Merge tag 'x86_mtrr_for_v6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 mtrr updates from Borislav Petkov:
 "A serious scrubbing of the MTRR code including adding a new map
  mechanism in order to look up the memory type of a region easily.

  Also address memory range lookup issues like returning an invalid
  memory type. Furthermore, this handles the decoupling of PAT from MTRR
  more naturally.

  All work by Juergen Gross"

* tag 'x86_mtrr_for_v6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/xen: Set default memory type for PV guests to WB
  x86/mtrr: Unify debugging printing
  x86/mtrr: Remove unused code
  x86/mm: Only check uniform after calling mtrr_type_lookup()
  x86/mtrr: Don't let mtrr_type_lookup() return MTRR_TYPE_INVALID
  x86/mtrr: Use new cache_map in mtrr_type_lookup()
  x86/mtrr: Add mtrr=debug command line option
  x86/mtrr: Construct a memory map with cache modes
  x86/mtrr: Add get_effective_type() service function
  x86/mtrr: Allocate mtrr_value array dynamically
  x86/mtrr: Move 32-bit code from mtrr.c to legacy.c
  x86/mtrr: Have only one set_mtrr() variant
  x86/mtrr: Replace vendor tests in MTRR code
  x86/xen: Set MTRR state when running as Xen PV initial domain
  x86/hyperv: Set MTRR state when running as SEV-SNP Hyper-V guest
  x86/mtrr: Support setting MTRR state for software defined MTRRs
  x86/mtrr: Replace size_or_mask and size_and_mask with a much easier concept
  x86/mtrr: Remove physical address size calculation

Documentation/admin-guide/kernel-parameters.txt

@@ -3433,6 +3433,10 @@
 			[HW] Make the MicroTouch USB driver use raw coordinates
 			('y', default) or cooked coordinates ('n')
 
+	mtrr=debug	[X86]
+			Enable printing debug information related to MTRR
+			registers at boot time.
+
 	mtrr_chunk_size=nn[KMG] [X86]
 			used for mtrr cleanup. It is largest continuous chunk
 			that could hold holes aka. UC entries.

arch/x86/hyperv/ivm.c

@@ -17,6 +17,7 @@
 #include <asm/mem_encrypt.h>
 #include <asm/mshyperv.h>
 #include <asm/hypervisor.h>
+#include <asm/mtrr.h>
 
 #ifdef CONFIG_AMD_MEM_ENCRYPT
 
@@ -372,6 +373,9 @@ void __init hv_vtom_init(void)
 	x86_platform.guest.enc_cache_flush_required = hv_vtom_cache_flush_required;
 	x86_platform.guest.enc_tlb_flush_required = hv_vtom_tlb_flush_required;
 	x86_platform.guest.enc_status_change_finish = hv_vtom_set_host_visibility;
+
+	/* Set WB as the default cache mode. */
+	mtrr_overwrite_state(NULL, 0, MTRR_TYPE_WRBACK);
 }
 
 #endif /* CONFIG_AMD_MEM_ENCRYPT */

arch/x86/include/asm/mtrr.h

@@ -23,14 +23,43 @@
 #ifndef _ASM_X86_MTRR_H
 #define _ASM_X86_MTRR_H
 
+#include <linux/bits.h>
 #include <uapi/asm/mtrr.h>
 
+/* Defines for hardware MTRR registers. */
+#define MTRR_CAP_VCNT		GENMASK(7, 0)
+#define MTRR_CAP_FIX		BIT_MASK(8)
+#define MTRR_CAP_WC		BIT_MASK(10)
+
+#define MTRR_DEF_TYPE_TYPE	GENMASK(7, 0)
+#define MTRR_DEF_TYPE_FE	BIT_MASK(10)
+#define MTRR_DEF_TYPE_E		BIT_MASK(11)
+
+#define MTRR_DEF_TYPE_ENABLE	(MTRR_DEF_TYPE_FE | MTRR_DEF_TYPE_E)
+#define MTRR_DEF_TYPE_DISABLE	~(MTRR_DEF_TYPE_TYPE | MTRR_DEF_TYPE_ENABLE)
+
+#define MTRR_PHYSBASE_TYPE	GENMASK(7, 0)
+#define MTRR_PHYSBASE_RSVD	GENMASK(11, 8)
+
+#define MTRR_PHYSMASK_RSVD	GENMASK(10, 0)
+#define MTRR_PHYSMASK_V		BIT_MASK(11)
+
+struct mtrr_state_type {
+	struct mtrr_var_range var_ranges[MTRR_MAX_VAR_RANGES];
+	mtrr_type fixed_ranges[MTRR_NUM_FIXED_RANGES];
+	unsigned char enabled;
+	bool have_fixed;
+	mtrr_type def_type;
+};
+
 /*
  * The following functions are for use by other drivers that cannot use
  * arch_phys_wc_add and arch_phys_wc_del.
  */
 # ifdef CONFIG_MTRR
 void mtrr_bp_init(void);
+void mtrr_overwrite_state(struct mtrr_var_range *var, unsigned int num_var,
+			  mtrr_type def_type);
 extern u8 mtrr_type_lookup(u64 addr, u64 end, u8 *uniform);
 extern void mtrr_save_fixed_ranges(void *);
 extern void mtrr_save_state(void);
@@ -40,7 +69,6 @@ extern int mtrr_add_page(unsigned long base, unsigned long size,
 			 unsigned int type, bool increment);
 extern int mtrr_del(int reg, unsigned long base, unsigned long size);
 extern int mtrr_del_page(int reg, unsigned long base, unsigned long size);
-extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi);
 extern void mtrr_bp_restore(void);
 extern int mtrr_trim_uncached_memory(unsigned long end_pfn);
 extern int amd_special_default_mtrr(void);
@@ -48,12 +76,21 @@ void mtrr_disable(void);
 void mtrr_enable(void);
 void mtrr_generic_set_state(void);
 #  else
+static inline void mtrr_overwrite_state(struct mtrr_var_range *var,
+					unsigned int num_var,
+					mtrr_type def_type)
+{
+}
+
 static inline u8 mtrr_type_lookup(u64 addr, u64 end, u8 *uniform)
 {
 	/*
-	 * Return no-MTRRs:
+	 * Return the default MTRR type, without any known other types in
+	 * that range.
 	 */
-	return MTRR_TYPE_INVALID;
+	*uniform = 1;
+
+	return MTRR_TYPE_UNCACHABLE;
 }
 #define mtrr_save_fixed_ranges(arg) do {} while (0)
 #define mtrr_save_state() do {} while (0)
@@ -79,9 +116,6 @@ static inline int mtrr_trim_uncached_memory(unsigned long end_pfn)
 {
 	return 0;
 }
-static inline void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi)
-{
-}
 #define mtrr_bp_init() do {} while (0)
 #define mtrr_bp_restore() do {} while (0)
 #define mtrr_disable() do {} while (0)
@@ -121,7 +155,8 @@ struct mtrr_gentry32 {
 #endif /* CONFIG_COMPAT */
 
 /* Bit fields for enabled in struct mtrr_state_type */
-#define MTRR_STATE_MTRR_FIXED_ENABLED	0x01
-#define MTRR_STATE_MTRR_ENABLED		0x02
+#define MTRR_STATE_SHIFT		10
+#define MTRR_STATE_MTRR_FIXED_ENABLED	(MTRR_DEF_TYPE_FE >> MTRR_STATE_SHIFT)
+#define MTRR_STATE_MTRR_ENABLED		(MTRR_DEF_TYPE_E >> MTRR_STATE_SHIFT)
 
 #endif /* _ASM_X86_MTRR_H */

arch/x86/include/uapi/asm/mtrr.h

@@ -81,14 +81,6 @@ typedef __u8 mtrr_type;
 #define MTRR_NUM_FIXED_RANGES 88
 #define MTRR_MAX_VAR_RANGES 256
 
-struct mtrr_state_type {
-	struct mtrr_var_range var_ranges[MTRR_MAX_VAR_RANGES];
-	mtrr_type fixed_ranges[MTRR_NUM_FIXED_RANGES];
-	unsigned char enabled;
-	unsigned char have_fixed;
-	mtrr_type def_type;
-};
-
 #define MTRRphysBase_MSR(reg) (0x200 + 2 * (reg))
 #define MTRRphysMask_MSR(reg) (0x200 + 2 * (reg) + 1)
 
@@ -115,9 +107,9 @@ struct mtrr_state_type {
 #define MTRR_NUM_TYPES       7
 
 /*
- * Invalid MTRR memory type.  mtrr_type_lookup() returns this value when
- * MTRRs are disabled.  Note, this value is allocated from the reserved
- * values (0x7-0xff) of the MTRR memory types.
+ * Invalid MTRR memory type.  No longer used outside of MTRR code.
+ * Note, this value is allocated from the reserved values (0x7-0xff) of
+ * the MTRR memory types.
  */
 #define MTRR_TYPE_INVALID    0xff
 

arch/x86/kernel/cpu/mtrr/Makefile

 # SPDX-License-Identifier: GPL-2.0-only
 obj-y		:= mtrr.o if.o generic.o cleanup.o
-obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o
+obj-$(CONFIG_X86_32) += amd.o cyrix.o centaur.o legacy.o
 

arch/x86/kernel/cpu/mtrr/amd.c

@@ -110,7 +110,7 @@ amd_validate_add_page(unsigned long base, unsigned long size, unsigned int type)
 }
 
 const struct mtrr_ops amd_mtrr_ops = {
-	.vendor            = X86_VENDOR_AMD,
+	.var_regs          = 2,
 	.set               = amd_set_mtrr,
 	.get               = amd_get_mtrr,
 	.get_free_region   = generic_get_free_region,

arch/x86/kernel/cpu/mtrr/centaur.c

@@ -45,15 +45,6 @@ centaur_get_free_region(unsigned long base, unsigned long size, int replace_reg)
 	return -ENOSPC;
 }
 
-/*
- * Report boot time MCR setups
- */
-void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi)
-{
-	centaur_mcr[mcr].low = lo;
-	centaur_mcr[mcr].high = hi;
-}
-
 static void
 centaur_get_mcr(unsigned int reg, unsigned long *base,
 		unsigned long *size, mtrr_type * type)
@@ -112,7 +103,7 @@ centaur_validate_add_page(unsigned long base, unsigned long size, unsigned int t
 }
 
 const struct mtrr_ops centaur_mtrr_ops = {
-	.vendor            = X86_VENDOR_CENTAUR,
+	.var_regs          = 8,
 	.set               = centaur_set_mcr,
 	.get               = centaur_get_mcr,
 	.get_free_region   = centaur_get_free_region,

arch/x86/kernel/cpu/mtrr/cleanup.c

@@ -55,9 +55,6 @@ static int __initdata				nr_range;
 
 static struct var_mtrr_range_state __initdata	range_state[RANGE_NUM];
 
-static int __initdata debug_print;
-#define Dprintk(x...) do { if (debug_print) pr_debug(x); } while (0)
-
 #define BIOS_BUG_MSG \
 	"WARNING: BIOS bug: VAR MTRR %d contains strange UC entry under 1M, check with your system vendor!\n"
 
@@ -79,12 +76,11 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 		nr_range = add_range_with_merge(range, RANGE_NUM, nr_range,
 						base, base + size);
 	}
-	if (debug_print) {
-		pr_debug("After WB checking\n");
-		for (i = 0; i < nr_range; i++)
-			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
-				 range[i].start, range[i].end);
-	}
+
+	Dprintk("After WB checking\n");
+	for (i = 0; i < nr_range; i++)
+		Dprintk("MTRR MAP PFN: %016llx - %016llx\n",
+			 range[i].start, range[i].end);
 
 	/* Take out UC ranges: */
 	for (i = 0; i < num_var_ranges; i++) {
@@ -112,24 +108,22 @@ x86_get_mtrr_mem_range(struct range *range, int nr_range,
 		subtract_range(range, RANGE_NUM, extra_remove_base,
 				 extra_remove_base + extra_remove_size);
 
-	if  (debug_print) {
-		pr_debug("After UC checking\n");
-		for (i = 0; i < RANGE_NUM; i++) {
-			if (!range[i].end)
-				continue;
-			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
-				 range[i].start, range[i].end);
-		}
+	Dprintk("After UC checking\n");
+	for (i = 0; i < RANGE_NUM; i++) {
+		if (!range[i].end)
+			continue;
+
+		Dprintk("MTRR MAP PFN: %016llx - %016llx\n",
+			 range[i].start, range[i].end);
 	}
 
 	/* sort the ranges */
 	nr_range = clean_sort_range(range, RANGE_NUM);
-	if  (debug_print) {
-		pr_debug("After sorting\n");
-		for (i = 0; i < nr_range; i++)
-			pr_debug("MTRR MAP PFN: %016llx - %016llx\n",
-				 range[i].start, range[i].end);
-	}
+
+	Dprintk("After sorting\n");
+	for (i = 0; i < nr_range; i++)
+		Dprintk("MTRR MAP PFN: %016llx - %016llx\n",
+			range[i].start, range[i].end);
 
 	return nr_range;
 }
@@ -164,16 +158,9 @@ static int __init enable_mtrr_cleanup_setup(char *str)
 }
 early_param("enable_mtrr_cleanup", enable_mtrr_cleanup_setup);
 
-static int __init mtrr_cleanup_debug_setup(char *str)
-{
-	debug_print = 1;
-	return 0;
-}
-early_param("mtrr_cleanup_debug", mtrr_cleanup_debug_setup);
-
 static void __init
 set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek,
-	     unsigned char type, unsigned int address_bits)
+	     unsigned char type)
 {
 	u32 base_lo, base_hi, mask_lo, mask_hi;
 	u64 base, mask;
@@ -183,7 +170,7 @@ set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek,
 		return;
 	}
 
-	mask = (1ULL << address_bits) - 1;
+	mask = (1ULL << boot_cpu_data.x86_phys_bits) - 1;
 	mask &= ~((((u64)sizek) << 10) - 1);
 
 	base = ((u64)basek) << 10;
@@ -209,7 +196,7 @@ save_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek,
 	range_state[reg].type = type;
 }
 
-static void __init set_var_mtrr_all(unsigned int address_bits)
+static void __init set_var_mtrr_all(void)
 {
 	unsigned long basek, sizek;
 	unsigned char type;
@@ -220,7 +207,7 @@ static void __init set_var_mtrr_all(unsigned int address_bits)
 		sizek = range_state[reg].size_pfn << (PAGE_SHIFT - 10);
 		type = range_state[reg].type;
 
-		set_var_mtrr(reg, basek, sizek, type, address_bits);
+		set_var_mtrr(reg, basek, sizek, type);
 	}
 }
 
@@ -267,7 +254,7 @@ range_to_mtrr(unsigned int reg, unsigned long range_startk,
 			align = max_align;
 
 		sizek = 1UL << align;
-		if (debug_print) {
+		if (mtrr_debug) {
 			char start_factor = 'K', size_factor = 'K';
 			unsigned long start_base, size_base;
 
@@ -542,7 +529,7 @@ static void __init print_out_mtrr_range_state(void)
 		start_base = to_size_factor(start_base, &start_factor);
 		type = range_state[i].type;
 
-		pr_debug("reg %d, base: %ld%cB, range: %ld%cB, type %s\n",
+		Dprintk("reg %d, base: %ld%cB, range: %ld%cB, type %s\n",
 			i, start_base, start_factor,
 			size_base, size_factor,
 			(type == MTRR_TYPE_UNCACHABLE) ? "UC" :
@@ -680,7 +667,7 @@ static int __init mtrr_search_optimal_index(void)
 	return index_good;
 }
 
-int __init mtrr_cleanup(unsigned address_bits)
+int __init mtrr_cleanup(void)
 {
 	unsigned long x_remove_base, x_remove_size;
 	unsigned long base, size, def, dummy;
@@ -689,7 +676,10 @@ int __init mtrr_cleanup(unsigned address_bits)
 	int index_good;
 	int i;
 
-	if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1)
+	if (!mtrr_enabled())
+		return 0;
+
+	if (!cpu_feature_enabled(X86_FEATURE_MTRR) || enable_mtrr_cleanup < 1)
 		return 0;
 
 	rdmsr(MSR_MTRRdefType, def, dummy);
@@ -711,7 +701,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 		return 0;
 
 	/* Print original var MTRRs at first, for debugging: */
-	pr_debug("original variable MTRRs\n");
+	Dprintk("original variable MTRRs\n");
 	print_out_mtrr_range_state();
 
 	memset(range, 0, sizeof(range));
@@ -742,8 +732,8 @@ int __init mtrr_cleanup(unsigned address_bits)
 		mtrr_print_out_one_result(i);
 
 		if (!result[i].bad) {
-			set_var_mtrr_all(address_bits);
-			pr_debug("New variable MTRRs\n");
+			set_var_mtrr_all();
+			Dprintk("New variable MTRRs\n");
 			print_out_mtrr_range_state();
 			return 1;
 		}
@@ -763,7 +753,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 
 			mtrr_calc_range_state(chunk_size, gran_size,
 				      x_remove_base, x_remove_size, i);
-			if (debug_print) {
+			if (mtrr_debug) {
 				mtrr_print_out_one_result(i);
 				pr_info("\n");
 			}
@@ -786,8 +776,8 @@ int __init mtrr_cleanup(unsigned address_bits)
 		gran_size = result[i].gran_sizek;
 		gran_size <<= 10;
 		x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size);
-		set_var_mtrr_all(address_bits);
-		pr_debug("New variable MTRRs\n");
+		set_var_mtrr_all();
+		Dprintk("New variable MTRRs\n");
 		print_out_mtrr_range_state();
 		return 1;
 	} else {
@@ -802,7 +792,7 @@ int __init mtrr_cleanup(unsigned address_bits)
 	return 0;
 }
 #else
-int __init mtrr_cleanup(unsigned address_bits)
+int __init mtrr_cleanup(void)
 {
 	return 0;
 }
@@ -882,15 +872,18 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn)
 	/* extra one for all 0 */
 	int num[MTRR_NUM_TYPES + 1];
 
+	if (!mtrr_enabled())
+		return 0;
+
 	/*
 	 * Make sure we only trim uncachable memory on machines that
 	 * support the Intel MTRR architecture:
 	 */
-	if (!is_cpu(INTEL) || disable_mtrr_trim)
+	if (!cpu_feature_enabled(X86_FEATURE_MTRR) || disable_mtrr_trim)
 		return 0;
 
 	rdmsr(MSR_MTRRdefType, def, dummy);
-	def &= 0xff;
+	def &= MTRR_DEF_TYPE_TYPE;
 	if (def != MTRR_TYPE_UNCACHABLE)
 		return 0;
 

arch/x86/kernel/cpu/mtrr/cyrix.c

@@ -235,7 +235,7 @@ static void cyrix_set_arr(unsigned int reg, unsigned long base,
 }
 
 const struct mtrr_ops cyrix_mtrr_ops = {
-	.vendor            = X86_VENDOR_CYRIX,
+	.var_regs          = 8,
 	.set               = cyrix_set_arr,
 	.get               = cyrix_get_arr,
 	.get_free_region   = cyrix_get_free_region,

arch/x86/kernel/cpu/mtrr/legacy.c

+// SPDX-License-Identifier: GPL-2.0-only
+
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/syscore_ops.h>
+#include <asm/cpufeature.h>
+#include <asm/mtrr.h>
+#include <asm/processor.h>
+#include "mtrr.h"
+
+void mtrr_set_if(void)
+{
+	switch (boot_cpu_data.x86_vendor) {
+	case X86_VENDOR_AMD:
+		/* Pre-Athlon (K6) AMD CPU MTRRs */
+		if (cpu_feature_enabled(X86_FEATURE_K6_MTRR))
+			mtrr_if = &amd_mtrr_ops;
+		break;
+	case X86_VENDOR_CENTAUR:
+		if (cpu_feature_enabled(X86_FEATURE_CENTAUR_MCR))
+			mtrr_if = &centaur_mtrr_ops;
+		break;
+	case X86_VENDOR_CYRIX:
+		if (cpu_feature_enabled(X86_FEATURE_CYRIX_ARR))
+			mtrr_if = &cyrix_mtrr_ops;
+		break;
+	default:
+		break;
+	}
+}
+
+/*
+ * The suspend/resume methods are only for CPUs without MTRR. CPUs using generic
+ * MTRR driver don't require this.
+ */
+struct mtrr_value {
+	mtrr_type	ltype;
+	unsigned long	lbase;
+	unsigned long	lsize;
+};
+
+static struct mtrr_value *mtrr_value;
+
+static int mtrr_save(void)
+{
+	int i;
+
+	if (!mtrr_value)
+		return -ENOMEM;
+
+	for (i = 0; i < num_var_ranges; i++) {
+		mtrr_if->get(i, &mtrr_value[i].lbase,
+				&mtrr_value[i].lsize,
+				&mtrr_value[i].ltype);
+	}
+	return 0;
+}
+
+static void mtrr_restore(void)
+{
+	int i;
+
+	for (i = 0; i < num_var_ranges; i++) {
+		if (mtrr_value[i].lsize) {
+			mtrr_if->set(i, mtrr_value[i].lbase,
+				     mtrr_value[i].lsize,
+				     mtrr_value[i].ltype);
+		}
+	}
+}
+
+static struct syscore_ops mtrr_syscore_ops = {
+	.suspend	= mtrr_save,
+	.resume		= mtrr_restore,
+};
+
+void mtrr_register_syscore(void)
+{
+	mtrr_value = kcalloc(num_var_ranges, sizeof(*mtrr_value), GFP_KERNEL);
+
+	/*
+	 * The CPU has no MTRR and seems to not support SMP. They have
+	 * specific drivers, we use a tricky method to support
+	 * suspend/resume for them.
+	 *
+	 * TBD: is there any system with such CPU which supports
+	 * suspend/resume? If no, we should remove the code.
+	 */
+	register_syscore_ops(&mtrr_syscore_ops);
+}

arch/x86/kernel/cpu/mtrr/mtrr.c

@@ -59,15 +59,9 @@
 #define MTRR_TO_PHYS_WC_OFFSET 1000
 
 u32 num_var_ranges;
-static bool mtrr_enabled(void)
-{
-	return !!mtrr_if;
-}
 
 unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES];
-static DEFINE_MUTEX(mtrr_mutex);
-
-u64 size_or_mask, size_and_mask;
+DEFINE_MUTEX(mtrr_mutex);
 
 const struct mtrr_ops *mtrr_if;
 
@@ -105,21 +99,6 @@ static int have_wrcomb(void)
 	return mtrr_if->have_wrcomb ? mtrr_if->have_wrcomb() : 0;
 }
 
-/*  This function returns the number of variable MTRRs  */
-static void __init set_num_var_ranges(bool use_generic)
-{
-	unsigned long config = 0, dummy;
-
-	if (use_generic)
-		rdmsr(MSR_MTRRcap, config, dummy);
-	else if (is_cpu(AMD) || is_cpu(HYGON))
-		config = 2;
-	else if (is_cpu(CYRIX) || is_cpu(CENTAUR))
-		config = 8;
-
-	num_var_ranges = config & 0xff;
-}
-
 static void __init init_table(void)
 {
 	int i, max;
@@ -194,20 +173,8 @@ static inline int types_compatible(mtrr_type type1, mtrr_type type2)
  * Note that the mechanism is the same for UP systems, too; all the SMP stuff
  * becomes nops.
  */
-static void
-set_mtrr(unsigned int reg, unsigned long base, unsigned long size, mtrr_type type)
-{
-	struct set_mtrr_data data = { .smp_reg = reg,
-				      .smp_base = base,
-				      .smp_size = size,
-				      .smp_type = type
-				    };
-
-	stop_machine(mtrr_rendezvous_handler, &data, cpu_online_mask);
-}
-
-static void set_mtrr_cpuslocked(unsigned int reg, unsigned long base,
-				unsigned long size, mtrr_type type)
+static void set_mtrr(unsigned int reg, unsigned long base, unsigned long size,
+		     mtrr_type type)
 {
 	struct set_mtrr_data data = { .smp_reg = reg,
 				      .smp_base = base,
@@ -216,6 +183,8 @@ static void set_mtrr_cpuslocked(unsigned int reg, unsigned long base,
 				    };
 
 	stop_machine_cpuslocked(mtrr_rendezvous_handler, &data, cpu_online_mask);
+
+	generic_rebuild_map();
 }
 
 /**
@@ -337,7 +306,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 	/* Search for an empty MTRR */
 	i = mtrr_if->get_free_region(base, size, replace);
 	if (i >= 0) {
-		set_mtrr_cpuslocked(i, base, size, type);
+		set_mtrr(i, base, size, type);
 		if (likely(replace < 0)) {
 			mtrr_usage_table[i] = 1;
 		} else {
@@ -345,7 +314,7 @@ int mtrr_add_page(unsigned long base, unsigned long size,
 			if (increment)
 				mtrr_usage_table[i]++;
 			if (unlikely(replace != i)) {
-				set_mtrr_cpuslocked(replace, 0, 0, 0);
+				set_mtrr(replace, 0, 0, 0);
 				mtrr_usage_table[replace] = 0;
 			}
 		}
@@ -363,7 +332,7 @@ static int mtrr_check(unsigned long base, unsigned long size)
 {
 	if ((base & (PAGE_SIZE - 1)) || (size & (PAGE_SIZE - 1))) {
 		pr_warn("size and base must be multiples of 4 kiB\n");
-		pr_debug("size: 0x%lx  base: 0x%lx\n", size, base);
+		Dprintk("size: 0x%lx  base: 0x%lx\n", size, base);
 		dump_stack();
 		return -1;
 	}
@@ -454,8 +423,7 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size)
 			}
 		}
 		if (reg < 0) {
-			pr_debug("no MTRR for %lx000,%lx000 found\n",
-				 base, size);
+			Dprintk("no MTRR for %lx000,%lx000 found\n", base, size);
 			goto out;
 		}
 	}
@@ -473,7 +441,7 @@ int mtrr_del_page(int reg, unsigned long base, unsigned long size)
 		goto out;
 	}
 	if (--mtrr_usage_table[reg] < 1)
-		set_mtrr_cpuslocked(reg, 0, 0, 0);
+		set_mtrr(reg, 0, 0, 0);
 	error = reg;
  out:
 	mutex_unlock(&mtrr_mutex);
@@ -574,136 +542,54 @@ int arch_phys_wc_index(int handle)
 }
 EXPORT_SYMBOL_GPL(arch_phys_wc_index);
 
-/* The suspend/resume methods are only for CPU without MTRR. CPU using generic
- * MTRR driver doesn't require this
- */
-struct mtrr_value {
-	mtrr_type	ltype;
-	unsigned long	lbase;
-	unsigned long	lsize;
-};
-
-static struct mtrr_value mtrr_value[MTRR_MAX_VAR_RANGES];
-
-static int mtrr_save(void)
-{
-	int i;
-
-	for (i = 0; i < num_var_ranges; i++) {
-		mtrr_if->get(i, &mtrr_value[i].lbase,
-				&mtrr_value[i].lsize,
-				&mtrr_value[i].ltype);
-	}
-	return 0;
-}
-
-static void mtrr_restore(void)
-{
-	int i;
-
-	for (i = 0; i < num_var_ranges; i++) {
-		if (mtrr_value[i].lsize) {
-			set_mtrr(i, mtrr_value[i].lbase,
-				    mtrr_value[i].lsize,
-				    mtrr_value[i].ltype);
-		}
-	}
-}
-
-
-
-static struct syscore_ops mtrr_syscore_ops = {
-	.suspend	= mtrr_save,
-	.resume		= mtrr_restore,
-};
-
 int __initdata changed_by_mtrr_cleanup;
 
-#define SIZE_OR_MASK_BITS(n)  (~((1ULL << ((n) - PAGE_SHIFT)) - 1))
 /**
- * mtrr_bp_init - initialize mtrrs on the boot CPU
+ * mtrr_bp_init - initialize MTRRs on the boot CPU
  *
  * This needs to be called early; before any of the other CPUs are
  * initialized (i.e. before smp_init()).
- *
  */
 void __init mtrr_bp_init(void)
 {
+	bool generic_mtrrs = cpu_feature_enabled(X86_FEATURE_MTRR);
 	const char *why = "(not available)";
-	u32 phys_addr;
-
-	phys_addr = 32;
+	unsigned long config, dummy;
 
-	if (boot_cpu_has(X86_FEATURE_MTRR)) {
-		mtrr_if = &generic_mtrr_ops;
-		size_or_mask = SIZE_OR_MASK_BITS(36);
-		size_and_mask = 0x00f00000;
-		phys_addr = 36;
+	phys_hi_rsvd = GENMASK(31, boot_cpu_data.x86_phys_bits - 32);
 
+	if (!generic_mtrrs && mtrr_state.enabled) {
 		/*
-		 * This is an AMD specific MSR, but we assume(hope?) that
-		 * Intel will implement it too when they extend the address
-		 * bus of the Xeon.
+		 * Software overwrite of MTRR state, only for generic case.
+		 * Note that X86_FEATURE_MTRR has been reset in this case.
 		 */
-		if (cpuid_eax(0x80000000) >= 0x80000008) {
-			phys_addr = cpuid_eax(0x80000008) & 0xff;
-			/* CPUID workaround for Intel 0F33/0F34 CPU */
-			if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
-			    boot_cpu_data.x86 == 0xF &&
-			    boot_cpu_data.x86_model == 0x3 &&
-			    (boot_cpu_data.x86_stepping == 0x3 ||
-			     boot_cpu_data.x86_stepping == 0x4))
-				phys_addr = 36;
-
-			size_or_mask = SIZE_OR_MASK_BITS(phys_addr);
-			size_and_mask = ~size_or_mask & 0xfffff00000ULL;
-		} else if (boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR &&
-			   boot_cpu_data.x86 == 6) {
-			/*
-			 * VIA C* family have Intel style MTRRs,
-			 * but don't support PAE
-			 */
-			size_or_mask = SIZE_OR_MASK_BITS(32);
-			size_and_mask = 0;
-			phys_addr = 32;
-		}
-	} else {
-		switch (boot_cpu_data.x86_vendor) {
-		case X86_VENDOR_AMD:
-			if (cpu_feature_enabled(X86_FEATURE_K6_MTRR)) {
-				/* Pre-Athlon (K6) AMD CPU MTRRs */
-				mtrr_if = &amd_mtrr_ops;
-				size_or_mask = SIZE_OR_MASK_BITS(32);
-				size_and_mask = 0;
-			}
-			break;
-		case X86_VENDOR_CENTAUR:
-			if (cpu_feature_enabled(X86_FEATURE_CENTAUR_MCR)) {
-				mtrr_if = &centaur_mtrr_ops;
-				size_or_mask = SIZE_OR_MASK_BITS(32);
-				size_and_mask = 0;
-			}
-			break;
-		case X86_VENDOR_CYRIX:
-			if (cpu_feature_enabled(X86_FEATURE_CYRIX_ARR)) {
-				mtrr_if = &cyrix_mtrr_ops;
-				size_or_mask = SIZE_OR_MASK_BITS(32);
-				size_and_mask = 0;
-			}
-			break;
-		default:
-			break;
-		}
+		init_table();
+		mtrr_build_map();
+		pr_info("MTRRs set to read-only\n");
+
+		return;
 	}
 
+	if (generic_mtrrs)
+		mtrr_if = &generic_mtrr_ops;
+	else
+		mtrr_set_if();
+
 	if (mtrr_enabled()) {
-		set_num_var_ranges(mtrr_if == &generic_mtrr_ops);
+		/* Get the number of variable MTRR ranges. */
+		if (mtrr_if == &generic_mtrr_ops)
+			rdmsr(MSR_MTRRcap, config, dummy);
+		else
+			config = mtrr_if->var_regs;
+		num_var_ranges = config & MTRR_CAP_VCNT;
+
 		init_table();
 		if (mtrr_if == &generic_mtrr_ops) {
 			/* BIOS may override */
 			if (get_mtrr_state()) {
 				memory_caching_control |= CACHE_MTRR;
-				changed_by_mtrr_cleanup = mtrr_cleanup(phys_addr);
+				changed_by_mtrr_cleanup = mtrr_cleanup();
+				mtrr_build_map();
 			} else {
 				mtrr_if = NULL;
 				why = "by BIOS";
@@ -730,8 +616,14 @@ void mtrr_save_state(void)
 	smp_call_function_single(first_cpu, mtrr_save_fixed_ranges, NULL, 1);
 }
 
-static int __init mtrr_init_finialize(void)
+static int __init mtrr_init_finalize(void)
 {
+	/*
+	 * Map might exist if mtrr_overwrite_state() has been called or if
+	 * mtrr_enabled() returns true.
+	 */
+	mtrr_copy_map();
+
 	if (!mtrr_enabled())
 		return 0;
 
@@ -741,16 +633,8 @@ static int __init mtrr_init_finialize(void)
 		return 0;
 	}
 
-	/*
-	 * The CPU has no MTRR and seems to not support SMP. They have
-	 * specific drivers, we use a tricky method to support
-	 * suspend/resume for them.
-	 *
-	 * TBD: is there any system with such CPU which supports
-	 * suspend/resume? If no, we should remove the code.
-	 */
-	register_syscore_ops(&mtrr_syscore_ops);
+	mtrr_register_syscore();
 
 	return 0;
 }
-subsys_initcall(mtrr_init_finialize);
+subsys_initcall(mtrr_init_finalize);

arch/x86/kernel/cpu/mtrr/mtrr.h

@@ -10,10 +10,13 @@
 #define MTRR_CHANGE_MASK_VARIABLE  0x02
 #define MTRR_CHANGE_MASK_DEFTYPE   0x04
 
+extern bool mtrr_debug;
+#define Dprintk(x...) do { if (mtrr_debug) pr_info(x); } while (0)
+
 extern unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES];
 
 struct mtrr_ops {
-	u32	vendor;
+	u32	var_regs;
 	void	(*set)(unsigned int reg, unsigned long base,
 		       unsigned long size, mtrr_type type);
 	void	(*get)(unsigned int reg, unsigned long *base,
@@ -51,18 +54,26 @@ void fill_mtrr_var_range(unsigned int index,
 		u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi);
 bool get_mtrr_state(void);
 
-extern u64 size_or_mask, size_and_mask;
 extern const struct mtrr_ops *mtrr_if;
-
-#define is_cpu(vnd)	(mtrr_if && mtrr_if->vendor == X86_VENDOR_##vnd)
+extern struct mutex mtrr_mutex;
 
 extern unsigned int num_var_ranges;
 extern u64 mtrr_tom2;
 extern struct mtrr_state_type mtrr_state;
+extern u32 phys_hi_rsvd;
 
 void mtrr_state_warn(void);
 const char *mtrr_attrib_to_str(int x);
 void mtrr_wrmsr(unsigned, unsigned, unsigned);
+#ifdef CONFIG_X86_32
+void mtrr_set_if(void);
+void mtrr_register_syscore(void);
+#else
+static inline void mtrr_set_if(void) { }
+static inline void mtrr_register_syscore(void) { }
+#endif
+void mtrr_build_map(void);
+void mtrr_copy_map(void);
 
 /* CPU specific mtrr_ops vectors. */
 extern const struct mtrr_ops amd_mtrr_ops;
@@ -70,4 +81,14 @@ extern const struct mtrr_ops cyrix_mtrr_ops;
 extern const struct mtrr_ops centaur_mtrr_ops;
 
 extern int changed_by_mtrr_cleanup;
-extern int mtrr_cleanup(unsigned address_bits);
+extern int mtrr_cleanup(void);
+
+/*
+ * Must be used by code which uses mtrr_if to call platform-specific
+ * MTRR manipulation functions.
+ */
+static inline bool mtrr_enabled(void)
+{
+	return !!mtrr_if;
+}
+void generic_rebuild_map(void);

arch/x86/kernel/setup.c

@@ -1037,6 +1037,8 @@ void __init setup_arch(char **cmdline_p)
 	/*
 	 * VMware detection requires dmi to be available, so this
 	 * needs to be done after dmi_setup(), for the boot CPU.
+	 * For some guest types (Xen PV, SEV-SNP, TDX) it is required to be
+	 * called before cache_bp_init() for setting up MTRR state.
 	 */
 	init_hypervisor_platform();
 

arch/x86/mm/pgtable.c

@@ -702,14 +702,8 @@ void p4d_clear_huge(p4d_t *p4d)
  * pud_set_huge - setup kernel PUD mapping
  *
  * MTRRs can override PAT memory types with 4KiB granularity. Therefore, this
- * function sets up a huge page only if any of the following conditions are met:
- *
- * - MTRRs are disabled, or
- *
- * - MTRRs are enabled and the range is completely covered by a single MTRR, or
- *
- * - MTRRs are enabled and the corresponding MTRR memory type is WB, which
- *   has no effect on the requested PAT memory type.
+ * function sets up a huge page only if the complete range has the same MTRR
+ * caching mode.
  *
  * Callers should try to decrease page size (1GB -> 2MB -> 4K) if the bigger
  * page mapping attempt fails.
@@ -718,11 +712,10 @@ void p4d_clear_huge(p4d_t *p4d)
  */
 int pud_set_huge(pud_t *pud, phys_addr_t addr, pgprot_t prot)
 {
-	u8 mtrr, uniform;
+	u8 uniform;
 
-	mtrr = mtrr_type_lookup(addr, addr + PUD_SIZE, &uniform);
-	if ((mtrr != MTRR_TYPE_INVALID) && (!uniform) &&
-	    (mtrr != MTRR_TYPE_WRBACK))
+	mtrr_type_lookup(addr, addr + PUD_SIZE, &uniform);
+	if (!uniform)
 		return 0;
 
 	/* Bail out if we are we on a populated non-leaf entry: */
@@ -745,11 +738,10 @@ int pud_set_huge(pud_t *pud, phys_addr_t addr, pgprot_t prot)
  */
 int pmd_set_huge(pmd_t *pmd, phys_addr_t addr, pgprot_t prot)
 {
-	u8 mtrr, uniform;
+	u8 uniform;
 
-	mtrr = mtrr_type_lookup(addr, addr + PMD_SIZE, &uniform);
-	if ((mtrr != MTRR_TYPE_INVALID) && (!uniform) &&
-	    (mtrr != MTRR_TYPE_WRBACK)) {
+	mtrr_type_lookup(addr, addr + PMD_SIZE, &uniform);
+	if (!uniform) {
 		pr_warn_once("%s: Cannot satisfy [mem %#010llx-%#010llx] with a huge-page mapping due to MTRR override.\n",
 			     __func__, addr, addr + PMD_SIZE);
 		return 0;

arch/x86/xen/enlighten_pv.c

@@ -68,6 +68,7 @@
 #include <asm/reboot.h>
 #include <asm/hypervisor.h>
 #include <asm/mach_traps.h>
+#include <asm/mtrr.h>
 #include <asm/mwait.h>
 #include <asm/pci_x86.h>
 #include <asm/cpu.h>
@@ -119,6 +120,54 @@ static int __init parse_xen_msr_safe(char *str)
 }
 early_param("xen_msr_safe", parse_xen_msr_safe);
 
+/* Get MTRR settings from Xen and put them into mtrr_state. */
+static void __init xen_set_mtrr_data(void)
+{
+#ifdef CONFIG_MTRR
+	struct xen_platform_op op = {
+		.cmd = XENPF_read_memtype,
+		.interface_version = XENPF_INTERFACE_VERSION,
+	};
+	unsigned int reg;
+	unsigned long mask;
+	uint32_t eax, width;
+	static struct mtrr_var_range var[MTRR_MAX_VAR_RANGES] __initdata;
+
+	/* Get physical address width (only 64-bit cpus supported). */
+	width = 36;
+	eax = cpuid_eax(0x80000000);
+	if ((eax >> 16) == 0x8000 && eax >= 0x80000008) {
+		eax = cpuid_eax(0x80000008);
+		width = eax & 0xff;
+	}
+
+	for (reg = 0; reg < MTRR_MAX_VAR_RANGES; reg++) {
+		op.u.read_memtype.reg = reg;
+		if (HYPERVISOR_platform_op(&op))
+			break;
+
+		/*
+		 * Only called in dom0, which has all RAM PFNs mapped at
+		 * RAM MFNs, and all PCI space etc. is identity mapped.
+		 * This means we can treat MFN == PFN regarding MTRR settings.
+		 */
+		var[reg].base_lo = op.u.read_memtype.type;
+		var[reg].base_lo |= op.u.read_memtype.mfn << PAGE_SHIFT;
+		var[reg].base_hi = op.u.read_memtype.mfn >> (32 - PAGE_SHIFT);
+		mask = ~((op.u.read_memtype.nr_mfns << PAGE_SHIFT) - 1);
+		mask &= (1UL << width) - 1;
+		if (mask)
+			mask |= MTRR_PHYSMASK_V;
+		var[reg].mask_lo = mask;
+		var[reg].mask_hi = mask >> 32;
+	}
+
+	/* Only overwrite MTRR state if any MTRR could be got from Xen. */
+	if (reg)
+		mtrr_overwrite_state(var, reg, MTRR_TYPE_UNCACHABLE);
+#endif
+}
+
 static void __init xen_pv_init_platform(void)
 {
 	/* PV guests can't operate virtio devices without grants. */
@@ -135,6 +184,11 @@ static void __init xen_pv_init_platform(void)
 
 	/* pvclock is in shared info area */
 	xen_init_time_ops();
+
+	if (xen_initial_domain())
+		xen_set_mtrr_data();
+	else
+		mtrr_overwrite_state(NULL, 0, MTRR_TYPE_WRBACK);
 }
 
 static void __init xen_pv_guest_late_init(void)