Merge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] fix personality(PER_LINUX32) performance issue
  [IA64] Properly unregister legacy interrupts
  [IA64] Remove NULL pointer check for argument never passed as NULL.
  [IA64] trivial cleanup for perfmon.c
  [IA64] trivial cleanup for entry.S
  [IA64] fix interrupt masking for pending works on kernel leave
  [IA64] allow user to force_pal_cache_flush
  [IA64] Don't reserve crashkernel memory > 4 GB
  [IA64] machvec support for SGI UV platform
  [IA64] Add header files for SGI UV platform

Documentation/kernel-parameters.txt

@@ -686,6 +686,12 @@ and is between 256 and 4096 characters. It is defined in the file
 	floppy=		[HW]
 			See Documentation/floppy.txt.
 
+	force_pal_cache_flush
+			[IA-64] Avoid check_sal_cache_flush which may hang on
+			buggy SAL_CACHE_FLUSH implementations. Using this
+			parameter will force ia64_sal_cache_flush to call
+			ia64_pal_cache_flush instead of SAL_CACHE_FLUSH.
+
 	gamecon.map[2|3]=
 			[HW,JOY] Multisystem joystick and NES/SNES/PSX pad
 			support via parallel port (up to 5 devices per port)

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

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

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;

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
@@ -622,6 +627,9 @@ 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);
 }
 

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)

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);
 }
 

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 @@ pfm_handle_work(void)
 	/*
 	 * 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);

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;

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);
 

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",

arch/ia64/mm/init.c

@@ -719,3 +719,28 @@ int remove_memory(u64 start, u64 size)
 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);

arch/ia64/uv/Makefile

+# 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/

arch/ia64/uv/kernel/Makefile

+# 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

arch/ia64/uv/kernel/machvec.c

+/*
+ * 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>

arch/ia64/uv/kernel/setup.c

+/*
+ * 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);
+	}
+}
+

include/asm-ia64/machvec.h

@@ -126,6 +126,8 @@ extern void machvec_tlb_migrate_finish (struct mm_struct *);
 #  include <asm/machvec_hpzx1_swiotlb.h>
 # elif defined (CONFIG_IA64_SGI_SN2)
 #  include <asm/machvec_sn2.h>
+# elif defined (CONFIG_IA64_SGI_UV)
+#  include <asm/machvec_uv.h>
 # elif defined (CONFIG_IA64_GENERIC)
 
 # ifdef MACHVEC_PLATFORM_HEADER

include/asm-ia64/machvec_uv.h

+/*
+ * 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.
+ */
+
+#ifndef _ASM_IA64_MACHVEC_UV_H
+#define _ASM_IA64_MACHVEC_UV_H
+
+extern ia64_mv_setup_t uv_setup;
+
+/*
+ * This stuff has dual use!
+ *
+ * For a generic kernel, the macros are used to initialize the
+ * platform's machvec structure.  When compiling a non-generic kernel,
+ * the macros are used directly.
+ */
+#define platform_name			"uv"
+#define platform_setup			uv_setup
+
+#endif /* _ASM_IA64_MACHVEC_UV_H */

include/asm-ia64/uv/uv_hub.h

+/*
+ * 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 architectural definitions
+ *
+ * Copyright (C) 2008 Silicon Graphics, Inc. All rights reserved.
+ */
+
+#ifndef __ASM_IA64_UV_HUB_H__
+#define __ASM_IA64_UV_HUB_H__
+
+#include <linux/numa.h>
+#include <linux/percpu.h>
+#include <asm/types.h>
+#include <asm/percpu.h>
+
+
+/*
+ * Addressing Terminology
+ *
+ *	M       - The low M bits of a physical address represent the offset
+ *		  into the blade local memory. RAM memory on a blade is physically
+ *		  contiguous (although various IO spaces may punch holes in
+ *		  it)..
+ *
+ * 	N	- Number of bits in the node portion of a socket physical
+ * 		  address.
+ *
+ * 	NASID   - network ID of a router, Mbrick or Cbrick. Nasid values of
+ * 	 	  routers always have low bit of 1, C/MBricks have low bit
+ * 		  equal to 0. Most addressing macros that target UV hub chips
+ * 		  right shift the NASID by 1 to exclude the always-zero bit.
+ * 		  NASIDs contain up to 15 bits.
+ *
+ *	GNODE   - NASID right shifted by 1 bit. Most mmrs contain gnodes instead
+ *		  of nasids.
+ *
+ * 	PNODE   - the low N bits of the GNODE. The PNODE is the most useful variant
+ * 		  of the nasid for socket usage.
+ *
+ *
+ *  NumaLink Global Physical Address Format:
+ *  +--------------------------------+---------------------+
+ *  |00..000|      GNODE             |      NodeOffset     |
+ *  +--------------------------------+---------------------+
+ *          |<-------53 - M bits --->|<--------M bits ----->
+ *
+ *	M - number of node offset bits (35 .. 40)
+ *
+ *
+ *  Memory/UV-HUB Processor Socket Address Format:
+ *  +----------------+---------------+---------------------+
+ *  |00..000000000000|   PNODE       |      NodeOffset     |
+ *  +----------------+---------------+---------------------+
+ *                   <--- N bits --->|<--------M bits ----->
+ *
+ *	M - number of node offset bits (35 .. 40)
+ *	N - number of PNODE bits (0 .. 10)
+ *
+ *		Note: M + N cannot currently exceed 44 (x86_64) or 46 (IA64).
+ *		The actual values are configuration dependent and are set at
+ *		boot time. M & N values are set by the hardware/BIOS at boot.
+ */
+
+
+/*
+ * Maximum number of bricks in all partitions and in all coherency domains.
+ * This is the total number of bricks accessible in the numalink fabric. It
+ * includes all C & M bricks. Routers are NOT included.
+ *
+ * This value is also the value of the maximum number of non-router NASIDs
+ * in the numalink fabric.
+ *
+ * NOTE: a brick may contain 1 or 2 OS nodes. Don't get these confused.
+ */
+#define UV_MAX_NUMALINK_BLADES	16384
+
+/*
+ * Maximum number of C/Mbricks within a software SSI (hardware may support
+ * more).
+ */
+#define UV_MAX_SSI_BLADES	1
+
+/*
+ * The largest possible NASID of a C or M brick (+ 2)
+ */
+#define UV_MAX_NASID_VALUE	(UV_MAX_NUMALINK_NODES * 2)
+
+/*
+ * The following defines attributes of the HUB chip. These attributes are
+ * frequently referenced and are kept in the per-cpu data areas of each cpu.
+ * They are kept together in a struct to minimize cache misses.
+ */
+struct uv_hub_info_s {
+	unsigned long	global_mmr_base;
+	unsigned long	gpa_mask;
+	unsigned long	gnode_upper;
+	unsigned long	lowmem_remap_top;
+	unsigned long	lowmem_remap_base;
+	unsigned short	pnode;
+	unsigned short	pnode_mask;
+	unsigned short	coherency_domain_number;
+	unsigned short	numa_blade_id;
+	unsigned char	blade_processor_id;
+	unsigned char	m_val;
+	unsigned char	n_val;
+};
+DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info);
+#define uv_hub_info 		(&__get_cpu_var(__uv_hub_info))
+#define uv_cpu_hub_info(cpu)	(&per_cpu(__uv_hub_info, cpu))
+
+/*
+ * Local & Global MMR space macros.
+ * 	Note: macros are intended to be used ONLY by inline functions
+ * 	in this file - not by other kernel code.
+ * 		n -  NASID (full 15-bit global nasid)
+ * 		g -  GNODE (full 15-bit global nasid, right shifted 1)
+ * 		p -  PNODE (local part of nsids, right shifted 1)
+ */
+#define UV_NASID_TO_PNODE(n)		(((n) >> 1) & uv_hub_info->pnode_mask)
+#define UV_PNODE_TO_NASID(p)		(((p) << 1) | uv_hub_info->gnode_upper)
+
+#define UV_LOCAL_MMR_BASE		0xf4000000UL
+#define UV_GLOBAL_MMR32_BASE		0xf8000000UL
+#define UV_GLOBAL_MMR64_BASE		(uv_hub_info->global_mmr_base)
+
+#define UV_GLOBAL_MMR32_PNODE_SHIFT	15
+#define UV_GLOBAL_MMR64_PNODE_SHIFT	26
+
+#define UV_GLOBAL_MMR32_PNODE_BITS(p)	((p) << (UV_GLOBAL_MMR32_PNODE_SHIFT))
+
+#define UV_GLOBAL_MMR64_PNODE_BITS(p)					\
+	((unsigned long)(p) << UV_GLOBAL_MMR64_PNODE_SHIFT)
+
+/*
+ * Macros for converting between kernel virtual addresses, socket local physical
+ * addresses, and UV global physical addresses.
+ * 	Note: use the standard __pa() & __va() macros for converting
+ * 	      between socket virtual and socket physical addresses.
+ */
+
+/* socket phys RAM --> UV global physical address */
+static inline unsigned long uv_soc_phys_ram_to_gpa(unsigned long paddr)
+{
+	if (paddr < uv_hub_info->lowmem_remap_top)
+		paddr += uv_hub_info->lowmem_remap_base;
+	return paddr | uv_hub_info->gnode_upper;
+}
+
+
+/* socket virtual --> UV global physical address */
+static inline unsigned long uv_gpa(void *v)
+{
+	return __pa(v) | uv_hub_info->gnode_upper;
+}
+
+/* socket virtual --> UV global physical address */
+static inline void *uv_vgpa(void *v)
+{
+	return (void *)uv_gpa(v);
+}
+
+/* UV global physical address --> socket virtual */
+static inline void *uv_va(unsigned long gpa)
+{
+	return __va(gpa & uv_hub_info->gpa_mask);
+}
+
+/* pnode, offset --> socket virtual */
+static inline void *uv_pnode_offset_to_vaddr(int pnode, unsigned long offset)
+{
+	return __va(((unsigned long)pnode << uv_hub_info->m_val) | offset);
+}
+
+
+/*
+ * Access global MMRs using the low memory MMR32 space. This region supports
+ * faster MMR access but not all MMRs are accessible in this space.
+ */
+static inline unsigned long *uv_global_mmr32_address(int pnode,
+				unsigned long offset)
+{
+	return __va(UV_GLOBAL_MMR32_BASE |
+		       UV_GLOBAL_MMR32_PNODE_BITS(pnode) | offset);
+}
+
+static inline void uv_write_global_mmr32(int pnode, unsigned long offset,
+				 unsigned long val)
+{
+	*uv_global_mmr32_address(pnode, offset) = val;
+}
+
+static inline unsigned long uv_read_global_mmr32(int pnode,
+						 unsigned long offset)
+{
+	return *uv_global_mmr32_address(pnode, offset);
+}
+
+/*
+ * Access Global MMR space using the MMR space located at the top of physical
+ * memory.
+ */
+static inline unsigned long *uv_global_mmr64_address(int pnode,
+				unsigned long offset)
+{
+	return __va(UV_GLOBAL_MMR64_BASE |
+		    UV_GLOBAL_MMR64_PNODE_BITS(pnode) | offset);
+}
+
+static inline void uv_write_global_mmr64(int pnode, unsigned long offset,
+				unsigned long val)
+{
+	*uv_global_mmr64_address(pnode, offset) = val;
+}
+
+static inline unsigned long uv_read_global_mmr64(int pnode,
+						 unsigned long offset)
+{
+	return *uv_global_mmr64_address(pnode, offset);
+}
+
+/*
+ * Access hub local MMRs. Faster than using global space but only local MMRs
+ * are accessible.
+ */
+static inline unsigned long *uv_local_mmr_address(unsigned long offset)
+{
+	return __va(UV_LOCAL_MMR_BASE | offset);
+}
+
+static inline unsigned long uv_read_local_mmr(unsigned long offset)
+{
+	return *uv_local_mmr_address(offset);
+}
+
+static inline void uv_write_local_mmr(unsigned long offset, unsigned long val)
+{
+	*uv_local_mmr_address(offset) = val;
+}
+
+/*
+ * Structures and definitions for converting between cpu, node, pnode, and blade
+ * numbers.
+ */
+
+/* Blade-local cpu number of current cpu. Numbered 0 .. <# cpus on the blade> */
+static inline int uv_blade_processor_id(void)
+{
+	return smp_processor_id();
+}
+
+/* Blade number of current cpu. Numnbered 0 .. <#blades -1> */
+static inline int uv_numa_blade_id(void)
+{
+	return 0;
+}
+
+/* Convert a cpu number to the the UV blade number */
+static inline int uv_cpu_to_blade_id(int cpu)
+{
+	return 0;
+}
+
+/* Convert linux node number to the UV blade number */
+static inline int uv_node_to_blade_id(int nid)
+{
+	return 0;
+}
+
+/* Convert a blade id to the PNODE of the blade */
+static inline int uv_blade_to_pnode(int bid)
+{
+	return 0;
+}
+
+/* Determine the number of possible cpus on a blade */
+static inline int uv_blade_nr_possible_cpus(int bid)
+{
+	return num_possible_cpus();
+}
+
+/* Determine the number of online cpus on a blade */
+static inline int uv_blade_nr_online_cpus(int bid)
+{
+	return num_online_cpus();
+}
+
+/* Convert a cpu id to the PNODE of the blade containing the cpu */
+static inline int uv_cpu_to_pnode(int cpu)
+{
+	return 0;
+}
+
+/* Convert a linux node number to the PNODE of the blade */
+static inline int uv_node_to_pnode(int nid)
+{
+	return 0;
+}
+
+/* Maximum possible number of blades */
+static inline int uv_num_possible_blades(void)
+{
+	return 1;
+}
+
+#endif /* __ASM_IA64_UV_HUB__ */
+