[PATCH] arm64: errata: Minimize tlb flush due to vttbr writes on AmpereOne

[Ganapatrao Kulkarni]

AmpereOne implementation is doing tlb flush when ever there is
a write to vttbr_el2. As per KVM implementation, vttbr_el2 is updated
with VM's S2-MMU while return to VM. This is not necessary when there
is no VM context switch and a just return to same Guest.

Adding a check to avoid the vttbr_el2 write if the same value
already exist to prevent needless tlb flush.

Signed-off-by: Ganapatrao Kulkarni <gankulkarni@xxxxxxxxxxxxxxxxxxxxxx>
---
 Documentation/arch/arm64/silicon-errata.rst |  2 ++
 arch/arm64/Kconfig                          | 13 +++++++++++++
 arch/arm64/include/asm/kvm_mmu.h            |  8 +++++++-
 arch/arm64/kernel/cpu_errata.c              |  7 +++++++
 arch/arm64/tools/cpucaps                    |  1 +
 5 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/Documentation/arch/arm64/silicon-errata.rst b/Documentation/arch/arm64/silicon-errata.rst
index e8c2ce1f9df6..8924e84358c9 100644
--- a/Documentation/arch/arm64/silicon-errata.rst
+++ b/Documentation/arch/arm64/silicon-errata.rst
@@ -54,6 +54,8 @@ stable kernels.
 +----------------+-----------------+-----------------+-----------------------------+
 | Ampere         | AmpereOne       | AC03_CPU_38     | AMPERE_ERRATUM_AC03_CPU_38  |
 +----------------+-----------------+-----------------+-----------------------------+
+| Ampere         | AmpereOne       | N/A             | AMPERE_AC03_REDUCE_TLB_FLUSH|
++----------------+-----------------+-----------------+-----------------------------+
 +----------------+-----------------+-----------------+-----------------------------+
 | ARM            | Cortex-A510     | #2457168        | ARM64_ERRATUM_2457168       |
 +----------------+-----------------+-----------------+-----------------------------+
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index aa7c1d435139..77485d0322e4 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -436,6 +436,19 @@ config AMPERE_ERRATUM_AC03_CPU_38
 
 	  If unsure, say Y.
 
+config AMPERE_AC03_REDUCE_TLB_FLUSH
+	bool "AmpereOne: Minimize the writes to vttbr_el2 register"
+	default y
+	help
+	  On AmpereOne, any writes to vttbr_el2 results in TLB flush.
+	  It can be avoided to improve the performance when there is no VM
+	  context switches and a just return to same VM from the hypervisor.
+
+	  This option adds a check to avoid rewrite of the same value
+	  to vttbr_el2.
+
+	  If unsure, say Y.
+
 config ARM64_WORKAROUND_CLEAN_CACHE
 	bool
 
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
index e3e793d0ec30..da39e4749434 100644
--- a/arch/arm64/include/asm/kvm_mmu.h
+++ b/arch/arm64/include/asm/kvm_mmu.h
@@ -317,8 +317,14 @@ static __always_inline u64 kvm_get_vttbr(struct kvm_s2_mmu *mmu)
 static __always_inline void __load_stage2(struct kvm_s2_mmu *mmu,
 					  struct kvm_arch *arch)
 {
+	u64 vttbr;
+
+	vttbr = kvm_get_vttbr(mmu);
 	write_sysreg(mmu->vtcr, vtcr_el2);
-	write_sysreg(kvm_get_vttbr(mmu), vttbr_el2);
+
+	if (!cpus_have_final_cap(ARM64_WORKAROUND_AMPERE_AC03_TLB_FLUSH) ||
+	    read_sysreg(vttbr_el2) != vttbr)
+		write_sysreg(vttbr, vttbr_el2);
 
 	/*
 	 * ARM errata 1165522 and 1530923 require the actual execution of the
diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c
index 967c7c7a4e7d..f612975e0cb5 100644
--- a/arch/arm64/kernel/cpu_errata.c
+++ b/arch/arm64/kernel/cpu_errata.c
@@ -740,6 +740,13 @@ const struct arm64_cpu_capabilities arm64_errata[] = {
 		.capability = ARM64_WORKAROUND_AMPERE_AC03_CPU_38,
 		ERRATA_MIDR_ALL_VERSIONS(MIDR_AMPERE1),
 	},
+#endif
+#ifdef CONFIG_AMPERE_AC03_REDUCE_TLB_FLUSH
+	{
+		.desc = "AmpereOne, minimize tlb flush due to vttbr write",
+		.capability = ARM64_WORKAROUND_AMPERE_AC03_TLB_FLUSH,
+		ERRATA_MIDR_ALL_VERSIONS(MIDR_AMPERE1),
+	},
 #endif
 	{
 	}
diff --git a/arch/arm64/tools/cpucaps b/arch/arm64/tools/cpucaps
index b912b1409fc0..b4bee37d0527 100644
--- a/arch/arm64/tools/cpucaps
+++ b/arch/arm64/tools/cpucaps
@@ -85,6 +85,7 @@ WORKAROUND_2457168
 WORKAROUND_2645198
 WORKAROUND_2658417
 WORKAROUND_AMPERE_AC03_CPU_38
+WORKAROUND_AMPERE_AC03_TLB_FLUSH
 WORKAROUND_TRBE_OVERWRITE_FILL_MODE
 WORKAROUND_TSB_FLUSH_FAILURE
 WORKAROUND_TRBE_WRITE_OUT_OF_RANGE
-- 
2.41.0