Merge branch 'perfcounters-core-for-linus' of...

Merge branch 'perfcounters-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'perfcounters-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (60 commits)
  perf tools: Avoid unnecessary work in directory lookups
  perf stat: Clean up statistics calculations a bit more
  perf stat: More advanced variance computation
  perf stat: Use stddev_mean in stead of stddev
  perf stat: Remove the limit on repeat
  perf stat: Change noise calculation to use stddev
  x86, perf_counter, bts: Do not allow kernel BTS tracing for now
  x86, perf_counter, bts: Correct pointer-to-u64 casts
  x86, perf_counter, bts: Fail if BTS is not available
  perf_counter: Fix output-sharing error path
  perf trace: Fix read_string()
  perf trace: Print out in nanoseconds
  perf tools: Seek to the end of the header area
  perf trace: Fix parsing of perf.data
  perf trace: Sample timestamps as well
  perf_counter: Introduce new (non-)paranoia level to allow raw tracepoint access
  perf trace: Sample the CPU too
  perf tools: Work around strict aliasing related warnings
  perf tools: Clean up warnings list in the Makefile
  perf tools: Complete support for dynamic strings
  ...

arch/powerpc/include/asm/pgtable.h

@@ -104,8 +104,8 @@ static inline void __set_pte_at(struct mm_struct *mm, unsigned long addr,
 	else
 		pte_update(ptep, ~_PAGE_HASHPTE, pte_val(pte));
 
-#elif defined(CONFIG_PPC32) && defined(CONFIG_PTE_64BIT) && defined(CONFIG_SMP)
-	/* Second case is 32-bit with 64-bit PTE in SMP mode. In this case, we
+#elif defined(CONFIG_PPC32) && defined(CONFIG_PTE_64BIT)
+	/* Second case is 32-bit with 64-bit PTE.  In this case, we
 	 * can just store as long as we do the two halves in the right order
 	 * with a barrier in between. This is possible because we take care,
 	 * in the hash code, to pre-invalidate if the PTE was already hashed,
@@ -140,7 +140,7 @@ static inline void __set_pte_at(struct mm_struct *mm, unsigned long addr,
 
 #else
 	/* Anything else just stores the PTE normally. That covers all 64-bit
-	 * cases, and 32-bit non-hash with 64-bit PTEs in UP mode
+	 * cases, and 32-bit non-hash with 32-bit PTEs.
 	 */
 	*ptep = pte;
 #endif

arch/powerpc/kernel/Makefile

@@ -97,7 +97,7 @@ obj64-$(CONFIG_AUDIT)		+= compat_audit.o
 
 obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
 obj-$(CONFIG_FUNCTION_GRAPH_TRACER)	+= ftrace.o
-obj-$(CONFIG_PPC_PERF_CTRS)	+= perf_counter.o
+obj-$(CONFIG_PPC_PERF_CTRS)	+= perf_counter.o perf_callchain.o
 obj64-$(CONFIG_PPC_PERF_CTRS)	+= power4-pmu.o ppc970-pmu.o power5-pmu.o \
 				   power5+-pmu.o power6-pmu.o power7-pmu.o
 obj32-$(CONFIG_PPC_PERF_CTRS)	+= mpc7450-pmu.o

arch/powerpc/kernel/asm-offsets.c

@@ -67,6 +67,8 @@ int main(void)
 	DEFINE(MMCONTEXTID, offsetof(struct mm_struct, context.id));
 #ifdef CONFIG_PPC64
 	DEFINE(AUDITCONTEXT, offsetof(struct task_struct, audit_context));
+	DEFINE(SIGSEGV, SIGSEGV);
+	DEFINE(NMI_MASK, NMI_MASK);
 #else
 	DEFINE(THREAD_INFO, offsetof(struct task_struct, stack));
 #endif /* CONFIG_PPC64 */

arch/powerpc/kernel/exceptions-64s.S

@@ -729,6 +729,11 @@ BEGIN_FTR_SECTION
 	bne-	do_ste_alloc		/* If so handle it */
 END_FTR_SECTION_IFCLR(CPU_FTR_SLB)
 
+	clrrdi	r11,r1,THREAD_SHIFT
+	lwz	r0,TI_PREEMPT(r11)	/* If we're in an "NMI" */
+	andis.	r0,r0,NMI_MASK@h	/* (i.e. an irq when soft-disabled) */
+	bne	77f			/* then don't call hash_page now */
+
 	/*
 	 * On iSeries, we soft-disable interrupts here, then
 	 * hard-enable interrupts so that the hash_page code can spin on
@@ -833,6 +838,20 @@ handle_page_fault:
 	bl	.low_hash_fault
 	b	.ret_from_except
 
+/*
+ * We come here as a result of a DSI at a point where we don't want
+ * to call hash_page, such as when we are accessing memory (possibly
+ * user memory) inside a PMU interrupt that occurred while interrupts
+ * were soft-disabled.  We want to invoke the exception handler for
+ * the access, or panic if there isn't a handler.
+ */
+77:	bl	.save_nvgprs
+	mr	r4,r3
+	addi	r3,r1,STACK_FRAME_OVERHEAD
+	li	r5,SIGSEGV
+	bl	.bad_page_fault
+	b	.ret_from_except
+
 	/* here we have a segment miss */
 do_ste_alloc:
 	bl	.ste_allocate		/* try to insert stab entry */

arch/powerpc/kernel/perf_callchain.c

+/*
+ * Performance counter callchain support - powerpc architecture code
+ *
+ * Copyright © 2009 Paul Mackerras, IBM Corporation.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version
+ * 2 of the License, or (at your option) any later version.
+ */
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/perf_counter.h>
+#include <linux/percpu.h>
+#include <linux/uaccess.h>
+#include <linux/mm.h>
+#include <asm/ptrace.h>
+#include <asm/pgtable.h>
+#include <asm/sigcontext.h>
+#include <asm/ucontext.h>
+#include <asm/vdso.h>
+#ifdef CONFIG_PPC64
+#include "ppc32.h"
+#endif
+
+/*
+ * Store another value in a callchain_entry.
+ */
+static inline void callchain_store(struct perf_callchain_entry *entry, u64 ip)
+{
+	unsigned int nr = entry->nr;
+
+	if (nr < PERF_MAX_STACK_DEPTH) {
+		entry->ip[nr] = ip;
+		entry->nr = nr + 1;
+	}
+}
+
+/*
+ * Is sp valid as the address of the next kernel stack frame after prev_sp?
+ * The next frame may be in a different stack area but should not go
+ * back down in the same stack area.
+ */
+static int valid_next_sp(unsigned long sp, unsigned long prev_sp)
+{
+	if (sp & 0xf)
+		return 0;		/* must be 16-byte aligned */
+	if (!validate_sp(sp, current, STACK_FRAME_OVERHEAD))
+		return 0;
+	if (sp >= prev_sp + STACK_FRAME_OVERHEAD)
+		return 1;
+	/*
+	 * sp could decrease when we jump off an interrupt stack
+	 * back to the regular process stack.
+	 */
+	if ((sp & ~(THREAD_SIZE - 1)) != (prev_sp & ~(THREAD_SIZE - 1)))
+		return 1;
+	return 0;
+}
+
+static void perf_callchain_kernel(struct pt_regs *regs,
+				  struct perf_callchain_entry *entry)
+{
+	unsigned long sp, next_sp;
+	unsigned long next_ip;
+	unsigned long lr;
+	long level = 0;
+	unsigned long *fp;
+
+	lr = regs->link;
+	sp = regs->gpr[1];
+	callchain_store(entry, PERF_CONTEXT_KERNEL);
+	callchain_store(entry, regs->nip);
+
+	if (!validate_sp(sp, current, STACK_FRAME_OVERHEAD))
+		return;
+
+	for (;;) {
+		fp = (unsigned long *) sp;
+		next_sp = fp[0];
+
+		if (next_sp == sp + STACK_INT_FRAME_SIZE &&
+		    fp[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) {
+			/*
+			 * This looks like an interrupt frame for an
+			 * interrupt that occurred in the kernel
+			 */
+			regs = (struct pt_regs *)(sp + STACK_FRAME_OVERHEAD);
+			next_ip = regs->nip;
+			lr = regs->link;
+			level = 0;
+			callchain_store(entry, PERF_CONTEXT_KERNEL);
+
+		} else {
+			if (level == 0)
+				next_ip = lr;
+			else
+				next_ip = fp[STACK_FRAME_LR_SAVE];
+
+			/*
+			 * We can't tell which of the first two addresses
+			 * we get are valid, but we can filter out the
+			 * obviously bogus ones here.  We replace them
+			 * with 0 rather than removing them entirely so
+			 * that userspace can tell which is which.
+			 */
+			if ((level == 1 && next_ip == lr) ||
+			    (level <= 1 && !kernel_text_address(next_ip)))
+				next_ip = 0;
+
+			++level;
+		}
+
+		callchain_store(entry, next_ip);
+		if (!valid_next_sp(next_sp, sp))
+			return;
+		sp = next_sp;
+	}
+}
+
+#ifdef CONFIG_PPC64
+
+#ifdef CONFIG_HUGETLB_PAGE
+#define is_huge_psize(pagesize)	(HPAGE_SHIFT && mmu_huge_psizes[pagesize])
+#else
+#define is_huge_psize(pagesize)	0
+#endif
+
+/*
+ * On 64-bit we don't want to invoke hash_page on user addresses from
+ * interrupt context, so if the access faults, we read the page tables
+ * to find which page (if any) is mapped and access it directly.
+ */
+static int read_user_stack_slow(void __user *ptr, void *ret, int nb)
+{
+	pgd_t *pgdir;
+	pte_t *ptep, pte;
+	int pagesize;
+	unsigned long addr = (unsigned long) ptr;
+	unsigned long offset;
+	unsigned long pfn;
+	void *kaddr;
+
+	pgdir = current->mm->pgd;
+	if (!pgdir)
+		return -EFAULT;
+
+	pagesize = get_slice_psize(current->mm, addr);
+
+	/* align address to page boundary */
+	offset = addr & ((1ul << mmu_psize_defs[pagesize].shift) - 1);
+	addr -= offset;
+
+	if (is_huge_psize(pagesize))
+		ptep = huge_pte_offset(current->mm, addr);
+	else
+		ptep = find_linux_pte(pgdir, addr);
+
+	if (ptep == NULL)
+		return -EFAULT;
+	pte = *ptep;
+	if (!pte_present(pte) || !(pte_val(pte) & _PAGE_USER))
+		return -EFAULT;
+	pfn = pte_pfn(pte);
+	if (!page_is_ram(pfn))
+		return -EFAULT;
+
+	/* no highmem to worry about here */
+	kaddr = pfn_to_kaddr(pfn);
+	memcpy(ret, kaddr + offset, nb);
+	return 0;
+}
+
+static int read_user_stack_64(unsigned long __user *ptr, unsigned long *ret)
+{
+	if ((unsigned long)ptr > TASK_SIZE - sizeof(unsigned long) ||
+	    ((unsigned long)ptr & 7))
+		return -EFAULT;
+
+	if (!__get_user_inatomic(*ret, ptr))
+		return 0;
+
+	return read_user_stack_slow(ptr, ret, 8);
+}
+
+static int read_user_stack_32(unsigned int __user *ptr, unsigned int *ret)
+{
+	if ((unsigned long)ptr > TASK_SIZE - sizeof(unsigned int) ||
+	    ((unsigned long)ptr & 3))
+		return -EFAULT;
+
+	if (!__get_user_inatomic(*ret, ptr))
+		return 0;
+
+	return read_user_stack_slow(ptr, ret, 4);
+}
+
+static inline int valid_user_sp(unsigned long sp, int is_64)
+{
+	if (!sp || (sp & 7) || sp > (is_64 ? TASK_SIZE : 0x100000000UL) - 32)
+		return 0;
+	return 1;
+}
+
+/*
+ * 64-bit user processes use the same stack frame for RT and non-RT signals.
+ */
+struct signal_frame_64 {
+	char		dummy[__SIGNAL_FRAMESIZE];
+	struct ucontext	uc;
+	unsigned long	unused[2];
+	unsigned int	tramp[6];
+	struct siginfo	*pinfo;
+	void		*puc;
+	struct siginfo	info;
+	char		abigap[288];
+};
+
+static int is_sigreturn_64_address(unsigned long nip, unsigned long fp)
+{
+	if (nip == fp + offsetof(struct signal_frame_64, tramp))
+		return 1;
+	if (vdso64_rt_sigtramp && current->mm->context.vdso_base &&
+	    nip == current->mm->context.vdso_base + vdso64_rt_sigtramp)
+		return 1;
+	return 0;
+}
+
+/*
+ * Do some sanity checking on the signal frame pointed to by sp.
+ * We check the pinfo and puc pointers in the frame.
+ */
+static int sane_signal_64_frame(unsigned long sp)
+{
+	struct signal_frame_64 __user *sf;
+	unsigned long pinfo, puc;
+
+	sf = (struct signal_frame_64 __user *) sp;
+	if (read_user_stack_64((unsigned long __user *) &sf->pinfo, &pinfo) ||
+	    read_user_stack_64((unsigned long __user *) &sf->puc, &puc))
+		return 0;
+	return pinfo == (unsigned long) &sf->info &&
+		puc == (unsigned long) &sf->uc;
+}
+
+static void perf_callchain_user_64(struct pt_regs *regs,
+				   struct perf_callchain_entry *entry)
+{
+	unsigned long sp, next_sp;
+	unsigned long next_ip;
+	unsigned long lr;
+	long level = 0;
+	struct signal_frame_64 __user *sigframe;
+	unsigned long __user *fp, *uregs;
+
+	next_ip = regs->nip;
+	lr = regs->link;
+	sp = regs->gpr[1];
+	callchain_store(entry, PERF_CONTEXT_USER);
+	callchain_store(entry, next_ip);
+
+	for (;;) {
+		fp = (unsigned long __user *) sp;
+		if (!valid_user_sp(sp, 1) || read_user_stack_64(fp, &next_sp))
+			return;
+		if (level > 0 && read_user_stack_64(&fp[2], &next_ip))
+			return;
+
+		/*
+		 * Note: the next_sp - sp >= signal frame size check
+		 * is true when next_sp < sp, which can happen when
+		 * transitioning from an alternate signal stack to the
+		 * normal stack.
+		 */
+		if (next_sp - sp >= sizeof(struct signal_frame_64) &&
+		    (is_sigreturn_64_address(next_ip, sp) ||
+		     (level <= 1 && is_sigreturn_64_address(lr, sp))) &&
+		    sane_signal_64_frame(sp)) {
+			/*
+			 * This looks like an signal frame
+			 */
+			sigframe = (struct signal_frame_64 __user *) sp;
+			uregs = sigframe->uc.uc_mcontext.gp_regs;
+			if (read_user_stack_64(&uregs[PT_NIP], &next_ip) ||
+			    read_user_stack_64(&uregs[PT_LNK], &lr) ||
+			    read_user_stack_64(&uregs[PT_R1], &sp))
+				return;
+			level = 0;
+			callchain_store(entry, PERF_CONTEXT_USER);
+			callchain_store(entry, next_ip);
+			continue;
+		}
+
+		if (level == 0)
+			next_ip = lr;
+		callchain_store(entry, next_ip);
+		++level;
+		sp = next_sp;
+	}
+}
+
+static inline int current_is_64bit(void)
+{
+	/*
+	 * We can't use test_thread_flag() here because we may be on an
+	 * interrupt stack, and the thread flags don't get copied over
+	 * from the thread_info on the main stack to the interrupt stack.
+	 */
+	return !test_ti_thread_flag(task_thread_info(current), TIF_32BIT);
+}
+
+#else  /* CONFIG_PPC64 */
+/*
+ * On 32-bit we just access the address and let hash_page create a
+ * HPTE if necessary, so there is no need to fall back to reading
+ * the page tables.  Since this is called at interrupt level,
+ * do_page_fault() won't treat a DSI as a page fault.
+ */
+static int read_user_stack_32(unsigned int __user *ptr, unsigned int *ret)
+{
+	if ((unsigned long)ptr > TASK_SIZE - sizeof(unsigned int) ||
+	    ((unsigned long)ptr & 3))
+		return -EFAULT;
+
+	return __get_user_inatomic(*ret, ptr);
+}
+
+static inline void perf_callchain_user_64(struct pt_regs *regs,
+					  struct perf_callchain_entry *entry)
+{
+}
+
+static inline int current_is_64bit(void)
+{
+	return 0;
+}
+
+static inline int valid_user_sp(unsigned long sp, int is_64)
+{
+	if (!sp || (sp & 7) || sp > TASK_SIZE - 32)
+		return 0;
+	return 1;
+}
+
+#define __SIGNAL_FRAMESIZE32	__SIGNAL_FRAMESIZE
+#define sigcontext32		sigcontext
+#define mcontext32		mcontext
+#define ucontext32		ucontext
+#define compat_siginfo_t	struct siginfo
+
+#endif /* CONFIG_PPC64 */
+
+/*
+ * Layout for non-RT signal frames
+ */
+struct signal_frame_32 {
+	char			dummy[__SIGNAL_FRAMESIZE32];
+	struct sigcontext32	sctx;
+	struct mcontext32	mctx;
+	int			abigap[56];
+};
+
+/*
+ * Layout for RT signal frames
+ */
+struct rt_signal_frame_32 {
+	char			dummy[__SIGNAL_FRAMESIZE32 + 16];
+	compat_siginfo_t	info;
+	struct ucontext32	uc;
+	int			abigap[56];
+};
+
+static int is_sigreturn_32_address(unsigned int nip, unsigned int fp)
+{
+	if (nip == fp + offsetof(struct signal_frame_32, mctx.mc_pad))
+		return 1;
+	if (vdso32_sigtramp && current->mm->context.vdso_base &&
+	    nip == current->mm->context.vdso_base + vdso32_sigtramp)
+		return 1;
+	return 0;
+}
+
+static int is_rt_sigreturn_32_address(unsigned int nip, unsigned int fp)
+{
+	if (nip == fp + offsetof(struct rt_signal_frame_32,
+				 uc.uc_mcontext.mc_pad))
+		return 1;
+	if (vdso32_rt_sigtramp && current->mm->context.vdso_base &&
+	    nip == current->mm->context.vdso_base + vdso32_rt_sigtramp)
+		return 1;
+	return 0;
+}
+
+static int sane_signal_32_frame(unsigned int sp)
+{
+	struct signal_frame_32 __user *sf;
+	unsigned int regs;
+
+	sf = (struct signal_frame_32 __user *) (unsigned long) sp;
+	if (read_user_stack_32((unsigned int __user *) &sf->sctx.regs, &regs))
+		return 0;
+	return regs == (unsigned long) &sf->mctx;
+}
+
+static int sane_rt_signal_32_frame(unsigned int sp)
+{
+	struct rt_signal_frame_32 __user *sf;
+	unsigned int regs;
+
+	sf = (struct rt_signal_frame_32 __user *) (unsigned long) sp;
+	if (read_user_stack_32((unsigned int __user *) &sf->uc.uc_regs, &regs))
+		return 0;
+	return regs == (unsigned long) &sf->uc.uc_mcontext;
+}
+
+static unsigned int __user *signal_frame_32_regs(unsigned int sp,
+				unsigned int next_sp, unsigned int next_ip)
+{
+	struct mcontext32 __user *mctx = NULL;
+	struct signal_frame_32 __user *sf;
+	struct rt_signal_frame_32 __user *rt_sf;
+
+	/*
+	 * Note: the next_sp - sp >= signal frame size check
+	 * is true when next_sp < sp, for example, when
+	 * transitioning from an alternate signal stack to the
+	 * normal stack.
+	 */
+	if (next_sp - sp >= sizeof(struct signal_frame_32) &&
+	    is_sigreturn_32_address(next_ip, sp) &&
+	    sane_signal_32_frame(sp)) {
+		sf = (struct signal_frame_32 __user *) (unsigned long) sp;
+		mctx = &sf->mctx;
+	}
+
+	if (!mctx && next_sp - sp >= sizeof(struct rt_signal_frame_32) &&
+	    is_rt_sigreturn_32_address(next_ip, sp) &&
+	    sane_rt_signal_32_frame(sp)) {
+		rt_sf = (struct rt_signal_frame_32 __user *) (unsigned long) sp;
+		mctx = &rt_sf->uc.uc_mcontext;
+	}
+
+	if (!mctx)
+		return NULL;
+	return mctx->mc_gregs;
+}
+
+static void perf_callchain_user_32(struct pt_regs *regs,
+				   struct perf_callchain_entry *entry)
+{
+	unsigned int sp, next_sp;
+	unsigned int next_ip;
+	unsigned int lr;
+	long level = 0;
+	unsigned int __user *fp, *uregs;
+
+	next_ip = regs->nip;
+	lr = regs->link;
+	sp = regs->gpr[1];
+	callchain_store(entry, PERF_CONTEXT_USER);
+	callchain_store(entry, next_ip);
+
+	while (entry->nr < PERF_MAX_STACK_DEPTH) {
+		fp = (unsigned int __user *) (unsigned long) sp;
+		if (!valid_user_sp(sp, 0) || read_user_stack_32(fp, &next_sp))
+			return;
+		if (level > 0 && read_user_stack_32(&fp[1], &next_ip))
+			return;
+
+		uregs = signal_frame_32_regs(sp, next_sp, next_ip);
+		if (!uregs && level <= 1)
+			uregs = signal_frame_32_regs(sp, next_sp, lr);
+		if (uregs) {
+			/*
+			 * This looks like an signal frame, so restart
+			 * the stack trace with the values in it.
+			 */
+			if (read_user_stack_32(&uregs[PT_NIP], &next_ip) ||
+			    read_user_stack_32(&uregs[PT_LNK], &lr) ||
+			    read_user_stack_32(&uregs[PT_R1], &sp))
+				return;
+			level = 0;
+			callchain_store(entry, PERF_CONTEXT_USER);
+			callchain_store(entry, next_ip);
+			continue;
+		}
+
+		if (level == 0)
+			next_ip = lr;
+		callchain_store(entry, next_ip);
+		++level;
+		sp = next_sp;
+	}
+}
+
+/*
+ * Since we can't get PMU interrupts inside a PMU interrupt handler,
+ * we don't need separate irq and nmi entries here.
+ */
+static DEFINE_PER_CPU(struct perf_callchain_entry, callchain);
+
+struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
+{
+	struct perf_callchain_entry *entry = &__get_cpu_var(callchain);
+
+	entry->nr = 0;
+
+	if (current->pid == 0)		/* idle task? */
+		return entry;
+
+	if (!user_mode(regs)) {
+		perf_callchain_kernel(regs, entry);
+		if (current->mm)
+			regs = task_pt_regs(current);
+		else
+			regs = NULL;
+	}
+
+	if (regs) {
+		if (current_is_64bit())
+			perf_callchain_user_64(regs, entry);
+		else
+			perf_callchain_user_32(regs, entry);
+	}
+
+	return entry;
+}

arch/powerpc/mm/slb.c

@@ -92,15 +92,13 @@ static inline void create_shadowed_slbe(unsigned long ea, int ssize,
 		     : "memory" );
 }
 
-void slb_flush_and_rebolt(void)
+static void __slb_flush_and_rebolt(void)
 {
 	/* If you change this make sure you change SLB_NUM_BOLTED
 	 * appropriately too. */
 	unsigned long linear_llp, vmalloc_llp, lflags, vflags;
 	unsigned long ksp_esid_data, ksp_vsid_data;
 
-	WARN_ON(!irqs_disabled());
-
 	linear_llp = mmu_psize_defs[mmu_linear_psize].sllp;
 	vmalloc_llp = mmu_psize_defs[mmu_vmalloc_psize].sllp;
 	lflags = SLB_VSID_KERNEL | linear_llp;
@@ -117,12 +115,6 @@ void slb_flush_and_rebolt(void)
 		ksp_vsid_data = get_slb_shadow()->save_area[2].vsid;
 	}
 
-	/*
-	 * We can't take a PMU exception in the following code, so hard
-	 * disable interrupts.
-	 */
-	hard_irq_disable();
-
 	/* We need to do this all in asm, so we're sure we don't touch
 	 * the stack between the slbia and rebolting it. */
 	asm volatile("isync\n"
@@ -139,6 +131,21 @@ void slb_flush_and_rebolt(void)
 		     : "memory");
 }
 
+void slb_flush_and_rebolt(void)
+{
+
+	WARN_ON(!irqs_disabled());
+
+	/*
+	 * We can't take a PMU exception in the following code, so hard
+	 * disable interrupts.
+	 */
+	hard_irq_disable();
+
+	__slb_flush_and_rebolt();
+	get_paca()->slb_cache_ptr = 0;
+}
+
 void slb_vmalloc_update(void)
 {
 	unsigned long vflags;
@@ -180,12 +187,20 @@ static inline int esids_match(unsigned long addr1, unsigned long addr2)
 /* Flush all user entries from the segment table of the current processor. */
 void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
 {
-	unsigned long offset = get_paca()->slb_cache_ptr;
+	unsigned long offset;
 	unsigned long slbie_data = 0;
 	unsigned long pc = KSTK_EIP(tsk);
 	unsigned long stack = KSTK_ESP(tsk);
 	unsigned long unmapped_base;
 
+	/*
+	 * We need interrupts hard-disabled here, not just soft-disabled,
+	 * so that a PMU interrupt can't occur, which might try to access
+	 * user memory (to get a stack trace) and possible cause an SLB miss
+	 * which would update the slb_cache/slb_cache_ptr fields in the PACA.
+	 */
+	hard_irq_disable();
+	offset = get_paca()->slb_cache_ptr;
 	if (!cpu_has_feature(CPU_FTR_NO_SLBIE_B) &&
 	    offset <= SLB_CACHE_ENTRIES) {
 		int i;
@@ -200,7 +215,7 @@ void switch_slb(struct task_struct *tsk, struct mm_struct *mm)
 		}
 		asm volatile("isync" : : : "memory");
 	} else {
-		slb_flush_and_rebolt();
+		__slb_flush_and_rebolt();
 	}
 
 	/* Workaround POWER5 < DD2.1 issue */

arch/powerpc/mm/stab.c

@@ -164,7 +164,7 @@ void switch_stab(struct task_struct *tsk, struct mm_struct *mm)
 {
 	struct stab_entry *stab = (struct stab_entry *) get_paca()->stab_addr;
 	struct stab_entry *ste;
-	unsigned long offset = __get_cpu_var(stab_cache_ptr);
+	unsigned long offset;
 	unsigned long pc = KSTK_EIP(tsk);
 	unsigned long stack = KSTK_ESP(tsk);
 	unsigned long unmapped_base;
@@ -172,6 +172,15 @@ void switch_stab(struct task_struct *tsk, struct mm_struct *mm)
 	/* Force previous translations to complete. DRENG */
 	asm volatile("isync" : : : "memory");
 
+	/*
+	 * We need interrupts hard-disabled here, not just soft-disabled,
+	 * so that a PMU interrupt can't occur, which might try to access
+	 * user memory (to get a stack trace) and possible cause an STAB miss
+	 * which would update the stab_cache/stab_cache_ptr per-cpu variables.
+	 */
+	hard_irq_disable();
+
+	offset = __get_cpu_var(stab_cache_ptr);
 	if (offset <= NR_STAB_CACHE_ENTRIES) {
 		int i;
 

arch/x86/include/asm/perf_counter.h

@@ -84,6 +84,16 @@ union cpuid10_edx {
 #define MSR_ARCH_PERFMON_FIXED_CTR2			0x30b
 #define X86_PMC_IDX_FIXED_BUS_CYCLES			(X86_PMC_IDX_FIXED + 2)
 
+/*
+ * We model BTS tracing as another fixed-mode PMC.
+ *
+ * We choose a value in the middle of the fixed counter range, since lower
+ * values are used by actual fixed counters and higher values are used
+ * to indicate other overflow conditions in the PERF_GLOBAL_STATUS msr.
+ */
+#define X86_PMC_IDX_FIXED_BTS				(X86_PMC_IDX_FIXED + 16)
+
+
 #ifdef CONFIG_PERF_COUNTERS
 extern void init_hw_perf_counters(void);
 extern void perf_counters_lapic_init(void);

arch/x86/kernel/cpu/perf_counter.c

@@ -6,6 +6,7 @@
  *  Copyright (C) 2009 Jaswinder Singh Rajput
  *  Copyright (C) 2009 Advanced Micro Devices, Inc., Robert Richter
  *  Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
+ *  Copyright (C) 2009 Intel Corporation, <markus.t.metzger@intel.com>
  *
  *  For licencing details see kernel-base/COPYING
  */
@@ -20,6 +21,7 @@
 #include <linux/sched.h>
 #include <linux/uaccess.h>
 #include <linux/highmem.h>
+#include <linux/cpu.h>
 
 #include <asm/apic.h>
 #include <asm/stacktrace.h>
@@ -27,12 +29,52 @@
 
 static u64 perf_counter_mask __read_mostly;
 
+/* The maximal number of PEBS counters: */
+#define MAX_PEBS_COUNTERS	4
+
+/* The size of a BTS record in bytes: */
+#define BTS_RECORD_SIZE		24
+
+/* The size of a per-cpu BTS buffer in bytes: */
+#define BTS_BUFFER_SIZE		(BTS_RECORD_SIZE * 1024)
+
+/* The BTS overflow threshold in bytes from the end of the buffer: */
+#define BTS_OVFL_TH		(BTS_RECORD_SIZE * 64)
+
+
+/*
+ * Bits in the debugctlmsr controlling branch tracing.
+ */
+#define X86_DEBUGCTL_TR			(1 << 6)
+#define X86_DEBUGCTL_BTS		(1 << 7)
+#define X86_DEBUGCTL_BTINT		(1 << 8)
+#define X86_DEBUGCTL_BTS_OFF_OS		(1 << 9)
+#define X86_DEBUGCTL_BTS_OFF_USR	(1 << 10)
+
+/*
+ * A debug store configuration.
+ *
+ * We only support architectures that use 64bit fields.
+ */
+struct debug_store {
+	u64	bts_buffer_base;
+	u64	bts_index;
+	u64	bts_absolute_maximum;
+	u64	bts_interrupt_threshold;
+	u64	pebs_buffer_base;
+	u64	pebs_index;
+	u64	pebs_absolute_maximum;
+	u64	pebs_interrupt_threshold;
+	u64	pebs_counter_reset[MAX_PEBS_COUNTERS];
+};
+
 struct cpu_hw_counters {
 	struct perf_counter	*counters[X86_PMC_IDX_MAX];
 	unsigned long		used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
 	unsigned long		active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)];
 	unsigned long		interrupts;
 	int			enabled;
+	struct debug_store	*ds;
 };
 
 /*
@@ -58,6 +100,8 @@ struct x86_pmu {
 	int		apic;
 	u64		max_period;
 	u64		intel_ctrl;
+	void		(*enable_bts)(u64 config);
+	void		(*disable_bts)(void);
 };
 
 static struct x86_pmu x86_pmu __read_mostly;
@@ -577,6 +621,9 @@ x86_perf_counter_update(struct perf_counter *counter,
 	u64 prev_raw_count, new_raw_count;
 	s64 delta;
 
+	if (idx == X86_PMC_IDX_FIXED_BTS)
+		return 0;
+
 	/*
 	 * Careful: an NMI might modify the previous counter value.
 	 *
@@ -666,10 +713,110 @@ static void release_pmc_hardware(void)
 #endif
 }
 
+static inline bool bts_available(void)
+{
+	return x86_pmu.enable_bts != NULL;
+}
+
+static inline void init_debug_store_on_cpu(int cpu)
+{
+	struct debug_store *ds = per_cpu(cpu_hw_counters, cpu).ds;
+
+	if (!ds)
+		return;
+
+	wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA,
+		     (u32)((u64)(unsigned long)ds),
+		     (u32)((u64)(unsigned long)ds >> 32));
+}
+
+static inline void fini_debug_store_on_cpu(int cpu)
+{
+	if (!per_cpu(cpu_hw_counters, cpu).ds)
+		return;
+
+	wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, 0, 0);
+}
+
+static void release_bts_hardware(void)
+{
+	int cpu;
+
+	if (!bts_available())
+		return;
+
+	get_online_cpus();
+
+	for_each_online_cpu(cpu)
+		fini_debug_store_on_cpu(cpu);
+
+	for_each_possible_cpu(cpu) {
+		struct debug_store *ds = per_cpu(cpu_hw_counters, cpu).ds;
+
+		if (!ds)
+			continue;
+
+		per_cpu(cpu_hw_counters, cpu).ds = NULL;
+
+		kfree((void *)(unsigned long)ds->bts_buffer_base);
+		kfree(ds);
+	}
+
+	put_online_cpus();
+}
+
+static int reserve_bts_hardware(void)
+{
+	int cpu, err = 0;
+
+	if (!bts_available())
+		return 0;
+
+	get_online_cpus();
+
+	for_each_possible_cpu(cpu) {
+		struct debug_store *ds;
+		void *buffer;
+
+		err = -ENOMEM;
+		buffer = kzalloc(BTS_BUFFER_SIZE, GFP_KERNEL);
+		if (unlikely(!buffer))
+			break;
+
+		ds = kzalloc(sizeof(*ds), GFP_KERNEL);
+		if (unlikely(!ds)) {
+			kfree(buffer);
+			break;
+		}
+
+		ds->bts_buffer_base = (u64)(unsigned long)buffer;
+		ds->bts_index = ds->bts_buffer_base;
+		ds->bts_absolute_maximum =
+			ds->bts_buffer_base + BTS_BUFFER_SIZE;
+		ds->bts_interrupt_threshold =
+			ds->bts_absolute_maximum - BTS_OVFL_TH;
+
+		per_cpu(cpu_hw_counters, cpu).ds = ds;
+		err = 0;
+	}
+
+	if (err)
+		release_bts_hardware();
+	else {
+		for_each_online_cpu(cpu)
+			init_debug_store_on_cpu(cpu);
+	}
+
+	put_online_cpus();
+
+	return err;
+}
+
 static void hw_perf_counter_destroy(struct perf_counter *counter)
 {
 	if (atomic_dec_and_mutex_lock(&active_counters, &pmc_reserve_mutex)) {
 		release_pmc_hardware();
+		release_bts_hardware();
 		mutex_unlock(&pmc_reserve_mutex);
 	}
 }
@@ -712,6 +859,42 @@ set_ext_hw_attr(struct hw_perf_counter *hwc, struct perf_counter_attr *attr)
 	return 0;
 }
 
+static void intel_pmu_enable_bts(u64 config)
+{
+	unsigned long debugctlmsr;
+
+	debugctlmsr = get_debugctlmsr();
+
+	debugctlmsr |= X86_DEBUGCTL_TR;
+	debugctlmsr |= X86_DEBUGCTL_BTS;
+	debugctlmsr |= X86_DEBUGCTL_BTINT;
+
+	if (!(config & ARCH_PERFMON_EVENTSEL_OS))
+		debugctlmsr |= X86_DEBUGCTL_BTS_OFF_OS;
+
+	if (!(config & ARCH_PERFMON_EVENTSEL_USR))
+		debugctlmsr |= X86_DEBUGCTL_BTS_OFF_USR;
+
+	update_debugctlmsr(debugctlmsr);
+}
+
+static void intel_pmu_disable_bts(void)
+{
+	struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters);
+	unsigned long debugctlmsr;
+
+	if (!cpuc->ds)
+		return;
+
+	debugctlmsr = get_debugctlmsr();
+
+	debugctlmsr &=
+		~(X86_DEBUGCTL_TR | X86_DEBUGCTL_BTS | X86_DEBUGCTL_BTINT |
+		  X86_DEBUGCTL_BTS_OFF_OS | X86_DEBUGCTL_BTS_OFF_USR);
+
+	update_debugctlmsr(debugctlmsr);
+}
+
 /*
  * Setup the hardware configuration for a given attr_type
  */
@@ -728,9 +911,13 @@ static int __hw_perf_counter_init(struct perf_counter *counter)
 	err = 0;
 	if (!atomic_inc_not_zero(&active_counters)) {
 		mutex_lock(&pmc_reserve_mutex);
-		if (atomic_read(&active_counters) == 0 && !reserve_pmc_hardware())
-			err = -EBUSY;
-		else
+		if (atomic_read(&active_counters) == 0) {
+			if (!reserve_pmc_hardware())
+				err = -EBUSY;
+			else
+				err = reserve_bts_hardware();
+		}
+		if (!err)
 			atomic_inc(&active_counters);
 		mutex_unlock(&pmc_reserve_mutex);
 	}
@@ -793,6 +980,20 @@ static int __hw_perf_counter_init(struct perf_counter *counter)
 	if (config == -1LL)
 		return -EINVAL;
 
+	/*
+	 * Branch tracing:
+	 */
+	if ((attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS) &&
+	    (hwc->sample_period == 1)) {
+		/* BTS is not supported by this architecture. */
+		if (!bts_available())
+			return -EOPNOTSUPP;
+
+		/* BTS is currently only allowed for user-mode. */
+		if (hwc->config & ARCH_PERFMON_EVENTSEL_OS)
+			return -EOPNOTSUPP;
+	}
+
 	hwc->config |= config;
 
 	return 0;
@@ -817,7 +1018,18 @@ static void p6_pmu_disable_all(void)
 
 static void intel_pmu_disable_all(void)
 {
+	struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters);
+
+	if (!cpuc->enabled)
+		return;
+
+	cpuc->enabled = 0;
+	barrier();
+
 	wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0);
+
+	if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask))
+		intel_pmu_disable_bts();
 }
 
 static void amd_pmu_disable_all(void)
@@ -875,7 +1087,25 @@ static void p6_pmu_enable_all(void)
 
 static void intel_pmu_enable_all(void)
 {
+	struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters);
+
+	if (cpuc->enabled)
+		return;
+
+	cpuc->enabled = 1;
+	barrier();
+
 	wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl);
+
+	if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask)) {
+		struct perf_counter *counter =
+			cpuc->counters[X86_PMC_IDX_FIXED_BTS];
+
+		if (WARN_ON_ONCE(!counter))
+			return;
+
+		intel_pmu_enable_bts(counter->hw.config);
+	}
 }
 
 static void amd_pmu_enable_all(void)
@@ -962,6 +1192,11 @@ p6_pmu_disable_counter(struct hw_perf_counter *hwc, int idx)
 static inline void
 intel_pmu_disable_counter(struct hw_perf_counter *hwc, int idx)
 {
+	if (unlikely(idx == X86_PMC_IDX_FIXED_BTS)) {
+		intel_pmu_disable_bts();
+		return;
+	}
+
 	if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) {
 		intel_pmu_disable_fixed(hwc, idx);
 		return;
@@ -990,6 +1225,9 @@ x86_perf_counter_set_period(struct perf_counter *counter,
 	s64 period = hwc->sample_period;
 	int err, ret = 0;
 
+	if (idx == X86_PMC_IDX_FIXED_BTS)
+		return 0;
+
 	/*
 	 * If we are way outside a reasoable range then just skip forward:
 	 */
@@ -1072,6 +1310,14 @@ static void p6_pmu_enable_counter(struct hw_perf_counter *hwc, int idx)
 
 static void intel_pmu_enable_counter(struct hw_perf_counter *hwc, int idx)
 {
+	if (unlikely(idx == X86_PMC_IDX_FIXED_BTS)) {
+		if (!__get_cpu_var(cpu_hw_counters).enabled)
+			return;
+
+		intel_pmu_enable_bts(hwc->config);
+		return;
+	}
+
 	if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) {
 		intel_pmu_enable_fixed(hwc, idx);
 		return;
@@ -1093,11 +1339,16 @@ fixed_mode_idx(struct perf_counter *counter, struct hw_perf_counter *hwc)
 {
 	unsigned int event;
 
+	event = hwc->config & ARCH_PERFMON_EVENT_MASK;
+
+	if (unlikely((event ==
+		      x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS)) &&
+		     (hwc->sample_period == 1)))
+		return X86_PMC_IDX_FIXED_BTS;
+
 	if (!x86_pmu.num_counters_fixed)
 		return -1;
 
-	event = hwc->config & ARCH_PERFMON_EVENT_MASK;
-
 	if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_INSTRUCTIONS)))
 		return X86_PMC_IDX_FIXED_INSTRUCTIONS;
 	if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_CPU_CYCLES)))
@@ -1118,7 +1369,15 @@ static int x86_pmu_enable(struct perf_counter *counter)
 	int idx;
 
 	idx = fixed_mode_idx(counter, hwc);
-	if (idx >= 0) {
+	if (idx == X86_PMC_IDX_FIXED_BTS) {
+		/* BTS is already occupied. */
+		if (test_and_set_bit(idx, cpuc->used_mask))
+			return -EAGAIN;
+
+		hwc->config_base	= 0;
+		hwc->counter_base	= 0;
+		hwc->idx		= idx;
+	} else if (idx >= 0) {
 		/*
 		 * Try to get the fixed counter, if that is already taken
 		 * then try to get a generic counter:
@@ -1229,6 +1488,44 @@ void perf_counter_print_debug(void)
 	local_irq_restore(flags);
 }
 
+static void intel_pmu_drain_bts_buffer(struct cpu_hw_counters *cpuc,
+				       struct perf_sample_data *data)
+{
+	struct debug_store *ds = cpuc->ds;
+	struct bts_record {
+		u64	from;
+		u64	to;
+		u64	flags;
+	};
+	struct perf_counter *counter = cpuc->counters[X86_PMC_IDX_FIXED_BTS];
+	unsigned long orig_ip = data->regs->ip;
+	struct bts_record *at, *top;
+
+	if (!counter)
+		return;
+
+	if (!ds)
+		return;
+
+	at  = (struct bts_record *)(unsigned long)ds->bts_buffer_base;
+	top = (struct bts_record *)(unsigned long)ds->bts_index;
+
+	ds->bts_index = ds->bts_buffer_base;
+
+	for (; at < top; at++) {
+		data->regs->ip	= at->from;
+		data->addr	= at->to;
+
+		perf_counter_output(counter, 1, data);
+	}
+
+	data->regs->ip	= orig_ip;
+	data->addr	= 0;
+
+	/* There's new data available. */
+	counter->pending_kill = POLL_IN;
+}
+
 static void x86_pmu_disable(struct perf_counter *counter)
 {
 	struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters);
@@ -1253,6 +1550,15 @@ static void x86_pmu_disable(struct perf_counter *counter)
 	 * that we are disabling:
 	 */
 	x86_perf_counter_update(counter, hwc, idx);
+
+	/* Drain the remaining BTS records. */
+	if (unlikely(idx == X86_PMC_IDX_FIXED_BTS)) {
+		struct perf_sample_data data;
+		struct pt_regs regs;
+
+		data.regs = &regs;
+		intel_pmu_drain_bts_buffer(cpuc, &data);
+	}
 	cpuc->counters[idx] = NULL;
 	clear_bit(idx, cpuc->used_mask);
 
@@ -1280,6 +1586,7 @@ static int intel_pmu_save_and_restart(struct perf_counter *counter)
 
 static void intel_pmu_reset(void)
 {
+	struct debug_store *ds = __get_cpu_var(cpu_hw_counters).ds;
 	unsigned long flags;
 	int idx;
 
@@ -1297,6 +1604,8 @@ static void intel_pmu_reset(void)
 	for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) {
 		checking_wrmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, 0ull);
 	}
+	if (ds)
+		ds->bts_index = ds->bts_buffer_base;
 
 	local_irq_restore(flags);
 }
@@ -1362,6 +1671,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs)
 	cpuc = &__get_cpu_var(cpu_hw_counters);
 
 	perf_disable();
+	intel_pmu_drain_bts_buffer(cpuc, &data);
 	status = intel_pmu_get_status();
 	if (!status) {
 		perf_enable();
@@ -1571,6 +1881,8 @@ static struct x86_pmu intel_pmu = {
 	 * the generic counter period:
 	 */
 	.max_period		= (1ULL << 31) - 1,
+	.enable_bts		= intel_pmu_enable_bts,
+	.disable_bts		= intel_pmu_disable_bts,
 };
 
 static struct x86_pmu amd_pmu = {
@@ -1962,3 +2274,8 @@ struct perf_callchain_entry *perf_callchain(struct pt_regs *regs)
 
 	return entry;
 }
+
+void hw_perf_counter_setup_online(int cpu)
+{
+	init_debug_store_on_cpu(cpu);
+}

include/linux/perf_counter.h

@@ -216,6 +216,7 @@ struct perf_counter_attr {
 #define PERF_COUNTER_IOC_REFRESH	_IO ('$', 2)
 #define PERF_COUNTER_IOC_RESET		_IO ('$', 3)
 #define PERF_COUNTER_IOC_PERIOD		_IOW('$', 4, u64)
+#define PERF_COUNTER_IOC_SET_OUTPUT	_IO ('$', 5)
 
 enum perf_counter_ioc_flags {
 	PERF_IOC_FLAG_GROUP		= 1U << 0,
@@ -415,6 +416,9 @@ enum perf_callchain_context {
 	PERF_CONTEXT_MAX		= (__u64)-4095,
 };
 
+#define PERF_FLAG_FD_NO_GROUP	(1U << 0)
+#define PERF_FLAG_FD_OUTPUT	(1U << 1)
+
 #ifdef __KERNEL__
 /*
  * Kernel-internal data types and definitions:
@@ -536,6 +540,7 @@ struct perf_counter {
 	struct list_head		sibling_list;
 	int				nr_siblings;
 	struct perf_counter		*group_leader;
+	struct perf_counter		*output;
 	const struct pmu		*pmu;
 
 	enum perf_counter_active_state	state;

kernel/perf_counter.c

@@ -46,12 +46,18 @@ static atomic_t nr_task_counters __read_mostly;
 
 /*
  * perf counter paranoia level:
- *  0 - not paranoid
- *  1 - disallow cpu counters to unpriv
- *  2 - disallow kernel profiling to unpriv
+ *  -1 - not paranoid at all
+ *   0 - disallow raw tracepoint access for unpriv
+ *   1 - disallow cpu counters for unpriv
+ *   2 - disallow kernel profiling for unpriv
  */
 int sysctl_perf_counter_paranoid __read_mostly = 1;
 
+static inline bool perf_paranoid_tracepoint_raw(void)
+{
+	return sysctl_perf_counter_paranoid > -1;
+}
+
 static inline bool perf_paranoid_cpu(void)
 {
 	return sysctl_perf_counter_paranoid > 0;
@@ -469,7 +475,8 @@ static void update_counter_times(struct perf_counter *counter)
 	struct perf_counter_context *ctx = counter->ctx;
 	u64 run_end;
 
-	if (counter->state < PERF_COUNTER_STATE_INACTIVE)
+	if (counter->state < PERF_COUNTER_STATE_INACTIVE ||
+	    counter->group_leader->state < PERF_COUNTER_STATE_INACTIVE)
 		return;
 
 	counter->total_time_enabled = ctx->time - counter->tstamp_enabled;
@@ -518,7 +525,7 @@ static void __perf_counter_disable(void *info)
 	 */
 	if (counter->state >= PERF_COUNTER_STATE_INACTIVE) {
 		update_context_time(ctx);
-		update_counter_times(counter);
+		update_group_times(counter);
 		if (counter == counter->group_leader)
 			group_sched_out(counter, cpuctx, ctx);
 		else
@@ -573,7 +580,7 @@ static void perf_counter_disable(struct perf_counter *counter)
 	 * in, so we can change the state safely.
 	 */
 	if (counter->state == PERF_COUNTER_STATE_INACTIVE) {
-		update_counter_times(counter);
+		update_group_times(counter);
 		counter->state = PERF_COUNTER_STATE_OFF;
 	}
 
@@ -850,6 +857,27 @@ perf_install_in_context(struct perf_counter_context *ctx,
 	spin_unlock_irq(&ctx->lock);
 }
 
+/*
+ * Put a counter into inactive state and update time fields.
+ * Enabling the leader of a group effectively enables all
+ * the group members that aren't explicitly disabled, so we
+ * have to update their ->tstamp_enabled also.
+ * Note: this works for group members as well as group leaders
+ * since the non-leader members' sibling_lists will be empty.
+ */
+static void __perf_counter_mark_enabled(struct perf_counter *counter,
+					struct perf_counter_context *ctx)
+{
+	struct perf_counter *sub;
+
+	counter->state = PERF_COUNTER_STATE_INACTIVE;
+	counter->tstamp_enabled = ctx->time - counter->total_time_enabled;
+	list_for_each_entry(sub, &counter->sibling_list, list_entry)
+		if (sub->state >= PERF_COUNTER_STATE_INACTIVE)
+			sub->tstamp_enabled =
+				ctx->time - sub->total_time_enabled;
+}
+
 /*
  * Cross CPU call to enable a performance counter
  */
@@ -877,8 +905,7 @@ static void __perf_counter_enable(void *info)
 
 	if (counter->state >= PERF_COUNTER_STATE_INACTIVE)
 		goto unlock;
-	counter->state = PERF_COUNTER_STATE_INACTIVE;
-	counter->tstamp_enabled = ctx->time - counter->total_time_enabled;
+	__perf_counter_mark_enabled(counter, ctx);
 
 	/*
 	 * If the counter is in a group and isn't the group leader,
@@ -971,11 +998,9 @@ static void perf_counter_enable(struct perf_counter *counter)
 	 * Since we have the lock this context can't be scheduled
 	 * in, so we can change the state safely.
 	 */
-	if (counter->state == PERF_COUNTER_STATE_OFF) {
-		counter->state = PERF_COUNTER_STATE_INACTIVE;
-		counter->tstamp_enabled =
-			ctx->time - counter->total_time_enabled;
-	}
+	if (counter->state == PERF_COUNTER_STATE_OFF)
+		__perf_counter_mark_enabled(counter, ctx);
+
  out:
 	spin_unlock_irq(&ctx->lock);
 }
@@ -1479,9 +1504,7 @@ static void perf_counter_enable_on_exec(struct task_struct *task)
 		counter->attr.enable_on_exec = 0;
 		if (counter->state >= PERF_COUNTER_STATE_INACTIVE)
 			continue;
-		counter->state = PERF_COUNTER_STATE_INACTIVE;
-		counter->tstamp_enabled =
-			ctx->time - counter->total_time_enabled;
+		__perf_counter_mark_enabled(counter, ctx);
 		enabled = 1;
 	}
 
@@ -1675,6 +1698,11 @@ static void free_counter(struct perf_counter *counter)
 			atomic_dec(&nr_task_counters);
 	}
 
+	if (counter->output) {
+		fput(counter->output->filp);
+		counter->output = NULL;
+	}
+
 	if (counter->destroy)
 		counter->destroy(counter);
 
@@ -1960,6 +1988,8 @@ static int perf_counter_period(struct perf_counter *counter, u64 __user *arg)
 	return ret;
 }
 
+int perf_counter_set_output(struct perf_counter *counter, int output_fd);
+
 static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
 	struct perf_counter *counter = file->private_data;
@@ -1983,6 +2013,9 @@ static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	case PERF_COUNTER_IOC_PERIOD:
 		return perf_counter_period(counter, (u64 __user *)arg);
 
+	case PERF_COUNTER_IOC_SET_OUTPUT:
+		return perf_counter_set_output(counter, arg);
+
 	default:
 		return -ENOTTY;
 	}
@@ -2253,6 +2286,11 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma)
 
 	WARN_ON_ONCE(counter->ctx->parent_ctx);
 	mutex_lock(&counter->mmap_mutex);
+	if (counter->output) {
+		ret = -EINVAL;
+		goto unlock;
+	}
+
 	if (atomic_inc_not_zero(&counter->mmap_count)) {
 		if (nr_pages != counter->data->nr_pages)
 			ret = -EINVAL;
@@ -2638,6 +2676,7 @@ static int perf_output_begin(struct perf_output_handle *handle,
 			     struct perf_counter *counter, unsigned int size,
 			     int nmi, int sample)
 {
+	struct perf_counter *output_counter;
 	struct perf_mmap_data *data;
 	unsigned int offset, head;
 	int have_lost;
@@ -2647,13 +2686,17 @@ static int perf_output_begin(struct perf_output_handle *handle,
 		u64			 lost;
 	} lost_event;
 
+	rcu_read_lock();
 	/*
 	 * For inherited counters we send all the output towards the parent.
 	 */
 	if (counter->parent)
 		counter = counter->parent;
 
-	rcu_read_lock();
+	output_counter = rcu_dereference(counter->output);
+	if (output_counter)
+		counter = output_counter;
+
 	data = rcu_dereference(counter->data);
 	if (!data)
 		goto out;
@@ -3934,6 +3977,7 @@ static const struct pmu *tp_perf_counter_init(struct perf_counter *counter)
 	 * have these.
 	 */
 	if ((counter->attr.sample_type & PERF_SAMPLE_RAW) &&
+			perf_paranoid_tracepoint_raw() &&
 			!capable(CAP_SYS_ADMIN))
 		return ERR_PTR(-EPERM);
 
@@ -4202,6 +4246,57 @@ static int perf_copy_attr(struct perf_counter_attr __user *uattr,
 	goto out;
 }
 
+int perf_counter_set_output(struct perf_counter *counter, int output_fd)
+{
+	struct perf_counter *output_counter = NULL;
+	struct file *output_file = NULL;
+	struct perf_counter *old_output;
+	int fput_needed = 0;
+	int ret = -EINVAL;
+
+	if (!output_fd)
+		goto set;
+
+	output_file = fget_light(output_fd, &fput_needed);
+	if (!output_file)
+		return -EBADF;
+
+	if (output_file->f_op != &perf_fops)
+		goto out;
+
+	output_counter = output_file->private_data;
+
+	/* Don't chain output fds */
+	if (output_counter->output)
+		goto out;
+
+	/* Don't set an output fd when we already have an output channel */
+	if (counter->data)
+		goto out;
+
+	atomic_long_inc(&output_file->f_count);
+
+set:
+	mutex_lock(&counter->mmap_mutex);
+	old_output = counter->output;
+	rcu_assign_pointer(counter->output, output_counter);
+	mutex_unlock(&counter->mmap_mutex);
+
+	if (old_output) {
+		/*
+		 * we need to make sure no existing perf_output_*()
+		 * is still referencing this counter.
+		 */
+		synchronize_rcu();
+		fput(old_output->filp);
+	}
+
+	ret = 0;
+out:
+	fput_light(output_file, fput_needed);
+	return ret;
+}
+
 /**
  * sys_perf_counter_open - open a performance counter, associate it to a task/cpu
  *
@@ -4221,15 +4316,15 @@ SYSCALL_DEFINE5(perf_counter_open,
 	struct file *group_file = NULL;
 	int fput_needed = 0;
 	int fput_needed2 = 0;
-	int ret;
+	int err;
 
 	/* for future expandability... */
-	if (flags)
+	if (flags & ~(PERF_FLAG_FD_NO_GROUP | PERF_FLAG_FD_OUTPUT))
 		return -EINVAL;
 
-	ret = perf_copy_attr(attr_uptr, &attr);
-	if (ret)
-		return ret;
+	err = perf_copy_attr(attr_uptr, &attr);
+	if (err)
+		return err;
 
 	if (!attr.exclude_kernel) {
 		if (perf_paranoid_kernel() && !capable(CAP_SYS_ADMIN))
@@ -4252,8 +4347,8 @@ SYSCALL_DEFINE5(perf_counter_open,
 	 * Look up the group leader (we will attach this counter to it):
 	 */
 	group_leader = NULL;
-	if (group_fd != -1) {
-		ret = -EINVAL;
+	if (group_fd != -1 && !(flags & PERF_FLAG_FD_NO_GROUP)) {
+		err = -EINVAL;
 		group_file = fget_light(group_fd, &fput_needed);
 		if (!group_file)
 			goto err_put_context;
@@ -4282,18 +4377,24 @@ SYSCALL_DEFINE5(perf_counter_open,
 
 	counter = perf_counter_alloc(&attr, cpu, ctx, group_leader,
 				     NULL, GFP_KERNEL);
-	ret = PTR_ERR(counter);
+	err = PTR_ERR(counter);
 	if (IS_ERR(counter))
 		goto err_put_context;
 
-	ret = anon_inode_getfd("[perf_counter]", &perf_fops, counter, 0);
-	if (ret < 0)
+	err = anon_inode_getfd("[perf_counter]", &perf_fops, counter, 0);
+	if (err < 0)
 		goto err_free_put_context;
 
-	counter_file = fget_light(ret, &fput_needed2);
+	counter_file = fget_light(err, &fput_needed2);
 	if (!counter_file)
 		goto err_free_put_context;
 
+	if (flags & PERF_FLAG_FD_OUTPUT) {
+		err = perf_counter_set_output(counter, group_fd);
+		if (err)
+			goto err_fput_free_put_context;
+	}
+
 	counter->filp = counter_file;
 	WARN_ON_ONCE(ctx->parent_ctx);
 	mutex_lock(&ctx->mutex);
@@ -4307,20 +4408,20 @@ SYSCALL_DEFINE5(perf_counter_open,
 	list_add_tail(&counter->owner_entry, &current->perf_counter_list);
 	mutex_unlock(&current->perf_counter_mutex);
 
+err_fput_free_put_context:
 	fput_light(counter_file, fput_needed2);
 
-out_fput:
-	fput_light(group_file, fput_needed);
-
-	return ret;
-
 err_free_put_context:
-	kfree(counter);
+	if (err < 0)
+		kfree(counter);
 
 err_put_context:
-	put_ctx(ctx);
+	if (err < 0)
+		put_ctx(ctx);
+
+	fput_light(group_file, fput_needed);
 
-	goto out_fput;
+	return err;
 }
 
 /*

tools/perf/Documentation/perf-record.txt

@@ -91,6 +91,10 @@ OPTIONS
 --no-samples::
 	Don't sample.
 
+-R::
+--raw-samples::
+Collect raw sample records from all opened counters (typically for tracepoint counters).
+
 SEE ALSO
 --------
 linkperf:perf-stat[1], linkperf:perf-list[1]

tools/perf/Documentation/perf-report.txt

@@ -27,6 +27,9 @@ OPTIONS
 -n
 --show-nr-samples
 	Show the number of samples for each symbol
+-T
+--threads
+	Show per-thread event counters
 -C::
 --comms=::
 	Only consider symbols in these comms. CSV that understands
@@ -48,6 +51,16 @@ OPTIONS
 	all occurances of this separator in symbol names (and other output)
 	with a '.' character, that thus it's the only non valid separator.
 
+-g [type,min]::
+--call-graph::
+        Display callchains using type and min percent threshold.
+	type can be either:
+	- flat: single column, linear exposure of callchains.
+	- graph: use a graph tree, displaying absolute overhead rates.
+	- fractal: like graph, but displays relative rates. Each branch of
+		 the tree is considered as a new profiled object. +
+	Default: fractal,0.5.
+
 SEE ALSO
 --------
 linkperf:perf-stat[1]

tools/perf/Makefile

@@ -166,7 +166,35 @@ endif
 
 # CFLAGS and LDFLAGS are for the users to override from the command line.
 
-CFLAGS = $(M64) -ggdb3 -Wall -Wextra -Wstrict-prototypes -Wmissing-declarations -Wmissing-prototypes -std=gnu99 -Wdeclaration-after-statement -Werror -O6
+#
+# Include saner warnings here, which can catch bugs:
+#
+
+EXTRA_WARNINGS := -Wcast-align
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-security
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-y2k
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wshadow
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Winit-self
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wpacked
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wredundant-decls
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstack-protector
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-aliasing=3
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-default
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-enum
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wno-system-headers
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wundef
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wvolatile-register-var
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wwrite-strings
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wbad-function-cast
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-declarations
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-prototypes
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wnested-externs
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes
+EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement
+
+CFLAGS = $(M64) -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS)
 LDFLAGS = -lpthread -lrt -lelf -lm
 ALL_CFLAGS = $(CFLAGS)
 ALL_LDFLAGS = $(LDFLAGS)
@@ -310,6 +338,7 @@ LIB_H += util/sigchain.h
 LIB_H += util/symbol.h
 LIB_H += util/module.h
 LIB_H += util/color.h
+LIB_H += util/values.h
 
 LIB_OBJS += util/abspath.o
 LIB_OBJS += util/alias.o
@@ -337,6 +366,13 @@ LIB_OBJS += util/color.o
 LIB_OBJS += util/pager.o
 LIB_OBJS += util/header.o
 LIB_OBJS += util/callchain.o
+LIB_OBJS += util/values.o
+LIB_OBJS += util/debug.o
+LIB_OBJS += util/map.o
+LIB_OBJS += util/thread.o
+LIB_OBJS += util/trace-event-parse.o
+LIB_OBJS += util/trace-event-read.o
+LIB_OBJS += util/trace-event-info.o
 
 BUILTIN_OBJS += builtin-annotate.o
 BUILTIN_OBJS += builtin-help.o
@@ -345,6 +381,7 @@ BUILTIN_OBJS += builtin-record.o
 BUILTIN_OBJS += builtin-report.o
 BUILTIN_OBJS += builtin-stat.o
 BUILTIN_OBJS += builtin-top.o
+BUILTIN_OBJS += builtin-trace.o
 
 PERFLIBS = $(LIB_FILE)
 

tools/perf/builtin-annotate.c

@@ -17,16 +17,13 @@
 #include "util/string.h"
 
 #include "perf.h"
+#include "util/debug.h"
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
-
-#define SHOW_KERNEL	1
-#define SHOW_USER	2
-#define SHOW_HV		4
+#include "util/thread.h"
 
 static char		const *input_name = "perf.data";
-static char		*vmlinux = "vmlinux";
 
 static char		default_sort_order[] = "comm,symbol";
 static char		*sort_order = default_sort_order;
@@ -35,13 +32,6 @@ static int		force;
 static int		input;
 static int		show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
 
-static int		dump_trace = 0;
-#define dprintf(x...)	do { if (dump_trace) printf(x); } while (0)
-
-static int		verbose;
-
-static int		modules;
-
 static int		full_paths;
 
 static int		print_line;
@@ -49,39 +39,8 @@ static int		print_line;
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
-struct ip_event {
-	struct perf_event_header header;
-	u64 ip;
-	u32 pid, tid;
-};
-
-struct mmap_event {
-	struct perf_event_header header;
-	u32 pid, tid;
-	u64 start;
-	u64 len;
-	u64 pgoff;
-	char filename[PATH_MAX];
-};
-
-struct comm_event {
-	struct perf_event_header header;
-	u32 pid, tid;
-	char comm[16];
-};
-
-struct fork_event {
-	struct perf_event_header header;
-	u32 pid, ppid;
-};
-
-typedef union event_union {
-	struct perf_event_header	header;
-	struct ip_event			ip;
-	struct mmap_event		mmap;
-	struct comm_event		comm;
-	struct fork_event		fork;
-} event_t;
+static struct rb_root	threads;
+static struct thread	*last_match;
 
 
 struct sym_ext {
@@ -90,323 +49,6 @@ struct sym_ext {
 	char		*path;
 };
 
-static LIST_HEAD(dsos);
-static struct dso *kernel_dso;
-static struct dso *vdso;
-
-
-static void dsos__add(struct dso *dso)
-{
-	list_add_tail(&dso->node, &dsos);
-}
-
-static struct dso *dsos__find(const char *name)
-{
-	struct dso *pos;
-
-	list_for_each_entry(pos, &dsos, node)
-		if (strcmp(pos->name, name) == 0)
-			return pos;
-	return NULL;
-}
-
-static struct dso *dsos__findnew(const char *name)
-{
-	struct dso *dso = dsos__find(name);
-	int nr;
-
-	if (dso)
-		return dso;
-
-	dso = dso__new(name, 0);
-	if (!dso)
-		goto out_delete_dso;
-
-	nr = dso__load(dso, NULL, verbose);
-	if (nr < 0) {
-		if (verbose)
-			fprintf(stderr, "Failed to open: %s\n", name);
-		goto out_delete_dso;
-	}
-	if (!nr && verbose) {
-		fprintf(stderr,
-		"No symbols found in: %s, maybe install a debug package?\n",
-				name);
-	}
-
-	dsos__add(dso);
-
-	return dso;
-
-out_delete_dso:
-	dso__delete(dso);
-	return NULL;
-}
-
-static void dsos__fprintf(FILE *fp)
-{
-	struct dso *pos;
-
-	list_for_each_entry(pos, &dsos, node)
-		dso__fprintf(pos, fp);
-}
-
-static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip)
-{
-	return dso__find_symbol(dso, ip);
-}
-
-static int load_kernel(void)
-{
-	int err;
-
-	kernel_dso = dso__new("[kernel]", 0);
-	if (!kernel_dso)
-		return -1;
-
-	err = dso__load_kernel(kernel_dso, vmlinux, NULL, verbose, modules);
-	if (err <= 0) {
-		dso__delete(kernel_dso);
-		kernel_dso = NULL;
-	} else
-		dsos__add(kernel_dso);
-
-	vdso = dso__new("[vdso]", 0);
-	if (!vdso)
-		return -1;
-
-	vdso->find_symbol = vdso__find_symbol;
-
-	dsos__add(vdso);
-
-	return err;
-}
-
-struct map {
-	struct list_head node;
-	u64	 start;
-	u64	 end;
-	u64	 pgoff;
-	u64	 (*map_ip)(struct map *, u64);
-	struct dso	 *dso;
-};
-
-static u64 map__map_ip(struct map *map, u64 ip)
-{
-	return ip - map->start + map->pgoff;
-}
-
-static u64 vdso__map_ip(struct map *map __used, u64 ip)
-{
-	return ip;
-}
-
-static struct map *map__new(struct mmap_event *event)
-{
-	struct map *self = malloc(sizeof(*self));
-
-	if (self != NULL) {
-		const char *filename = event->filename;
-
-		self->start = event->start;
-		self->end   = event->start + event->len;
-		self->pgoff = event->pgoff;
-
-		self->dso = dsos__findnew(filename);
-		if (self->dso == NULL)
-			goto out_delete;
-
-		if (self->dso == vdso)
-			self->map_ip = vdso__map_ip;
-		else
-			self->map_ip = map__map_ip;
-	}
-	return self;
-out_delete:
-	free(self);
-	return NULL;
-}
-
-static struct map *map__clone(struct map *self)
-{
-	struct map *map = malloc(sizeof(*self));
-
-	if (!map)
-		return NULL;
-
-	memcpy(map, self, sizeof(*self));
-
-	return map;
-}
-
-static int map__overlap(struct map *l, struct map *r)
-{
-	if (l->start > r->start) {
-		struct map *t = l;
-		l = r;
-		r = t;
-	}
-
-	if (l->end > r->start)
-		return 1;
-
-	return 0;
-}
-
-static size_t map__fprintf(struct map *self, FILE *fp)
-{
-	return fprintf(fp, " %Lx-%Lx %Lx %s\n",
-		       self->start, self->end, self->pgoff, self->dso->name);
-}
-
-
-struct thread {
-	struct rb_node	 rb_node;
-	struct list_head maps;
-	pid_t		 pid;
-	char		 *comm;
-};
-
-static struct thread *thread__new(pid_t pid)
-{
-	struct thread *self = malloc(sizeof(*self));
-
-	if (self != NULL) {
-		self->pid = pid;
-		self->comm = malloc(32);
-		if (self->comm)
-			snprintf(self->comm, 32, ":%d", self->pid);
-		INIT_LIST_HEAD(&self->maps);
-	}
-
-	return self;
-}
-
-static int thread__set_comm(struct thread *self, const char *comm)
-{
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(comm);
-	return self->comm ? 0 : -ENOMEM;
-}
-
-static size_t thread__fprintf(struct thread *self, FILE *fp)
-{
-	struct map *pos;
-	size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm);
-
-	list_for_each_entry(pos, &self->maps, node)
-		ret += map__fprintf(pos, fp);
-
-	return ret;
-}
-
-
-static struct rb_root threads;
-static struct thread *last_match;
-
-static struct thread *threads__findnew(pid_t pid)
-{
-	struct rb_node **p = &threads.rb_node;
-	struct rb_node *parent = NULL;
-	struct thread *th;
-
-	/*
-	 * Font-end cache - PID lookups come in blocks,
-	 * so most of the time we dont have to look up
-	 * the full rbtree:
-	 */
-	if (last_match && last_match->pid == pid)
-		return last_match;
-
-	while (*p != NULL) {
-		parent = *p;
-		th = rb_entry(parent, struct thread, rb_node);
-
-		if (th->pid == pid) {
-			last_match = th;
-			return th;
-		}
-
-		if (pid < th->pid)
-			p = &(*p)->rb_left;
-		else
-			p = &(*p)->rb_right;
-	}
-
-	th = thread__new(pid);
-	if (th != NULL) {
-		rb_link_node(&th->rb_node, parent, p);
-		rb_insert_color(&th->rb_node, &threads);
-		last_match = th;
-	}
-
-	return th;
-}
-
-static void thread__insert_map(struct thread *self, struct map *map)
-{
-	struct map *pos, *tmp;
-
-	list_for_each_entry_safe(pos, tmp, &self->maps, node) {
-		if (map__overlap(pos, map)) {
-			list_del_init(&pos->node);
-			/* XXX leaks dsos */
-			free(pos);
-		}
-	}
-
-	list_add_tail(&map->node, &self->maps);
-}
-
-static int thread__fork(struct thread *self, struct thread *parent)
-{
-	struct map *map;
-
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(parent->comm);
-	if (!self->comm)
-		return -ENOMEM;
-
-	list_for_each_entry(map, &parent->maps, node) {
-		struct map *new = map__clone(map);
-		if (!new)
-			return -ENOMEM;
-		thread__insert_map(self, new);
-	}
-
-	return 0;
-}
-
-static struct map *thread__find_map(struct thread *self, u64 ip)
-{
-	struct map *pos;
-
-	if (self == NULL)
-		return NULL;
-
-	list_for_each_entry(pos, &self->maps, node)
-		if (ip >= pos->start && ip <= pos->end)
-			return pos;
-
-	return NULL;
-}
-
-static size_t threads__fprintf(FILE *fp)
-{
-	size_t ret = 0;
-	struct rb_node *nd;
-
-	for (nd = rb_first(&threads); nd; nd = rb_next(nd)) {
-		struct thread *pos = rb_entry(nd, struct thread, rb_node);
-
-		ret += thread__fprintf(pos, fp);
-	}
-
-	return ret;
-}
-
 /*
  * histogram, sorted on item, collects counts
  */
@@ -433,7 +75,7 @@ struct hist_entry {
 struct sort_entry {
 	struct list_head list;
 
-	char *header;
+	const char *header;
 
 	int64_t (*cmp)(struct hist_entry *, struct hist_entry *);
 	int64_t (*collapse)(struct hist_entry *, struct hist_entry *);
@@ -577,7 +219,7 @@ static struct sort_entry sort_sym = {
 static int sort__need_collapse = 0;
 
 struct sort_dimension {
-	char			*name;
+	const char		*name;
 	struct sort_entry	*entry;
 	int			taken;
 };
@@ -830,17 +472,6 @@ static void output__resort(void)
 	}
 }
 
-static void register_idle_thread(void)
-{
-	struct thread *thread = threads__findnew(0);
-
-	if (thread == NULL ||
-			thread__set_comm(thread, "[idle]")) {
-		fprintf(stderr, "problem inserting idle task.\n");
-		exit(-1);
-	}
-}
-
 static unsigned long total = 0,
 		     total_mmap = 0,
 		     total_comm = 0,
@@ -853,18 +484,20 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	char level;
 	int show = 0;
 	struct dso *dso = NULL;
-	struct thread *thread = threads__findnew(event->ip.pid);
+	struct thread *thread;
 	u64 ip = event->ip.ip;
 	struct map *map = NULL;
 
-	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
+	dump_printf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->header.misc,
 		event->ip.pid,
 		(void *)(long)ip);
 
-	dprintf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
 
 	if (thread == NULL) {
 		fprintf(stderr, "problem processing %d event, skipping it.\n",
@@ -878,7 +511,7 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 
 		dso = kernel_dso;
 
-		dprintf(" ...... dso: %s\n", dso->name);
+		dump_printf(" ...... dso: %s\n", dso->name);
 
 	} else if (event->header.misc & PERF_EVENT_MISC_USER) {
 
@@ -899,12 +532,12 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 			if ((long long)ip < 0)
 				dso = kernel_dso;
 		}
-		dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
+		dump_printf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
 
 	} else {
 		show = SHOW_HV;
 		level = 'H';
-		dprintf(" ...... dso: [hypervisor]\n");
+		dump_printf(" ...... dso: [hypervisor]\n");
 	}
 
 	if (show & show_mask) {
@@ -927,10 +560,12 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->mmap.pid);
-	struct map *map = map__new(&event->mmap);
+	struct thread *thread;
+	struct map *map = map__new(&event->mmap, NULL, 0);
+
+	thread = threads__findnew(event->mmap.pid, &threads, &last_match);
 
-	dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n",
+	dump_printf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->mmap.pid,
@@ -940,7 +575,7 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 		event->mmap.filename);
 
 	if (thread == NULL || map == NULL) {
-		dprintf("problem processing PERF_EVENT_MMAP, skipping event.\n");
+		dump_printf("problem processing PERF_EVENT_MMAP, skipping event.\n");
 		return 0;
 	}
 
@@ -953,16 +588,17 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
 
-	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+	dump_printf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->comm.comm, event->comm.pid);
 
 	if (thread == NULL ||
 	    thread__set_comm(thread, event->comm.comm)) {
-		dprintf("problem processing PERF_EVENT_COMM, skipping event.\n");
+		dump_printf("problem processing PERF_EVENT_COMM, skipping event.\n");
 		return -1;
 	}
 	total_comm++;
@@ -973,10 +609,12 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
 
-	dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
+	thread = threads__findnew(event->fork.pid, &threads, &last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
+	dump_printf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
 		event->fork.pid, event->fork.ppid);
@@ -989,7 +627,7 @@ process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 		return 0;
 
 	if (!thread || !parent || thread__fork(thread, parent)) {
-		dprintf("problem processing PERF_EVENT_FORK, skipping event.\n");
+		dump_printf("problem processing PERF_EVENT_FORK, skipping event.\n");
 		return -1;
 	}
 	total_fork++;
@@ -1075,7 +713,7 @@ parse_line(FILE *file, struct symbol *sym, u64 start, u64 len)
 		const char *path = NULL;
 		unsigned int hits = 0;
 		double percent = 0.0;
-		char *color;
+		const char *color;
 		struct sym_ext *sym_ext = sym->priv;
 
 		offset = line_ip - start;
@@ -1157,7 +795,7 @@ static void free_source_line(struct symbol *sym, int len)
 
 /* Get the filename:line for the colored entries */
 static void
-get_source_line(struct symbol *sym, u64 start, int len, char *filename)
+get_source_line(struct symbol *sym, u64 start, int len, const char *filename)
 {
 	int i;
 	char cmd[PATH_MAX * 2];
@@ -1203,7 +841,7 @@ get_source_line(struct symbol *sym, u64 start, int len, char *filename)
 	}
 }
 
-static void print_summary(char *filename)
+static void print_summary(const char *filename)
 {
 	struct sym_ext *sym_ext;
 	struct rb_node *node;
@@ -1219,7 +857,7 @@ static void print_summary(char *filename)
 	node = rb_first(&root_sym_ext);
 	while (node) {
 		double percent;
-		char *color;
+		const char *color;
 		char *path;
 
 		sym_ext = rb_entry(node, struct sym_ext, node);
@@ -1234,7 +872,7 @@ static void print_summary(char *filename)
 
 static void annotate_sym(struct dso *dso, struct symbol *sym)
 {
-	char *filename = dso->name, *d_filename;
+	const char *filename = dso->name, *d_filename;
 	u64 start, end, len;
 	char command[PATH_MAX*2];
 	FILE *file;
@@ -1244,7 +882,7 @@ static void annotate_sym(struct dso *dso, struct symbol *sym)
 	if (sym->module)
 		filename = sym->module->path;
 	else if (dso == kernel_dso)
-		filename = vmlinux;
+		filename = vmlinux_name;
 
 	start = sym->obj_start;
 	if (!start)
@@ -1316,12 +954,12 @@ static int __cmd_annotate(void)
 	int ret, rc = EXIT_FAILURE;
 	unsigned long offset = 0;
 	unsigned long head = 0;
-	struct stat stat;
+	struct stat input_stat;
 	event_t *event;
 	uint32_t size;
 	char *buf;
 
-	register_idle_thread();
+	register_idle_thread(&threads, &last_match);
 
 	input = open(input_name, O_RDONLY);
 	if (input < 0) {
@@ -1329,18 +967,18 @@ static int __cmd_annotate(void)
 		exit(-1);
 	}
 
-	ret = fstat(input, &stat);
+	ret = fstat(input, &input_stat);
 	if (ret < 0) {
 		perror("failed to stat file");
 		exit(-1);
 	}
 
-	if (!force && (stat.st_uid != geteuid())) {
-		fprintf(stderr, "file: %s not owned by current user\n", input_name);
+	if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) {
+		fprintf(stderr, "file: %s not owned by current user or root\n", input_name);
 		exit(-1);
 	}
 
-	if (!stat.st_size) {
+	if (!input_stat.st_size) {
 		fprintf(stderr, "zero-sized file, nothing to do!\n");
 		exit(0);
 	}
@@ -1367,10 +1005,10 @@ static int __cmd_annotate(void)
 
 	if (head + event->header.size >= page_size * mmap_window) {
 		unsigned long shift = page_size * (head / page_size);
-		int ret;
+		int munmap_ret;
 
-		ret = munmap(buf, page_size * mmap_window);
-		assert(ret == 0);
+		munmap_ret = munmap(buf, page_size * mmap_window);
+		assert(munmap_ret == 0);
 
 		offset += shift;
 		head -= shift;
@@ -1379,14 +1017,14 @@ static int __cmd_annotate(void)
 
 	size = event->header.size;
 
-	dprintf("%p [%p]: event: %d\n",
+	dump_printf("%p [%p]: event: %d\n",
 			(void *)(offset + head),
 			(void *)(long)event->header.size,
 			event->header.type);
 
 	if (!size || process_event(event, offset, head) < 0) {
 
-		dprintf("%p [%p]: skipping unknown header type: %d\n",
+		dump_printf("%p [%p]: skipping unknown header type: %d\n",
 			(void *)(offset + head),
 			(void *)(long)(event->header.size),
 			event->header.type);
@@ -1406,23 +1044,23 @@ static int __cmd_annotate(void)
 
 	head += size;
 
-	if (offset + head < (unsigned long)stat.st_size)
+	if (offset + head < (unsigned long)input_stat.st_size)
 		goto more;
 
 	rc = EXIT_SUCCESS;
 	close(input);
 
-	dprintf("      IP events: %10ld\n", total);
-	dprintf("    mmap events: %10ld\n", total_mmap);
-	dprintf("    comm events: %10ld\n", total_comm);
-	dprintf("    fork events: %10ld\n", total_fork);
-	dprintf(" unknown events: %10ld\n", total_unknown);
+	dump_printf("      IP events: %10ld\n", total);
+	dump_printf("    mmap events: %10ld\n", total_mmap);
+	dump_printf("    comm events: %10ld\n", total_comm);
+	dump_printf("    fork events: %10ld\n", total_fork);
+	dump_printf(" unknown events: %10ld\n", total_unknown);
 
 	if (dump_trace)
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
@@ -1450,7 +1088,7 @@ static const struct option options[] = {
 		    "be more verbose (show symbol address, etc)"),
 	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
 		    "dump raw trace in ASCII"),
-	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
+	OPT_STRING('k', "vmlinux", &vmlinux_name, "file", "vmlinux pathname"),
 	OPT_BOOLEAN('m', "modules", &modules,
 		    "load module symbols - WARNING: use only with -k and LIVE kernel"),
 	OPT_BOOLEAN('l', "print-line", &print_line,

tools/perf/builtin-help.c

@@ -456,6 +456,7 @@ int cmd_help(int argc, const char **argv, const char *prefix __used)
 		break;
 	case HELP_FORMAT_WEB:
 		show_html_page(argv[0]);
+	default:
 		break;
 	}
 

tools/perf/builtin-record.c

@@ -15,6 +15,9 @@
 #include "util/string.h"
 
 #include "util/header.h"
+#include "util/event.h"
+#include "util/debug.h"
+#include "util/trace-event.h"
 
 #include <unistd.h>
 #include <sched.h>
@@ -42,7 +45,6 @@ static int			inherit				= 1;
 static int			force				= 0;
 static int			append_file			= 0;
 static int			call_graph			= 0;
-static int			verbose				= 0;
 static int			inherit_stat			= 0;
 static int			no_samples			= 0;
 static int			sample_address			= 0;
@@ -62,24 +64,6 @@ static int			file_new = 1;
 
 struct perf_header		*header;
 
-struct mmap_event {
-	struct perf_event_header	header;
-	u32				pid;
-	u32				tid;
-	u64				start;
-	u64				len;
-	u64				pgoff;
-	char				filename[PATH_MAX];
-};
-
-struct comm_event {
-	struct perf_event_header	header;
-	u32				pid;
-	u32				tid;
-	char				comm[16];
-};
-
-
 struct mmap_data {
 	int			counter;
 	void			*base;
@@ -419,8 +403,11 @@ static void create_counter(int counter, int cpu, pid_t pid)
 	if (call_graph)
 		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
 
-	if (raw_samples)
+	if (raw_samples) {
+		attr->sample_type	|= PERF_SAMPLE_TIME;
 		attr->sample_type	|= PERF_SAMPLE_RAW;
+		attr->sample_type	|= PERF_SAMPLE_CPU;
+	}
 
 	attr->mmap		= track;
 	attr->comm		= track;
@@ -563,6 +550,17 @@ static int __cmd_record(int argc, const char **argv)
 	else
 		header = perf_header__new();
 
+
+	if (raw_samples) {
+		read_tracing_data(attrs, nr_counters);
+	} else {
+		for (i = 0; i < nr_counters; i++) {
+			if (attrs[i].sample_type & PERF_SAMPLE_RAW) {
+				read_tracing_data(attrs, nr_counters);
+				break;
+			}
+		}
+	}
 	atexit(atexit_header);
 
 	if (!system_wide) {

tools/perf/builtin-stat.c

@@ -42,6 +42,8 @@
 #include "util/util.h"
 #include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/event.h"
+#include "util/debug.h"
 
 #include <sys/prctl.h>
 #include <math.h>
@@ -60,10 +62,7 @@ static struct perf_counter_attr default_attrs[] = {
 
 };
 
-#define MAX_RUN			100
-
 static int			system_wide			=  0;
-static int			verbose				=  0;
 static unsigned int		nr_cpus				=  0;
 static int			run_idx				=  0;
 
@@ -75,26 +74,56 @@ static int			null_run			=  0;
 
 static int			fd[MAX_NR_CPUS][MAX_COUNTERS];
 
-static u64			runtime_nsecs[MAX_RUN];
-static u64			walltime_nsecs[MAX_RUN];
-static u64			runtime_cycles[MAX_RUN];
+static int			event_scaled[MAX_COUNTERS];
 
-static u64			event_res[MAX_RUN][MAX_COUNTERS][3];
-static u64			event_scaled[MAX_RUN][MAX_COUNTERS];
+struct stats
+{
+	double n, mean, M2;
+};
 
-static u64			event_res_avg[MAX_COUNTERS][3];
-static u64			event_res_noise[MAX_COUNTERS][3];
+static void update_stats(struct stats *stats, u64 val)
+{
+	double delta;
 
-static u64			event_scaled_avg[MAX_COUNTERS];
+	stats->n++;
+	delta = val - stats->mean;
+	stats->mean += delta / stats->n;
+	stats->M2 += delta*(val - stats->mean);
+}
 
-static u64			runtime_nsecs_avg;
-static u64			runtime_nsecs_noise;
+static double avg_stats(struct stats *stats)
+{
+	return stats->mean;
+}
 
-static u64			walltime_nsecs_avg;
-static u64			walltime_nsecs_noise;
+/*
+ * http://en.wikipedia.org/wiki/Algorithms_for_calculating_variance
+ *
+ *       (\Sum n_i^2) - ((\Sum n_i)^2)/n
+ * s^2 = -------------------------------
+ *                  n - 1
+ *
+ * http://en.wikipedia.org/wiki/Stddev
+ *
+ * The std dev of the mean is related to the std dev by:
+ *
+ *             s
+ * s_mean = -------
+ *          sqrt(n)
+ *
+ */
+static double stddev_stats(struct stats *stats)
+{
+	double variance = stats->M2 / (stats->n - 1);
+	double variance_mean = variance / stats->n;
+
+	return sqrt(variance_mean);
+}
 
-static u64			runtime_cycles_avg;
-static u64			runtime_cycles_noise;
+struct stats			event_res_stats[MAX_COUNTERS][3];
+struct stats			runtime_nsecs_stats;
+struct stats			walltime_nsecs_stats;
+struct stats			runtime_cycles_stats;
 
 #define MATCH_EVENT(t, c, counter)			\
 	(attrs[counter].type == PERF_TYPE_##t &&	\
@@ -149,12 +178,11 @@ static inline int nsec_counter(int counter)
  */
 static void read_counter(int counter)
 {
-	u64 *count, single_count[3];
+	u64 count[3], single_count[3];
 	unsigned int cpu;
 	size_t res, nv;
 	int scaled;
-
-	count = event_res[run_idx][counter];
+	int i;
 
 	count[0] = count[1] = count[2] = 0;
 
@@ -179,24 +207,33 @@ static void read_counter(int counter)
 	scaled = 0;
 	if (scale) {
 		if (count[2] == 0) {
-			event_scaled[run_idx][counter] = -1;
+			event_scaled[counter] = -1;
 			count[0] = 0;
 			return;
 		}
 
 		if (count[2] < count[1]) {
-			event_scaled[run_idx][counter] = 1;
+			event_scaled[counter] = 1;
 			count[0] = (unsigned long long)
 				((double)count[0] * count[1] / count[2] + 0.5);
 		}
 	}
+
+	for (i = 0; i < 3; i++)
+		update_stats(&event_res_stats[counter][i], count[i]);
+
+	if (verbose) {
+		fprintf(stderr, "%s: %Ld %Ld %Ld\n", event_name(counter),
+				count[0], count[1], count[2]);
+	}
+
 	/*
 	 * Save the full runtime - to allow normalization during printout:
 	 */
 	if (MATCH_EVENT(SOFTWARE, SW_TASK_CLOCK, counter))
-		runtime_nsecs[run_idx] = count[0];
+		update_stats(&runtime_nsecs_stats, count[0]);
 	if (MATCH_EVENT(HARDWARE, HW_CPU_CYCLES, counter))
-		runtime_cycles[run_idx] = count[0];
+		update_stats(&runtime_cycles_stats, count[0]);
 }
 
 static int run_perf_stat(int argc __used, const char **argv)
@@ -270,7 +307,7 @@ static int run_perf_stat(int argc __used, const char **argv)
 
 	t1 = rdclock();
 
-	walltime_nsecs[run_idx] = t1 - t0;
+	update_stats(&walltime_nsecs_stats, t1 - t0);
 
 	for (counter = 0; counter < nr_counters; counter++)
 		read_counter(counter);
@@ -278,42 +315,38 @@ static int run_perf_stat(int argc __used, const char **argv)
 	return WEXITSTATUS(status);
 }
 
-static void print_noise(u64 *count, u64 *noise)
+static void print_noise(int counter, double avg)
 {
-	if (run_count > 1)
-		fprintf(stderr, "   ( +- %7.3f%% )",
-			(double)noise[0]/(count[0]+1)*100.0);
+	if (run_count == 1)
+		return;
+
+	fprintf(stderr, "   ( +- %7.3f%% )",
+			100 * stddev_stats(&event_res_stats[counter][0]) / avg);
 }
 
-static void nsec_printout(int counter, u64 *count, u64 *noise)
+static void nsec_printout(int counter, double avg)
 {
-	double msecs = (double)count[0] / 1000000;
+	double msecs = avg / 1e6;
 
 	fprintf(stderr, " %14.6f  %-24s", msecs, event_name(counter));
 
 	if (MATCH_EVENT(SOFTWARE, SW_TASK_CLOCK, counter)) {
-		if (walltime_nsecs_avg)
-			fprintf(stderr, " # %10.3f CPUs ",
-				(double)count[0] / (double)walltime_nsecs_avg);
+		fprintf(stderr, " # %10.3f CPUs ",
+				avg / avg_stats(&walltime_nsecs_stats));
 	}
-	print_noise(count, noise);
 }
 
-static void abs_printout(int counter, u64 *count, u64 *noise)
+static void abs_printout(int counter, double avg)
 {
-	fprintf(stderr, " %14Ld  %-24s", count[0], event_name(counter));
+	fprintf(stderr, " %14.0f  %-24s", avg, event_name(counter));
 
-	if (runtime_cycles_avg &&
-	    MATCH_EVENT(HARDWARE, HW_INSTRUCTIONS, counter)) {
+	if (MATCH_EVENT(HARDWARE, HW_INSTRUCTIONS, counter)) {
 		fprintf(stderr, " # %10.3f IPC  ",
-			(double)count[0] / (double)runtime_cycles_avg);
+				avg / avg_stats(&runtime_cycles_stats));
 	} else {
-		if (runtime_nsecs_avg) {
-			fprintf(stderr, " # %10.3f M/sec",
-				(double)count[0]/runtime_nsecs_avg*1000.0);
-		}
+		fprintf(stderr, " # %10.3f M/sec",
+				1000.0 * avg / avg_stats(&runtime_nsecs_stats));
 	}
-	print_noise(count, noise);
 }
 
 /*
@@ -321,12 +354,8 @@ static void abs_printout(int counter, u64 *count, u64 *noise)
  */
 static void print_counter(int counter)
 {
-	u64 *count, *noise;
-	int scaled;
-
-	count = event_res_avg[counter];
-	noise = event_res_noise[counter];
-	scaled = event_scaled_avg[counter];
+	double avg = avg_stats(&event_res_stats[counter][0]);
+	int scaled = event_scaled[counter];
 
 	if (scaled == -1) {
 		fprintf(stderr, " %14s  %-24s\n",
@@ -335,110 +364,29 @@ static void print_counter(int counter)
 	}
 
 	if (nsec_counter(counter))
-		nsec_printout(counter, count, noise);
+		nsec_printout(counter, avg);
 	else
-		abs_printout(counter, count, noise);
-
-	if (scaled)
-		fprintf(stderr, "  (scaled from %.2f%%)",
-			(double) count[2] / count[1] * 100);
-
-	fprintf(stderr, "\n");
-}
+		abs_printout(counter, avg);
 
-/*
- * normalize_noise noise values down to stddev:
- */
-static void normalize_noise(u64 *val)
-{
-	double res;
+	print_noise(counter, avg);
 
-	res = (double)*val / (run_count * sqrt((double)run_count));
+	if (scaled) {
+		double avg_enabled, avg_running;
 
-	*val = (u64)res;
-}
+		avg_enabled = avg_stats(&event_res_stats[counter][1]);
+		avg_running = avg_stats(&event_res_stats[counter][2]);
 
-static void update_avg(const char *name, int idx, u64 *avg, u64 *val)
-{
-	*avg += *val;
-
-	if (verbose > 1)
-		fprintf(stderr, "debug: %20s[%d]: %Ld\n", name, idx, *val);
-}
-/*
- * Calculate the averages and noises:
- */
-static void calc_avg(void)
-{
-	int i, j;
-
-	if (verbose > 1)
-		fprintf(stderr, "\n");
-
-	for (i = 0; i < run_count; i++) {
-		update_avg("runtime", 0, &runtime_nsecs_avg, runtime_nsecs + i);
-		update_avg("walltime", 0, &walltime_nsecs_avg, walltime_nsecs + i);
-		update_avg("runtime_cycles", 0, &runtime_cycles_avg, runtime_cycles + i);
-
-		for (j = 0; j < nr_counters; j++) {
-			update_avg("counter/0", j,
-				event_res_avg[j]+0, event_res[i][j]+0);
-			update_avg("counter/1", j,
-				event_res_avg[j]+1, event_res[i][j]+1);
-			update_avg("counter/2", j,
-				event_res_avg[j]+2, event_res[i][j]+2);
-			if (event_scaled[i][j] != (u64)-1)
-				update_avg("scaled", j,
-					event_scaled_avg + j, event_scaled[i]+j);
-			else
-				event_scaled_avg[j] = -1;
-		}
-	}
-	runtime_nsecs_avg /= run_count;
-	walltime_nsecs_avg /= run_count;
-	runtime_cycles_avg /= run_count;
-
-	for (j = 0; j < nr_counters; j++) {
-		event_res_avg[j][0] /= run_count;
-		event_res_avg[j][1] /= run_count;
-		event_res_avg[j][2] /= run_count;
-	}
-
-	for (i = 0; i < run_count; i++) {
-		runtime_nsecs_noise +=
-			abs((s64)(runtime_nsecs[i] - runtime_nsecs_avg));
-		walltime_nsecs_noise +=
-			abs((s64)(walltime_nsecs[i] - walltime_nsecs_avg));
-		runtime_cycles_noise +=
-			abs((s64)(runtime_cycles[i] - runtime_cycles_avg));
-
-		for (j = 0; j < nr_counters; j++) {
-			event_res_noise[j][0] +=
-				abs((s64)(event_res[i][j][0] - event_res_avg[j][0]));
-			event_res_noise[j][1] +=
-				abs((s64)(event_res[i][j][1] - event_res_avg[j][1]));
-			event_res_noise[j][2] +=
-				abs((s64)(event_res[i][j][2] - event_res_avg[j][2]));
-		}
+		fprintf(stderr, "  (scaled from %.2f%%)",
+				100 * avg_running / avg_enabled);
 	}
 
-	normalize_noise(&runtime_nsecs_noise);
-	normalize_noise(&walltime_nsecs_noise);
-	normalize_noise(&runtime_cycles_noise);
-
-	for (j = 0; j < nr_counters; j++) {
-		normalize_noise(&event_res_noise[j][0]);
-		normalize_noise(&event_res_noise[j][1]);
-		normalize_noise(&event_res_noise[j][2]);
-	}
+	fprintf(stderr, "\n");
 }
 
 static void print_stat(int argc, const char **argv)
 {
 	int i, counter;
 
-	calc_avg();
-
 	fflush(stdout);
 
 	fprintf(stderr, "\n");
@@ -457,10 +405,11 @@ static void print_stat(int argc, const char **argv)
 
 	fprintf(stderr, "\n");
 	fprintf(stderr, " %14.9f  seconds time elapsed",
-			(double)walltime_nsecs_avg/1e9);
+			avg_stats(&walltime_nsecs_stats)/1e9);
 	if (run_count > 1) {
 		fprintf(stderr, "   ( +- %7.3f%% )",
-			100.0*(double)walltime_nsecs_noise/(double)walltime_nsecs_avg);
+				100*stddev_stats(&walltime_nsecs_stats) /
+				avg_stats(&walltime_nsecs_stats));
 	}
 	fprintf(stderr, "\n\n");
 }
@@ -515,7 +464,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 		PARSE_OPT_STOP_AT_NON_OPTION);
 	if (!argc)
 		usage_with_options(stat_usage, options);
-	if (run_count <= 0 || run_count > MAX_RUN)
+	if (run_count <= 0)
 		usage_with_options(stat_usage, options);
 
 	/* Set attrs and nr_counters if no event is selected and !null_run */

tools/perf/builtin-top.c

@@ -27,6 +27,8 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 
+#include "util/debug.h"
+
 #include <assert.h>
 #include <fcntl.h>
 
@@ -68,8 +70,6 @@ static int			group				=  0;
 static unsigned int		page_size;
 static unsigned int		mmap_pages			= 16;
 static int			freq				=  0;
-static int			verbose				=  0;
-static char			*vmlinux			=  NULL;
 
 static int			delay_secs			=  2;
 static int			zero;
@@ -122,7 +122,8 @@ static void parse_source(struct sym_entry *syme)
 	struct module *module;
 	struct section *section = NULL;
 	FILE *file;
-	char command[PATH_MAX*2], *path = vmlinux;
+	char command[PATH_MAX*2];
+	const char *path = vmlinux_name;
 	u64 start, end, len;
 
 	if (!syme)
@@ -338,8 +339,6 @@ static void show_details(struct sym_entry *syme)
 		printf("%d lines not displayed, maybe increase display entries [e]\n", more);
 }
 
-struct dso			*kernel_dso;
-
 /*
  * Symbols will be added here in record_ip and will get out
  * after decayed.
@@ -484,17 +483,24 @@ static void print_sym_table(void)
 	if (nr_counters == 1)
 		printf("             samples    pcnt");
 	else
-		printf("  weight     samples    pcnt");
+		printf("   weight    samples    pcnt");
 
-	printf("         RIP          kernel function\n"
-	       	       "  ______     _______   _____   ________________   _______________\n\n"
-	);
+	if (verbose)
+		printf("         RIP       ");
+	printf("   kernel function\n");
+	printf("   %s    _______   _____",
+	       nr_counters == 1 ? "      " : "______");
+	if (verbose)
+		printf("   ________________");
+	printf("   _______________\n\n");
 
 	for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) {
-		struct sym_entry *syme = rb_entry(nd, struct sym_entry, rb_node);
-		struct symbol *sym = (struct symbol *)(syme + 1);
+		struct symbol *sym;
 		double pcnt;
 
+		syme = rb_entry(nd, struct sym_entry, rb_node);
+		sym = (struct symbol *)(syme + 1);
+
 		if (++printed > print_entries || (int)syme->snap_count < count_filter)
 			continue;
 
@@ -507,7 +513,9 @@ static void print_sym_table(void)
 			printf("%9.1f %10ld - ", syme->weight, syme->snap_count);
 
 		percent_color_fprintf(stdout, "%4.1f%%", pcnt);
-		printf(" - %016llx : %s", sym->start, sym->name);
+		if (verbose)
+			printf(" - %016llx", sym->start);
+		printf(" : %s", sym->name);
 		if (sym->module)
 			printf("\t[%s]", sym->module->name);
 		printf("\n");
@@ -613,7 +621,7 @@ static void print_mapped_keys(void)
 
 	fprintf(stdout, "\t[f]     profile display filter (count).    \t(%d)\n", count_filter);
 
-	if (vmlinux) {
+	if (vmlinux_name) {
 		fprintf(stdout, "\t[F]     annotate display filter (percent). \t(%d%%)\n", sym_pcnt_filter);
 		fprintf(stdout, "\t[s]     annotate symbol.                   \t(%s)\n", name?: "NULL");
 		fprintf(stdout, "\t[S]     stop annotation.\n");
@@ -642,7 +650,9 @@ static int key_mapped(int c)
 		case 'F':
 		case 's':
 		case 'S':
-			return vmlinux ? 1 : 0;
+			return vmlinux_name ? 1 : 0;
+		default:
+			break;
 	}
 
 	return 0;
@@ -728,6 +738,8 @@ static void handle_keypress(int c)
 		case 'z':
 			zero = ~zero;
 			break;
+		default:
+			break;
 	}
 }
 
@@ -816,13 +828,13 @@ static int parse_symbols(void)
 {
 	struct rb_node *node;
 	struct symbol  *sym;
-	int modules = vmlinux ? 1 : 0;
+	int use_modules = vmlinux_name ? 1 : 0;
 
 	kernel_dso = dso__new("[kernel]", sizeof(struct sym_entry));
 	if (kernel_dso == NULL)
 		return -1;
 
-	if (dso__load_kernel(kernel_dso, vmlinux, symbol_filter, verbose, modules) <= 0)
+	if (dso__load_kernel(kernel_dso, vmlinux_name, symbol_filter, verbose, use_modules) <= 0)
 		goto out_delete_dso;
 
 	node = rb_first(&kernel_dso->syms);
@@ -937,26 +949,6 @@ static void mmap_read_counter(struct mmap_data *md)
 	last_read = this_read;
 
 	for (; old != head;) {
-		struct ip_event {
-			struct perf_event_header header;
-			u64 ip;
-			u32 pid, target_pid;
-		};
-		struct mmap_event {
-			struct perf_event_header header;
-			u32 pid, target_pid;
-			u64 start;
-			u64 len;
-			u64 pgoff;
-			char filename[PATH_MAX];
-		};
-
-		typedef union event_union {
-			struct perf_event_header header;
-			struct ip_event ip;
-			struct mmap_event mmap;
-		} event_t;
-
 		event_t *event = (event_t *)&data[old & md->mask];
 
 		event_t event_copy;
@@ -1138,7 +1130,7 @@ static const struct option options[] = {
 			    "system-wide collection from all CPUs"),
 	OPT_INTEGER('C', "CPU", &profile_cpu,
 		    "CPU to profile on"),
-	OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"),
+	OPT_STRING('k', "vmlinux", &vmlinux_name, "file", "vmlinux pathname"),
 	OPT_INTEGER('m', "mmap-pages", &mmap_pages,
 		    "number of mmap data pages"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,

tools/perf/builtin-trace.c

+#include "builtin.h"
+
+#include "util/util.h"
+#include "util/cache.h"
+#include "util/symbol.h"
+#include "util/thread.h"
+#include "util/header.h"
+
+#include "util/parse-options.h"
+
+#include "perf.h"
+#include "util/debug.h"
+
+#include "util/trace-event.h"
+
+static char		const *input_name = "perf.data";
+static int		input;
+static unsigned long	page_size;
+static unsigned long	mmap_window = 32;
+
+static unsigned long	total = 0;
+static unsigned long	total_comm = 0;
+
+static struct rb_root	threads;
+static struct thread	*last_match;
+
+static struct perf_header *header;
+static u64		sample_type;
+
+
+static int
+process_comm_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	struct thread *thread;
+
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
+
+	dump_printf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
+		(void *)(offset + head),
+		(void *)(long)(event->header.size),
+		event->comm.comm, event->comm.pid);
+
+	if (thread == NULL ||
+	    thread__set_comm(thread, event->comm.comm)) {
+		dump_printf("problem processing PERF_EVENT_COMM, skipping event.\n");
+		return -1;
+	}
+	total_comm++;
+
+	return 0;
+}
+
+static int
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	char level;
+	int show = 0;
+	struct dso *dso = NULL;
+	struct thread *thread;
+	u64 ip = event->ip.ip;
+	u64 timestamp = -1;
+	u32 cpu = -1;
+	u64 period = 1;
+	void *more_data = event->ip.__more_data;
+	int cpumode;
+
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
+	if (sample_type & PERF_SAMPLE_TIME) {
+		timestamp = *(u64 *)more_data;
+		more_data += sizeof(u64);
+	}
+
+	if (sample_type & PERF_SAMPLE_CPU) {
+		cpu = *(u32 *)more_data;
+		more_data += sizeof(u32);
+		more_data += sizeof(u32); /* reserved */
+	}
+
+	if (sample_type & PERF_SAMPLE_PERIOD) {
+		period = *(u64 *)more_data;
+		more_data += sizeof(u64);
+	}
+
+	dump_printf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d/%d: %p period: %Ld\n",
+		(void *)(offset + head),
+		(void *)(long)(event->header.size),
+		event->header.misc,
+		event->ip.pid, event->ip.tid,
+		(void *)(long)ip,
+		(long long)period);
+
+	dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid);
+
+	if (thread == NULL) {
+		eprintf("problem processing %d event, skipping it.\n",
+			event->header.type);
+		return -1;
+	}
+
+	cpumode = event->header.misc & PERF_EVENT_MISC_CPUMODE_MASK;
+
+	if (cpumode == PERF_EVENT_MISC_KERNEL) {
+		show = SHOW_KERNEL;
+		level = 'k';
+
+		dso = kernel_dso;
+
+		dump_printf(" ...... dso: %s\n", dso->name);
+
+	} else if (cpumode == PERF_EVENT_MISC_USER) {
+
+		show = SHOW_USER;
+		level = '.';
+
+	} else {
+		show = SHOW_HV;
+		level = 'H';
+
+		dso = hypervisor_dso;
+
+		dump_printf(" ...... dso: [hypervisor]\n");
+	}
+
+	if (sample_type & PERF_SAMPLE_RAW) {
+		struct {
+			u32 size;
+			char data[0];
+		} *raw = more_data;
+
+		/*
+		 * FIXME: better resolve from pid from the struct trace_entry
+		 * field, although it should be the same than this perf
+		 * event pid
+		 */
+		print_event(cpu, raw->data, raw->size, timestamp, thread->comm);
+	}
+	total += period;
+
+	return 0;
+}
+
+static int
+process_event(event_t *event, unsigned long offset, unsigned long head)
+{
+	trace_event(event);
+
+	switch (event->header.type) {
+	case PERF_EVENT_MMAP ... PERF_EVENT_LOST:
+		return 0;
+
+	case PERF_EVENT_COMM:
+		return process_comm_event(event, offset, head);
+
+	case PERF_EVENT_EXIT ... PERF_EVENT_READ:
+		return 0;
+
+	case PERF_EVENT_SAMPLE:
+		return process_sample_event(event, offset, head);
+
+	case PERF_EVENT_MAX:
+	default:
+		return -1;
+	}
+
+	return 0;
+}
+
+static int __cmd_trace(void)
+{
+	int ret, rc = EXIT_FAILURE;
+	unsigned long offset = 0;
+	unsigned long head = 0;
+	struct stat perf_stat;
+	event_t *event;
+	uint32_t size;
+	char *buf;
+
+	trace_report();
+	register_idle_thread(&threads, &last_match);
+
+	input = open(input_name, O_RDONLY);
+	if (input < 0) {
+		perror("failed to open file");
+		exit(-1);
+	}
+
+	ret = fstat(input, &perf_stat);
+	if (ret < 0) {
+		perror("failed to stat file");
+		exit(-1);
+	}
+
+	if (!perf_stat.st_size) {
+		fprintf(stderr, "zero-sized file, nothing to do!\n");
+		exit(0);
+	}
+	header = perf_header__read(input);
+	head = header->data_offset;
+	sample_type = perf_header__sample_type(header);
+
+	if (!(sample_type & PERF_SAMPLE_RAW))
+		die("No trace sample to read. Did you call perf record "
+		    "without -R?");
+
+	if (load_kernel() < 0) {
+		perror("failed to load kernel symbols");
+		return EXIT_FAILURE;
+	}
+
+remap:
+	buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
+			   MAP_SHARED, input, offset);
+	if (buf == MAP_FAILED) {
+		perror("failed to mmap file");
+		exit(-1);
+	}
+
+more:
+	event = (event_t *)(buf + head);
+
+	size = event->header.size;
+	if (!size)
+		size = 8;
+
+	if (head + event->header.size >= page_size * mmap_window) {
+		unsigned long shift = page_size * (head / page_size);
+		int res;
+
+		res = munmap(buf, page_size * mmap_window);
+		assert(res == 0);
+
+		offset += shift;
+		head -= shift;
+		goto remap;
+	}
+
+	size = event->header.size;
+
+
+	if (!size || process_event(event, offset, head) < 0) {
+
+		/*
+		 * assume we lost track of the stream, check alignment, and
+		 * increment a single u64 in the hope to catch on again 'soon'.
+		 */
+
+		if (unlikely(head & 7))
+			head &= ~7ULL;
+
+		size = 8;
+	}
+
+	head += size;
+
+	if (offset + head < (unsigned long)perf_stat.st_size)
+		goto more;
+
+	rc = EXIT_SUCCESS;
+	close(input);
+
+	return rc;
+}
+
+static const char * const annotate_usage[] = {
+	"perf trace [<options>] <command>",
+	NULL
+};
+
+static const struct option options[] = {
+	OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace,
+		    "dump raw trace in ASCII"),
+	OPT_BOOLEAN('v', "verbose", &verbose,
+		    "be more verbose (show symbol address, etc)"),
+	OPT_END()
+};
+
+int cmd_trace(int argc, const char **argv, const char *prefix __used)
+{
+	symbol__init();
+	page_size = getpagesize();
+
+	argc = parse_options(argc, argv, options, annotate_usage, 0);
+	if (argc) {
+		/*
+		 * Special case: if there's an argument left then assume tha
+		 * it's a symbol filter:
+		 */
+		if (argc > 1)
+			usage_with_options(annotate_usage, options);
+	}
+
+
+	setup_pager();
+
+	return __cmd_trace();
+}

tools/perf/builtin.h

@@ -22,5 +22,6 @@ extern int cmd_stat(int argc, const char **argv, const char *prefix);
 extern int cmd_top(int argc, const char **argv, const char *prefix);
 extern int cmd_version(int argc, const char **argv, const char *prefix);
 extern int cmd_list(int argc, const char **argv, const char *prefix);
+extern int cmd_trace(int argc, const char **argv, const char *prefix);
 
 #endif

tools/perf/perf.c

@@ -292,6 +292,7 @@ static void handle_internal_command(int argc, const char **argv)
 		{ "top", cmd_top, 0 },
 		{ "annotate", cmd_annotate, 0 },
 		{ "version", cmd_version, 0 },
+		{ "trace", cmd_trace, 0 },
 	};
 	unsigned int i;
 	static const char ext[] = STRIP_EXTENSION;

tools/perf/util/abspath.c

@@ -50,7 +50,8 @@ const char *make_absolute_path(const char *path)
 			die ("Could not get current working directory");
 
 		if (last_elem) {
-			int len = strlen(buf);
+			len = strlen(buf);
+
 			if (len + strlen(last_elem) + 2 > PATH_MAX)
 				die ("Too long path name: '%s/%s'",
 						buf, last_elem);

tools/perf/util/cache.h

@@ -52,7 +52,6 @@ extern const char *perf_mailmap_file;
 extern void maybe_flush_or_die(FILE *, const char *);
 extern int copy_fd(int ifd, int ofd);
 extern int copy_file(const char *dst, const char *src, int mode);
-extern ssize_t read_in_full(int fd, void *buf, size_t count);
 extern ssize_t write_in_full(int fd, const void *buf, size_t count);
 extern void write_or_die(int fd, const void *buf, size_t count);
 extern int write_or_whine(int fd, const void *buf, size_t count, const char *msg);

tools/perf/util/callchain.c

@@ -50,6 +50,7 @@ rb_insert_callchain(struct rb_root *root, struct callchain_node *chain,
 			else
 				p = &(*p)->rb_right;
 			break;
+		case CHAIN_NONE:
 		default:
 			break;
 		}
@@ -143,6 +144,7 @@ int register_callchain_param(struct callchain_param *param)
 	case CHAIN_FLAT:
 		param->sort = sort_chain_flat;
 		break;
+	case CHAIN_NONE:
 	default:
 		return -1;
 	}

tools/perf/util/callchain.h

@@ -4,6 +4,7 @@
 #include "../perf.h"
 #include <linux/list.h>
 #include <linux/rbtree.h>
+#include "util.h"
 #include "symbol.h"
 
 enum chain_mode {

tools/perf/util/color.c

@@ -166,7 +166,7 @@ int perf_color_default_config(const char *var, const char *value, void *cb)
 	return perf_default_config(var, value, cb);
 }
 
-static int color_vfprintf(FILE *fp, const char *color, const char *fmt,
+static int __color_vfprintf(FILE *fp, const char *color, const char *fmt,
 		va_list args, const char *trail)
 {
 	int r = 0;
@@ -191,6 +191,10 @@ static int color_vfprintf(FILE *fp, const char *color, const char *fmt,
 	return r;
 }
 
+int color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args)
+{
+	return __color_vfprintf(fp, color, fmt, args, NULL);
+}
 
 
 int color_fprintf(FILE *fp, const char *color, const char *fmt, ...)
@@ -199,7 +203,7 @@ int color_fprintf(FILE *fp, const char *color, const char *fmt, ...)
 	int r;
 
 	va_start(args, fmt);
-	r = color_vfprintf(fp, color, fmt, args, NULL);
+	r = color_vfprintf(fp, color, fmt, args);
 	va_end(args);
 	return r;
 }
@@ -209,7 +213,7 @@ int color_fprintf_ln(FILE *fp, const char *color, const char *fmt, ...)
 	va_list args;
 	int r;
 	va_start(args, fmt);
-	r = color_vfprintf(fp, color, fmt, args, "\n");
+	r = __color_vfprintf(fp, color, fmt, args, "\n");
 	va_end(args);
 	return r;
 }
@@ -242,9 +246,9 @@ int color_fwrite_lines(FILE *fp, const char *color,
 	return 0;
 }
 
-char *get_percent_color(double percent)
+const char *get_percent_color(double percent)
 {
-	char *color = PERF_COLOR_NORMAL;
+	const char *color = PERF_COLOR_NORMAL;
 
 	/*
 	 * We color high-overhead entries in red, mid-overhead
@@ -263,7 +267,7 @@ char *get_percent_color(double percent)
 int percent_color_fprintf(FILE *fp, const char *fmt, double percent)
 {
 	int r;
-	char *color;
+	const char *color;
 
 	color = get_percent_color(percent);
 	r = color_fprintf(fp, color, fmt, percent);

tools/perf/util/color.h

@@ -32,10 +32,11 @@ int perf_color_default_config(const char *var, const char *value, void *cb);
 int perf_config_colorbool(const char *var, const char *value, int stdout_is_tty);
 void color_parse(const char *value, const char *var, char *dst);
 void color_parse_mem(const char *value, int len, const char *var, char *dst);
+int color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args);
 int color_fprintf(FILE *fp, const char *color, const char *fmt, ...);
 int color_fprintf_ln(FILE *fp, const char *color, const char *fmt, ...);
 int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *buf);
 int percent_color_fprintf(FILE *fp, const char *fmt, double percent);
-char *get_percent_color(double percent);
+const char *get_percent_color(double percent);
 
 #endif /* COLOR_H */

tools/perf/util/config.c

@@ -160,17 +160,18 @@ static int get_extended_base_var(char *name, int baselen, int c)
 	name[baselen++] = '.';
 
 	for (;;) {
-		int c = get_next_char();
-		if (c == '\n')
+		int ch = get_next_char();
+
+		if (ch == '\n')
 			return -1;
-		if (c == '"')
+		if (ch == '"')
 			break;
-		if (c == '\\') {
-			c = get_next_char();
-			if (c == '\n')
+		if (ch == '\\') {
+			ch = get_next_char();
+			if (ch == '\n')
 				return -1;
 		}
-		name[baselen++] = c;
+		name[baselen++] = ch;
 		if (baselen > MAXNAME / 2)
 			return -1;
 	}
@@ -530,6 +531,8 @@ static int store_aux(const char* key, const char* value, void *cb __used)
 					store.offset[store.seen] = ftell(config_file);
 			}
 		}
+	default:
+		break;
 	}
 	return 0;
 }
@@ -619,6 +622,7 @@ static ssize_t find_beginning_of_line(const char* contents, size_t size,
 		switch (contents[offset]) {
 			case '=': equal_offset = offset; break;
 			case ']': bracket_offset = offset; break;
+			default: break;
 		}
 	if (offset > 0 && contents[offset-1] == '\\') {
 		offset_ = offset;
@@ -742,9 +746,9 @@ int perf_config_set_multivar(const char* key, const char* value,
 			goto write_err_out;
 	} else {
 		struct stat st;
-		char* contents;
+		char *contents;
 		ssize_t contents_sz, copy_begin, copy_end;
-		int i, new_line = 0;
+		int new_line = 0;
 
 		if (value_regex == NULL)
 			store.value_regex = NULL;

tools/perf/util/debug.c

+/* For general debugging purposes */
+
+#include "../perf.h"
+
+#include <string.h>
+#include <stdarg.h>
+#include <stdio.h>
+
+#include "color.h"
+#include "event.h"
+#include "debug.h"
+
+int verbose = 0;
+int dump_trace = 0;
+
+int eprintf(const char *fmt, ...)
+{
+	va_list args;
+	int ret = 0;
+
+	if (verbose) {
+		va_start(args, fmt);
+		ret = vfprintf(stderr, fmt, args);
+		va_end(args);
+	}
+
+	return ret;
+}
+
+int dump_printf(const char *fmt, ...)
+{
+	va_list args;
+	int ret = 0;
+
+	if (dump_trace) {
+		va_start(args, fmt);
+		ret = vprintf(fmt, args);
+		va_end(args);
+	}
+
+	return ret;
+}
+
+static int dump_printf_color(const char *fmt, const char *color, ...)
+{
+	va_list args;
+	int ret = 0;
+
+	if (dump_trace) {
+		va_start(args, color);
+		ret = color_vfprintf(stdout, color, fmt, args);
+		va_end(args);
+	}
+
+	return ret;
+}
+
+
+void trace_event(event_t *event)
+{
+	unsigned char *raw_event = (void *)event;
+	const char *color = PERF_COLOR_BLUE;
+	int i, j;
+
+	if (!dump_trace)
+		return;
+
+	dump_printf(".");
+	dump_printf_color("\n. ... raw event: size %d bytes\n", color,
+			  event->header.size);
+
+	for (i = 0; i < event->header.size; i++) {
+		if ((i & 15) == 0) {
+			dump_printf(".");
+			dump_printf_color("  %04x: ", color, i);
+		}
+
+		dump_printf_color(" %02x", color, raw_event[i]);
+
+		if (((i & 15) == 15) || i == event->header.size-1) {
+			dump_printf_color("  ", color);
+			for (j = 0; j < 15-(i & 15); j++)
+				dump_printf_color("   ", color);
+			for (j = 0; j < (i & 15); j++) {
+				if (isprint(raw_event[i-15+j]))
+					dump_printf_color("%c", color,
+							  raw_event[i-15+j]);
+				else
+					dump_printf_color(".", color);
+			}
+			dump_printf_color("\n", color);
+		}
+	}
+	dump_printf(".\n");
+}

tools/perf/util/debug.h

+/* For debugging general purposes */
+
+extern int verbose;
+extern int dump_trace;
+
+int eprintf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2)));
+void trace_event(event_t *event);

tools/perf/util/event.h

+#ifndef __PERF_EVENT_H
+#define __PERF_EVENT_H
+#include "../perf.h"
+#include "util.h"
+#include <linux/list.h>
+
+enum {
+	SHOW_KERNEL	= 1,
+	SHOW_USER	= 2,
+	SHOW_HV		= 4,
+};
+
+/*
+ * PERF_SAMPLE_IP | PERF_SAMPLE_TID | *
+ */
+struct ip_event {
+	struct perf_event_header header;
+	u64 ip;
+	u32 pid, tid;
+	unsigned char __more_data[];
+};
+
+struct mmap_event {
+	struct perf_event_header header;
+	u32 pid, tid;
+	u64 start;
+	u64 len;
+	u64 pgoff;
+	char filename[PATH_MAX];
+};
+
+struct comm_event {
+	struct perf_event_header header;
+	u32 pid, tid;
+	char comm[16];
+};
+
+struct fork_event {
+	struct perf_event_header header;
+	u32 pid, ppid;
+	u32 tid, ptid;
+};
+
+struct lost_event {
+	struct perf_event_header header;
+	u64 id;
+	u64 lost;
+};
+
+/*
+ * PERF_FORMAT_ENABLED | PERF_FORMAT_RUNNING | PERF_FORMAT_ID
+ */
+struct read_event {
+	struct perf_event_header header;
+	u32 pid,tid;
+	u64 value;
+	u64 time_enabled;
+	u64 time_running;
+	u64 id;
+};
+
+typedef union event_union {
+	struct perf_event_header	header;
+	struct ip_event			ip;
+	struct mmap_event		mmap;
+	struct comm_event		comm;
+	struct fork_event		fork;
+	struct lost_event		lost;
+	struct read_event		read;
+} event_t;
+
+struct map {
+	struct list_head	node;
+	u64			start;
+	u64			end;
+	u64			pgoff;
+	u64			(*map_ip)(struct map *, u64);
+	struct dso		*dso;
+};
+
+static inline u64 map__map_ip(struct map *map, u64 ip)
+{
+	return ip - map->start + map->pgoff;
+}
+
+static inline u64 vdso__map_ip(struct map *map __used, u64 ip)
+{
+	return ip;
+}
+
+struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen);
+struct map *map__clone(struct map *self);
+int map__overlap(struct map *l, struct map *r);
+size_t map__fprintf(struct map *self, FILE *fp);
+
+#endif

tools/perf/util/exec_cmd.c

@@ -6,7 +6,6 @@
 
 #define MAX_ARGS	32
 
-extern char **environ;
 static const char *argv_exec_path;
 static const char *argv0_path;
 

tools/perf/util/header.c

@@ -237,9 +237,44 @@ struct perf_header *perf_header__read(int fd)
 	self->data_offset = f_header.data.offset;
 	self->data_size   = f_header.data.size;
 
-	lseek(fd, self->data_offset + self->data_size, SEEK_SET);
+	lseek(fd, self->data_offset, SEEK_SET);
 
 	self->frozen = 1;
 
 	return self;
 }
+
+u64 perf_header__sample_type(struct perf_header *header)
+{
+	u64 type = 0;
+	int i;
+
+	for (i = 0; i < header->attrs; i++) {
+		struct perf_header_attr *attr = header->attr[i];
+
+		if (!type)
+			type = attr->attr.sample_type;
+		else if (type != attr->attr.sample_type)
+			die("non matching sample_type");
+	}
+
+	return type;
+}
+
+struct perf_counter_attr *
+perf_header__find_attr(u64 id, struct perf_header *header)
+{
+	int i;
+
+	for (i = 0; i < header->attrs; i++) {
+		struct perf_header_attr *attr = header->attr[i];
+		int j;
+
+		for (j = 0; j < attr->ids; j++) {
+			if (attr->id[j] == id)
+				return &attr->attr;
+		}
+	}
+
+	return NULL;
+}

tools/perf/util/header.h

@@ -31,6 +31,10 @@ struct perf_header_attr *
 perf_header_attr__new(struct perf_counter_attr *attr);
 void perf_header_attr__add_id(struct perf_header_attr *self, u64 id);
 
+u64 perf_header__sample_type(struct perf_header *header);
+struct perf_counter_attr *
+perf_header__find_attr(u64 id, struct perf_header *header);
+
 
 struct perf_header *perf_header__new(void);
 

tools/perf/util/map.c

+#include "event.h"
+#include "symbol.h"
+#include <stdlib.h>
+#include <string.h>
+#include <stdio.h>
+
+static inline int is_anon_memory(const char *filename)
+{
+	return strcmp(filename, "//anon") == 0;
+}
+
+static int strcommon(const char *pathname, char *cwd, int cwdlen)
+{
+	int n = 0;
+
+	while (n < cwdlen && pathname[n] == cwd[n])
+		++n;
+
+	return n;
+}
+
+ struct map *map__new(struct mmap_event *event, char *cwd, int cwdlen)
+{
+	struct map *self = malloc(sizeof(*self));
+
+	if (self != NULL) {
+		const char *filename = event->filename;
+		char newfilename[PATH_MAX];
+		int anon;
+
+		if (cwd) {
+			int n = strcommon(filename, cwd, cwdlen);
+
+			if (n == cwdlen) {
+				snprintf(newfilename, sizeof(newfilename),
+					 ".%s", filename + n);
+				filename = newfilename;
+			}
+		}
+
+		anon = is_anon_memory(filename);
+
+		if (anon) {
+			snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", event->pid);
+			filename = newfilename;
+		}
+
+		self->start = event->start;
+		self->end   = event->start + event->len;
+		self->pgoff = event->pgoff;
+
+		self->dso = dsos__findnew(filename);
+		if (self->dso == NULL)
+			goto out_delete;
+
+		if (self->dso == vdso || anon)
+			self->map_ip = vdso__map_ip;
+		else
+			self->map_ip = map__map_ip;
+	}
+	return self;
+out_delete:
+	free(self);
+	return NULL;
+}
+
+struct map *map__clone(struct map *self)
+{
+	struct map *map = malloc(sizeof(*self));
+
+	if (!map)
+		return NULL;
+
+	memcpy(map, self, sizeof(*self));
+
+	return map;
+}
+
+int map__overlap(struct map *l, struct map *r)
+{
+	if (l->start > r->start) {
+		struct map *t = l;
+		l = r;
+		r = t;
+	}
+
+	if (l->end > r->start)
+		return 1;
+
+	return 0;
+}
+
+size_t map__fprintf(struct map *self, FILE *fp)
+{
+	return fprintf(fp, " %Lx-%Lx %Lx %s\n",
+		       self->start, self->end, self->pgoff, self->dso->name);
+}

tools/perf/util/module.c

@@ -436,9 +436,9 @@ static int mod_dso__load_module_paths(struct mod_dso *self)
 		goto out_failure;
 
 	while (!feof(file)) {
-		char *path, *name, *tmp;
+		char *name, *tmp;
 		struct module *module;
-		int line_len, len;
+		int line_len;
 
 		line_len = getline(&line, &n, file);
 		if (line_len < 0)

tools/perf/util/parse-events.h

-
+#ifndef _PARSE_EVENTS_H
+#define _PARSE_EVENTS_H
 /*
  * Parse symbolic events/counts passed in as options:
  */
 
 struct option;
 
+struct tracepoint_path {
+	char *system;
+	char *name;
+	struct tracepoint_path *next;
+};
+
+extern struct tracepoint_path *tracepoint_id_to_path(u64 config);
+
 extern int			nr_counters;
 
 extern struct perf_counter_attr attrs[MAX_COUNTERS];
 
-extern char *event_name(int ctr);
-extern char *__event_name(int type, u64 config);
+extern const char *event_name(int ctr);
+extern const char *__event_name(int type, u64 config);
 
 extern int parse_events(const struct option *opt, const char *str, int unset);
 
@@ -21,3 +30,5 @@ extern void print_events(void);
 extern char debugfs_path[];
 extern int valid_debugfs_mount(const char *debugfs);
 
+
+#endif /* _PARSE_EVENTS_H */

tools/perf/util/parse-options.c

@@ -53,6 +53,12 @@ static int get_value(struct parse_opt_ctx_t *p,
 		case OPTION_SET_INT:
 		case OPTION_SET_PTR:
 			return opterror(opt, "takes no value", flags);
+		case OPTION_END:
+		case OPTION_ARGUMENT:
+		case OPTION_GROUP:
+		case OPTION_STRING:
+		case OPTION_INTEGER:
+		case OPTION_LONG:
 		default:
 			break;
 		}
@@ -130,6 +136,9 @@ static int get_value(struct parse_opt_ctx_t *p,
 			return opterror(opt, "expects a numerical value", flags);
 		return 0;
 
+	case OPTION_END:
+	case OPTION_ARGUMENT:
+	case OPTION_GROUP:
 	default:
 		die("should not happen, someone must be hit on the forehead");
 	}
@@ -296,6 +305,8 @@ int parse_options_step(struct parse_opt_ctx_t *ctx,
 				return parse_options_usage(usagestr, options);
 			case -2:
 				goto unknown;
+			default:
+				break;
 			}
 			if (ctx->opt)
 				check_typos(arg + 1, options);
@@ -314,6 +325,8 @@ int parse_options_step(struct parse_opt_ctx_t *ctx,
 					ctx->argv[0] = strdup(ctx->opt - 1);
 					*(char *)ctx->argv[0] = '-';
 					goto unknown;
+				default:
+					break;
 				}
 			}
 			continue;
@@ -336,6 +349,8 @@ int parse_options_step(struct parse_opt_ctx_t *ctx,
 			return parse_options_usage(usagestr, options);
 		case -2:
 			goto unknown;
+		default:
+			break;
 		}
 		continue;
 unknown:
@@ -456,6 +471,13 @@ int usage_with_options_internal(const char * const *usagestr,
 			}
 			break;
 		default: /* OPTION_{BIT,BOOLEAN,SET_INT,SET_PTR} */
+		case OPTION_END:
+		case OPTION_GROUP:
+		case OPTION_BIT:
+		case OPTION_BOOLEAN:
+		case OPTION_SET_INT:
+		case OPTION_SET_PTR:
+		case OPTION_LONG:
 			break;
 		}
 

tools/perf/util/path.c

@@ -17,7 +17,7 @@ static char bad_path[] = "/bad-path/";
  * Two hacks:
  */
 
-static char *get_perf_dir(void)
+static const char *get_perf_dir(void)
 {
 	return ".";
 }
@@ -38,8 +38,9 @@ size_t strlcpy(char *dest, const char *src, size_t size)
 static char *get_pathname(void)
 {
 	static char pathname_array[4][PATH_MAX];
-	static int index;
-	return pathname_array[3 & ++index];
+	static int idx;
+
+	return pathname_array[3 & ++idx];
 }
 
 static char *cleanup_path(char *path)
@@ -161,20 +162,24 @@ int perf_mkstemp(char *path, size_t len, const char *template)
 }
 
 
-const char *make_relative_path(const char *abs, const char *base)
+const char *make_relative_path(const char *abs_path, const char *base)
 {
 	static char buf[PATH_MAX + 1];
 	int baselen;
+
 	if (!base)
-		return abs;
+		return abs_path;
+
 	baselen = strlen(base);
-	if (prefixcmp(abs, base))
-		return abs;
-	if (abs[baselen] == '/')
+	if (prefixcmp(abs_path, base))
+		return abs_path;
+	if (abs_path[baselen] == '/')
 		baselen++;
 	else if (base[baselen - 1] != '/')
-		return abs;
-	strcpy(buf, abs + baselen);
+		return abs_path;
+
+	strcpy(buf, abs_path + baselen);
+
 	return buf;
 }
 

tools/perf/util/run-command.c

@@ -262,7 +262,7 @@ int run_hook(const char *index_file, const char *name, ...)
 {
 	struct child_process hook;
 	const char **argv = NULL, *env[2];
-	char index[PATH_MAX];
+	char idx[PATH_MAX];
 	va_list args;
 	int ret;
 	size_t i = 0, alloc = 0;
@@ -284,8 +284,8 @@ int run_hook(const char *index_file, const char *name, ...)
 	hook.no_stdin = 1;
 	hook.stdout_to_stderr = 1;
 	if (index_file) {
-		snprintf(index, sizeof(index), "PERF_INDEX_FILE=%s", index_file);
-		env[0] = index;
+		snprintf(idx, sizeof(idx), "PERF_INDEX_FILE=%s", index_file);
+		env[0] = idx;
 		env[1] = NULL;
 		hook.env = env;
 	}

tools/perf/util/symbol.h

@@ -6,6 +6,7 @@
 #include <linux/list.h>
 #include <linux/rbtree.h>
 #include "module.h"
+#include "event.h"
 
 #ifdef HAVE_CPLUS_DEMANGLE
 extern char *cplus_demangle(const char *, int);
@@ -54,7 +55,7 @@ struct dso {
 	char		 name[0];
 };
 
-const char *sym_hist_filter;
+extern const char *sym_hist_filter;
 
 typedef int (*symbol_filter_t)(struct dso *self, struct symbol *sym);
 
@@ -72,9 +73,20 @@ int dso__load_kernel(struct dso *self, const char *vmlinux,
 		     symbol_filter_t filter, int verbose, int modules);
 int dso__load_modules(struct dso *self, symbol_filter_t filter, int verbose);
 int dso__load(struct dso *self, symbol_filter_t filter, int verbose);
+struct dso *dsos__findnew(const char *name);
+void dsos__fprintf(FILE *fp);
 
 size_t dso__fprintf(struct dso *self, FILE *fp);
 char dso__symtab_origin(const struct dso *self);
 
+int load_kernel(void);
+
 void symbol__init(void);
+
+extern struct list_head dsos;
+extern struct dso *kernel_dso;
+extern struct dso *vdso;
+extern struct dso *hypervisor_dso;
+extern const char *vmlinux_name;
+extern int   modules;
 #endif /* _PERF_SYMBOL_ */

tools/perf/util/thread.h

+#include <linux/rbtree.h>
+#include <linux/list.h>
+#include <unistd.h>
+#include "symbol.h"
+
+struct thread {
+	struct rb_node	 rb_node;
+	struct list_head maps;
+	pid_t		 pid;
+	char		 *comm;
+};
+
+int thread__set_comm(struct thread *self, const char *comm);
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match);
+struct thread *
+register_idle_thread(struct rb_root *threads, struct thread **last_match);
+void thread__insert_map(struct thread *self, struct map *map);
+int thread__fork(struct thread *self, struct thread *parent);
+struct map *thread__find_map(struct thread *self, u64 ip);
+size_t threads__fprintf(FILE *fp, struct rb_root *threads);

tools/perf/util/util.h

@@ -39,10 +39,6 @@
 /* Approximation of the length of the decimal representation of this type. */
 #define decimal_length(x)	((int)(sizeof(x) * 2.56 + 0.5) + 1)
 
-#if !defined(__APPLE__) && !defined(__FreeBSD__)  && !defined(__USLC__) && !defined(_M_UNIX)
-#define _XOPEN_SOURCE 600 /* glibc2 and AIX 5.3L need 500, OpenBSD needs 600 for S_ISLNK() */
-#define _XOPEN_SOURCE_EXTENDED 1 /* AIX 5.3L needs this */
-#endif
 #define _ALL_SOURCE 1
 #define _GNU_SOURCE 1
 #define _BSD_SOURCE 1
@@ -83,6 +79,7 @@
 #include <inttypes.h>
 #include "../../../include/linux/magic.h"
 
+
 #ifndef NO_ICONV
 #include <iconv.h>
 #endif
@@ -310,6 +307,7 @@ static inline int has_extension(const char *filename, const char *ext)
 #undef isspace
 #undef isdigit
 #undef isalpha
+#undef isprint
 #undef isalnum
 #undef tolower
 #undef toupper