11 files changed, 79 insertions, 61 deletions

diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile
index 4e165342210a..bd87626c1f60 100644
--- a/arch/powerpc/Makefile
+++ b/arch/powerpc/Makefile
@@ -107,6 +107,9 @@ endif
 # No AltiVec instruction when building kernel
 KBUILD_CFLAGS += $(call cc-option,-mno-altivec)
 
+# No SPE instruction when building kernel
+KBUILD_CFLAGS += $(call cc-option,-mno-spe)
+
 # Enable unit-at-a-time mode when possible. It shrinks the
 # kernel considerably.
 KBUILD_CFLAGS += $(call cc-option,-funit-at-a-time)
diff --git a/arch/powerpc/kernel/btext.c b/arch/powerpc/kernel/btext.c
index 3ef51fb6f107..9c74fdf29eec 100644
--- a/arch/powerpc/kernel/btext.c
+++ b/arch/powerpc/kernel/btext.c
@@ -186,7 +186,9 @@ int btext_initialize(struct device_node *np)
                 pitch = *prop;
         if (pitch == 1)
                 pitch = 0x1000;
-        prop = of_get_property(np, "address", NULL);
+        prop = of_get_property(np, "linux,bootx-addr", NULL);
+        if (prop == NULL)
+                prop = of_get_property(np, "address", NULL);
         if (prop)
                 address = *prop;
 
diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index a7572cf464bd..69a91bd46115 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -251,6 +251,9 @@ syscall_exit_cont:
         bne-    2f
 1:
 #endif /* CONFIG_44x */
+BEGIN_FTR_SECTION
+        lwarx   r7,0,r1
+END_FTR_SECTION_IFSET(CPU_FTR_NEED_PAIRED_STWCX)
         stwcx.  r0,0,r1                 /* to clear the reservation */
         lwz     r4,_LINK(r1)
         lwz     r5,_CCR(r1)
@@ -717,6 +720,9 @@ restore:
         mtctr   r11
 
         PPC405_ERR77(0,r1)
+BEGIN_FTR_SECTION
+        lwarx   r11,0,r1
+END_FTR_SECTION_IFSET(CPU_FTR_NEED_PAIRED_STWCX)
         stwcx.  r0,0,r1                 /* to clear the reservation */
 
 #if !(defined(CONFIG_4xx) || defined(CONFIG_BOOKE))
diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c
index 214780798289..52e95c2158c0 100644
--- a/arch/powerpc/kernel/rtas.c
+++ b/arch/powerpc/kernel/rtas.c
@@ -19,6 +19,9 @@
 #include <linux/init.h>
 #include <linux/capability.h>
 #include <linux/delay.h>
+#include <linux/smp.h>
+#include <linux/completion.h>
+#include <linux/cpumask.h>
 
 #include <asm/prom.h>
 #include <asm/rtas.h>
@@ -34,6 +37,8 @@
 #include <asm/lmb.h>
 #include <asm/udbg.h>
 #include <asm/syscalls.h>
+#include <asm/smp.h>
+#include <asm/atomic.h>
 
 struct rtas_t rtas = {
         .lock = SPIN_LOCK_UNLOCKED
@@ -41,8 +46,10 @@ struct rtas_t rtas = {
 EXPORT_SYMBOL(rtas);
 
 struct rtas_suspend_me_data {
-        long waiting;
-        struct rtas_args *args;
+        atomic_t working; /* number of cpus accessing this struct */
+        int token; /* ibm,suspend-me */
+        int error;
+        struct completion *complete; /* wait on this until working == 0 */
 };
 
 DEFINE_SPINLOCK(rtas_data_buf_lock);
@@ -657,50 +664,62 @@ static int ibm_suspend_me_token = RTAS_UNKNOWN_SERVICE;
 #ifdef CONFIG_PPC_PSERIES
 static void rtas_percpu_suspend_me(void *info)
 {
-        int i;
         long rc;
-        long flags;
+        unsigned long msr_save;
+        int cpu;
         struct rtas_suspend_me_data *data =
                 (struct rtas_suspend_me_data *)info;
 
-        /*
-         * We use "waiting" to indicate our state.  As long
-         * as it is >0, we are still trying to all join up.
-         * If it goes to 0, we have successfully joined up and
-         * one thread got H_CONTINUE.  If any error happens,
-         * we set it to <0.
-         */
-        local_irq_save(flags);
-        do {
-                rc = plpar_hcall_norets(H_JOIN);
-                smp_rmb();
-        } while (rc == H_SUCCESS && data->waiting > 0);
-        if (rc == H_SUCCESS)
-                goto out;
+        atomic_inc(&data->working);
+
+        /* really need to ensure MSR.EE is off for H_JOIN */
+        msr_save = mfmsr();
+        mtmsr(msr_save & ~(MSR_EE));
+
+        rc = plpar_hcall_norets(H_JOIN);
+
+        mtmsr(msr_save);
 
-        if (rc == H_CONTINUE) {
-                data->waiting = 0;
-                data->args->args[data->args->nargs] =
-                        rtas_call(ibm_suspend_me_token, 0, 1, NULL);
-                for_each_possible_cpu(i)
-                        plpar_hcall_norets(H_PROD,i);
+        if (rc == H_SUCCESS) {
+                /* This cpu was prodded and the suspend is complete. */
+                goto out;
+        } else if (rc == H_CONTINUE) {
+                /* All other cpus are in H_JOIN, this cpu does
+                 * the suspend.
+                 */
+                printk(KERN_DEBUG "calling ibm,suspend-me on cpu %i\n",
+                       smp_processor_id());
+                data->error = rtas_call(data->token, 0, 1, NULL);
+
+                if (data->error)
+                        printk(KERN_DEBUG "ibm,suspend-me returned %d\n",
+                               data->error);
         } else {
-                data->waiting = -EBUSY;
-                printk(KERN_ERR "Error on H_JOIN hypervisor call\n");
+                printk(KERN_ERR "H_JOIN on cpu %i failed with rc = %ld\n",
+                       smp_processor_id(), rc);
+                data->error = rc;
         }
-
+        /* This cpu did the suspend or got an error; in either case,
+         * we need to prod all other other cpus out of join state.
+         * Extra prods are harmless.
+         */
+        for_each_online_cpu(cpu)
+                plpar_hcall_norets(H_PROD, get_hard_smp_processor_id(cpu));
 out:
-        local_irq_restore(flags);
-        return;
+        if (atomic_dec_return(&data->working) == 0)
+                complete(data->complete);
 }
 
 static int rtas_ibm_suspend_me(struct rtas_args *args)
 {
-        int i;
         long state;
         long rc;
         unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
         struct rtas_suspend_me_data data;
+        DECLARE_COMPLETION_ONSTACK(done);
+
+        if (!rtas_service_present("ibm,suspend-me"))
+                return -ENOSYS;
 
         /* Make sure the state is valid */
         rc = plpar_hcall(H_VASI_STATE, retbuf,
@@ -721,25 +740,23 @@ static int rtas_ibm_suspend_me(struct rtas_args *args)
                 return 0;
         }
 
-        data.waiting = 1;
-        data.args = args;
+        atomic_set(&data.working, 0);
+        data.token = rtas_token("ibm,suspend-me");
+        data.error = 0;
+        data.complete = &done;
 
         /* Call function on all CPUs.  One of us will make the
          * rtas call
          */
         if (on_each_cpu(rtas_percpu_suspend_me, &data, 1, 0))
-                data.waiting = -EINVAL;
+                data.error = -EINVAL;
 
-        if (data.waiting != 0)
-                printk(KERN_ERR "Error doing global join\n");
+        wait_for_completion(&done);
 
-        /* Prod each CPU.  This won't hurt, and will wake
-         * anyone we successfully put to sleep with H_JOIN.
-         */
-        for_each_possible_cpu(i)
-                plpar_hcall_norets(H_PROD, i);
+        if (data.error != 0)
+                printk(KERN_ERR "Error doing global join\n");
 
-        return data.waiting;
+        return data.error;
 }
 #else /* CONFIG_PPC_PSERIES */
 static int rtas_ibm_suspend_me(struct rtas_args *args)
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 4beb6329dfb7..c0d77723ba11 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -829,7 +829,7 @@ static void register_decrementer_clockevent(int cpu)
         *dec = decrementer_clockevent;
         dec->cpumask = cpumask_of_cpu(cpu);
 
-        printk(KERN_INFO "clockevent: %s mult[%lx] shift[%d] cpu[%d]\n",
+        printk(KERN_DEBUG "clockevent: %s mult[%lx] shift[%d] cpu[%d]\n",
                dec->name, dec->mult, dec->shift, cpu);
 
         clockevents_register_device(dec);
diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c
index d9c82d3d6482..c0f5cff77035 100644
--- a/arch/powerpc/mm/init_64.c
+++ b/arch/powerpc/mm/init_64.c
@@ -19,8 +19,6 @@
  *
  */
 
-#undef DEBUG
-
 #include <linux/signal.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
@@ -66,12 +64,6 @@
 
 #include "mmu_decl.h"
 
-#ifdef DEBUG
-#define DBG(fmt...) printk(fmt)
-#else
-#define DBG(fmt...)
-#endif
-
 #if PGTABLE_RANGE > USER_VSID_RANGE
 #warning Limited user VSID range means pagetable space is wasted
 #endif
@@ -175,8 +167,8 @@ void pgtable_cache_init(void)
                 int size = pgtable_cache_size[i];
                 const char *name = pgtable_cache_name[i];
 
-                DBG("Allocating page table cache %s (#%d) "
-                    "for size: %08x...\n", name, i, size);
+                pr_debug("Allocating page table cache %s (#%d) "
+                        "for size: %08x...\n", name, i, size);
                 pgtable_cache[i] = kmem_cache_create(name,
                                                      size, size,
                                                      SLAB_PANIC,
@@ -239,8 +231,8 @@ int __meminit vmemmap_populate(struct page *start_page,
                 if (!p)
                         return -ENOMEM;
 
-                printk(KERN_WARNING "vmemmap %08lx allocated at %p, "
-                                    "physical %08lx.\n", start, p, __pa(p));
+                pr_debug("vmemmap %08lx allocated at %p, physical %08lx.\n",
+                        start, p, __pa(p));
 
                 mapped = htab_bolt_mapping(start, start + page_size,
                                         __pa(p), mode_rw, mmu_linear_psize,
diff --git a/arch/powerpc/mm/stab.c b/arch/powerpc/mm/stab.c
index 9e85bda76216..50448d5de9d2 100644
--- a/arch/powerpc/mm/stab.c
+++ b/arch/powerpc/mm/stab.c
@@ -20,6 +20,7 @@
 #include <asm/lmb.h>
 #include <asm/abs_addr.h>
 #include <asm/firmware.h>
+#include <asm/iseries/hv_call.h>
 
 struct stab_entry {
         unsigned long esid_data;
diff --git a/arch/powerpc/mm/tlb_64.c b/arch/powerpc/mm/tlb_64.c
index eafbca52bff9..e2d867ce1c7e 100644
--- a/arch/powerpc/mm/tlb_64.c
+++ b/arch/powerpc/mm/tlb_64.c
@@ -54,12 +54,10 @@ unsigned long pte_freelist_forced_free;
         ((PAGE_SIZE - sizeof(struct pte_freelist_batch)) \
           / sizeof(pgtable_free_t))
 
-#ifdef CONFIG_SMP
 static void pte_free_smp_sync(void *arg)
 {
         /* Do nothing, just ensure we sync with all CPUs */
 }
-#endif
 
 /* This is only called when we are critically out of memory
  * (and fail to get a page in pte_free_tlb).
diff --git a/arch/powerpc/platforms/embedded6xx/prpmc2800.c b/arch/powerpc/platforms/embedded6xx/prpmc2800.c
index e484cac75095..653a5eb91c90 100644
--- a/arch/powerpc/platforms/embedded6xx/prpmc2800.c
+++ b/arch/powerpc/platforms/embedded6xx/prpmc2800.c
@@ -144,6 +144,7 @@ static int __init prpmc2800_probe(void)
                 strncpy(prpmc2800_platform_name, m,
                         min((int)len, PLATFORM_NAME_MAX - 1));
 
+        _set_L2CR(_get_L2CR() | L2CR_L2E);
         return 1;
 }
 
diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig
index 16e4e401b820..306a9d07491d 100644
--- a/arch/powerpc/platforms/pseries/Kconfig
+++ b/arch/powerpc/platforms/pseries/Kconfig
@@ -21,7 +21,7 @@ config PPC_SPLPAR
 
 config EEH
         bool "PCI Extended Error Handling (EEH)" if EMBEDDED
-        depends on PPC_PSERIES
+        depends on PPC_PSERIES && PCI
         default y if !EMBEDDED
 
 config SCANLOG
diff --git a/arch/powerpc/sysdev/cpm_common.c b/arch/powerpc/sysdev/cpm_common.c
index 66c8ad4cfce6..165981c87786 100644
--- a/arch/powerpc/sysdev/cpm_common.c
+++ b/arch/powerpc/sysdev/cpm_common.c
@@ -77,8 +77,6 @@ int __init cpm_muram_init(void)
         int i = 0;
         int ret = 0;
 
-        printk("cpm_muram_init\n");
-
         spin_lock_init(&cpm_muram_lock);
         /* initialize the info header */
         rh_init(&cpm_muram_info, 1,
@@ -193,7 +191,7 @@ void __iomem *cpm_muram_addr(unsigned long offset)
 EXPORT_SYMBOL(cpm_muram_addr);
 
 /**
- * cpm_muram_phys - turn a muram virtual address into a DMA address
+ * cpm_muram_dma - turn a muram virtual address into a DMA address
  * @offset: virtual address from cpm_muram_addr() to convert
  */
 dma_addr_t cpm_muram_dma(void __iomem *addr)