1 files changed, 181 insertions, 98 deletions

diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c
index 4aa9eaea76c..d86669bcdfb 100644
--- a/arch/ia64/kernel/setup.c
+++ b/arch/ia64/kernel/setup.c
@@ -46,11 +46,12 @@
 #include <linux/kexec.h>
 #include <linux/crash_dump.h>
 
-#include <asm/ia32.h>
 #include <asm/machvec.h>
 #include <asm/mca.h>
 #include <asm/meminit.h>
 #include <asm/page.h>
+#include <asm/paravirt.h>
+#include <asm/paravirt_patch.h>
 #include <asm/patch.h>
 #include <asm/pgtable.h>
 #include <asm/processor.h>
@@ -58,7 +59,7 @@
 #include <asm/sections.h>
 #include <asm/setup.h>
 #include <asm/smp.h>
-#include <asm/system.h>
+#include <asm/tlbflush.h>
 #include <asm/unistd.h>
 #include <asm/hpsim.h>
 
@@ -71,7 +72,7 @@ unsigned long __per_cpu_offset[NR_CPUS];
 EXPORT_SYMBOL(__per_cpu_offset);
 #endif
 
-DEFINE_PER_CPU(struct cpuinfo_ia64, cpu_info);
+DEFINE_PER_CPU(struct cpuinfo_ia64, ia64_cpu_info);
 DEFINE_PER_CPU(unsigned long, local_per_cpu_offset);
 unsigned long ia64_cycles_per_usec;
 struct ia64_boot_param *ia64_boot_param;
@@ -96,12 +97,6 @@ static struct resource bss_resource = {
 
 unsigned long ia64_max_cacheline_size;
 
-int dma_get_cache_alignment(void)
-{
-        return ia64_max_cacheline_size;
-}
-EXPORT_SYMBOL(dma_get_cache_alignment);
-
 unsigned long ia64_iobase;	/* virtual address for I/O accesses */
 EXPORT_SYMBOL(ia64_iobase);
 struct io_space io_space[MAX_IO_SPACES];
@@ -114,6 +109,13 @@ unsigned int num_io_spaces;
  */
 #define	I_CACHE_STRIDE_SHIFT	5	/* Safest way to go: 32 bytes by 32 bytes */
 unsigned long ia64_i_cache_stride_shift = ~0;
+/*
+ * "clflush_cache_range()" needs to know what processor dependent stride size to
+ * use when it flushes cache lines including both d-cache and i-cache.
+ */
+/* Safest way to go: 32 bytes by 32 bytes */
+#define	CACHE_STRIDE_SHIFT	5
+unsigned long ia64_cache_stride_shift = ~0;
 
 /*
  * The merge_mask variable needs to be set to (max(iommu_page_size(iommu)) - 1).  This
@@ -141,9 +143,9 @@ int num_rsvd_regions __initdata;
  * This routine does not assume the incoming segments are sorted.
  */
 int __init
-filter_rsvd_memory (unsigned long start, unsigned long end, void *arg)
+filter_rsvd_memory (u64 start, u64 end, void *arg)
 {
-	unsigned long range_start, range_end, prev_start;
+	u64 range_start, range_end, prev_start;
 	void (*func)(unsigned long, unsigned long, int);
 	int i;
 
@@ -176,6 +178,29 @@ filter_rsvd_memory (unsigned long start, unsigned long end, void *arg)
 	return 0;
 }
 
+/*
+ * Similar to "filter_rsvd_memory()", but the reserved memory ranges
+ * are not filtered out.
+ */
+int __init
+filter_memory(u64 start, u64 end, void *arg)
+{
+	void (*func)(unsigned long, unsigned long, int);
+
+#if IGNORE_PFN0
+	if (start == PAGE_OFFSET) {
+		printk(KERN_WARNING "warning: skipping physical page 0\n");
+		start += PAGE_SIZE;
+		if (start >= end)
+			return 0;
+	}
+#endif
+	func = arg;
+	if (start < end)
+		call_pernode_memory(__pa(start), end - start, func);
+	return 0;
+}
+
 static void __init
 sort_regions (struct rsvd_region *rsvd_region, int max)
 {
@@ -194,6 +219,23 @@ sort_regions (struct rsvd_region *rsvd_region, int max)
 	}
 }
 
+/* merge overlaps */
+static int __init
+merge_regions (struct rsvd_region *rsvd_region, int max)
+{
+	int i;
+	for (i = 1; i < max; ++i) {
+		if (rsvd_region[i].start >= rsvd_region[i-1].end)
+			continue;
+		if (rsvd_region[i].end > rsvd_region[i-1].end)
+			rsvd_region[i-1].end = rsvd_region[i].end;
+		--max;
+		memmove(&rsvd_region[i], &rsvd_region[i+1],
+			(max - i) * sizeof(struct rsvd_region));
+	}
+	return max;
+}
+
 /*
  * Request address space for all standard resources
  */
@@ -215,6 +257,25 @@ __initcall(register_memory);
 
 
 #ifdef CONFIG_KEXEC
+
+/*
+ * This function checks if the reserved crashkernel is allowed on the specific
+ * IA64 machine flavour. Machines without an IO TLB use swiotlb and require
+ * some memory below 4 GB (i.e. in 32 bit area), see the implementation of
+ * lib/swiotlb.c. The hpzx1 architecture has an IO TLB but cannot use that
+ * in kdump case. See the comment in sba_init() in sba_iommu.c.
+ *
+ * So, the only machvec that really supports loading the kdump kernel
+ * over 4 GB is "sn2".
+ */
+static int __init check_crashkernel_memory(unsigned long pbase, size_t size)
+{
+	if (ia64_platform_is("sn2") || ia64_platform_is("uv"))
+		return 1;
+	else
+		return pbase < (1UL << 32);
+}
+
 static void __init setup_crashkernel(unsigned long total, int *n)
 {
 	unsigned long long base = 0, size = 0;
@@ -225,9 +286,20 @@ static void __init setup_crashkernel(unsigned long total, int *n)
 	if (ret == 0 && size > 0) {
 		if (!base) {
 			sort_regions(rsvd_region, *n);
+			*n = merge_regions(rsvd_region, *n);
 			base = kdump_find_rsvd_region(size,
 					rsvd_region, *n);
 		}
+
+		if (!check_crashkernel_memory(base, size)) {
+			pr_warning("crashkernel: There would be kdump memory "
+				"at %ld GB but this is unusable because it "
+				"must\nbe below 4 GB. Change the memory "
+				"configuration of the machine.\n",
+				(unsigned long)(base >> 30));
+			return;
+		}
+
 		if (base != ~0UL) {
 			printk(KERN_INFO "Reserving %ldMB of memory at %ldMB "
 					"for crashkernel (System RAM: %ldMB)\n",
@@ -260,7 +332,7 @@ static inline void __init setup_crashkernel(unsigned long total, int *n)
  *
  * Setup the reserved memory areas set aside for the boot parameters,
  * initrd, etc.  There are currently %IA64_MAX_RSVD_REGIONS defined,
- * see include/asm-ia64/meminit.h if you need to define more.
+ * see arch/ia64/include/asm/meminit.h if you need to define more.
  */
 void __init
 reserve_memory (void)
@@ -288,6 +360,8 @@ reserve_memory (void)
 	rsvd_region[n].end   = (unsigned long) ia64_imva(_end);
 	n++;
 
+	n += paravirt_reserve_memory(&rsvd_region[n]);
+
 #ifdef CONFIG_BLK_DEV_INITRD
 	if (ia64_boot_param->initrd_start) {
 		rsvd_region[n].start = (unsigned long)__va(ia64_boot_param->initrd_start);
@@ -296,7 +370,7 @@ reserve_memory (void)
 	}
 #endif
 
-#ifdef CONFIG_PROC_VMCORE
+#ifdef CONFIG_CRASH_DUMP
 	if (reserve_elfcorehdr(&rsvd_region[n].start,
 			       &rsvd_region[n].end) == 0)
 		n++;
@@ -316,6 +390,7 @@ reserve_memory (void)
 	BUG_ON(IA64_MAX_RSVD_REGIONS + 1 < n);
 
 	sort_regions(rsvd_region, num_rsvd_regions);
+	num_rsvd_regions = merge_regions(rsvd_region, num_rsvd_regions);
 }
 
 
@@ -333,7 +408,7 @@ find_initrd (void)
 		initrd_start = (unsigned long)__va(ia64_boot_param->initrd_start);
 		initrd_end   = initrd_start+ia64_boot_param->initrd_size;
 
-		printk(KERN_INFO "Initial ramdisk at: 0x%lx (%lu bytes)\n",
+		printk(KERN_INFO "Initial ramdisk at: 0x%lx (%llu bytes)\n",
 		       initrd_start, ia64_boot_param->initrd_size);
 	}
 #endif
@@ -410,7 +485,7 @@ mark_bsp_online (void)
 {
 #ifdef CONFIG_SMP
 	/* If we register an early console, allow CPU 0 to printk */
-	cpu_set(smp_processor_id(), cpu_online_map);
+	set_cpu_online(smp_processor_id(), true);
 #endif
 }
 
@@ -422,23 +497,10 @@ static __init int setup_nomca(char *s)
 }
 early_param("nomca", setup_nomca);
 
-#ifdef CONFIG_PROC_VMCORE
-/* elfcorehdr= specifies the location of elf core header
- * stored by the crashed kernel.
- */
-static int __init parse_elfcorehdr(char *arg)
+#ifdef CONFIG_CRASH_DUMP
+int __init reserve_elfcorehdr(u64 *start, u64 *end)
 {
-	if (!arg)
-		return -EINVAL;
-
-        elfcorehdr_addr = memparse(arg, &arg);
-	return 0;
-}
-early_param("elfcorehdr", parse_elfcorehdr);
-
-int __init reserve_elfcorehdr(unsigned long *start, unsigned long *end)
-{
-	unsigned long length;
+	u64 length;
 
 	/* We get the address using the kernel command line,
 	 * but the size is extracted from the EFI tables.
@@ -446,11 +508,11 @@ int __init reserve_elfcorehdr(unsigned long *start, unsigned long *end)
 	 * to work properly.
 	 */
 
-	if (elfcorehdr_addr >= ELFCORE_ADDR_MAX)
+	if (!is_vmcore_usable())
 		return -EINVAL;
 
 	if ((length = vmcore_find_descriptor_size(elfcorehdr_addr)) == 0) {
-		elfcorehdr_addr = ELFCORE_ADDR_MAX;
+		vmcore_unusable();
 		return -EINVAL;
 	}
 
@@ -466,7 +528,10 @@ setup_arch (char **cmdline_p)
 {
 	unw_init();
 
+	paravirt_arch_setup_early();
+
 	ia64_patch_vtop((u64) __start___vtop_patchlist, (u64) __end___vtop_patchlist);
+	paravirt_patch_apply();
 
 	*cmdline_p = __va(ia64_boot_param->command_line);
 	strlcpy(boot_command_line, *cmdline_p, COMMAND_LINE_SIZE);
@@ -491,20 +556,37 @@ setup_arch (char **cmdline_p)
 #ifdef CONFIG_ACPI
 	/* Initialize the ACPI boot-time table parser */
 	acpi_table_init();
+	early_acpi_boot_init();
 # ifdef CONFIG_ACPI_NUMA
 	acpi_numa_init();
-# endif
-#else
-# ifdef CONFIG_SMP
-	smp_build_cpu_map();	/* happens, e.g., with the Ski simulator */
+#  ifdef CONFIG_ACPI_HOTPLUG_CPU
+	prefill_possible_map();
+#  endif
+	per_cpu_scan_finalize((cpus_weight(early_cpu_possible_map) == 0 ?
+		32 : cpus_weight(early_cpu_possible_map)),
+		additional_cpus > 0 ? additional_cpus : 0);
 # endif
 #endif /* CONFIG_APCI_BOOT */
 
+#ifdef CONFIG_SMP
+	smp_build_cpu_map();
+#endif
 	find_memory();
 
 	/* process SAL system table: */
 	ia64_sal_init(__va(efi.sal_systab));
 
+#ifdef CONFIG_ITANIUM
+	ia64_patch_rse((u64) __start___rse_patchlist, (u64) __end___rse_patchlist);
+#else
+	{
+		unsigned long num_phys_stacked;
+
+		if (ia64_pal_rse_info(&num_phys_stacked, 0) == 0 && num_phys_stacked > 96)
+			ia64_patch_rse((u64) __start___rse_patchlist, (u64) __end___rse_patchlist);
+	}
+#endif
+
 #ifdef CONFIG_SMP
 	cpu_physical_id(0) = hard_smp_processor_id();
 #endif
@@ -512,11 +594,8 @@ setup_arch (char **cmdline_p)
 	cpu_init();	/* initialize the bootstrap CPU */
 	mmu_context_init();	/* initialize context_id bitmap */
 
-	check_sal_cache_flush();
-
-#ifdef CONFIG_ACPI
-	acpi_boot_init();
-#endif
+	paravirt_banner();
+	paravirt_arch_setup_console(cmdline_p);
 
 #ifdef CONFIG_VT
 	if (!conswitchp) {
@@ -537,10 +616,15 @@ setup_arch (char **cmdline_p)
 #endif
 
 	/* enable IA-64 Machine Check Abort Handling unless disabled */
+	if (paravirt_arch_setup_nomca())
+		nomca = 1;
 	if (!nomca)
 		ia64_mca_init();
 
 	platform_setup(cmdline_p);
+#ifndef CONFIG_IA64_HP_SIM
+	check_sal_cache_flush();
+#endif
 	paging_init();
 }
 
@@ -636,10 +720,10 @@ static void *
 c_start (struct seq_file *m, loff_t *pos)
 {
 #ifdef CONFIG_SMP
-	while (*pos < NR_CPUS && !cpu_isset(*pos, cpu_online_map))
+	while (*pos < nr_cpu_ids && !cpu_online(*pos))
 		++*pos;
 #endif
-	return *pos < NR_CPUS ? cpu_data(*pos) : NULL;
+	return *pos < nr_cpu_ids ? cpu_data(*pos) : NULL;
 }
 
 static void *
@@ -664,7 +748,7 @@ const struct seq_operations cpuinfo_op = {
 #define MAX_BRANDS	8
 static char brandname[MAX_BRANDS][128];
 
-static char * __cpuinit
+static char *
 get_model_name(__u8 family, __u8 model)
 {
 	static int overflow;
@@ -694,7 +778,7 @@ get_model_name(__u8 family, __u8 model)
 	return "Unknown";
 }
 
-static void __cpuinit
+static void
 identify_cpu (struct cpuinfo_ia64 *c)
 {
 	union {
@@ -759,28 +843,20 @@ identify_cpu (struct cpuinfo_ia64 *c)
 	c->unimpl_pa_mask = ~((1L<<63) | ((1L << phys_addr_size) - 1));
 }
 
-void __init
-setup_per_cpu_areas (void)
-{
-	/* start_kernel() requires this... */
-#ifdef CONFIG_ACPI_HOTPLUG_CPU
-	prefill_possible_map();
-#endif
-}
-
 /*
- * Calculate the max. cache line size.
+ * Do the following calculations:
  *
- * In addition, the minimum of the i-cache stride sizes is calculated for
- * "flush_icache_range()".
+ * 1. the max. cache line size.
+ * 2. the minimum of the i-cache stride sizes for "flush_icache_range()".
+ * 3. the minimum of the cache stride sizes for "clflush_cache_range()".
  */
-static void __cpuinit
-get_max_cacheline_size (void)
+static void
+get_cache_info(void)
 {
 	unsigned long line_size, max = 1;
-	u64 l, levels, unique_caches;
-        pal_cache_config_info_t cci;
-        s64 status;
+	unsigned long l, levels, unique_caches;
+	pal_cache_config_info_t cci;
+	long status;
 
         status = ia64_pal_cache_summary(&levels, &unique_caches);
         if (status != 0) {
@@ -789,33 +865,41 @@ get_max_cacheline_size (void)
                 max = SMP_CACHE_BYTES;
 		/* Safest setup for "flush_icache_range()" */
 		ia64_i_cache_stride_shift = I_CACHE_STRIDE_SHIFT;
+		/* Safest setup for "clflush_cache_range()" */
+		ia64_cache_stride_shift = CACHE_STRIDE_SHIFT;
 		goto out;
         }
 
 	for (l = 0; l < levels; ++l) {
-		status = ia64_pal_cache_config_info(l, /* cache_type (data_or_unified)= */ 2,
-						    &cci);
+		/* cache_type (data_or_unified)=2 */
+		status = ia64_pal_cache_config_info(l, 2, &cci);
 		if (status != 0) {
-			printk(KERN_ERR
-			       "%s: ia64_pal_cache_config_info(l=%lu, 2) failed (status=%ld)\n",
-			       __func__, l, status);
+			printk(KERN_ERR "%s: ia64_pal_cache_config_info"
+				"(l=%lu, 2) failed (status=%ld)\n",
+				__func__, l, status);
 			max = SMP_CACHE_BYTES;
 			/* The safest setup for "flush_icache_range()" */
 			cci.pcci_stride = I_CACHE_STRIDE_SHIFT;
+			/* The safest setup for "clflush_cache_range()" */
+			ia64_cache_stride_shift = CACHE_STRIDE_SHIFT;
 			cci.pcci_unified = 1;
+		} else {
+			if (cci.pcci_stride < ia64_cache_stride_shift)
+				ia64_cache_stride_shift = cci.pcci_stride;
+
+			line_size = 1 << cci.pcci_line_size;
+			if (line_size > max)
+				max = line_size;
 		}
-		line_size = 1 << cci.pcci_line_size;
-		if (line_size > max)
-			max = line_size;
+
 		if (!cci.pcci_unified) {
-			status = ia64_pal_cache_config_info(l,
-						    /* cache_type (instruction)= */ 1,
-						    &cci);
+			/* cache_type (instruction)=1*/
+			status = ia64_pal_cache_config_info(l, 1, &cci);
 			if (status != 0) {
-				printk(KERN_ERR
-				"%s: ia64_pal_cache_config_info(l=%lu, 1) failed (status=%ld)\n",
+				printk(KERN_ERR "%s: ia64_pal_cache_config_info"
+					"(l=%lu, 1) failed (status=%ld)\n",
 					__func__, l, status);
-				/* The safest setup for "flush_icache_range()" */
+				/* The safest setup for flush_icache_range() */
 				cci.pcci_stride = I_CACHE_STRIDE_SHIFT;
 			}
 		}
@@ -831,10 +915,10 @@ get_max_cacheline_size (void)
  * cpu_init() initializes state that is per-CPU.  This function acts
  * as a 'CPU state barrier', nothing should get across.
  */
-void __cpuinit
+void
 cpu_init (void)
 {
-	extern void __cpuinit ia64_mmu_init (void *);
+	extern void ia64_mmu_init(void *);
 	static unsigned long max_num_phys_stacked = IA64_NUM_PHYS_STACK_REG;
 	unsigned long num_phys_stacked;
 	pal_vm_info_2_u_t vmi;
@@ -851,18 +935,20 @@ cpu_init (void)
 	if (smp_processor_id() == 0) {
 		cpu_set(0, per_cpu(cpu_sibling_map, 0));
 		cpu_set(0, cpu_core_map[0]);
+	} else {
+		/*
+		 * Set ar.k3 so that assembly code in MCA handler can compute
+		 * physical addresses of per cpu variables with a simple:
+		 *   phys = ar.k3 + &per_cpu_var
+		 * and the alt-dtlb-miss handler can set per-cpu mapping into
+		 * the TLB when needed. head.S already did this for cpu0.
+		 */
+		ia64_set_kr(IA64_KR_PER_CPU_DATA,
+			    ia64_tpa(cpu_data) - (long) __per_cpu_start);
 	}
 #endif
 
-	/*
-	 * We set ar.k3 so that assembly code in MCA handler can compute
-	 * physical addresses of per cpu variables with a simple:
-	 *   phys = ar.k3 + &per_cpu_var
-	 */
-	ia64_set_kr(IA64_KR_PER_CPU_DATA,
-		    ia64_tpa(cpu_data) - (long) __per_cpu_start);
-
-	get_max_cacheline_size();
+	get_cache_info();
 
 	/*
 	 * We can't pass "local_cpu_data" to identify_cpu() because we haven't called
@@ -870,7 +956,7 @@ cpu_init (void)
 	 * depends on the data returned by identify_cpu().  We break the dependency by
 	 * accessing cpu_data() through the canonical per-CPU address.
 	 */
-	cpu_info = cpu_data + ((char *) &__ia64_per_cpu_var(cpu_info) - __per_cpu_start);
+	cpu_info = cpu_data + ((char *) &__ia64_per_cpu_var(ia64_cpu_info) - __per_cpu_start);
 	identify_cpu(cpu_info);
 
 #ifdef CONFIG_MCKINLEY
@@ -914,16 +1000,11 @@ cpu_init (void)
 					| IA64_DCR_DA | IA64_DCR_DD | IA64_DCR_LC));
 	atomic_inc(&init_mm.mm_count);
 	current->active_mm = &init_mm;
-	if (current->mm)
-		BUG();
+	BUG_ON(current->mm);
 
 	ia64_mmu_init(ia64_imva(cpu_data));
 	ia64_mca_cpu_init(ia64_imva(cpu_data));
 
-#ifdef CONFIG_IA32_SUPPORT
-	ia32_cpu_init();
-#endif
-
 	/* Clear ITC to eliminate sched_clock() overflows in human time.  */
 	ia64_set_itc(0);
 
@@ -946,9 +1027,10 @@ cpu_init (void)
 #endif
 
 	/* set ia64_ctx.max_rid to the maximum RID that is supported by all CPUs: */
-	if (ia64_pal_vm_summary(NULL, &vmi) == 0)
+	if (ia64_pal_vm_summary(NULL, &vmi) == 0) {
 		max_ctx = (1U << (vmi.pal_vm_info_2_s.rid_size - 3)) - 1;
-	else {
+		setup_ptcg_sem(vmi.pal_vm_info_2_s.max_purges, NPTCG_FROM_PAL);
+	} else {
 		printk(KERN_WARNING "cpu_init: PAL VM summary failed, assuming 18 RID bits\n");
 		max_ctx = (1U << 15) - 1;	/* use architected minimum */
 	}
@@ -969,7 +1051,6 @@ cpu_init (void)
 		max_num_phys_stacked = num_phys_stacked;
 	}
 	platform_cpu_init();
-	pm_idle = default_idle;
 }
 
 void __init
@@ -982,6 +1063,8 @@ check_bugs (void)
 static int __init run_dmi_scan(void)
 {
 	dmi_scan_machine();
+	dmi_memdev_walk();
+	dmi_set_dump_stack_arch_desc();
 	return 0;
 }
 core_initcall(run_dmi_scan);