arm64: prefetch: add alternative pattern for CPUs without a prefetcher

Most CPUs have a hardware prefetcher which generally performs better
without explicit prefetch instructions issued by software, however
some CPUs (e.g. Cavium ThunderX) rely solely on explicit prefetch
instructions.

This patch adds an alternative pattern (ARM64_HAS_NO_HW_PREFETCH) to
allow our library code to make use of explicit prefetch instructions
during things like copy routines only when the CPU does not have the
capability to perform the prefetching itself.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Tested-by: Andrew Pinski <apinski@cavium.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

arch/arm64/include/asm/cpufeature.h

@@ -30,8 +30,9 @@
 #define ARM64_HAS_LSE_ATOMICS			5
 #define ARM64_WORKAROUND_CAVIUM_23154		6
 #define ARM64_WORKAROUND_834220			7
+#define ARM64_HAS_NO_HW_PREFETCH		8
 
-#define ARM64_NCAPS				8
+#define ARM64_NCAPS				9
 
 #ifndef __ASSEMBLY__
 

arch/arm64/include/asm/cputype.h

@@ -57,11 +57,22 @@
 #define MIDR_IMPLEMENTOR(midr)	\
 	(((midr) & MIDR_IMPLEMENTOR_MASK) >> MIDR_IMPLEMENTOR_SHIFT)
 
-#define MIDR_CPU_PART(imp, partnum) \
+#define MIDR_CPU_MODEL(imp, partnum) \
 	(((imp)			<< MIDR_IMPLEMENTOR_SHIFT) | \
 	(0xf			<< MIDR_ARCHITECTURE_SHIFT) | \
 	((partnum)		<< MIDR_PARTNUM_SHIFT))
 
+#define MIDR_CPU_MODEL_MASK (MIDR_IMPLEMENTOR_MASK | MIDR_PARTNUM_MASK | \
+			     MIDR_ARCHITECTURE_MASK)
+
+#define MIDR_IS_CPU_MODEL_RANGE(midr, model, rv_min, rv_max)		\
+({									\
+	u32 _model = (midr) & MIDR_CPU_MODEL_MASK;			\
+	u32 rv = (midr) & (MIDR_REVISION_MASK | MIDR_VARIANT_MASK);	\
+									\
+	_model == (model) && rv >= (rv_min) && rv <= (rv_max);		\
+ })
+
 #define ARM_CPU_IMP_ARM			0x41
 #define ARM_CPU_IMP_APM			0x50
 #define ARM_CPU_IMP_CAVIUM		0x43
@@ -75,6 +86,10 @@
 
 #define CAVIUM_CPU_PART_THUNDERX	0x0A1
 
+#define MIDR_CORTEX_A53 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A53)
+#define MIDR_CORTEX_A57 MIDR_CPU_MODEL(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A57)
+#define MIDR_THUNDERX	MIDR_CPU_MODEL(ARM_CPU_IMP_CAVIUM, CAVIUM_CPU_PART_THUNDERX)
+
 #ifndef __ASSEMBLY__
 
 /*

arch/arm64/kernel/cpu_errata.c

@@ -21,24 +21,12 @@
 #include <asm/cputype.h>
 #include <asm/cpufeature.h>
 
-#define MIDR_CORTEX_A53 MIDR_CPU_PART(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A53)
-#define MIDR_CORTEX_A57 MIDR_CPU_PART(ARM_CPU_IMP_ARM, ARM_CPU_PART_CORTEX_A57)
-#define MIDR_THUNDERX	MIDR_CPU_PART(ARM_CPU_IMP_CAVIUM, CAVIUM_CPU_PART_THUNDERX)
-
-#define CPU_MODEL_MASK (MIDR_IMPLEMENTOR_MASK | MIDR_PARTNUM_MASK | \
-			MIDR_ARCHITECTURE_MASK)
-
 static bool __maybe_unused
 is_affected_midr_range(const struct arm64_cpu_capabilities *entry)
 {
-	u32 midr = read_cpuid_id();
-
-	if ((midr & CPU_MODEL_MASK) != entry->midr_model)
-		return false;
-
-	midr &= MIDR_REVISION_MASK | MIDR_VARIANT_MASK;
-
-	return (midr >= entry->midr_range_min && midr <= entry->midr_range_max);
+	return MIDR_IS_CPU_MODEL_RANGE(read_cpuid_id(), entry->midr_model,
+				       entry->midr_range_min,
+				       entry->midr_range_max);
 }
 
 #define MIDR_RANGE(model, min, max) \

arch/arm64/kernel/cpufeature.c

@@ -621,6 +621,18 @@ static bool has_useable_gicv3_cpuif(const struct arm64_cpu_capabilities *entry)
 	return has_sre;
 }
 
+static bool has_no_hw_prefetch(const struct arm64_cpu_capabilities *entry)
+{
+	u32 midr = read_cpuid_id();
+	u32 rv_min, rv_max;
+
+	/* Cavium ThunderX pass 1.x and 2.x */
+	rv_min = 0;
+	rv_max = (1 << MIDR_VARIANT_SHIFT) | MIDR_REVISION_MASK;
+
+	return MIDR_IS_CPU_MODEL_RANGE(midr, MIDR_THUNDERX, rv_min, rv_max);
+}
+
 static const struct arm64_cpu_capabilities arm64_features[] = {
 	{
 		.desc = "GIC system register CPU interface",
@@ -651,6 +663,11 @@ static const struct arm64_cpu_capabilities arm64_features[] = {
 		.min_field_value = 2,
 	},
 #endif /* CONFIG_AS_LSE && CONFIG_ARM64_LSE_ATOMICS */
+	{
+		.desc = "Software prefetching using PRFM",
+		.capability = ARM64_HAS_NO_HW_PREFETCH,
+		.matches = has_no_hw_prefetch,
+	},
 	{},
 };