Merge branch 'linus' into x86/nmitip-x86-nmi-2008-06-16_09.20_Mon

Conflicts:

	arch/x86/kernel/nmi_32.c

24 files changed, 406 insertions, 90 deletions

diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index 0df5f6f75edf..16be41446b5b 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -135,6 +135,7 @@ config IA64_GENERIC
 	  HP-zx1/sx1000		For HP systems
 	  HP-zx1/sx1000+swiotlb	For HP systems with (broken) DMA-constrained devices.
 	  SGI-SN2		For SGI Altix systems
+	  SGI-UV		For SGI UV systems
 	  Ski-simulator		For the HP simulator <http://www.hpl.hp.com/research/linux/ski/>
 
 	  If you don't know what to do, choose "generic".
@@ -170,6 +171,18 @@ config IA64_SGI_SN2
 	  to select this option.  If in doubt, select ia64 generic support
 	  instead.
 
+config IA64_SGI_UV`
+	bool "SGI-UV`"
+	select NUMA
+	select ACPI_NUMA
+	select SWIOTLB
+	help
+	  Selecting this option will optimize the kernel for use on UV based
+	  systems, but the resulting kernel binary will not run on other
+	  types of ia64 systems.  If you have an SGI UV system, it's safe
+	  to select this option.  If in doubt, select ia64 generic support
+	  instead.
+
 config IA64_HP_SIM
 	bool "Ski-simulator"
 	select SWIOTLB
diff --git a/arch/ia64/Makefile b/arch/ia64/Makefile
index ec4cca477f49..88f1a55c6c94 100644
--- a/arch/ia64/Makefile
+++ b/arch/ia64/Makefile
@@ -63,7 +63,7 @@ drivers-$(CONFIG_PCI)		+= arch/ia64/pci/
 drivers-$(CONFIG_IA64_HP_SIM)	+= arch/ia64/hp/sim/
 drivers-$(CONFIG_IA64_HP_ZX1)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/
 drivers-$(CONFIG_IA64_HP_ZX1_SWIOTLB) += arch/ia64/hp/common/ arch/ia64/hp/zx1/
-drivers-$(CONFIG_IA64_GENERIC)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/ arch/ia64/sn/
+drivers-$(CONFIG_IA64_GENERIC)	+= arch/ia64/hp/common/ arch/ia64/hp/zx1/ arch/ia64/hp/sim/ arch/ia64/sn/ arch/ia64/uv/
 drivers-$(CONFIG_OPROFILE)	+= arch/ia64/oprofile/
 
 boot := arch/ia64/hp/sim/boot
diff --git a/arch/ia64/ia32/ia32_support.c b/arch/ia64/ia32/ia32_support.c
index 896b1ebbfb26..a6965ddafc46 100644
--- a/arch/ia64/ia32/ia32_support.c
+++ b/arch/ia64/ia32/ia32_support.c
@@ -15,7 +15,6 @@
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/mm.h>
-#include <linux/personality.h>
 #include <linux/sched.h>
 
 #include <asm/intrinsics.h>
@@ -29,7 +28,6 @@
 
 extern int die_if_kernel (char *str, struct pt_regs *regs, long err);
 
-struct exec_domain ia32_exec_domain;
 struct page *ia32_shared_page[NR_CPUS];
 unsigned long *ia32_boot_gdt;
 unsigned long *cpu_gdt_table[NR_CPUS];
@@ -240,14 +238,6 @@ ia32_cpu_init (void)
 static int __init
 ia32_init (void)
 {
-	ia32_exec_domain.name = "Linux/x86";
-	ia32_exec_domain.handler = NULL;
-	ia32_exec_domain.pers_low = PER_LINUX32;
-	ia32_exec_domain.pers_high = PER_LINUX32;
-	ia32_exec_domain.signal_map = default_exec_domain.signal_map;
-	ia32_exec_domain.signal_invmap = default_exec_domain.signal_invmap;
-	register_exec_domain(&ia32_exec_domain);
-
 #if PAGE_SHIFT > IA32_PAGE_SHIFT
 	{
 		extern struct kmem_cache *ia64_partial_page_cachep;
diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c
index 19709a079635..43687cc60dfb 100644
--- a/arch/ia64/kernel/acpi.c
+++ b/arch/ia64/kernel/acpi.c
@@ -117,7 +117,10 @@ acpi_get_sysname(void)
 	if (!strcmp(hdr->oem_id, "HP")) {
 		return "hpzx1";
 	} else if (!strcmp(hdr->oem_id, "SGI")) {
-		return "sn2";
+		if (!strcmp(hdr->oem_table_id + 4, "UV"))
+			return "uv";
+		else
+			return "sn2";
 	}
 
 	return "dig";
@@ -130,6 +133,8 @@ acpi_get_sysname(void)
 	return "hpzx1_swiotlb";
 # elif defined (CONFIG_IA64_SGI_SN2)
 	return "sn2";
+# elif defined (CONFIG_IA64_SGI_UV)
+	return "uv";
 # elif defined (CONFIG_IA64_DIG)
 	return "dig";
 # else
@@ -460,7 +465,6 @@ void __init acpi_numa_slit_init(struct acpi_table_slit *slit)
 		printk(KERN_ERR
 		       "ACPI 2.0 SLIT: size mismatch: %d expected, %d actual\n",
 		       len, slit->header.length);
-		memset(numa_slit, 10, sizeof(numa_slit));
 		return;
 	}
 	slit_table = slit;
@@ -569,8 +573,14 @@ void __init acpi_numa_arch_fixup(void)
 	printk(KERN_INFO "Number of memory chunks in system = %d\n",
 	       num_node_memblks);
 
-	if (!slit_table)
+	if (!slit_table) {
+		for (i = 0; i < MAX_NUMNODES; i++)
+			for (j = 0; j < MAX_NUMNODES; j++)
+				node_distance(i, j) = i == j ? LOCAL_DISTANCE :
+							REMOTE_DISTANCE;
 		return;
+	}
+
 	memset(numa_slit, -1, sizeof(numa_slit));
 	for (i = 0; i < slit_table->locality_count; i++) {
 		if (!pxm_bit_test(i))
@@ -622,6 +632,9 @@ void acpi_unregister_gsi(u32 gsi)
 	if (acpi_irq_model == ACPI_IRQ_MODEL_PLATFORM)
 		return;
 
+	if (has_8259 && gsi < 16)
+		return;
+
 	iosapic_unregister_intr(gsi);
 }
 
diff --git a/arch/ia64/kernel/entry.S b/arch/ia64/kernel/entry.S
index e49ad8c5dc69..ca2bb95726de 100644
--- a/arch/ia64/kernel/entry.S
+++ b/arch/ia64/kernel/entry.S
@@ -1156,6 +1156,9 @@ skip_rbs_switch:
 	 *	r31 = current->thread_info->flags
 	 * On exit:
 	 *	p6 = TRUE if work-pending-check needs to be redone
+	 *
+	 * Interrupts are disabled on entry, reenabled depend on work, and
+	 * disabled on exit.
 	 */
 .work_pending_syscall:
 	add r2=-8,r2
@@ -1164,16 +1167,16 @@ skip_rbs_switch:
 	st8 [r2]=r8
 	st8 [r3]=r10
 .work_pending:
-	tbit.z p6,p0=r31,TIF_NEED_RESCHED		// current_thread_info()->need_resched==0?
+	tbit.z p6,p0=r31,TIF_NEED_RESCHED	// is resched not needed?
 (p6)	br.cond.sptk.few .notify
 #ifdef CONFIG_PREEMPT
 (pKStk) dep r21=-1,r0,PREEMPT_ACTIVE_BIT,1
 	;;
 (pKStk) st4 [r20]=r21
-	ssm psr.i		// enable interrupts
 #endif
+	ssm psr.i		// enable interrupts
 	br.call.spnt.many rp=schedule
-.ret9:	cmp.eq p6,p0=r0,r0				// p6 <- 1
+.ret9:	cmp.eq p6,p0=r0,r0	// p6 <- 1 (re-check)
 	rsm psr.i		// disable interrupts
 	;;
 #ifdef CONFIG_PREEMPT
@@ -1182,13 +1185,13 @@ skip_rbs_switch:
 (pKStk)	st4 [r20]=r0		// preempt_count() <- 0
 #endif
 (pLvSys)br.cond.sptk.few  .work_pending_syscall_end
-	br.cond.sptk.many .work_processed_kernel	// re-check
+	br.cond.sptk.many .work_processed_kernel
 
 .notify:
 (pUStk)	br.call.spnt.many rp=notify_resume_user
-.ret10:	cmp.ne p6,p0=r0,r0				// p6 <- 0
+.ret10:	cmp.ne p6,p0=r0,r0	// p6 <- 0 (don't re-check)
 (pLvSys)br.cond.sptk.few  .work_pending_syscall_end
-	br.cond.sptk.many .work_processed_kernel	// don't re-check
+	br.cond.sptk.many .work_processed_kernel
 
 .work_pending_syscall_end:
 	adds r2=PT(R8)+16,r12
@@ -1196,7 +1199,7 @@ skip_rbs_switch:
 	;;
 	ld8 r8=[r2]
 	ld8 r10=[r3]
-	br.cond.sptk.many .work_processed_syscall	// re-check
+	br.cond.sptk.many .work_processed_syscall
 
 END(ia64_leave_kernel)
 
@@ -1234,9 +1237,12 @@ GLOBAL_ENTRY(ia64_invoke_schedule_tail)
 END(ia64_invoke_schedule_tail)
 
 	/*
-	 * Setup stack and call do_notify_resume_user().  Note that pSys and pNonSys need to
-	 * be set up by the caller.  We declare 8 input registers so the system call
-	 * args get preserved, in case we need to restart a system call.
+	 * Setup stack and call do_notify_resume_user(), keeping interrupts
+	 * disabled.
+	 *
+	 * Note that pSys and pNonSys need to be set up by the caller.
+	 * We declare 8 input registers so the system call args get preserved,
+	 * in case we need to restart a system call.
 	 */
 ENTRY(notify_resume_user)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8)
diff --git a/arch/ia64/kernel/init_task.c b/arch/ia64/kernel/init_task.c
index bc8efcad28b8..9d7e1c66faf4 100644
--- a/arch/ia64/kernel/init_task.c
+++ b/arch/ia64/kernel/init_task.c
@@ -18,7 +18,6 @@
 #include <asm/pgtable.h>
 
 static struct fs_struct init_fs = INIT_FS;
-static struct files_struct init_files = INIT_FILES;
 static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
 static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
 struct mm_struct init_mm = INIT_MM(init_mm);
diff --git a/arch/ia64/kernel/ivt.S b/arch/ia64/kernel/ivt.S
index 6678c49daba3..80b44ea052d7 100644
--- a/arch/ia64/kernel/ivt.S
+++ b/arch/ia64/kernel/ivt.S
@@ -1076,48 +1076,6 @@ END(ia64_syscall_setup)
 	DBG_FAULT(15)
 	FAULT(15)
 
-	/*
-	 * Squatting in this space ...
-	 *
-	 * This special case dispatcher for illegal operation faults allows preserved
-	 * registers to be modified through a callback function (asm only) that is handed
-	 * back from the fault handler in r8. Up to three arguments can be passed to the
-	 * callback function by returning an aggregate with the callback as its first
-	 * element, followed by the arguments.
-	 */
-ENTRY(dispatch_illegal_op_fault)
-	.prologue
-	.body
-	SAVE_MIN_WITH_COVER
-	ssm psr.ic | PSR_DEFAULT_BITS
-	;;
-	srlz.i		// guarantee that interruption collection is on
-	;;
-(p15)	ssm psr.i	// restore psr.i
-	adds r3=8,r2	// set up second base pointer for SAVE_REST
-	;;
-	alloc r14=ar.pfs,0,0,1,0	// must be first in insn group
-	mov out0=ar.ec
-	;;
-	SAVE_REST
-	PT_REGS_UNWIND_INFO(0)
-	;;
-	br.call.sptk.many rp=ia64_illegal_op_fault
-.ret0:	;;
-	alloc r14=ar.pfs,0,0,3,0	// must be first in insn group
-	mov out0=r9
-	mov out1=r10
-	mov out2=r11
-	movl r15=ia64_leave_kernel
-	;;
-	mov rp=r15
-	mov b6=r8
-	;;
-	cmp.ne p6,p0=0,r8
-(p6)	br.call.dpnt.many b6=b6		// call returns to ia64_leave_kernel
-	br.sptk.many ia64_leave_kernel
-END(dispatch_illegal_op_fault)
-
 	.org ia64_ivt+0x4000
 /////////////////////////////////////////////////////////////////////////////////////////
 // 0x4000 Entry 16 (size 64 bundles) Reserved
@@ -1715,6 +1673,48 @@ END(ia32_interrupt)
 	DBG_FAULT(67)
 	FAULT(67)
 
+	/*
+	 * Squatting in this space ...
+	 *
+	 * This special case dispatcher for illegal operation faults allows preserved
+	 * registers to be modified through a callback function (asm only) that is handed
+	 * back from the fault handler in r8. Up to three arguments can be passed to the
+	 * callback function by returning an aggregate with the callback as its first
+	 * element, followed by the arguments.
+	 */
+ENTRY(dispatch_illegal_op_fault)
+	.prologue
+	.body
+	SAVE_MIN_WITH_COVER
+	ssm psr.ic | PSR_DEFAULT_BITS
+	;;
+	srlz.i		// guarantee that interruption collection is on
+	;;
+(p15)	ssm psr.i	// restore psr.i
+	adds r3=8,r2	// set up second base pointer for SAVE_REST
+	;;
+	alloc r14=ar.pfs,0,0,1,0	// must be first in insn group
+	mov out0=ar.ec
+	;;
+	SAVE_REST
+	PT_REGS_UNWIND_INFO(0)
+	;;
+	br.call.sptk.many rp=ia64_illegal_op_fault
+.ret0:	;;
+	alloc r14=ar.pfs,0,0,3,0	// must be first in insn group
+	mov out0=r9
+	mov out1=r10
+	mov out2=r11
+	movl r15=ia64_leave_kernel
+	;;
+	mov rp=r15
+	mov b6=r8
+	;;
+	cmp.ne p6,p0=0,r8
+(p6)	br.call.dpnt.many b6=b6		// call returns to ia64_leave_kernel
+	br.sptk.many ia64_leave_kernel
+END(dispatch_illegal_op_fault)
+
 #ifdef CONFIG_IA32_SUPPORT
 
 	/*
diff --git a/arch/ia64/kernel/minstate.h b/arch/ia64/kernel/minstate.h
index 7c548ac52bbc..74b6d670aaef 100644
--- a/arch/ia64/kernel/minstate.h
+++ b/arch/ia64/kernel/minstate.h
@@ -15,6 +15,9 @@
 #define ACCOUNT_SYS_ENTER
 #endif
 
+.section ".data.patch.rse", "a"
+.previous
+
 /*
  * DO_SAVE_MIN switches to the kernel stacks (if necessary) and saves
  * the minimum state necessary that allows us to turn psr.ic back
@@ -40,7 +43,7 @@
  * Note that psr.ic is NOT turned on by this macro.  This is so that
  * we can pass interruption state as arguments to a handler.
  */
-#define DO_SAVE_MIN(COVER,SAVE_IFS,EXTRA)							\
+#define DO_SAVE_MIN(COVER,SAVE_IFS,EXTRA,WORKAROUND)						\
 	mov r16=IA64_KR(CURRENT);	/* M */							\
 	mov r27=ar.rsc;			/* M */							\
 	mov r20=r1;			/* A */							\
@@ -87,6 +90,7 @@
 	tbit.nz p15,p0=r29,IA64_PSR_I_BIT;							\
 	mov r29=b0										\
 	;;											\
+	WORKAROUND;										\
 	adds r16=PT(R8),r1;	/* initialize first base pointer */				\
 	adds r17=PT(R9),r1;	/* initialize second base pointer */				\
 (pKStk)	mov r18=r0;		/* make sure r18 isn't NaT */					\
@@ -206,6 +210,40 @@
 	st8 [r25]=r10;      	/* ar.ssd */	\
 	;;
 
-#define SAVE_MIN_WITH_COVER	DO_SAVE_MIN(cover, mov r30=cr.ifs,)
-#define SAVE_MIN_WITH_COVER_R19	DO_SAVE_MIN(cover, mov r30=cr.ifs, mov r15=r19)
-#define SAVE_MIN		DO_SAVE_MIN(     , mov r30=r0, )
+#define RSE_WORKAROUND				\
+(pUStk) extr.u r17=r18,3,6;			\
+(pUStk)	sub r16=r18,r22;			\
+[1:](pKStk)	br.cond.sptk.many 1f;		\
+	.xdata4 ".data.patch.rse",1b-.		\
+	;;					\
+	cmp.ge p6,p7 = 33,r17;			\
+	;;					\
+(p6)	mov r17=0x310;				\
+(p7)	mov r17=0x308;				\
+	;;					\
+	cmp.leu p1,p0=r16,r17;			\
+(p1)	br.cond.sptk.many 1f;			\
+	dep.z r17=r26,0,62;			\
+	movl r16=2f;				\
+	;;					\
+	mov ar.pfs=r17;				\
+	dep r27=r0,r27,16,14;			\
+	mov b0=r16;				\
+	;;					\
+	br.ret.sptk b0;				\
+	;;					\
+2:						\
+	mov ar.rsc=r0				\
+	;;					\
+	flushrs;				\
+	;;					\
+	mov ar.bspstore=r22			\
+	;;					\
+	mov r18=ar.bsp;				\
+	;;					\
+1:						\
+	.pred.rel "mutex", pKStk, pUStk
+
+#define SAVE_MIN_WITH_COVER	DO_SAVE_MIN(cover, mov r30=cr.ifs, , RSE_WORKAROUND)
+#define SAVE_MIN_WITH_COVER_R19	DO_SAVE_MIN(cover, mov r30=cr.ifs, mov r15=r19, RSE_WORKAROUND)
+#define SAVE_MIN			DO_SAVE_MIN(     , mov r30=r0, , )
diff --git a/arch/ia64/kernel/palinfo.c b/arch/ia64/kernel/palinfo.c
index 4547a2092af9..9dc00f7fe10e 100644
--- a/arch/ia64/kernel/palinfo.c
+++ b/arch/ia64/kernel/palinfo.c
@@ -900,12 +900,6 @@ static void
 palinfo_smp_call(void *info)
 {
 	palinfo_smp_data_t *data = (palinfo_smp_data_t *)info;
-	if (data == NULL) {
-		printk(KERN_ERR "palinfo: data pointer is NULL\n");
-		data->ret = 0; /* no output */
-		return;
-	}
-	/* does this actual call */
 	data->ret = (*data->func)(data->page);
 }
 
diff --git a/arch/ia64/kernel/patch.c b/arch/ia64/kernel/patch.c
index e0dca8743dbb..b83b2c516008 100644
--- a/arch/ia64/kernel/patch.c
+++ b/arch/ia64/kernel/patch.c
@@ -115,6 +115,29 @@ ia64_patch_vtop (unsigned long start, unsigned long end)
 	ia64_srlz_i();
 }
 
+/*
+ * Disable the RSE workaround by turning the conditional branch
+ * that we tagged in each place the workaround was used into an
+ * unconditional branch.
+ */
+void __init
+ia64_patch_rse (unsigned long start, unsigned long end)
+{
+	s32 *offp = (s32 *) start;
+	u64 ip, *b;
+
+	while (offp < (s32 *) end) {
+		ip = (u64) offp + *offp;
+
+		b = (u64 *)(ip & -16);
+		b[1] &= ~0xf800000L;
+		ia64_fc((void *) ip);
+		++offp;
+	}
+	ia64_sync_i();
+	ia64_srlz_i();
+}
+
 void __init
 ia64_patch_mckinley_e9 (unsigned long start, unsigned long end)
 {
diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index c1ad27de2dd2..71d05133f556 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -5013,12 +5013,13 @@ pfm_context_force_terminate(pfm_context_t *ctx, struct pt_regs *regs)
 }
 
 static int pfm_ovfl_notify_user(pfm_context_t *ctx, unsigned long ovfl_pmds);
+
  /*
   * pfm_handle_work() can be called with interrupts enabled
   * (TIF_NEED_RESCHED) or disabled. The down_interruptible
   * call may sleep, therefore we must re-enable interrupts
   * to avoid deadlocks. It is safe to do so because this function
-  * is called ONLY when returning to user level (PUStk=1), in which case
+  * is called ONLY when returning to user level (pUStk=1), in which case
   * there is no risk of kernel stack overflow due to deep
   * interrupt nesting.
   */
@@ -5034,7 +5035,8 @@ pfm_handle_work(void)
 
 	ctx = PFM_GET_CTX(current);
 	if (ctx == NULL) {
-		printk(KERN_ERR "perfmon: [%d] has no PFM context\n", task_pid_nr(current));
+		printk(KERN_ERR "perfmon: [%d] has no PFM context\n",
+			task_pid_nr(current));
 		return;
 	}
 
@@ -5058,11 +5060,12 @@ pfm_handle_work(void)
 	/*
 	 * must be done before we check for simple-reset mode
 	 */
-	if (ctx->ctx_fl_going_zombie || ctx->ctx_state == PFM_CTX_ZOMBIE) goto do_zombie;
-
+	if (ctx->ctx_fl_going_zombie || ctx->ctx_state == PFM_CTX_ZOMBIE)
+		goto do_zombie;
 
 	//if (CTX_OVFL_NOBLOCK(ctx)) goto skip_blocking;
-	if (reason == PFM_TRAP_REASON_RESET) goto skip_blocking;
+	if (reason == PFM_TRAP_REASON_RESET)
+		goto skip_blocking;
 
 	/*
 	 * restore interrupt mask to what it was on entry.
@@ -5110,7 +5113,8 @@ do_zombie:
 	/*
 	 * in case of interruption of down() we don't restart anything
 	 */
-	if (ret < 0) goto nothing_to_do;
+	if (ret < 0)
+		goto nothing_to_do;
 
 skip_blocking:
 	pfm_resume_after_ovfl(ctx, ovfl_regs, regs);
diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c
index 58dcfac5ea88..a3a34b4eb038 100644
--- a/arch/ia64/kernel/process.c
+++ b/arch/ia64/kernel/process.c
@@ -167,11 +167,18 @@ void tsk_clear_notify_resume(struct task_struct *tsk)
 	clear_ti_thread_flag(task_thread_info(tsk), TIF_NOTIFY_RESUME);
 }
 
+/*
+ * do_notify_resume_user():
+ *	Called from notify_resume_user at entry.S, with interrupts disabled.
+ */
 void
-do_notify_resume_user (sigset_t *unused, struct sigscratch *scr, long in_syscall)
+do_notify_resume_user(sigset_t *unused, struct sigscratch *scr, long in_syscall)
 {
 	if (fsys_mode(current, &scr->pt)) {
-		/* defer signal-handling etc. until we return to privilege-level 0.  */
+		/*
+		 * defer signal-handling etc. until we return to
+		 * privilege-level 0.
+		 */
 		if (!ia64_psr(&scr->pt)->lp)
 			ia64_psr(&scr->pt)->lp = 1;
 		return;
@@ -179,16 +186,26 @@ do_notify_resume_user (sigset_t *unused, struct sigscratch *scr, long in_syscall
 
 #ifdef CONFIG_PERFMON
 	if (current->thread.pfm_needs_checking)
+		/*
+		 * Note: pfm_handle_work() allow us to call it with interrupts
+		 * disabled, and may enable interrupts within the function.
+		 */
 		pfm_handle_work();
 #endif
 
 	/* deal with pending signal delivery */
-	if (test_thread_flag(TIF_SIGPENDING))
+	if (test_thread_flag(TIF_SIGPENDING)) {
+		local_irq_enable();	/* force interrupt enable */
 		ia64_do_signal(scr, in_syscall);
+	}
 
 	/* copy user rbs to kernel rbs */
-	if (unlikely(test_thread_flag(TIF_RESTORE_RSE)))
+	if (unlikely(test_thread_flag(TIF_RESTORE_RSE))) {
+		local_irq_enable();	/* force interrupt enable */
 		ia64_sync_krbs();
+	}
+
+	local_irq_disable();	/* force interrupt disable */
 }
 
 static int pal_halt        = 1;
diff --git a/arch/ia64/kernel/sal.c b/arch/ia64/kernel/sal.c
index a3022dc48ef8..7e0259709c04 100644
--- a/arch/ia64/kernel/sal.c
+++ b/arch/ia64/kernel/sal.c
@@ -229,6 +229,14 @@ static void __init sal_desc_ap_wakeup(void *p) { }
  */
 static int sal_cache_flush_drops_interrupts;
 
+static int __init
+force_pal_cache_flush(char *str)
+{
+	sal_cache_flush_drops_interrupts = 1;
+	return 0;
+}
+early_param("force_pal_cache_flush", force_pal_cache_flush);
+
 void __init
 check_sal_cache_flush (void)
 {
@@ -237,6 +245,9 @@ check_sal_cache_flush (void)
 	u64 vector, cache_type = 3;
 	struct ia64_sal_retval isrv;
 
+	if (sal_cache_flush_drops_interrupts)
+		return;
+
 	cpu = get_cpu();
 	local_irq_save(flags);
 
diff --git a/arch/ia64/kernel/setup.c b/arch/ia64/kernel/setup.c
index 5015ca1275ca..f48a809c686d 100644
--- a/arch/ia64/kernel/setup.c
+++ b/arch/ia64/kernel/setup.c
@@ -239,6 +239,25 @@ __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;
@@ -252,6 +271,16 @@ static void __init setup_crashkernel(unsigned long total, int *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",
@@ -531,6 +560,17 @@ setup_arch (char **cmdline_p)
 	/* 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
+	{
+		u64 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
diff --git a/arch/ia64/kernel/vmlinux.lds.S b/arch/ia64/kernel/vmlinux.lds.S
index 80622acc95de..5929ab10a289 100644
--- a/arch/ia64/kernel/vmlinux.lds.S
+++ b/arch/ia64/kernel/vmlinux.lds.S
@@ -156,6 +156,13 @@ SECTIONS
 	  __end___vtop_patchlist = .;
 	}
 
+  .data.patch.rse : AT(ADDR(.data.patch.rse) - LOAD_OFFSET)
+	{
+	  __start___rse_patchlist = .;
+	  *(.data.patch.rse)
+	  __end___rse_patchlist = .;
+	}
+
   .data.patch.mckinley_e9 : AT(ADDR(.data.patch.mckinley_e9) - LOAD_OFFSET)
 	{
 	  __start___mckinley_e9_bundles = .;
diff --git a/arch/ia64/kvm/Makefile b/arch/ia64/kvm/Makefile
index 52353397a1a4..112791dd2542 100644
--- a/arch/ia64/kvm/Makefile
+++ b/arch/ia64/kvm/Makefile
@@ -7,7 +7,6 @@ offsets-file := asm-offsets.h
 always  := $(offsets-file)
 targets := $(offsets-file)
 targets += arch/ia64/kvm/asm-offsets.s
-clean-files := $(addprefix $(objtree)/,$(targets) $(obj)/memcpy.S $(obj)/memset.S)
 
 # Default sed regexp - multiline due to syntax constraints
 define sed-y
@@ -54,5 +53,5 @@ EXTRA_CFLAGS_vcpu.o += -mfixed-range=f2-f5,f12-f127
 kvm-intel-objs = vmm.o vmm_ivt.o trampoline.o vcpu.o optvfault.o mmio.o \
 	vtlb.o process.o
 #Add link memcpy and memset to avoid possible structure assignment error
-kvm-intel-objs += ../lib/memset.o ../lib/memcpy.o
+kvm-intel-objs += memcpy.o memset.o
 obj-$(CONFIG_KVM_INTEL) += kvm-intel.o
diff --git a/arch/ia64/kvm/memcpy.S b/arch/ia64/kvm/memcpy.S
new file mode 100644
index 000000000000..c04cdbe9f80f
--- /dev/null
+++ b/arch/ia64/kvm/memcpy.S
@@ -0,0 +1 @@
+#include "../lib/memcpy.S"
diff --git a/arch/ia64/kvm/memset.S b/arch/ia64/kvm/memset.S
new file mode 100644
index 000000000000..83c3066d844a
--- /dev/null
+++ b/arch/ia64/kvm/memset.S
@@ -0,0 +1 @@
+#include "../lib/memset.S"
diff --git a/arch/ia64/kvm/mmio.c b/arch/ia64/kvm/mmio.c
index 351bf70da463..7f1a858bc69f 100644
--- a/arch/ia64/kvm/mmio.c
+++ b/arch/ia64/kvm/mmio.c
@@ -159,7 +159,8 @@ static void mmio_access(struct kvm_vcpu *vcpu, u64 src_pa, u64 *dest,
 
 	if (p->u.ioreq.state == STATE_IORESP_READY) {
 		if (dir == IOREQ_READ)
-			*dest = p->u.ioreq.data;
+			/* it's necessary to ensure zero extending */
+			*dest = p->u.ioreq.data & (~0UL >> (64-(s*8)));
 	} else
 		panic_vm(vcpu);
 out:
diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c
index fc6c6636ffda..200100ea7610 100644
--- a/arch/ia64/mm/init.c
+++ b/arch/ia64/mm/init.c
@@ -719,3 +719,28 @@ out:
 EXPORT_SYMBOL_GPL(remove_memory);
 #endif /* CONFIG_MEMORY_HOTREMOVE */
 #endif
+
+/*
+ * Even when CONFIG_IA32_SUPPORT is not enabled it is
+ * useful to have the Linux/x86 domain registered to
+ * avoid an attempted module load when emulators call
+ * personality(PER_LINUX32). This saves several milliseconds
+ * on each such call.
+ */
+static struct exec_domain ia32_exec_domain;
+
+static int __init
+per_linux32_init(void)
+{
+	ia32_exec_domain.name = "Linux/x86";
+	ia32_exec_domain.handler = NULL;
+	ia32_exec_domain.pers_low = PER_LINUX32;
+	ia32_exec_domain.pers_high = PER_LINUX32;
+	ia32_exec_domain.signal_map = default_exec_domain.signal_map;
+	ia32_exec_domain.signal_invmap = default_exec_domain.signal_invmap;
+	register_exec_domain(&ia32_exec_domain);
+
+	return 0;
+}
+
+__initcall(per_linux32_init);
diff --git a/arch/ia64/uv/Makefile b/arch/ia64/uv/Makefile
new file mode 100644
index 000000000000..aa9f91947c49
--- /dev/null
+++ b/arch/ia64/uv/Makefile
@@ -0,0 +1,12 @@
+# arch/ia64/uv/Makefile
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+# Copyright (C) 2008 Silicon Graphics, Inc.  All Rights Reserved.
+#
+# Makefile for the sn uv subplatform
+#
+
+obj-y += kernel/
diff --git a/arch/ia64/uv/kernel/Makefile b/arch/ia64/uv/kernel/Makefile
new file mode 100644
index 000000000000..8d92b4684d8e
--- /dev/null
+++ b/arch/ia64/uv/kernel/Makefile
@@ -0,0 +1,13 @@
+# arch/ia64/uv/kernel/Makefile
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+# Copyright (C) 2008 Silicon Graphics, Inc.  All Rights Reserved.
+#
+
+EXTRA_CFLAGS += -Iarch/ia64/sn/include
+
+obj-y				+= setup.o
+obj-$(CONFIG_IA64_GENERIC)      += machvec.o
diff --git a/arch/ia64/uv/kernel/machvec.c b/arch/ia64/uv/kernel/machvec.c
new file mode 100644
index 000000000000..50737a9dca74
--- /dev/null
+++ b/arch/ia64/uv/kernel/machvec.c
@@ -0,0 +1,11 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (c) 2008 Silicon Graphics, Inc.  All Rights Reserved.
+ */
+
+#define MACHVEC_PLATFORM_NAME	uv
+#define MACHVEC_PLATFORM_HEADER	<asm/machvec_uv.h>
+#include <asm/machvec_init.h>
diff --git a/arch/ia64/uv/kernel/setup.c b/arch/ia64/uv/kernel/setup.c
new file mode 100644
index 000000000000..9aa743203c3c
--- /dev/null
+++ b/arch/ia64/uv/kernel/setup.c
@@ -0,0 +1,98 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * SGI UV Core Functions
+ *
+ * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved.
+ */
+
+#include <linux/module.h>
+#include <linux/percpu.h>
+#include <asm/sn/simulator.h>
+#include <asm/uv/uv_mmrs.h>
+#include <asm/uv/uv_hub.h>
+
+DEFINE_PER_CPU(struct uv_hub_info_s, __uv_hub_info);
+EXPORT_PER_CPU_SYMBOL_GPL(__uv_hub_info);
+
+
+struct redir_addr {
+	unsigned long redirect;
+	unsigned long alias;
+};
+
+#define DEST_SHIFT UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_SHFT
+
+static __initdata struct redir_addr redir_addrs[] = {
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR, UVH_SI_ALIAS0_OVERLAY_CONFIG},
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR, UVH_SI_ALIAS1_OVERLAY_CONFIG},
+	{UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR, UVH_SI_ALIAS2_OVERLAY_CONFIG},
+};
+
+static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size)
+{
+	union uvh_si_alias0_overlay_config_u alias;
+	union uvh_rh_gam_alias210_redirect_config_2_mmr_u redirect;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(redir_addrs); i++) {
+		alias.v = uv_read_local_mmr(redir_addrs[i].alias);
+		if (alias.s.base == 0) {
+			*size = (1UL << alias.s.m_alias);
+			redirect.v = uv_read_local_mmr(redir_addrs[i].redirect);
+			*base = (unsigned long)redirect.s.dest_base << DEST_SHIFT;
+			return;
+		}
+	}
+	BUG();
+}
+
+void __init uv_setup(char **cmdline_p)
+{
+	union uvh_si_addr_map_config_u m_n_config;
+	union uvh_node_id_u node_id;
+	unsigned long gnode_upper;
+	int nid, cpu, m_val, n_val;
+	unsigned long mmr_base, lowmem_redir_base, lowmem_redir_size;
+
+	if (IS_MEDUSA()) {
+		lowmem_redir_base = 0;
+		lowmem_redir_size = 0;
+		node_id.v = 0;
+		m_n_config.s.m_skt = 37;
+		m_n_config.s.n_skt = 0;
+		mmr_base = 0;
+	} else {
+		get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size);
+		node_id.v = uv_read_local_mmr(UVH_NODE_ID);
+		m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG);
+		mmr_base =
+			uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) &
+				~UV_MMR_ENABLE;
+	}
+
+	m_val = m_n_config.s.m_skt;
+	n_val = m_n_config.s.n_skt;
+	printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base);
+
+	gnode_upper = (((unsigned long)node_id.s.node_id) &
+		       ~((1 << n_val) - 1)) << m_val;
+
+	for_each_present_cpu(cpu) {
+		nid = cpu_to_node(cpu);
+		uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base;
+		uv_cpu_hub_info(cpu)->lowmem_remap_top =
+			lowmem_redir_base + lowmem_redir_size;
+		uv_cpu_hub_info(cpu)->m_val = m_val;
+		uv_cpu_hub_info(cpu)->n_val = m_val;
+		uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) -1;
+		uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1;
+		uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper;
+		uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base;
+		uv_cpu_hub_info(cpu)->coherency_domain_number = 0;/* ZZZ */
+		printk(KERN_DEBUG "UV cpu %d, nid %d\n", cpu, nid);
+	}
+}
+