Merge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 fixes from Thomas Gleixner:
 "The x86 updates contain:

   - A fix for a longstanding PAT bug, where PAT was reported on CPUs
     that do not support it, which leads to wrong caching attributes and
     missing MTRR updates

   - Prevent overwriting of the e820 firmware table, which causes kexec
     kernels to lose the fake mptable which is stored there.

   - Cleanup of the UV/BAU code, removing unused code and making local
     functions static"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/boot/e820: Introduce the bootloader provided e820_table_firmware[] table
  x86/boot/e820: Rename the e820_table_firmware to e820_table_kexec
  x86/boot/e820: Avoid overwriting e820_table_firmware
  x86/mm/pat: Don't report PAT on CPUs that don't support it
  x86/platform/uv/BAU: Minor cleanup, make some local functions static

Author: torvalds

arch/x86/include/asm/e820/api.h

@@ -4,6 +4,7 @@
 #include <asm/e820/types.h>
 
 extern struct e820_table *e820_table;
+extern struct e820_table *e820_table_kexec;
 extern struct e820_table *e820_table_firmware;
 
 extern unsigned long pci_mem_start;

arch/x86/include/asm/pat.h

@@ -7,6 +7,7 @@
 bool pat_enabled(void);
 void pat_disable(const char *reason);
 extern void pat_init(void);
+extern void init_cache_modes(void);
 
 extern int reserve_memtype(u64 start, u64 end,
 		enum page_cache_mode req_pcm, enum page_cache_mode *ret_pcm);

arch/x86/kernel/e820.c

@@ -20,17 +20,27 @@
 #include <asm/setup.h>
 
 /*
- * We organize the E820 table into two main data structures:
+ * We organize the E820 table into three main data structures:
  *
  * - 'e820_table_firmware': the original firmware version passed to us by the
- *   bootloader - not modified by the kernel. We use this to:
+ *   bootloader - not modified by the kernel. It is composed of two parts:
+ *   the first 128 E820 memory entries in boot_params.e820_table and the remaining
+ *   (if any) entries of the SETUP_E820_EXT nodes. We use this to:
  *
  *       - inform the user about the firmware's notion of memory layout
  *         via /sys/firmware/memmap
  *
  *       - the hibernation code uses it to generate a kernel-independent MD5
  *         fingerprint of the physical memory layout of a system.
  *
+ * - 'e820_table_kexec': a slightly modified (by the kernel) firmware version
+ *   passed to us by the bootloader - the major difference between
+ *   e820_table_firmware[] and this one is that, the latter marks the setup_data
+ *   list created by the EFI boot stub as reserved, so that kexec can reuse the
+ *   setup_data information in the second kernel. Besides, e820_table_kexec[]
+ *   might also be modified by the kexec itself to fake a mptable.
+ *   We use this to:
+ *
  *       - kexec, which is a bootloader in disguise, uses the original E820
  *         layout to pass to the kexec-ed kernel. This way the original kernel
  *         can have a restricted E820 map while the kexec()-ed kexec-kernel
@@ -46,9 +56,11 @@
  * specific memory layout data during early bootup.
  */
 static struct e820_table e820_table_init		__initdata;
+static struct e820_table e820_table_kexec_init		__initdata;
 static struct e820_table e820_table_firmware_init	__initdata;
 
 struct e820_table *e820_table __refdata			= &e820_table_init;
+struct e820_table *e820_table_kexec __refdata		= &e820_table_kexec_init;
 struct e820_table *e820_table_firmware __refdata	= &e820_table_firmware_init;
 
 /* For PCI or other memory-mapped resources */
@@ -470,9 +482,9 @@ u64 __init e820__range_update(u64 start, u64 size, enum e820_type old_type, enum
 	return __e820__range_update(e820_table, start, size, old_type, new_type);
 }
 
-static u64 __init e820__range_update_firmware(u64 start, u64 size, enum e820_type old_type, enum e820_type  new_type)
+static u64 __init e820__range_update_kexec(u64 start, u64 size, enum e820_type old_type, enum e820_type  new_type)
 {
-	return __e820__range_update(e820_table_firmware, start, size, old_type, new_type);
+	return __e820__range_update(e820_table_kexec, start, size, old_type, new_type);
 }
 
 /* Remove a range of memory from the E820 table: */
@@ -546,9 +558,9 @@ void __init e820__update_table_print(void)
 	e820__print_table("modified");
 }
 
-static void __init e820__update_table_firmware(void)
+static void __init e820__update_table_kexec(void)
 {
-	e820__update_table(e820_table_firmware);
+	e820__update_table(e820_table_kexec);
 }
 
 #define MAX_GAP_END 0x100000000ull
@@ -623,7 +635,7 @@ __init void e820__setup_pci_gap(void)
 /*
  * Called late during init, in free_initmem().
  *
- * Initial e820_table and e820_table_firmware are largish __initdata arrays.
+ * Initial e820_table and e820_table_kexec are largish __initdata arrays.
  *
  * Copy them to a (usually much smaller) dynamically allocated area that is
  * sized precisely after the number of e820 entries.
@@ -643,6 +655,12 @@ __init void e820__reallocate_tables(void)
 	memcpy(n, e820_table, size);
 	e820_table = n;
 
+	size = offsetof(struct e820_table, entries) + sizeof(struct e820_entry)*e820_table_kexec->nr_entries;
+	n = kmalloc(size, GFP_KERNEL);
+	BUG_ON(!n);
+	memcpy(n, e820_table_kexec, size);
+	e820_table_kexec = n;
+
 	size = offsetof(struct e820_table, entries) + sizeof(struct e820_entry)*e820_table_firmware->nr_entries;
 	n = kmalloc(size, GFP_KERNEL);
 	BUG_ON(!n);
@@ -669,6 +687,9 @@ void __init e820__memory_setup_extended(u64 phys_addr, u32 data_len)
 	__append_e820_table(extmap, entries);
 	e820__update_table(e820_table);
 
+	memcpy(e820_table_kexec, e820_table, sizeof(*e820_table_kexec));
+	memcpy(e820_table_firmware, e820_table, sizeof(*e820_table_firmware));
+
 	early_memunmap(sdata, data_len);
 	pr_info("e820: extended physical RAM map:\n");
 	e820__print_table("extended");
@@ -727,7 +748,7 @@ core_initcall(e820__register_nvs_regions);
 /*
  * Allocate the requested number of bytes with the requsted alignment
  * and return (the physical address) to the caller. Also register this
- * range in the 'firmware' E820 table as a reserved range.
+ * range in the 'kexec' E820 table as a reserved range.
  *
  * This allows kexec to fake a new mptable, as if it came from the real
  * system.
@@ -738,9 +759,9 @@ u64 __init e820__memblock_alloc_reserved(u64 size, u64 align)
 
 	addr = __memblock_alloc_base(size, align, MEMBLOCK_ALLOC_ACCESSIBLE);
 	if (addr) {
-		e820__range_update_firmware(addr, size, E820_TYPE_RAM, E820_TYPE_RESERVED);
-		pr_info("e820: update e820_table_firmware for e820__memblock_alloc_reserved()\n");
-		e820__update_table_firmware();
+		e820__range_update_kexec(addr, size, E820_TYPE_RAM, E820_TYPE_RESERVED);
+		pr_info("e820: update e820_table_kexec for e820__memblock_alloc_reserved()\n");
+		e820__update_table_kexec();
 	}
 
 	return addr;
@@ -923,13 +944,13 @@ void __init e820__reserve_setup_data(void)
 	while (pa_data) {
 		data = early_memremap(pa_data, sizeof(*data));
 		e820__range_update(pa_data, sizeof(*data)+data->len, E820_TYPE_RAM, E820_TYPE_RESERVED_KERN);
+		e820__range_update_kexec(pa_data, sizeof(*data)+data->len, E820_TYPE_RAM, E820_TYPE_RESERVED_KERN);
 		pa_data = data->next;
 		early_memunmap(data, sizeof(*data));
 	}
 
 	e820__update_table(e820_table);
-
-	memcpy(e820_table_firmware, e820_table, sizeof(*e820_table_firmware));
+	e820__update_table(e820_table_kexec);
 
 	pr_info("extended physical RAM map:\n");
 	e820__print_table("reserve setup_data");
@@ -1062,6 +1083,7 @@ void __init e820__reserve_resources(void)
 		res++;
 	}
 
+	/* Expose the bootloader-provided memory layout to the sysfs. */
 	for (i = 0; i < e820_table_firmware->nr_entries; i++) {
 		struct e820_entry *entry = e820_table_firmware->entries + i;
 
@@ -1175,6 +1197,7 @@ void __init e820__memory_setup(void)
 
 	who = x86_init.resources.memory_setup();
 
+	memcpy(e820_table_kexec, e820_table, sizeof(*e820_table_kexec));
 	memcpy(e820_table_firmware, e820_table, sizeof(*e820_table_firmware));
 
 	pr_info("e820: BIOS-provided physical RAM map:\n");

arch/x86/kernel/kexec-bzimage64.c

@@ -100,14 +100,14 @@ static int setup_e820_entries(struct boot_params *params)
 {
 	unsigned int nr_e820_entries;
 
-	nr_e820_entries = e820_table_firmware->nr_entries;
+	nr_e820_entries = e820_table_kexec->nr_entries;
 
 	/* TODO: Pass entries more than E820_MAX_ENTRIES_ZEROPAGE in bootparams setup data */
 	if (nr_e820_entries > E820_MAX_ENTRIES_ZEROPAGE)
 		nr_e820_entries = E820_MAX_ENTRIES_ZEROPAGE;
 
 	params->e820_entries = nr_e820_entries;
-	memcpy(&params->e820_table, &e820_table_firmware->entries, nr_e820_entries*sizeof(struct e820_entry));
+	memcpy(&params->e820_table, &e820_table_kexec->entries, nr_e820_entries*sizeof(struct e820_entry));
 
 	return 0;
 }

arch/x86/kernel/setup.c

@@ -1075,6 +1075,13 @@ void __init setup_arch(char **cmdline_p)
 
 	max_possible_pfn = max_pfn;
 
+	/*
+	 * This call is required when the CPU does not support PAT. If
+	 * mtrr_bp_init() invoked it already via pat_init() the call has no
+	 * effect.
+	 */
+	init_cache_modes();
+
 	/*
 	 * Define random base addresses for memory sections after max_pfn is
 	 * defined and before each memory section base is used.

arch/x86/mm/pat.c

@@ -37,25 +37,23 @@
 #undef pr_fmt
 #define pr_fmt(fmt) "" fmt
 
-static bool boot_cpu_done;
-
-static int __read_mostly __pat_enabled = IS_ENABLED(CONFIG_X86_PAT);
-static void init_cache_modes(void);
+static bool __read_mostly boot_cpu_done;
+static bool __read_mostly pat_disabled = !IS_ENABLED(CONFIG_X86_PAT);
+static bool __read_mostly pat_initialized;
+static bool __read_mostly init_cm_done;
 
 void pat_disable(const char *reason)
 {
-	if (!__pat_enabled)
+	if (pat_disabled)
 		return;
 
 	if (boot_cpu_done) {
 		WARN_ONCE(1, "x86/PAT: PAT cannot be disabled after initialization\n");
 		return;
 	}
 
-	__pat_enabled = 0;
+	pat_disabled = true;
 	pr_info("x86/PAT: %s\n", reason);
-
-	init_cache_modes();
 }
 
 static int __init nopat(char *str)
@@ -67,7 +65,7 @@ early_param("nopat", nopat);
 
 bool pat_enabled(void)
 {
-	return !!__pat_enabled;
+	return pat_initialized;
 }
 EXPORT_SYMBOL_GPL(pat_enabled);
 
@@ -205,6 +203,8 @@ static void __init_cache_modes(u64 pat)
 		update_cache_mode_entry(i, cache);
 	}
 	pr_info("x86/PAT: Configuration [0-7]: %s\n", pat_msg);
+
+	init_cm_done = true;
 }
 
 #define PAT(x, y)	((u64)PAT_ ## y << ((x)*8))
@@ -225,6 +225,7 @@ static void pat_bsp_init(u64 pat)
 	}
 
 	wrmsrl(MSR_IA32_CR_PAT, pat);
+	pat_initialized = true;
 
 	__init_cache_modes(pat);
 }
@@ -242,10 +243,9 @@ static void pat_ap_init(u64 pat)
 	wrmsrl(MSR_IA32_CR_PAT, pat);
 }
 
-static void init_cache_modes(void)
+void init_cache_modes(void)
 {
 	u64 pat = 0;
-	static int init_cm_done;
 
 	if (init_cm_done)
 		return;
@@ -287,8 +287,6 @@ static void init_cache_modes(void)
 	}
 
 	__init_cache_modes(pat);
-
-	init_cm_done = 1;
 }
 
 /**
@@ -306,10 +304,8 @@ void pat_init(void)
 	u64 pat;
 	struct cpuinfo_x86 *c = &boot_cpu_data;
 
-	if (!pat_enabled()) {
-		init_cache_modes();
+	if (pat_disabled)
 		return;
-	}
 
 	if ((c->x86_vendor == X86_VENDOR_INTEL) &&
 	    (((c->x86 == 0x6) && (c->x86_model <= 0xd)) ||

arch/x86/platform/uv/tlb_uv.c

@@ -587,32 +587,12 @@ static unsigned long uv2_3_read_status(unsigned long offset, int rshft, int desc
 	return ((read_lmmr(offset) >> rshft) & UV_ACT_STATUS_MASK) << 1;
 }
 
-/*
- * Return whether the status of the descriptor that is normally used for this
- * cpu (the one indexed by its hub-relative cpu number) is busy.
- * The status of the original 32 descriptors is always reflected in the 64
- * bits of UVH_LB_BAU_SB_ACTIVATION_STATUS_0.
- * The bit provided by the activation_status_2 register is irrelevant to
- * the status if it is only being tested for busy or not busy.
- */
-int normal_busy(struct bau_control *bcp)
-{
-	int cpu = bcp->uvhub_cpu;
-	int mmr_offset;
-	int right_shift;
-
-	mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_0;
-	right_shift = cpu * UV_ACT_STATUS_SIZE;
-	return (((((read_lmmr(mmr_offset) >> right_shift) &
-				UV_ACT_STATUS_MASK)) << 1) == UV2H_DESC_BUSY);
-}
-
 /*
  * Entered when a bau descriptor has gone into a permanent busy wait because
  * of a hardware bug.
  * Workaround the bug.
  */
-int handle_uv2_busy(struct bau_control *bcp)
+static int handle_uv2_busy(struct bau_control *bcp)
 {
 	struct ptc_stats *stat = bcp->statp;
 
@@ -917,8 +897,9 @@ static void handle_cmplt(int completion_status, struct bau_desc *bau_desc,
  * Returns 1 if it gives up entirely and the original cpu mask is to be
  * returned to the kernel.
  */
-int uv_flush_send_and_wait(struct cpumask *flush_mask, struct bau_control *bcp,
-	struct bau_desc *bau_desc)
+static int uv_flush_send_and_wait(struct cpumask *flush_mask,
+				  struct bau_control *bcp,
+				  struct bau_desc *bau_desc)
 {
 	int seq_number = 0;
 	int completion_stat = 0;
@@ -1212,8 +1193,8 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
  * Search the message queue for any 'other' unprocessed message with the
  * same software acknowledge resource bit vector as the 'msg' message.
  */
-struct bau_pq_entry *find_another_by_swack(struct bau_pq_entry *msg,
-					   struct bau_control *bcp)
+static struct bau_pq_entry *find_another_by_swack(struct bau_pq_entry *msg,
+						  struct bau_control *bcp)
 {
 	struct bau_pq_entry *msg_next = msg + 1;
 	unsigned char swack_vec = msg->swack_vec;