8 files changed, 97 insertions, 55 deletions

diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c
index a8e99c56a768..72dfd9e7de0f 100644
--- a/arch/ia64/kernel/acpi.c
+++ b/arch/ia64/kernel/acpi.c
@@ -779,7 +779,7 @@ acpi_map_iosapic (acpi_handle handle, u32 depth, void *context, void **ret)
         union acpi_object *obj;
         struct acpi_table_iosapic *iosapic;
         unsigned int gsi_base;
-        int node;
+        int pxm, node;
 
         /* Only care about objects w/ a method that returns the MADT */
         if (ACPI_FAILURE(acpi_evaluate_object(handle, "_MAT", NULL, &buffer)))
@@ -805,29 +805,16 @@ acpi_map_iosapic (acpi_handle handle, u32 depth, void *context, void **ret)
         gsi_base = iosapic->global_irq_base;
 
         acpi_os_free(buffer.pointer);
-        buffer.length = ACPI_ALLOCATE_BUFFER;
-        buffer.pointer = NULL;
 
         /*
-         * OK, it's an IOSAPIC MADT entry, look for a _PXM method to tell
+         * OK, it's an IOSAPIC MADT entry, look for a _PXM value to tell
          * us which node to associate this with.
          */
-        if (ACPI_FAILURE(acpi_evaluate_object(handle, "_PXM", NULL, &buffer)))
-                return AE_OK;
-
-        if (!buffer.length || !buffer.pointer)
-                return AE_OK;
-
-        obj = buffer.pointer;
-
-        if (obj->type != ACPI_TYPE_INTEGER ||
-            obj->integer.value >= MAX_PXM_DOMAINS) {
-                acpi_os_free(buffer.pointer);
+        pxm = acpi_get_pxm(handle);
+        if (pxm < 0)
                 return AE_OK;
-        }
 
-        node = pxm_to_nid_map[obj->integer.value];
-        acpi_os_free(buffer.pointer);
+        node = pxm_to_nid_map[pxm];
 
         if (node >= MAX_NUMNODES || !node_online(node) ||
             cpus_empty(node_to_cpumask(node)))
diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
index d3f093820bc7..81c45d447394 100644
--- a/arch/ia64/kernel/entry.S
+++ b/arch/ia64/kernel/entry.S
@@ -782,7 +782,7 @@ GLOBAL_ENTRY(ia64_ret_from_ia32_execve)
         st8.spill [r2]=r8       // store return value in slot for r8 and set unat bit
         .mem.offset 8,0
         st8.spill [r3]=r0       // clear error indication in slot for r10 and set unat bit
-END(ia64_ret_from_ia32_execve_syscall)
+END(ia64_ret_from_ia32_execve)
         // fall through
 #endif /* CONFIG_IA32_SUPPORT */
 GLOBAL_ENTRY(ia64_leave_kernel)
diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S
index 0d8650f7fce7..4f3cdef75797 100644
--- a/arch/ia64/kernel/fsys.S
+++ b/arch/ia64/kernel/fsys.S
@@ -611,8 +611,10 @@ GLOBAL_ENTRY(fsys_bubble_down)
         movl r2=ia64_ret_from_syscall
         ;;
         mov rp=r2                               // set the real return addr
-        tbit.z p8,p0=r3,TIF_SYSCALL_TRACE
+        and r3=_TIF_SYSCALL_TRACEAUDIT,r3
         ;;
+        cmp.eq p8,p0=r3,r0
+
 (p10)   br.cond.spnt.many ia64_ret_from_syscall // p10==true means out registers are more than 8
 (p8)    br.call.sptk.many b6=b6         // ignore this return addr
         br.cond.sptk ia64_trace_syscall
diff --git a/arch/ia64/kernel/mca_drv.c b/arch/ia64/kernel/mca_drv.c
index ab478172c349..abc0113a821d 100644
--- a/arch/ia64/kernel/mca_drv.c
+++ b/arch/ia64/kernel/mca_drv.c
@@ -132,8 +132,7 @@ mca_handler_bh(unsigned long paddr)
         spin_unlock(&mca_bh_lock);
 
         /* This process is about to be killed itself */
-        force_sig(SIGKILL, current);
-        schedule();
+        do_exit(SIGKILL);
 }
 
 /**
@@ -439,6 +438,7 @@ recover_from_read_error(slidx_table_t *slidx, peidx_table_t *peidx, pal_bus_chec
                         psr2 = (struct ia64_psr *)&pmsa->pmsa_ipsr;
                         psr2->cpl = 0;
                         psr2->ri  = 0;
+                        psr2->i  = 0;
 
                         return 1;
                 }
diff --git a/arch/ia64/kernel/mca_drv_asm.S b/arch/ia64/kernel/mca_drv_asm.S
index bcfa05acc561..2d7e0217638d 100644
--- a/arch/ia64/kernel/mca_drv_asm.S
+++ b/arch/ia64/kernel/mca_drv_asm.S
@@ -10,6 +10,7 @@
 
 #include <asm/asmmacro.h>
 #include <asm/processor.h>
+#include <asm/ptrace.h>
 
 GLOBAL_ENTRY(mca_handler_bhhook)
         invala                                          // clear RSE ?
@@ -20,12 +21,21 @@ GLOBAL_ENTRY(mca_handler_bhhook)
         ;;                                              
         alloc           r16=ar.pfs,0,2,1,0              // make a new frame
         ;;
+        mov             ar.rsc=0
+        ;;
         mov             r13=IA64_KR(CURRENT)            // current task pointer
         ;;
-        adds            r12=IA64_TASK_THREAD_KSP_OFFSET,r13
+        mov             r2=r13
+        ;;
+        addl            r22=IA64_RBS_OFFSET,r2
+        ;;
+        mov             ar.bspstore=r22
         ;;
-        ld8             r12=[r12]                       // stack pointer
+        addl            sp=IA64_STK_OFFSET-IA64_PT_REGS_SIZE,r2
         ;;
+        adds            r2=IA64_TASK_THREAD_ON_USTACK_OFFSET,r13
+        ;;
+        st1             [r2]=r0                         // clear current->thread.on_ustack flag
         mov             loc0=r16
         movl            loc1=mca_handler_bh             // recovery C function
         ;;
@@ -34,7 +44,9 @@ GLOBAL_ENTRY(mca_handler_bhhook)
         ;;
         mov             loc1=rp
         ;;
-        br.call.sptk.many    rp=b6                      // not return ...
+        ssm             psr.i
+        ;;
+        br.call.sptk.many    rp=b6                      // does not return ...
         ;;
         mov             ar.pfs=loc0
         mov             rp=loc1
diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index 376fcbc3f8da..71c101601e3e 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -1265,6 +1265,8 @@ out:
 }
 EXPORT_SYMBOL(pfm_unregister_buffer_fmt);
 
+extern void update_pal_halt_status(int);
+
 static int
 pfm_reserve_session(struct task_struct *task, int is_syswide, unsigned int cpu)
 {
@@ -1311,6 +1313,11 @@ pfm_reserve_session(struct task_struct *task, int is_syswide, unsigned int cpu)
                 is_syswide,
                 cpu));
 
+        /*
+         * disable default_idle() to go to PAL_HALT
+         */
+        update_pal_halt_status(0);
+
         UNLOCK_PFS(flags);
 
         return 0;
@@ -1366,6 +1373,12 @@ pfm_unreserve_session(pfm_context_t *ctx, int is_syswide, unsigned int cpu)
                 is_syswide,
                 cpu));
 
+        /*
+         * if possible, enable default_idle() to go into PAL_HALT
+         */
+        if (pfm_sessions.pfs_task_sessions == 0 && pfm_sessions.pfs_sys_sessions == 0)
+                update_pal_halt_status(1);
+
         UNLOCK_PFS(flags);
 
         return 0;
@@ -4202,7 +4215,7 @@ pfm_context_load(pfm_context_t *ctx, void *arg, int count, struct pt_regs *regs)
                 DPRINT(("cannot load to [%d], invalid ctx_state=%d\n",
                         req->load_pid,
                         ctx->ctx_state));
-                return -EINVAL;
+                return -EBUSY;
         }
 
         DPRINT(("load_pid [%d] using_dbreg=%d\n", req->load_pid, ctx->ctx_fl_using_dbreg));
@@ -4704,16 +4717,26 @@ recheck:
         if (task == current || ctx->ctx_fl_system) return 0;
 
         /*
-         * if context is UNLOADED we are safe to go
-         */
-        if (state == PFM_CTX_UNLOADED) return 0;
-
-        /*
-         * no command can operate on a zombie context
+         * we are monitoring another thread
          */
-        if (state == PFM_CTX_ZOMBIE) {
-                DPRINT(("cmd %d state zombie cannot operate on context\n", cmd));
-                return -EINVAL;
+        switch(state) {
+                case PFM_CTX_UNLOADED:
+                        /*
+                         * if context is UNLOADED we are safe to go
+                         */
+                        return 0;
+                case PFM_CTX_ZOMBIE:
+                        /*
+                         * no command can operate on a zombie context
+                         */
+                        DPRINT(("cmd %d state zombie cannot operate on context\n", cmd));
+                        return -EINVAL;
+                case PFM_CTX_MASKED:
+                        /*
+                         * PMU state has been saved to software even though
+                         * the thread may still be running.
+                         */
+                        if (cmd != PFM_UNLOAD_CONTEXT) return 0;
         }
 
         /*
diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c
index 7c43aea5f7f7..ebb71f3d6d19 100644
--- a/arch/ia64/kernel/process.c
+++ b/arch/ia64/kernel/process.c
@@ -50,7 +50,7 @@
 #include "sigframe.h"
 
 void (*ia64_mark_idle)(int);
-static cpumask_t cpu_idle_map;
+static DEFINE_PER_CPU(unsigned int, cpu_idle_state);
 
 unsigned long boot_option_idle_override = 0;
 EXPORT_SYMBOL(boot_option_idle_override);
@@ -173,7 +173,9 @@ do_notify_resume_user (sigset_t *oldset, struct sigscratch *scr, long in_syscall
                 ia64_do_signal(oldset, scr, in_syscall);
 }
 
-static int pal_halt = 1;
+static int pal_halt        = 1;
+static int can_do_pal_halt = 1;
+
 static int __init nohalt_setup(char * str)
 {
         pal_halt = 0;
@@ -181,16 +183,20 @@ static int __init nohalt_setup(char * str)
 }
 __setup("nohalt", nohalt_setup);
 
+void
+update_pal_halt_status(int status)
+{
+        can_do_pal_halt = pal_halt && status;
+}
+
 /*
  * We use this if we don't have any better idle routine..
  */
 void
 default_idle (void)
 {
-        unsigned long pmu_active = ia64_getreg(_IA64_REG_PSR) & (IA64_PSR_PP | IA64_PSR_UP);
-
         while (!need_resched())
-                if (pal_halt && !pmu_active)
+                if (can_do_pal_halt)
                         safe_halt();
                 else
                         cpu_relax();
@@ -223,20 +229,31 @@ static inline void play_dead(void)
 }
 #endif /* CONFIG_HOTPLUG_CPU */
 
-
 void cpu_idle_wait(void)
 {
-        int cpu;
-        cpumask_t map;
+        unsigned int cpu, this_cpu = get_cpu();
+        cpumask_t map;
 
-        for_each_online_cpu(cpu)
-                cpu_set(cpu, cpu_idle_map);
+        set_cpus_allowed(current, cpumask_of_cpu(this_cpu));
+        put_cpu();
 
-        wmb();
-        do {
-                ssleep(1);
-                cpus_and(map, cpu_idle_map, cpu_online_map);
-        } while (!cpus_empty(map));
+        cpus_clear(map);
+        for_each_online_cpu(cpu) {
+                per_cpu(cpu_idle_state, cpu) = 1;
+                cpu_set(cpu, map);
+        }
+
+        __get_cpu_var(cpu_idle_state) = 0;
+
+        wmb();
+        do {
+                ssleep(1);
+                for_each_online_cpu(cpu) {
+                        if (cpu_isset(cpu, map) && !per_cpu(cpu_idle_state, cpu))
+                                cpu_clear(cpu, map);
+                }
+                cpus_and(map, map, cpu_online_map);
+        } while (!cpus_empty(map));
 }
 EXPORT_SYMBOL_GPL(cpu_idle_wait);
 
@@ -244,7 +261,6 @@ void __attribute__((noreturn))
 cpu_idle (void)
 {
         void (*mark_idle)(int) = ia64_mark_idle;
-        int cpu = smp_processor_id();
 
         /* endless idle loop with no priority at all */
         while (1) {
@@ -255,12 +271,13 @@ cpu_idle (void)
                 while (!need_resched()) {
                         void (*idle)(void);
 
+                        if (__get_cpu_var(cpu_idle_state))
+                                __get_cpu_var(cpu_idle_state) = 0;
+
+                        rmb();
                         if (mark_idle)
                                 (*mark_idle)(1);
 
-                        if (cpu_isset(cpu, cpu_idle_map))
-                                cpu_clear(cpu, cpu_idle_map);
-                        rmb();
                         idle = pm_idle;
                         if (!idle)
                                 idle = default_idle;
diff --git a/arch/ia64/kernel/signal.c b/arch/ia64/kernel/signal.c
index 6891d86937d9..499b7e5317cf 100644
--- a/arch/ia64/kernel/signal.c
+++ b/arch/ia64/kernel/signal.c
@@ -224,7 +224,8 @@ ia64_rt_sigreturn (struct sigscratch *scr)
          * could be corrupted.
          */
         retval = (long) &ia64_leave_kernel;
-        if (test_thread_flag(TIF_SYSCALL_TRACE))
+        if (test_thread_flag(TIF_SYSCALL_TRACE)
+            || test_thread_flag(TIF_SYSCALL_AUDIT))
                 /*
                  * strace expects to be notified after sigreturn returns even though the
                  * context to which we return may not be in the middle of a syscall.