7 files changed, 210 insertions, 39 deletions

diff --git a/arch/sparc64/kernel/head.S b/arch/sparc64/kernel/head.S
index 3eadac5e171e..31c5892f5acc 100644
--- a/arch/sparc64/kernel/head.S
+++ b/arch/sparc64/kernel/head.S
@@ -10,6 +10,7 @@
 #include <linux/config.h>
 #include <linux/version.h>
 #include <linux/errno.h>
+#include <linux/threads.h>
 #include <asm/thread_info.h>
 #include <asm/asi.h>
 #include <asm/pstate.h>
@@ -493,6 +494,35 @@ tlb_fixup_done:
         call    prom_init
          mov    %l7, %o0                        ! OpenPROM cif handler
 
+        /* Initialize current_thread_info()->cpu as early as possible.
+         * In order to do that accurately we have to patch up the get_cpuid()
+         * assembler sequences.  And that, in turn, requires that we know
+         * if we are on a Starfire box or not.  While we're here, patch up
+         * the sun4v sequences as well.
+         */
+        call    check_if_starfire
+         nop
+        call    per_cpu_patch
+         nop
+        call    sun4v_patch
+         nop
+
+#ifdef CONFIG_SMP
+        call    hard_smp_processor_id
+         nop
+        cmp     %o0, NR_CPUS
+        blu,pt  %xcc, 1f
+         nop
+        call    boot_cpu_id_too_large
+         nop
+        /* Not reached... */
+
+1:
+#else
+        mov     0, %o0
+#endif
+        stb     %o0, [%g6 + TI_CPU]
+
         /* Off we go.... */
         call    start_kernel
          nop
diff --git a/arch/sparc64/kernel/pci_sun4v.c b/arch/sparc64/kernel/pci_sun4v.c
index 2b7a1f316a93..0c0895202970 100644
--- a/arch/sparc64/kernel/pci_sun4v.c
+++ b/arch/sparc64/kernel/pci_sun4v.c
@@ -599,18 +599,128 @@ struct pci_iommu_ops pci_sun4v_iommu_ops = {
 
 /* SUN4V PCI configuration space accessors. */
 
-static inline int pci_sun4v_out_of_range(struct pci_pbm_info *pbm, unsigned int bus, unsigned int device, unsigned int func)
+struct pdev_entry {
+        struct pdev_entry       *next;
+        u32                     devhandle;
+        unsigned int            bus;
+        unsigned int            device;
+        unsigned int            func;
+};
+
+#define PDEV_HTAB_SIZE  16
+#define PDEV_HTAB_MASK  (PDEV_HTAB_SIZE - 1)
+static struct pdev_entry *pdev_htab[PDEV_HTAB_SIZE];
+
+static inline unsigned int pdev_hashfn(u32 devhandle, unsigned int bus, unsigned int device, unsigned int func)
 {
-        if (bus == pbm->pci_first_busno) {
-                if (device == 0 && func == 0)
-                        return 0;
-                return 1;
+        unsigned int val;
+
+        val = (devhandle ^ (devhandle >> 4));
+        val ^= bus;
+        val ^= device;
+        val ^= func;
+
+        return val & PDEV_HTAB_MASK;
+}
+
+static int pdev_htab_add(u32 devhandle, unsigned int bus, unsigned int device, unsigned int func)
+{
+        struct pdev_entry *p = kmalloc(sizeof(*p), GFP_KERNEL);
+        struct pdev_entry **slot;
+
+        if (!p)
+                return -ENOMEM;
+
+        slot = &pdev_htab[pdev_hashfn(devhandle, bus, device, func)];
+        p->next = *slot;
+        *slot = p;
+
+        p->devhandle = devhandle;
+        p->bus = bus;
+        p->device = device;
+        p->func = func;
+
+        return 0;
+}
+
+/* Recursively descend into the OBP device tree, rooted at toplevel_node,
+ * looking for a PCI device matching bus and devfn.
+ */
+static int obp_find(struct linux_prom_pci_registers *pregs, int toplevel_node, unsigned int bus, unsigned int devfn)
+{
+        toplevel_node = prom_getchild(toplevel_node);
+
+        while (toplevel_node != 0) {
+                int ret = obp_find(pregs, toplevel_node, bus, devfn);
+
+                if (ret != 0)
+                        return ret;
+
+                ret = prom_getproperty(toplevel_node, "reg", (char *) pregs,
+                                       sizeof(*pregs) * PROMREG_MAX);
+                if (ret == 0 || ret == -1)
+                        goto next_sibling;
+
+                if (((pregs[0].phys_hi >> 16) & 0xff) == bus &&
+                    ((pregs[0].phys_hi >> 8) & 0xff) == devfn)
+                        break;
+
+        next_sibling:
+                toplevel_node = prom_getsibling(toplevel_node);
+        }
+
+        return toplevel_node;
+}
+
+static int pdev_htab_populate(struct pci_pbm_info *pbm)
+{
+        struct linux_prom_pci_registers pr[PROMREG_MAX];
+        u32 devhandle = pbm->devhandle;
+        unsigned int bus;
+
+        for (bus = pbm->pci_first_busno; bus <= pbm->pci_last_busno; bus++) {
+                unsigned int devfn;
+
+                for (devfn = 0; devfn < 256; devfn++) {
+                        unsigned int device = PCI_SLOT(devfn);
+                        unsigned int func = PCI_FUNC(devfn);
+
+                        if (obp_find(pr, pbm->prom_node, bus, devfn)) {
+                                int err = pdev_htab_add(devhandle, bus,
+                                                        device, func);
+                                if (err)
+                                        return err;
+                        }
+                }
+        }
+
+        return 0;
+}
+
+static struct pdev_entry *pdev_find(u32 devhandle, unsigned int bus, unsigned int device, unsigned int func)
+{
+        struct pdev_entry *p;
+
+        p = pdev_htab[pdev_hashfn(devhandle, bus, device, func)];
+        while (p) {
+                if (p->devhandle == devhandle &&
+                    p->bus == bus &&
+                    p->device == device &&
+                    p->func == func)
+                        break;
+
+                p = p->next;
         }
 
+        return p;
+}
+
+static inline int pci_sun4v_out_of_range(struct pci_pbm_info *pbm, unsigned int bus, unsigned int device, unsigned int func)
+{
         if (bus < pbm->pci_first_busno ||
             bus > pbm->pci_last_busno)
                 return 1;
-        return 0;
+        return pdev_find(pbm->devhandle, bus, device, func) == NULL;
 }
 
 static int pci_sun4v_read_pci_cfg(struct pci_bus *bus_dev, unsigned int devfn,
@@ -1063,6 +1173,8 @@ static void pci_sun4v_pbm_init(struct pci_controller_info *p, int prom_node, u32
 
         pci_sun4v_get_bus_range(pbm);
         pci_sun4v_iommu_init(pbm);
+
+        pdev_htab_populate(pbm);
 }
 
 void sun4v_pci_init(int node, char *model_name)
diff --git a/arch/sparc64/kernel/setup.c b/arch/sparc64/kernel/setup.c
index 005167f82419..9cf1c88cd774 100644
--- a/arch/sparc64/kernel/setup.c
+++ b/arch/sparc64/kernel/setup.c
@@ -220,7 +220,7 @@ char reboot_command[COMMAND_LINE_SIZE];
 
 static struct pt_regs fake_swapper_regs = { { 0, }, 0, 0, 0, 0 };
 
-static void __init per_cpu_patch(void)
+void __init per_cpu_patch(void)
 {
         struct cpuid_patch_entry *p;
         unsigned long ver;
@@ -280,7 +280,7 @@ static void __init per_cpu_patch(void)
         }
 }
 
-static void __init sun4v_patch(void)
+void __init sun4v_patch(void)
 {
         struct sun4v_1insn_patch_entry *p1;
         struct sun4v_2insn_patch_entry *p2;
@@ -315,6 +315,15 @@ static void __init sun4v_patch(void)
         }
 }
 
+#ifdef CONFIG_SMP
+void __init boot_cpu_id_too_large(int cpu)
+{
+        prom_printf("Serious problem, boot cpu id (%d) >= NR_CPUS (%d)\n",
+                    cpu, NR_CPUS);
+        prom_halt();
+}
+#endif
+
 void __init setup_arch(char **cmdline_p)
 {
         /* Initialize PROM console and command line. */
@@ -332,16 +341,6 @@ void __init setup_arch(char **cmdline_p)
         conswitchp = &prom_con;
 #endif
 
-        /* Work out if we are starfire early on */
-        check_if_starfire();
-
-        /* Now we know enough to patch the get_cpuid sequences
-         * used by trap code.
-         */
-        per_cpu_patch();
-
-        sun4v_patch();
-
         boot_flags_init(*cmdline_p);
 
         idprom_init();
diff --git a/arch/sparc64/kernel/smp.c b/arch/sparc64/kernel/smp.c
index 90eaca3ec9a6..f03d52d0b88d 100644
--- a/arch/sparc64/kernel/smp.c
+++ b/arch/sparc64/kernel/smp.c
@@ -1264,7 +1264,6 @@ void __init smp_tick_init(void)
         boot_cpu_id = hard_smp_processor_id();
         current_tick_offset = timer_tick_offset;
 
-        cpu_set(boot_cpu_id, cpu_online_map);
         prof_counter(boot_cpu_id) = prof_multiplier(boot_cpu_id) = 1;
 }
 
@@ -1288,6 +1287,40 @@ int setup_profiling_timer(unsigned int multiplier)
         return 0;
 }
 
+static void __init smp_tune_scheduling(void)
+{
+        int instance, node;
+        unsigned int def, smallest = ~0U;
+
+        def = ((tlb_type == hypervisor) ?
+               (3 * 1024 * 1024) :
+               (4 * 1024 * 1024));
+
+        instance = 0;
+        while (!cpu_find_by_instance(instance, &node, NULL)) {
+                unsigned int val;
+
+                val = prom_getintdefault(node, "ecache-size", def);
+                if (val < smallest)
+                        smallest = val;
+
+                instance++;
+        }
+
+        /* Any value less than 256K is nonsense.  */
+        if (smallest < (256U * 1024U))
+                smallest = 256 * 1024;
+
+        max_cache_size = smallest;
+
+        if (smallest < 1U * 1024U * 1024U)
+                printk(KERN_INFO "Using max_cache_size of %uKB\n",
+                       smallest / 1024U);
+        else
+                printk(KERN_INFO "Using max_cache_size of %uMB\n",
+                       smallest / 1024U / 1024U);
+}
+
 /* Constrain the number of cpus to max_cpus.  */
 void __init smp_prepare_cpus(unsigned int max_cpus)
 {
@@ -1323,6 +1356,7 @@ void __init smp_prepare_cpus(unsigned int max_cpus)
         }
 
         smp_store_cpu_info(boot_cpu_id);
+        smp_tune_scheduling();
 }
 
 /* Set this up early so that things like the scheduler can init
@@ -1345,18 +1379,6 @@ void __init smp_setup_cpu_possible_map(void)
 
 void __devinit smp_prepare_boot_cpu(void)
 {
-        int cpu = hard_smp_processor_id();
-
-        if (cpu >= NR_CPUS) {
-                prom_printf("Serious problem, boot cpu id >= NR_CPUS\n");
-                prom_halt();
-        }
-
-        current_thread_info()->cpu = cpu;
-        __local_per_cpu_offset = __per_cpu_offset(cpu);
-
-        cpu_set(smp_processor_id(), cpu_online_map);
-        cpu_set(smp_processor_id(), phys_cpu_present_map);
 }
 
 int __devinit __cpu_up(unsigned int cpu)
@@ -1433,4 +1455,7 @@ void __init setup_per_cpu_areas(void)
 
         for (i = 0; i < NR_CPUS; i++, ptr += size)
                 memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start);
+
+        /* Setup %g5 for the boot cpu.  */
+        __local_per_cpu_offset = __per_cpu_offset(smp_processor_id());
 }
diff --git a/arch/sparc64/kernel/traps.c b/arch/sparc64/kernel/traps.c
index 2793a5d82380..563db528e031 100644
--- a/arch/sparc64/kernel/traps.c
+++ b/arch/sparc64/kernel/traps.c
@@ -1797,7 +1797,9 @@ static const char *sun4v_err_type_to_str(u32 type)
         };
 }
 
-static void sun4v_log_error(struct sun4v_error_entry *ent, int cpu, const char *pfx, atomic_t *ocnt)
+extern void __show_regs(struct pt_regs * regs);
+
+static void sun4v_log_error(struct pt_regs *regs, struct sun4v_error_entry *ent, int cpu, const char *pfx, atomic_t *ocnt)
 {
         int cnt;
 
@@ -1830,6 +1832,8 @@ static void sun4v_log_error(struct sun4v_error_entry *ent, int cpu, const char *
                pfx,
                ent->err_raddr, ent->err_size, ent->err_cpu);
 
+        __show_regs(regs);
+
         if ((cnt = atomic_read(ocnt)) != 0) {
                 atomic_set(ocnt, 0);
                 wmb();
@@ -1862,7 +1866,7 @@ void sun4v_resum_error(struct pt_regs *regs, unsigned long offset)
 
         put_cpu();
 
-        sun4v_log_error(&local_copy, cpu,
+        sun4v_log_error(regs, &local_copy, cpu,
                         KERN_ERR "RESUMABLE ERROR",
                         &sun4v_resum_oflow_cnt);
 }
@@ -1910,7 +1914,7 @@ void sun4v_nonresum_error(struct pt_regs *regs, unsigned long offset)
         }
 #endif
 
-        sun4v_log_error(&local_copy, cpu,
+        sun4v_log_error(regs, &local_copy, cpu,
                         KERN_EMERG "NON-RESUMABLE ERROR",
                         &sun4v_nonresum_oflow_cnt);
 
@@ -2200,7 +2204,6 @@ static inline struct reg_window *kernel_stack_up(struct reg_window *rw)
 void die_if_kernel(char *str, struct pt_regs *regs)
 {
         static int die_counter;
-        extern void __show_regs(struct pt_regs * regs);
         extern void smp_report_regs(void);
         int count = 0;
         
diff --git a/arch/sparc64/lib/checksum.S b/arch/sparc64/lib/checksum.S
index ba9cd3ccc2b2..1d230f693dc4 100644
--- a/arch/sparc64/lib/checksum.S
+++ b/arch/sparc64/lib/checksum.S
@@ -165,8 +165,9 @@ csum_partial_end_cruft:
         sll             %g1, 8, %g1
         or              %o5, %g1, %o4
 
-1:      add             %o2, %o4, %o2
+1:      addcc           %o2, %o4, %o2
+        addc            %g0, %o2, %o2
 
 csum_partial_finish:
         retl
-         mov            %o2, %o0
+         srl            %o2, 0, %o0
diff --git a/arch/sparc64/lib/csum_copy.S b/arch/sparc64/lib/csum_copy.S
index 71af48839064..e566c770a0f6 100644
--- a/arch/sparc64/lib/csum_copy.S
+++ b/arch/sparc64/lib/csum_copy.S
@@ -221,11 +221,12 @@ FUNC_NAME:		/* %o0=src, %o1=dst, %o2=len, %o3=sum */
         sll             %g1, 8, %g1
         or              %o5, %g1, %o4
 
-1:      add             %o3, %o4, %o3
+1:      addcc           %o3, %o4, %o3
+        addc            %g0, %o3, %o3
 
 70:
         retl
-         mov            %o3, %o0
+         srl            %o3, 0, %o0
 
 95:     mov             0, GLOBAL_SPARE
         brlez,pn        %o2, 4f