Merge http://lia64.bkbits.net/to-linus-2.5

into ppc970.osdl.org:/home/torvalds/v2.6/linux

arch/ia64/Kconfig

@@ -388,7 +388,7 @@ config IA64_GRANULE_16MB
 
 config IA64_GRANULE_64MB
 	bool "64MB"
-	depends on !(IA64_GENERIC || IA64_HP_ZX1)
+	depends on !(IA64_GENERIC || IA64_HP_ZX1 || IA64_SGI_SN2)
 
 endchoice
 

arch/ia64/Makefile

@@ -43,7 +43,8 @@ endif
 
 ifeq ($(GCC_VERSION),3)
  ifeq ($(GCC_MINOR_VERSION),4)
-	cflags-$(CONFIG_ITANIUM)	+= -mtune=merced
+# Workaround Itanium 1 bugs in gcc 3.4.
+#	cflags-$(CONFIG_ITANIUM)	+= -mtune=merced
 	cflags-$(CONFIG_MCKINLEY)	+= -mtune=mckinley
  endif
 endif

arch/ia64/kernel/efi_stub.S

@@ -44,7 +44,7 @@
 
 GLOBAL_ENTRY(efi_call_phys)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8)
-	alloc loc1=ar.pfs,8,5,7,0
+	alloc loc1=ar.pfs,8,7,7,0
 	ld8 r2=[in0],8			// load EFI function's entry point
 	mov loc0=rp
 	.body
@@ -70,9 +70,13 @@ GLOBAL_ENTRY(efi_call_phys)
 	mov out3=in4
 	mov out5=in6
 	mov out6=in7
+	mov loc5=r19
+	mov loc6=r20
 	br.call.sptk.many rp=b6		// call the EFI function
 .ret1:	mov ar.rsc=0			// put RSE in enforced lazy, LE mode
 	mov r16=loc3
+	mov r19=loc5
+	mov r20=loc6
 	br.call.sptk.many rp=ia64_switch_mode_virt // return to virtual mode
 .ret2:	mov ar.rsc=loc4			// restore RSE configuration
 	mov ar.pfs=loc1

arch/ia64/kernel/entry.S

@@ -182,7 +182,7 @@ GLOBAL_ENTRY(ia64_switch_to)
 	movl r25=init_task
 	mov r27=IA64_KR(CURRENT_STACK)
 	adds r21=IA64_TASK_THREAD_KSP_OFFSET,in0
-	dep r20=0,in0,61,3		// physical address of "current"
+	dep r20=0,in0,61,3		// physical address of "next"
 	;;
 	st8 [r22]=sp			// save kernel stack pointer of old task
 	shr.u r26=r20,IA64_GRANULE_SHIFT
@@ -195,7 +195,7 @@ GLOBAL_ENTRY(ia64_switch_to)
 (p6)	br.cond.dpnt .map
 	;;
 .done:
-(p6)	ssm psr.ic			// if we we had to map, renable the psr.ic bit FIRST!!!
+(p6)	ssm psr.ic			// if we had to map, reenable the psr.ic bit FIRST!!!
 	;;
 (p6)	srlz.d
 	ld8 sp=[r21]			// load kernel stack pointer of new task

arch/ia64/kernel/head.S

@@ -67,7 +67,7 @@ start_ap:
 	 * Initialize kernel region registers:
 	 *	rr[5]: VHPT enabled, page size = PAGE_SHIFT
 	 *	rr[6]: VHPT disabled, page size = IA64_GRANULE_SHIFT
-	 *	rr[5]: VHPT disabled, page size = IA64_GRANULE_SHIFT
+	 *	rr[7]: VHPT disabled, page size = IA64_GRANULE_SHIFT
 	 */
 	mov r16=((ia64_rid(IA64_REGION_ID_KERNEL, (5<<61)) << 8) | (PAGE_SHIFT << 2) | 1)
 	movl r17=(5<<61)
@@ -154,8 +154,7 @@ start_ap:
 #endif
 	;;
 	tpa r3=r2		// r3 == phys addr of task struct
-	;;
-	shr.u r16=r3,IA64_GRANULE_SHIFT
+	mov r16=-1
 (isBP)	br.cond.dpnt .load_current // BP stack is on region 5 --- no need to map it
 
 	// load mapping for stack (virtaddr in r2, physaddr in r3)
@@ -169,6 +168,7 @@ start_ap:
 	dep r2=-1,r3,61,3	// IMVA of task
 	;;
 	mov r17=rr[r2]
+	shr.u r16=r3,IA64_GRANULE_SHIFT
 	;;
 	dep r17=0,r17,8,24
 	;;
@@ -706,6 +706,9 @@ END(__ia64_init_fpu)
  *
  * Inputs:
  *	r16 = new psr to establish
+ * Output:
+ *	r19 = old virtual address of ar.bsp
+ *	r20 = old virtual address of sp
  *
  * Note: RSE must already be in enforced lazy mode
  */
@@ -724,12 +727,13 @@ GLOBAL_ENTRY(ia64_switch_mode_phys)
 	mov cr.ipsr=r16			// set new PSR
 	add r3=1f-ia64_switch_mode_phys,r15
 
-	mov r17=ar.bsp
+	mov r19=ar.bsp
+	mov r20=sp
 	mov r14=rp			// get return address into a general register
 	;;
 
 	// going to physical mode, use tpa to translate virt->phys
-	tpa r17=r17
+	tpa r17=r19
 	tpa r3=r3
 	tpa sp=sp
 	tpa r14=r14
@@ -752,6 +756,8 @@ END(ia64_switch_mode_phys)
  *
  * Inputs:
  *	r16 = new psr to establish
+ *	r19 = new bspstore to establish
+ *	r20 = new sp to establish
  *
  * Note: RSE must already be in enforced lazy mode
  */
@@ -770,7 +776,6 @@ GLOBAL_ENTRY(ia64_switch_mode_virt)
 	mov cr.ipsr=r16			// set new PSR
 	add r3=1f-ia64_switch_mode_virt,r15
 
-	mov r17=ar.bsp
 	mov r14=rp			// get return address into a general register
 	;;
 
@@ -781,15 +786,14 @@ GLOBAL_ENTRY(ia64_switch_mode_virt)
 	movl r18=KERNEL_START
 	dep r3=0,r3,KERNEL_TR_PAGE_SHIFT,64-KERNEL_TR_PAGE_SHIFT
 	dep r14=0,r14,KERNEL_TR_PAGE_SHIFT,64-KERNEL_TR_PAGE_SHIFT
-	dep r17=-1,r17,61,3
-	dep sp=-1,sp,61,3
+	mov sp=r20
 	;;
 	or r3=r3,r18
 	or r14=r14,r18
 	;;
 
 	mov r18=ar.rnat			// save ar.rnat
-	mov ar.bspstore=r17		// this steps on ar.rnat
+	mov ar.bspstore=r19		// this steps on ar.rnat
 	mov cr.iip=r3
 	mov cr.ifs=r0
 	;;

arch/ia64/kernel/iosapic.c

@@ -217,10 +217,8 @@ set_rte (unsigned int vector, unsigned int dest, int mask)
 
 	spin_lock_irqsave(&iosapic_lock, flags);
 	{
-		writel(IOSAPIC_RTE_HIGH(rte_index), addr + IOSAPIC_REG_SELECT);
-		writel(high32, addr + IOSAPIC_WINDOW);
-		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-		writel(low32, addr + IOSAPIC_WINDOW);
+		iosapic_write(addr, IOSAPIC_RTE_HIGH(rte_index), high32);
+		iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32);
 		iosapic_intr_info[vector].low32 = low32;
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
@@ -249,12 +247,9 @@ mask_irq (unsigned int irq)
 
 	spin_lock_irqsave(&iosapic_lock, flags);
 	{
-		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-
 		/* set only the mask bit */
 		low32 = iosapic_intr_info[vec].low32 |= IOSAPIC_MASK;
-
-		writel(low32, addr + IOSAPIC_WINDOW);
+		iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32);
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
 }
@@ -275,9 +270,8 @@ unmask_irq (unsigned int irq)
 
 	spin_lock_irqsave(&iosapic_lock, flags);
 	{
-		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
 		low32 = iosapic_intr_info[vec].low32 &= ~IOSAPIC_MASK;
-		writel(low32, addr + IOSAPIC_WINDOW);
+		iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32);
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
 }
@@ -325,10 +319,8 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask)
 			low32 |= (IOSAPIC_FIXED << IOSAPIC_DELIVERY_SHIFT);
 
 		iosapic_intr_info[vec].low32 = low32;
-		writel(IOSAPIC_RTE_HIGH(rte_index), addr + IOSAPIC_REG_SELECT);
-		writel(high32, addr + IOSAPIC_WINDOW);
-		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-		writel(low32, addr + IOSAPIC_WINDOW);
+		iosapic_write(addr, IOSAPIC_RTE_HIGH(rte_index), high32);
+		iosapic_write(addr, IOSAPIC_RTE_LOW(rte_index), low32);
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
 #endif
@@ -351,7 +343,7 @@ iosapic_end_level_irq (unsigned int irq)
 	ia64_vector vec = irq_to_vector(irq);
 
 	move_irq(irq);
-	writel(vec, iosapic_intr_info[vec].addr + IOSAPIC_EOI);
+	iosapic_eoi(iosapic_intr_info[vec].addr, vec);
 }
 
 #define iosapic_shutdown_level_irq	mask_irq
@@ -428,8 +420,7 @@ iosapic_version (char *addr)
 	 *	unsigned int reserved2 : 8;
 	 * }
 	 */
-	writel(IOSAPIC_VERSION, addr + IOSAPIC_REG_SELECT);
-	return readl(IOSAPIC_WINDOW + addr);
+	return iosapic_read(addr, IOSAPIC_VERSION);
 }
 
 /*

arch/ia64/kernel/mca.c

@@ -256,7 +256,7 @@ ia64_mca_log_sal_error_record(int sal_info_type)
 	salinfo_log_wakeup(sal_info_type, buffer, size, irq_safe);
 
 	if (irq_safe)
-		printk(KERN_INFO "CPU %d: SAL log contains %s error record\n",
+		IA64_MCA_DEBUG("CPU %d: SAL log contains %s error record\n",
 			smp_processor_id(),
 			sal_info_type < ARRAY_SIZE(rec_name) ? rec_name[sal_info_type] : "UNKNOWN");
 

arch/ia64/kernel/module.c

@@ -656,8 +656,26 @@ do_reloc (struct module *mod, uint8_t r_type, Elf64_Sym *sym, uint64_t addend,
 	      case RV_PCREL:
 		switch (r_type) {
 		      case R_IA64_PCREL21B:
-			/* special because it can cross into other module/kernel-core.  */
-			if (!is_internal(mod, val))
+			if (in_init(mod, val)) {
+				/* Calls to init code from core are bad news */
+				if (in_core(mod, (uint64_t)location)) {
+					printk(KERN_ERR "%s: init symbol 0x%lx used in module code at %p\n",
+						mod->name, val, location);
+					return -ENOEXEC;
+				}
+			} else if (in_core(mod, val)) {
+				/*
+				 * Init section may have been allocated far away from core,
+				 * if the branch won't reach, then allocate a plt for it.
+				 */
+				if (in_init(mod, (uint64_t)location)) {
+					uint64_t delta = ((int64_t)val - (int64_t)location) / 16;
+					if (delta + (1 << 20) >= (1 << 21)) {
+						val = get_fdesc(mod, val, &ok);
+						val = get_plt(mod, location, val, &ok);
+					}
+				}
+			} else
 				val = get_plt(mod, location, val, &ok);
 			/* FALL THROUGH */
 		      default:

arch/ia64/kernel/pal.S

@@ -55,7 +55,7 @@ END(ia64_pal_default_handler)
  */
 GLOBAL_ENTRY(ia64_pal_call_static)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(6)
-	alloc loc1 = ar.pfs,6,90,0,0
+	alloc loc1 = ar.pfs,5,5,0,0
 	movl loc2 = pal_entry_point
 1:	{
 	  mov r28 = in0
@@ -66,7 +66,9 @@ GLOBAL_ENTRY(ia64_pal_call_static)
 	ld8 loc2 = [loc2]		// loc2 <- entry point
 	tbit.nz p6,p7 = in4, 0
 	adds r8 = 1f-1b,r8
+	mov loc4=ar.rsc			// save RSE configuration
 	;;
+	mov ar.rsc=0			// put RSE in enforced lazy, LE mode
 	mov loc3 = psr
 	mov loc0 = rp
 	.body
@@ -82,6 +84,7 @@ GLOBAL_ENTRY(ia64_pal_call_static)
 	mov rp = r8
 	br.cond.sptk.many b7
 1:	mov psr.l = loc3
+	mov ar.rsc = loc4		// restore RSE configuration
 	mov ar.pfs = loc1
 	mov rp = loc0
 	;;
@@ -98,7 +101,7 @@ END(ia64_pal_call_static)
  */
 GLOBAL_ENTRY(ia64_pal_call_stacked)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
-	alloc loc1 = ar.pfs,5,4,87,0
+	alloc loc1 = ar.pfs,4,4,4,0
 	movl loc2 = pal_entry_point
 
 	mov r28  = in0			// Index MUST be copied to r28
@@ -145,7 +148,7 @@ END(ia64_pal_call_stacked)
 
 GLOBAL_ENTRY(ia64_pal_call_phys_static)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(6)
-	alloc loc1 = ar.pfs,6,90,0,0
+	alloc loc1 = ar.pfs,4,7,0,0
 	movl loc2 = pal_entry_point
 1:	{
 	  mov r28  = in0		// copy procedure index
@@ -176,10 +179,14 @@ GLOBAL_ENTRY(ia64_pal_call_phys_static)
 	andcm r16=loc3,r16		// removes bits to clear from psr
 	br.call.sptk.many rp=ia64_switch_mode_phys
 .ret1:	mov rp = r8			// install return address (physical)
+	mov loc5 = r19
+	mov loc6 = r20
 	br.cond.sptk.many b7
 1:
 	mov ar.rsc=0			// put RSE in enforced lazy, LE mode
 	mov r16=loc3			// r16= original psr
+	mov r19=loc5
+	mov r20=loc6
 	br.call.sptk.many rp=ia64_switch_mode_virt // return to virtual mode
 .ret2:
 	mov psr.l = loc3		// restore init PSR
@@ -201,7 +208,7 @@ END(ia64_pal_call_phys_static)
  */
 GLOBAL_ENTRY(ia64_pal_call_phys_stacked)
 	.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(5)
-	alloc	loc1 = ar.pfs,5,5,86,0
+	alloc	loc1 = ar.pfs,5,7,4,0
 	movl	loc2 = pal_entry_point
 1:	{
 	  mov r28  = in0		// copy procedure index
@@ -230,10 +237,14 @@ GLOBAL_ENTRY(ia64_pal_call_phys_stacked)
 	andcm r16=loc3,r16		// removes bits to clear from psr
 	br.call.sptk.many rp=ia64_switch_mode_phys
 .ret6:
+	mov loc5 = r19
+	mov loc6 = r20
 	br.call.sptk.many rp=b7		// now make the call
 .ret7:
 	mov ar.rsc=0			// put RSE in enforced lazy, LE mode
 	mov r16=loc3			// r16= original psr
+	mov r19=loc5
+	mov r20=loc6
 	br.call.sptk.many rp=ia64_switch_mode_virt	// return to virtual mode
 
 .ret8:	mov psr.l  = loc3		// restore init PSR

arch/ia64/kernel/setup.c

@@ -277,6 +277,29 @@ setup_serial_legacy (void)
 }
 #endif
 
+/**
+ * early_console_setup - setup debugging console
+ *
+ * Consoles started here require little enough setup that we can start using
+ * them very early in the boot process, either right after the machine
+ * vector initialization, or even before if the drivers can detect their hw.
+ *
+ * Returns non-zero if a console couldn't be setup.
+ */
+static inline int __init
+early_console_setup (void)
+{
+#ifdef CONFIG_SERIAL_SGI_L1_CONSOLE
+	{
+		extern int sn_serial_console_early_setup(void);
+		if(!sn_serial_console_early_setup())
+			return 0;
+	}
+#endif
+
+	return -1;
+}
+
 void __init
 setup_arch (char **cmdline_p)
 {
@@ -294,6 +317,12 @@ setup_arch (char **cmdline_p)
 	machvec_init(acpi_get_sysname());
 #endif
 
+#ifdef CONFIG_SMP
+	/* If we register an early console, allow CPU 0 to printk */
+	if (!early_console_setup())
+		cpu_set(smp_processor_id(), cpu_online_map);
+#endif
+
 #ifdef CONFIG_ACPI_BOOT
 	/* Initialize the ACPI boot-time table parser */
 	acpi_table_init();

arch/ia64/mm/discontig.c

@@ -549,7 +549,7 @@ void call_pernode_memory(unsigned long start, unsigned long len, void *arg)
 	if (!num_node_memblks) {
 		/* No SRAT table, so assume one node (node 0) */
 		if (start < end)
-			(*func)(start, len, 0);
+			(*func)(start, end - start, 0);
 		return;
 	}
 

arch/ia64/sn/io/sn2/bte_error.c

@@ -3,7 +3,7 @@
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
- * Copyright (c) 2000-2003 Silicon Graphics, Inc.  All Rights Reserved.
+ * Copyright (c) 2000-2004 Silicon Graphics, Inc.  All Rights Reserved.
  */
 
 
@@ -199,7 +199,7 @@ bte_error_handler(unsigned long _nodepda)
 		err_nodepda->bte_if[i].cleanup_active = 0;
 		BTE_PRINTK(("eh:%p:%d Unlocked %d\n", err_nodepda,
 			    smp_processor_id(), i));
-		spin_unlock(&pda->cpu_bte_if[i]->spinlock);
+		spin_unlock(&err_nodepda->bte_if[i].spinlock);
 	}
 
 	del_timer(recovery_timer);

arch/ia64/sn/kernel/bte.c

@@ -7,6 +7,7 @@
  */
 
 #include <linux/config.h>
+#include <linux/module.h>
 #include <asm/sn/sgi.h>
 #include <asm/sn/nodepda.h>
 #include <asm/sn/addrs.h>
@@ -27,10 +28,18 @@
 #define L1_CACHE_MASK (L1_CACHE_BYTES - 1)
 #endif
 
-/*
- * The base address of for each set of bte registers.
- */
-static int bte_offsets[] = { IIO_IBLS0, IIO_IBLS1 };
+/* two interfaces on two btes */
+#define MAX_INTERFACES_TO_TRY		4
+
+static struct bteinfo_s *
+bte_if_on_node(nasid_t nasid, int interface)
+{
+	nodepda_t *tmp_nodepda;
+
+	tmp_nodepda = NODEPDA(nasid_to_cnodeid(nasid));
+	return &tmp_nodepda->bte_if[interface];
+
+}
 
 
 /************************************************************************
@@ -61,11 +70,12 @@ static int bte_offsets[] = { IIO_IBLS0, IIO_IBLS1 };
 bte_result_t
 bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 {
-	int bte_to_use;
 	u64 transfer_size;
 	struct bteinfo_s *bte;
 	bte_result_t bte_status;
 	unsigned long irq_flags;
+	struct bteinfo_s *btes_to_try[MAX_INTERFACES_TO_TRY];
+	int bte_if_index;
 
 
 	BTE_PRINTK(("bte_copy(0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%p)\n",
@@ -79,17 +89,57 @@ bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 		 (src & L1_CACHE_MASK) || (dest & L1_CACHE_MASK)));
 	ASSERT(len < ((BTE_LEN_MASK + 1) << L1_CACHE_SHIFT));
 
+	if (mode & BTE_USE_DEST) {
+		/* try remote then local */
+		btes_to_try[0] = bte_if_on_node(NASID_GET(dest), 0);
+		btes_to_try[1] = bte_if_on_node(NASID_GET(dest), 1);
+		if (mode & BTE_USE_ANY) {
+			btes_to_try[2] = bte_if_on_node(get_nasid(), 0);
+			btes_to_try[3] = bte_if_on_node(get_nasid(), 1);
+		} else {
+			btes_to_try[2] = NULL;
+			btes_to_try[3] = NULL;
+		}
+	} else {
+		/* try local then remote */
+		btes_to_try[0] = bte_if_on_node(get_nasid(), 0);
+		btes_to_try[1] = bte_if_on_node(get_nasid(), 1);
+		if (mode & BTE_USE_ANY) {
+			btes_to_try[2] = bte_if_on_node(NASID_GET(dest), 0);
+			btes_to_try[3] = bte_if_on_node(NASID_GET(dest), 1);
+		} else {
+			btes_to_try[2] = NULL;
+			btes_to_try[3] = NULL;
+		}
+	}
+
 	do {
 		local_irq_save(irq_flags);
 
-		bte_to_use = 0;
+		bte_if_index = 0;
+
 		/* Attempt to lock one of the BTE interfaces. */
-		while ((bte_to_use < BTES_PER_NODE) &&
-		       BTE_LOCK_IF_AVAIL(bte_to_use)) {
-			bte_to_use++;
+		while (bte_if_index < MAX_INTERFACES_TO_TRY) {
+			bte = btes_to_try[bte_if_index++];
+
+			if (bte == NULL) {
+				continue;
 			}
 
-		if (bte_to_use < BTES_PER_NODE) {
+			if (spin_trylock(&bte->spinlock)) {
+				if ((*bte->most_rcnt_na & BTE_ACTIVE) ||
+				    (BTE_LNSTAT_LOAD(bte) & BTE_ACTIVE)) {
+					/* Got the lock but BTE still busy */
+					spin_unlock(&bte->spinlock);
+					bte = NULL;
+				} else {
+					/* we got the lock and it's not busy */
+					break;
+				}
+			}
+		}
+
+		if (bte != NULL) {
 			break;
 		}
 
@@ -100,12 +150,9 @@ bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 		}
 
 		/* Wait until a bte is available. */
-		udelay(10);
+		udelay(1);
 	} while (1);
 
-	bte = pda->cpu_bte_if[bte_to_use];
-	BTE_PRINTKV(("Got a lock on bte %d\n", bte_to_use));
-
 
 	if (notification == NULL) {
 		/* User does not want to be notified. */
@@ -121,28 +168,24 @@ bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 	*bte->most_rcnt_na = -1L;
 
 	/* Set the status reg busy bit and transfer length */
-	BTE_PRINTKV(("IBLS - HUB_S(0x%p, 0x%lx)\n",
-		     BTEREG_LNSTAT_ADDR, IBLS_BUSY | transfer_size));
-	HUB_S(BTEREG_LNSTAT_ADDR, (IBLS_BUSY | transfer_size));
+	BTE_PRINTKV(("IBLS = 0x%lx\n", IBLS_BUSY | transfer_size));
+	BTE_LNSTAT_STORE(bte, IBLS_BUSY | transfer_size);
 
 	/* Set the source and destination registers */
-	BTE_PRINTKV(("IBSA - HUB_S(0x%p, 0x%lx)\n", BTEREG_SRC_ADDR,
-		     (TO_PHYS(src))));
-	HUB_S(BTEREG_SRC_ADDR, (TO_PHYS(src)));
-	BTE_PRINTKV(("IBDA - HUB_S(0x%p, 0x%lx)\n", BTEREG_DEST_ADDR,
-		     (TO_PHYS(dest))));
-	HUB_S(BTEREG_DEST_ADDR, (TO_PHYS(dest)));
+	BTE_PRINTKV(("IBSA = 0x%lx)\n", (TO_PHYS(src))));
+	BTE_SRC_STORE(bte, TO_PHYS(src));
+	BTE_PRINTKV(("IBDA = 0x%lx)\n", (TO_PHYS(dest))));
+	BTE_DEST_STORE(bte, TO_PHYS(dest));
 
 	/* Set the notification register */
-	BTE_PRINTKV(("IBNA - HUB_S(0x%p, 0x%lx)\n", BTEREG_NOTIF_ADDR,
-		     (TO_PHYS(ia64_tpa((unsigned long)bte->most_rcnt_na)))));
-	HUB_S(BTEREG_NOTIF_ADDR, (TO_PHYS(ia64_tpa((unsigned long)bte->most_rcnt_na))));
+	BTE_PRINTKV(("IBNA = 0x%lx)\n", 
+		     TO_PHYS(ia64_tpa((unsigned long)bte->most_rcnt_na))));
+	BTE_NOTIF_STORE(bte, TO_PHYS(ia64_tpa((unsigned long)bte->most_rcnt_na)));
 
 
 	/* Initiate the transfer */
-	BTE_PRINTK(("IBCT - HUB_S(0x%p, 0x%lx)\n", BTEREG_CTRL_ADDR,
-		     BTE_VALID_MODE(mode)));
-	HUB_S(BTEREG_CTRL_ADDR, BTE_VALID_MODE(mode));
+	BTE_PRINTK(("IBCT = 0x%lx)\n", BTE_VALID_MODE(mode)));
+	BTE_CTRL_STORE(bte, BTE_VALID_MODE(mode));
 
 	spin_unlock_irqrestore(&bte->spinlock, irq_flags);
 
@@ -156,7 +199,7 @@ bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 
 
 	BTE_PRINTKV((" Delay Done.  IBLS = 0x%lx, most_rcnt_na = 0x%lx\n",
-				HUB_L(BTEREG_LNSTAT_ADDR), *bte->most_rcnt_na));
+				BTE_LNSTAT_LOAD(bte), *bte->most_rcnt_na));
 
 	if (*bte->most_rcnt_na & IBLS_ERROR) {
 		bte_status = *bte->most_rcnt_na & ~IBLS_ERROR;
@@ -165,10 +208,11 @@ bte_copy(u64 src, u64 dest, u64 len, u64 mode, void *notification)
 		bte_status = BTE_SUCCESS;
 	}
 	BTE_PRINTK(("Returning status is 0x%lx and most_rcnt_na is 0x%lx\n",
-				HUB_L(BTEREG_LNSTAT_ADDR), *bte->most_rcnt_na));
+				BTE_LNSTAT_LOAD(bte), *bte->most_rcnt_na));
 
 	return bte_status;
 }
+EXPORT_SYMBOL(bte_copy);
 
 
 /*
@@ -201,14 +245,19 @@ bte_unaligned_copy(u64 src, u64 dest, u64 len, u64 mode)
 	u64 footBcopyDest;
 	u64 footBcopyLen;
 	bte_result_t rv;
-	char *bteBlock;
+	char *bteBlock, *bteBlock_unaligned;
 
 	if (len == 0) {
 		return BTE_SUCCESS;
 	}
 
 	/* temporary buffer used during unaligned transfers */
-	bteBlock = pda->cpu_bte_if[0]->scratch_buf;
+	bteBlock_unaligned = kmalloc(len + 3 * L1_CACHE_BYTES,
+				     GFP_KERNEL | GFP_DMA);
+	if (bteBlock_unaligned == NULL) {
+		return BTEFAIL_NOTAVAIL;
+	}
+	bteBlock = (char *) L1_CACHE_ALIGN((u64) bteBlock_unaligned);
 
 	headBcopySrcOffset = src & L1_CACHE_MASK;
 	destFirstCacheOffset = dest & L1_CACHE_MASK;
@@ -276,6 +325,7 @@ bte_unaligned_copy(u64 src, u64 dest, u64 len, u64 mode)
 					      ia64_tpa((unsigned long)bteBlock),
 					      footBteLen, mode, NULL);
 				if (rv != BTE_SUCCESS) {
+					kfree(bteBlock_unaligned);
 					return rv;
 				}
 
@@ -296,6 +346,7 @@ bte_unaligned_copy(u64 src, u64 dest, u64 len, u64 mode)
 				      (len - headBcopyLen -
 				       footBcopyLen), mode, NULL);
 			if (rv != BTE_SUCCESS) {
+				kfree(bteBlock_unaligned);
 				return rv;
 			}
 
@@ -325,6 +376,7 @@ bte_unaligned_copy(u64 src, u64 dest, u64 len, u64 mode)
 		rv = bte_copy(headBteSource,
 			      ia64_tpa((unsigned long)bteBlock), headBteLen, mode, NULL);
 		if (rv != BTE_SUCCESS) {
+			kfree(bteBlock_unaligned);
 			return rv;
 		}
 
@@ -332,8 +384,10 @@ bte_unaligned_copy(u64 src, u64 dest, u64 len, u64 mode)
 					     headBcopySrcOffset),
 		       headBcopyLen);
 	}
+	kfree(bteBlock_unaligned);
 	return BTE_SUCCESS;
 }
+EXPORT_SYMBOL(bte_unaligned_copy);
 
 
 /************************************************************************
@@ -370,9 +424,9 @@ bte_init_node(nodepda_t * mynodepda, cnodeid_t cnode)
 	mynodepda->bte_recovery_timer.data = (unsigned long) mynodepda;
 
 	for (i = 0; i < BTES_PER_NODE; i++) {
-		/* >>> Don't know why the 0x1800000L is here.  Robin */
-		mynodepda->bte_if[i].bte_base_addr =
-		    (char *) LOCAL_MMR_ADDR(bte_offsets[i] | 0x1800000L);
+		(u64) mynodepda->bte_if[i].bte_base_addr =
+		    REMOTE_HUB_ADDR(cnodeid_to_nasid(cnode),
+			(i == 0 ? IIO_IBLS0 : IIO_IBLS1));
 
 		/*
 		 * Initialize the notification and spinlock
@@ -383,8 +437,6 @@ bte_init_node(nodepda_t * mynodepda, cnodeid_t cnode)
 		mynodepda->bte_if[i].notify = 0L;
 		spin_lock_init(&mynodepda->bte_if[i].spinlock);
 
-		mynodepda->bte_if[i].scratch_buf =
-		    alloc_bootmem_node(NODE_DATA(cnode), BTE_MAX_XFER);
 		mynodepda->bte_if[i].bte_cnode = cnode;
 		mynodepda->bte_if[i].bte_error_count = 0;
 		mynodepda->bte_if[i].bte_num = i;
@@ -393,23 +445,3 @@ bte_init_node(nodepda_t * mynodepda, cnodeid_t cnode)
 	}
 
 }
-
-/*
- * bte_init_cpu()
- *
- * Initialize the cpupda structure with pointers to the
- * nodepda bte blocks.
- *
- */
-void
-bte_init_cpu(void)
-{
-	/* Called by setup.c as each cpu is being added to the nodepda */
-	if (local_node_data->active_cpu_count & 0x1) {
-		pda->cpu_bte_if[0] = &(nodepda->bte_if[0]);
-		pda->cpu_bte_if[1] = &(nodepda->bte_if[1]);
-	} else {
-		pda->cpu_bte_if[0] = &(nodepda->bte_if[1]);
-		pda->cpu_bte_if[1] = &(nodepda->bte_if[0]);
-	}
-}

arch/ia64/sn/kernel/setup.c

@@ -54,7 +54,6 @@ DEFINE_PER_CPU(struct pda_s, pda_percpu);
 #define MAX_PHYS_MEMORY		(1UL << 49)     /* 1 TB */
 
 extern void bte_init_node (nodepda_t *, cnodeid_t);
-extern void bte_init_cpu (void);
 extern void sn_timer_init(void);
 extern unsigned long last_time_offset;
 extern void init_platform_hubinfo(nodepda_t **nodepdaindr);
@@ -496,8 +495,6 @@ sn_cpu_init(void)
 		buddy_nasid = cnodeid_to_nasid(numa_node_id() == numnodes-1 ? 0 : numa_node_id()+ 1);
 		pda->pio_shub_war_cam_addr = (volatile unsigned long*)GLOBAL_MMR_ADDR(nasid, SH_PI_CAM_CONTROL);
 	}
-
-	bte_init_cpu();
 }
 
 /*

include/asm-ia64/iosapic.h

 #ifndef __ASM_IA64_IOSAPIC_H
 #define __ASM_IA64_IOSAPIC_H
 
-#define	IOSAPIC_DEFAULT_ADDR	0xFEC00000
-
 #define	IOSAPIC_REG_SELECT	0x0
 #define	IOSAPIC_WINDOW		0x10
 #define	IOSAPIC_EOI		0x40
@@ -55,6 +53,23 @@
 
 #define NR_IOSAPICS			256
 
+static inline unsigned int iosapic_read(char *iosapic, unsigned int reg)
+{
+	writel(reg, iosapic + IOSAPIC_REG_SELECT);
+	return readl(iosapic + IOSAPIC_WINDOW);
+}
+
+static inline void iosapic_write(char *iosapic, unsigned int reg, u32 val)
+{
+	writel(reg, iosapic + IOSAPIC_REG_SELECT);
+	writel(val, iosapic + IOSAPIC_WINDOW);
+}
+
+static inline void iosapic_eoi(char *iosapic, u32 vector)
+{
+	writel(vector, iosapic + IOSAPIC_EOI);
+}
+
 extern void __init iosapic_system_init (int pcat_compat);
 extern void __init iosapic_init (unsigned long address,
 				    unsigned int gsi_base);

include/asm-ia64/smp.h

@@ -123,5 +123,9 @@ extern void smp_send_reschedule (int cpu);
 extern void lock_ipi_calllock(void);
 extern void unlock_ipi_calllock(void);
 
+#else
+
+#define cpu_logical_id(cpuid)		0
+
 #endif /* CONFIG_SMP */
 #endif /* _ASM_IA64_SMP_H */

include/asm-ia64/sn/bte.h

@@ -3,7 +3,7 @@
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
- * Copyright (c) 2000-2003 Silicon Graphics, Inc.  All Rights Reserved.
+ * Copyright (c) 2000-2004 Silicon Graphics, Inc.  All Rights Reserved.
  */
 
 
@@ -48,35 +48,31 @@
 #define BTE_ZERO_FILL (BTE_NOTIFY | IBCT_ZFIL_MODE)
 /* Use a reserved bit to let the caller specify a wait for any BTE */
 #define BTE_WACQUIRE (0x4000)
+/* Use the BTE on the node with the destination memory */
+#define BTE_USE_DEST (BTE_WACQUIRE << 1)
+/* Use any available BTE interface on any node for the transfer */
+#define BTE_USE_ANY (BTE_USE_DEST << 1)
 /* macro to force the IBCT0 value valid */
 #define BTE_VALID_MODE(x) ((x) & (IBCT_NOTIFY | IBCT_ZFIL_MODE))
 
-
-/*
- * Handle locking of the bte interfaces.
- *
- * All transfers spinlock the interface before setting up the SHUB
- * registers.  Sync transfers hold the lock until all processing is
- * complete.  Async transfers release the lock as soon as the transfer
- * is initiated.
- *
- * To determine if an interface is available, we must check both the
- * busy bit and the spinlock for that interface.
- */
-#define BTE_LOCK_IF_AVAIL(_x) (\
-	(*pda->cpu_bte_if[_x]->most_rcnt_na & (IBLS_BUSY | IBLS_ERROR)) && \
-	(!(spin_trylock(&(pda->cpu_bte_if[_x]->spinlock)))) \
-	)
+#define BTE_ACTIVE	(IBLS_BUSY | IBLS_ERROR)
 
 /*
  * Some macros to simplify reading.
  * Start with macros to locate the BTE control registers.
  */
-#define BTEREG_LNSTAT_ADDR ((u64 *)(bte->bte_base_addr))
-#define BTEREG_SRC_ADDR ((u64 *)(bte->bte_base_addr + BTEOFF_SRC))
-#define BTEREG_DEST_ADDR ((u64 *)(bte->bte_base_addr + BTEOFF_DEST))
-#define BTEREG_CTRL_ADDR ((u64 *)(bte->bte_base_addr + BTEOFF_CTRL))
-#define BTEREG_NOTIF_ADDR ((u64 *)(bte->bte_base_addr + BTEOFF_NOTIFY))
+#define BTE_LNSTAT_LOAD(_bte)						\
+			HUB_L(_bte->bte_base_addr)
+#define BTE_LNSTAT_STORE(_bte, _x)					\
+			HUB_S(_bte->bte_base_addr, (_x))
+#define BTE_SRC_STORE(_bte, _x)						\
+			HUB_S(_bte->bte_base_addr + (BTEOFF_SRC/8), (_x))
+#define BTE_DEST_STORE(_bte, _x)					\
+			HUB_S(_bte->bte_base_addr + (BTEOFF_DEST/8), (_x))
+#define BTE_CTRL_STORE(_bte, _x)					\
+			HUB_S(_bte->bte_base_addr + (BTEOFF_CTRL/8), (_x))
+#define BTE_NOTIF_STORE(_bte, _x)					\
+			HUB_S(_bte->bte_base_addr + (BTEOFF_NOTIFY/8), (_x))
 
 
 /* Possible results from bte_copy and bte_unaligned_copy */
@@ -110,16 +106,15 @@ typedef enum {
  * to work with a BTE.
  */
 struct bteinfo_s {
-	u64 volatile notify ____cacheline_aligned;
-	char *bte_base_addr ____cacheline_aligned;
+	volatile u64 notify ____cacheline_aligned;
+	u64 *bte_base_addr ____cacheline_aligned;
 	spinlock_t spinlock;
 	cnodeid_t bte_cnode;	/* cnode                            */
 	int bte_error_count;	/* Number of errors encountered     */
 	int bte_num;		/* 0 --> BTE0, 1 --> BTE1           */
 	int cleanup_active;	/* Interface is locked for cleanup  */
 	volatile bte_result_t bh_error;	/* error while processing   */
-	u64 volatile *most_rcnt_na;
-	void *scratch_buf;	/* Node local scratch buffer        */
+	volatile u64 *most_rcnt_na;
 };
 
 
@@ -130,6 +125,8 @@ extern bte_result_t bte_copy(u64, u64, u64, u64, void *);
 extern bte_result_t bte_unaligned_copy(u64, u64, u64, u64);
 extern void bte_error_handler(unsigned long);
 
+#define bte_zero(dest, len, mode, notification) \
+	bte_copy(0, dest, len, ((mode) | BTE_ZERO_FILL), notification)
 
 /*
  * The following is the prefered way of calling bte_unaligned_copy

include/asm-ia64/sn/pda.h

@@ -49,8 +49,6 @@ typedef struct pda_s {
 	volatile unsigned long *pio_shub_war_cam_addr;
 	volatile unsigned long *mem_write_status_addr;
 
-	struct bteinfo_s *cpu_bte_if[BTES_PER_NODE];	/* cpu interface order */
-
 	unsigned long	sn_soft_irr[4];
 	unsigned long	sn_in_service_ivecs[4];
 	short		cnodeid_to_nasid_table[MAX_NUMNODES];

include/asm-ia64/sn/sn_cpuid.h

@@ -84,7 +84,6 @@
  */
 
 #ifndef CONFIG_SMP
-#define cpu_logical_id(cpu)				0
 #define cpu_physical_id(cpuid)			((ia64_getreg(_IA64_REG_CR_LID) >> 16) & 0xffff)
 #endif