (aa64dfr0 >> 20) & 0xf;
+	int brps = *host_data_ptr(debug_brps);
+	int wrps = *host_data_ptr(debug_wrps);
 
 	restore_debug(dbg->dbg_bcr, dbgbcr, brps);
 	restore_debug(dbg->dbg_bvr, dbgbvr, brps);