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

into ppc970.osdl.org:/home/torvalds/v2.5/linux

arch/ia64/Kconfig

@@ -288,39 +288,6 @@ config FORCE_MAX_ZONEORDER
 	int
 	default "18"
 
-choice
-	prompt "Huge TLB page size"
-	depends on HUGETLB_PAGE
-	default HUGETLB_PAGE_SIZE_16MB
-
-config HUGETLB_PAGE_SIZE_4GB
-	depends on MCKINLEY
-	bool "4GB"
-
-config HUGETLB_PAGE_SIZE_1GB
-	depends on MCKINLEY
-	bool "1GB"
-
-config HUGETLB_PAGE_SIZE_256MB
-	bool "256MB"
-
-config HUGETLB_PAGE_SIZE_64MB
-	bool "64MB"
-
-config HUGETLB_PAGE_SIZE_16MB
-	bool "16MB"
-
-config HUGETLB_PAGE_SIZE_4MB
-	bool "4MB"
-
-config HUGETLB_PAGE_SIZE_1MB
-	bool "1MB"
-
-config HUGETLB_PAGE_SIZE_256KB
-	bool "256KB"
-
-endchoice
-
 config IA64_PAL_IDLE
 	bool "Use PAL_HALT_LIGHT in idle loop"
 	help

arch/ia64/kernel/head.S

@@ -816,6 +816,19 @@ GLOBAL_ENTRY(ia64_delay_loop)
 	br.ret.sptk.many rp
 END(ia64_delay_loop)
 
+GLOBAL_ENTRY(ia64_invoke_kernel_thread_helper)
+	.prologue
+	.save rp, r0				// this is the end of the call-chain
+	.body
+	alloc r2 = ar.pfs, 0, 0, 2, 0
+	mov out0 = r9
+	mov out1 = r11;;
+	br.call.sptk.many rp = kernel_thread_helper;;
+	mov out0 = r8
+	br.call.sptk.many rp = sys_exit;;
+1:	br.sptk.few 1b				// not reached
+END(ia64_invoke_kernel_thread_helper)
+
 #ifdef CONFIG_IA64_BRL_EMU
 
 /*

arch/ia64/kernel/iosapic.c

@@ -103,6 +103,7 @@ static spinlock_t iosapic_lock = SPIN_LOCK_UNLOCKED;
 
 static struct iosapic_intr_info {
 	char		*addr;		/* base address of IOSAPIC */
+	u32		low32;		/* current value of low word of Redirection table entry */
 	unsigned int	gsi_base;	/* first GSI assigned to this IOSAPIC */
 	char		rte_index;	/* IOSAPIC RTE index (-1 => not an IOSAPIC interrupt) */
 	unsigned char	dmode	: 3;	/* delivery mode (see iosapic.h) */
@@ -213,6 +214,7 @@ set_rte (unsigned int vector, unsigned int dest)
 	writel(high32, addr + IOSAPIC_WINDOW);
 	writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
 	writel(low32, addr + IOSAPIC_WINDOW);
+	iosapic_intr_info[vector].low32 = low32;
 }
 
 static void
@@ -239,9 +241,10 @@ mask_irq (unsigned int irq)
 	spin_lock_irqsave(&iosapic_lock, flags);
 	{
 		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-		low32 = readl(addr + IOSAPIC_WINDOW);
 
-		low32 |= (1 << IOSAPIC_MASK_SHIFT);    /* set only the mask bit */
+		/* set only the mask bit */
+		low32 = iosapic_intr_info[vec].low32 |= IOSAPIC_MASK;
+
 		writel(low32, addr + IOSAPIC_WINDOW);
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
@@ -264,9 +267,7 @@ unmask_irq (unsigned int irq)
 	spin_lock_irqsave(&iosapic_lock, flags);
 	{
 		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-		low32 = readl(addr + IOSAPIC_WINDOW);
-
-		low32 &= ~(1 << IOSAPIC_MASK_SHIFT);    /* clear only the mask bit */
+		low32 = iosapic_intr_info[vec].low32 &= ~IOSAPIC_MASK;
 		writel(low32, addr + IOSAPIC_WINDOW);
 	}
 	spin_unlock_irqrestore(&iosapic_lock, flags);
@@ -307,9 +308,7 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask)
 	{
 		/* get current delivery mode by reading the low32 */
 		writel(IOSAPIC_RTE_LOW(rte_index), addr + IOSAPIC_REG_SELECT);
-		low32 = readl(addr + IOSAPIC_WINDOW);
-
-		low32 &= ~(7 << IOSAPIC_DELIVERY_SHIFT);
+		low32 = iosapic_intr_info[vec].low32 & ~(7 << IOSAPIC_DELIVERY_SHIFT);
 		if (redir)
 		        /* change delivery mode to lowest priority */
 			low32 |= (IOSAPIC_LOWEST_PRIORITY << IOSAPIC_DELIVERY_SHIFT);
@@ -317,6 +316,7 @@ iosapic_set_affinity (unsigned int irq, cpumask_t mask)
 		        /* change delivery mode to fixed */
 			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);

arch/ia64/kernel/irq.c

@@ -455,7 +455,6 @@ unsigned int do_IRQ(unsigned long irq, struct pt_regs *regs)
 	unsigned int status;
 	int cpu;
 
-	irq_enter();
 	cpu = smp_processor_id(); /* for CONFIG_PREEMPT, this must come after irq_enter()! */
 
 	kstat_cpu(cpu).irqs[irq]++;
@@ -525,7 +524,6 @@ unsigned int do_IRQ(unsigned long irq, struct pt_regs *regs)
 		desc->handler->end(irq);
 		spin_unlock(&desc->lock);
 	}
-	irq_exit();
 	return 1;
 }
 

arch/ia64/kernel/irq_ia64.c

@@ -120,6 +120,7 @@ ia64_handle_irq (ia64_vector vector, struct pt_regs *regs)
 	 * 16 (without this, it would be ~240, which could easily lead
 	 * to kernel stack overflows).
 	 */
+	irq_enter();
 	saved_tpr = ia64_getreg(_IA64_REG_CR_TPR);
 	ia64_srlz_d();
 	while (vector != IA64_SPURIOUS_INT_VECTOR) {
@@ -143,8 +144,7 @@ ia64_handle_irq (ia64_vector vector, struct pt_regs *regs)
 	 * handler needs to be able to wait for further keyboard interrupts, which can't
 	 * come through until ia64_eoi() has been done.
 	 */
-	if (local_softirq_pending())
-		do_softirq();
+	irq_exit();
 }
 
 #ifdef CONFIG_SMP

arch/ia64/kernel/ivt.S

@@ -118,10 +118,11 @@ ENTRY(vhpt_miss)
 #ifdef CONFIG_HUGETLB_PAGE
 	extr.u r26=r25,2,6
 	;;
-	cmp.eq p8,p0=HPAGE_SHIFT,r26
+	cmp.ne p8,p0=r18,r26
+	sub r27=r26,r18
 	;;
 (p8)	dep r25=r18,r25,2,6
-(p8)	shr r22=r22,HPAGE_SHIFT-PAGE_SHIFT
+(p8)	shr r22=r22,r27
 #endif
 	;;
 	cmp.eq p6,p7=5,r17			// is IFA pointing into to region 5?

arch/ia64/kernel/perfmon_default_smpl.c

@@ -178,6 +178,7 @@ default_handler(struct task_struct *task, void *buf, pfm_ovfl_arg_t *arg, struct
 	ent->tstamp    = stamp;
 	ent->cpu       = smp_processor_id();
 	ent->set       = arg->active_set;
+	ent->tgid      = current->tgid;
 
 	/*
 	 * selectively store PMDs in increasing index number

arch/ia64/kernel/process.c

@@ -259,10 +259,12 @@ ia64_load_extra (struct task_struct *task)
  *
  * We get here through the following  call chain:
  *
- *	<clone syscall>
- *	sys_clone
- *	do_fork
- *	copy_thread
+ *	from user-level:	from kernel:
+ *
+ *	<clone syscall>	        <some kernel call frames>
+ *	sys_clone		   :
+ *	do_fork			do_fork
+ *	copy_thread		copy_thread
  *
  * This means that the stack layout is as follows:
  *
@@ -276,9 +278,6 @@ ia64_load_extra (struct task_struct *task)
  *	|                     | <-- sp (lowest addr)
  *	+---------------------+
  *
- * Note: if we get called through kernel_thread() then the memory above "(highest addr)"
- * is valid kernel stack memory that needs to be copied as well.
- *
  * Observe that we copy the unat values that are in pt_regs and switch_stack.  Spilling an
  * integer to address X causes bit N in ar.unat to be set to the NaT bit of the register,
  * with N=(X & 0x1ff)/8.  Thus, copying the unat value preserves the NaT bits ONLY if the
@@ -291,9 +290,9 @@ copy_thread (int nr, unsigned long clone_flags,
 	     unsigned long user_stack_base, unsigned long user_stack_size,
 	     struct task_struct *p, struct pt_regs *regs)
 {
-	unsigned long rbs, child_rbs, rbs_size, stack_offset, stack_top, stack_used;
-	struct switch_stack *child_stack, *stack;
 	extern char ia64_ret_from_clone, ia32_ret_from_clone;
+	struct switch_stack *child_stack, *stack;
+	unsigned long rbs, child_rbs, rbs_size;
 	struct pt_regs *child_ptregs;
 	int retval = 0;
 
@@ -306,16 +305,13 @@ copy_thread (int nr, unsigned long clone_flags,
 		return 0;
 #endif
 
-	stack_top = (unsigned long) current + IA64_STK_OFFSET;
 	stack = ((struct switch_stack *) regs) - 1;
-	stack_used = stack_top - (unsigned long) stack;
-	stack_offset = IA64_STK_OFFSET - stack_used;
 
-	child_stack = (struct switch_stack *) ((unsigned long) p + stack_offset);
-	child_ptregs = (struct pt_regs *) (child_stack + 1);
+	child_ptregs = (struct pt_regs *) ((unsigned long) p + IA64_STK_OFFSET) - 1;
+	child_stack = (struct switch_stack *) child_ptregs - 1;
 
 	/* copy parent's switch_stack & pt_regs to child: */
-	memcpy(child_stack, stack, stack_used);
+	memcpy(child_stack, stack, sizeof(*child_ptregs) + sizeof(*child_stack));
 
 	rbs = (unsigned long) current + IA64_RBS_OFFSET;
 	child_rbs = (unsigned long) p + IA64_RBS_OFFSET;
@@ -324,7 +320,7 @@ copy_thread (int nr, unsigned long clone_flags,
 	/* copy the parent's register backing store to the child: */
 	memcpy((void *) child_rbs, (void *) rbs, rbs_size);
 
-	if (user_mode(child_ptregs)) {
+	if (likely(user_mode(child_ptregs))) {
 		if ((clone_flags & CLONE_SETTLS) && !IS_IA32_PROCESS(regs))
 			child_ptregs->r13 = regs->r16;	/* see sys_clone2() in entry.S */
 		if (user_stack_base) {
@@ -341,14 +337,14 @@ copy_thread (int nr, unsigned long clone_flags,
 		 * been taken care of by the caller of sys_clone()
 		 * already.
 		 */
-		child_ptregs->r12 = (unsigned long) (child_ptregs + 1); /* kernel sp */
+		child_ptregs->r12 = (unsigned long) child_ptregs - 16; /* kernel sp */
 		child_ptregs->r13 = (unsigned long) p;		/* set `current' pointer */
 	}
+	child_stack->ar_bspstore = child_rbs + rbs_size;
 	if (IS_IA32_PROCESS(regs))
 		child_stack->b0 = (unsigned long) &ia32_ret_from_clone;
 	else
 		child_stack->b0 = (unsigned long) &ia64_ret_from_clone;
-	child_stack->ar_bspstore = child_rbs + rbs_size;
 
 	/* copy parts of thread_struct: */
 	p->thread.ksp = (unsigned long) child_stack - 16;
@@ -358,8 +354,8 @@ copy_thread (int nr, unsigned long clone_flags,
 	 * therefore we must specify them explicitly here and not include them in
 	 * IA64_PSR_BITS_TO_CLEAR.
 	 */
-	child_ptregs->cr_ipsr =  ((child_ptregs->cr_ipsr | IA64_PSR_BITS_TO_SET)
-			      & ~(IA64_PSR_BITS_TO_CLEAR | IA64_PSR_PP | IA64_PSR_UP));
+	child_ptregs->cr_ipsr = ((child_ptregs->cr_ipsr | IA64_PSR_BITS_TO_SET)
+				 & ~(IA64_PSR_BITS_TO_CLEAR | IA64_PSR_PP | IA64_PSR_UP));
 
 	/*
 	 * NOTE: The calling convention considers all floating point
@@ -578,27 +574,43 @@ ia64_set_personality (struct elf64_hdr *elf_ex, int ibcs2_interpreter)
 pid_t
 kernel_thread (int (*fn)(void *), void *arg, unsigned long flags)
 {
-	struct task_struct *parent = current;
-	int result; 
-	pid_t tid;
+	extern void ia64_invoke_kernel_thread_helper (void);
+	unsigned long *helper_fptr = (unsigned long *) &ia64_invoke_kernel_thread_helper;
+	struct {
+		struct switch_stack sw;
+		struct pt_regs pt;
+	} regs;
+
+	memset(&regs, 0, sizeof(regs));
+	regs.pt.cr_iip = helper_fptr[0];	/* set entry point (IP) */
+	regs.pt.r1 = helper_fptr[1];		/* set GP */
+	regs.pt.r9 = (unsigned long) fn;	/* 1st argument */
+	regs.pt.r11 = (unsigned long) arg;	/* 2nd argument */
+	/* Preserve PSR bits, except for bits 32-34 and 37-45, which we can't read.  */
+	regs.pt.cr_ipsr = ia64_getreg(_IA64_REG_PSR) | IA64_PSR_BN;
+	regs.pt.cr_ifs = 1UL << 63;		/* mark as valid, empty frame */
+	regs.sw.ar_fpsr = regs.pt.ar_fpsr = ia64_getreg(_IA64_REG_AR_FPSR);
+	regs.sw.ar_bspstore = (unsigned long) current + IA64_RBS_OFFSET;
+
+	return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, &regs.pt, 0, NULL, NULL);
+}
+EXPORT_SYMBOL(kernel_thread);
 
-	tid = clone(flags | CLONE_VM | CLONE_UNTRACED, 0);
-	if (parent != current) {
+/* This gets called from kernel_thread() via ia64_invoke_thread_helper().  */
+int
+kernel_thread_helper (int (*fn)(void *), void *arg)
+{
 #ifdef CONFIG_IA32_SUPPORT
-		if (IS_IA32_PROCESS(ia64_task_regs(current))) {
-			/* A kernel thread is always a 64-bit process. */
-			current->thread.map_base  = DEFAULT_MAP_BASE;
-			current->thread.task_size = DEFAULT_TASK_SIZE;
-			ia64_set_kr(IA64_KR_IO_BASE, current->thread.old_iob);
-			ia64_set_kr(IA64_KR_TSSD, current->thread.old_k1);
-		}
-#endif
-		result = (*fn)(arg);
-		_exit(result);
+	if (IS_IA32_PROCESS(ia64_task_regs(current))) {
+		/* A kernel thread is always a 64-bit process. */
+		current->thread.map_base  = DEFAULT_MAP_BASE;
+		current->thread.task_size = DEFAULT_TASK_SIZE;
+		ia64_set_kr(IA64_KR_IO_BASE, current->thread.old_iob);
+		ia64_set_kr(IA64_KR_TSSD, current->thread.old_k1);
 	}
-	return tid;
+#endif
+	return (*fn)(arg);
 }
-EXPORT_SYMBOL(kernel_thread);
 
 /*
  * Flush thread state.  This is called when a thread does an execve().

arch/ia64/mm/hugetlbpage.c

 /*
  * IA-64 Huge TLB Page Support for Kernel.
  *
- * Copyright (C) 2002, Rohit Seth <rohit.seth@intel.com>
+ * Copyright (C) 2002-2004 Rohit Seth <rohit.seth@intel.com>
+ * Copyright (C) 2003-2004 Ken Chen <kenneth.w.chen@intel.com>
+ *
+ * Sep, 2003: add numa support
+ * Feb, 2004: dynamic hugetlb page size via boot parameter
  */
 
 #include <linux/config.h>
@@ -18,11 +22,10 @@
 #include <asm/tlb.h>
 #include <asm/tlbflush.h>
 
-#define TASK_HPAGE_BASE (REGION_HPAGE << REGION_SHIFT)
-
 static long	htlbpagemem;
 int		htlbpage_max;
 static long	htlbzone_pages;
+unsigned int	hpage_shift=HPAGE_SHIFT_DEFAULT;
 
 static struct list_head hugepage_freelists[MAX_NUMNODES];
 static spinlock_t htlbpage_lock = SPIN_LOCK_UNLOCKED;
@@ -407,7 +410,7 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr, u
 		return -EINVAL;
 	/* This code assumes that REGION_HPAGE != 0. */
 	if ((REGION_NUMBER(addr) != REGION_HPAGE) || (addr & (HPAGE_SIZE - 1)))
-		addr = TASK_HPAGE_BASE;
+		addr = HPAGE_REGION_BASE;
 	else
 		addr = ALIGN(addr, HPAGE_SIZE);
 	for (vmm = find_vma(current->mm, addr); ; vmm = vmm->vm_next) {
@@ -520,6 +523,35 @@ static int __init hugetlb_setup(char *s)
 }
 __setup("hugepages=", hugetlb_setup);
 
+static int __init hugetlb_setup_sz(char *str)
+{
+	u64 tr_pages;
+	unsigned long long size;
+
+	if (ia64_pal_vm_page_size(&tr_pages, NULL) != 0)
+		/*
+		 * shouldn't happen, but just in case.
+		 */
+		tr_pages = 0x15557000UL;
+
+	size = memparse(str, &str);
+	if (*str || (size & (size-1)) || !(tr_pages & size) ||
+		size <= PAGE_SIZE ||
+		size >= (1UL << PAGE_SHIFT << MAX_ORDER)) {
+		printk(KERN_WARNING "Invalid huge page size specified\n");
+		return 1;
+	}
+
+	hpage_shift = __ffs(size);
+	/*
+	 * boot cpu already executed ia64_mmu_init, and has HPAGE_SHIFT_DEFAULT
+	 * override here with new page shift.
+	 */
+	ia64_set_rr(HPAGE_REGION_BASE, hpage_shift << 2);
+	return 1;
+}
+__setup("hugepagesz=", hugetlb_setup_sz);
+
 static int __init hugetlb_init(void)
 {
 	int i;
@@ -540,7 +572,7 @@ static int __init hugetlb_init(void)
 	printk("Total HugeTLB memory allocated, %ld\n", htlbpagemem);
 	return 0;
 }
-module_init(hugetlb_init);
+__initcall(hugetlb_init);
 
 int hugetlb_report_meminfo(char *buf)
 {

arch/ia64/mm/init.c

@@ -342,6 +342,10 @@ ia64_mmu_init (void *my_cpu_data)
 
 	ia64_tlb_init();
 
+#ifdef	CONFIG_HUGETLB_PAGE
+	ia64_set_rr(HPAGE_REGION_BASE, HPAGE_SHIFT << 2);
+#endif
+
 #ifdef	CONFIG_IA64_MCA
 	cpu = smp_processor_id();
 

include/asm-ia64/iosapic.h

@@ -45,9 +45,9 @@
 /*
  * Mask bit
  */
+
 #define	IOSAPIC_MASK_SHIFT		16
-#define	IOSAPIC_UNMASK			0
-#define	IOSAPIC_MSAK			1
+#define	IOSAPIC_MASK			(1<<IOSAPIC_MASK_SHIFT)
 
 #ifndef __ASSEMBLY__
 

include/asm-ia64/mmu_context.h

@@ -140,8 +140,9 @@ reload_context (mm_context_t context)
 {
 	unsigned long rid;
 	unsigned long rid_incr = 0;
-	unsigned long rr0, rr1, rr2, rr3, rr4;
+	unsigned long rr0, rr1, rr2, rr3, rr4, old_rr4;
 
+	old_rr4 = ia64_get_rr(0x8000000000000000);
 	rid = context << 3;	/* make space for encoding the region number */
 	rid_incr = 1 << 8;
 
@@ -152,7 +153,7 @@ reload_context (mm_context_t context)
 	rr3 = rr0 + 3*rid_incr;
 	rr4 = rr0 + 4*rid_incr;
 #ifdef  CONFIG_HUGETLB_PAGE
-	rr4 = (rr4 & (~(0xfcUL))) | (HPAGE_SHIFT << 2);
+	rr4 = (rr4 & (~(0xfcUL))) | (old_rr4 & 0xfc);
 #endif
 
 	ia64_set_rr(0x0000000000000000, rr0);

include/asm-ia64/page.h

@@ -37,31 +37,14 @@
 #define RGN_MAP_LIMIT	((1UL << (4*PAGE_SHIFT - 12)) - PAGE_SIZE)	/* per region addr limit */
 
 #ifdef CONFIG_HUGETLB_PAGE
+# define REGION_HPAGE		(4UL)	/* note: this is hardcoded in reload_context()!*/
+# define REGION_SHIFT		61
+# define HPAGE_REGION_BASE	(REGION_HPAGE << REGION_SHIFT)
+# define HPAGE_SHIFT		hpage_shift
+# define HPAGE_SHIFT_DEFAULT	28	/* check ia64 SDM for architecture supported size */
+# define HPAGE_SIZE		(__IA64_UL_CONST(1) << HPAGE_SHIFT)
+# define HPAGE_MASK		(~(HPAGE_SIZE - 1))
 
-# if defined(CONFIG_HUGETLB_PAGE_SIZE_4GB)
-#  define HPAGE_SHIFT	32
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_1GB)
-#  define HPAGE_SHIFT	30
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_256MB)
-#  define HPAGE_SHIFT	28
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_64MB)
-#  define HPAGE_SHIFT	26
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_16MB)
-#  define HPAGE_SHIFT	24
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_4MB)
-#  define HPAGE_SHIFT	22
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_1MB)
-#  define HPAGE_SHIFT	20
-# elif defined(CONFIG_HUGETLB_PAGE_SIZE_256KB)
-#  define HPAGE_SHIFT	18
-# else
-#  error Unsupported IA-64 HugeTLB Page Size!
-# endif
-
-# define REGION_HPAGE	(4UL)	/* note: this is hardcoded in mmu_context.h:reload_context()!*/
-# define REGION_SHIFT	61
-# define HPAGE_SIZE	(__IA64_UL_CONST(1) << HPAGE_SHIFT)
-# define HPAGE_MASK	(~(HPAGE_SIZE - 1))
 # define HAVE_ARCH_HUGETLB_UNMAPPED_AREA
 # define ARCH_HAS_HUGEPAGE_ONLY_RANGE
 #endif /* CONFIG_HUGETLB_PAGE */
@@ -140,6 +123,7 @@ typedef union ia64_va {
 # define is_hugepage_only_range(addr, len)		\
 	 (REGION_NUMBER(addr) == REGION_HPAGE &&	\
 	  REGION_NUMBER((addr)+(len)) == REGION_HPAGE)
+extern unsigned int hpage_shift;
 #endif
 
 static __inline__ int

include/asm-ia64/perfmon_default_smpl.h

@@ -59,7 +59,7 @@ typedef struct {
  * last_reset_value member indicates the initial value of the overflowed PMD. 
  */
 typedef struct {
-        int             pid;                    /* active process at PMU interrupt point */
+        int             pid;                    /* thread id (for NPTL, this is gettid()) */
         unsigned char   reserved1[3];           /* reserved for future use */
         unsigned char   ovfl_pmd;               /* index of overflowed PMD */
 
@@ -69,7 +69,7 @@ typedef struct {
 
         unsigned short  cpu;                    /* cpu on which the overfow occured */
         unsigned short  set;                    /* event set active when overflow ocurred   */
-        unsigned int    reserved2;              /* for future use */
+        int    		tgid;              	/* thread group id (for NPTL, this is getpid()) */
 } pfm_default_smpl_entry_t;
 
 #define PFM_DEFAULT_MAX_PMDS		64 /* how many pmds supported by data structures (sizeof(unsigned long) */

include/asm-ia64/scatterlist.h

@@ -2,7 +2,7 @@
 #define _ASM_IA64_SCATTERLIST_H
 
 /*
- * Modified 1998-1999, 2001-2002
+ * Modified 1998-1999, 2001-2002, 2004
  *	David Mosberger-Tang <davidm@hpl.hp.com>, Hewlett-Packard Co
  */
 
@@ -15,6 +15,14 @@ struct scatterlist {
 	unsigned int dma_length;
 };
 
-#define ISA_DMA_THRESHOLD	(~0UL)
+/*
+ * It used to be that ISA_DMA_THRESHOLD had something to do with the
+ * DMA-limits of ISA-devices.  Nowadays, its only remaining use (apart
+ * from the aha1542.c driver, which isn't 64-bit clean anyhow) is to
+ * tell the block-layer (via BLK_BOUNCE_ISA) what the max. physical
+ * address of a page is that is allocated with GFP_DMA.  On IA-64,
+ * that's 4GB - 1.
+ */
+#define ISA_DMA_THRESHOLD	0xffffffff
 
 #endif /* _ASM_IA64_SCATTERLIST_H */