Merge tag 'arc-4.0-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc

Pull ARC fixes from Vineet Gupta:
 - Fix for /proc/<pid>/maps "stack" vma annotation
 - sched stats not printing correct sleeping task PC
 - perf not reporting page faults

* tag 'arc-4.0-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:
  ARC: Fix thread_saved_pc()
  ARC: Fix KSTK_ESP()
  ARC: perf: Enable generic software events
  ARC: Make arc_unwind_core accessible externally

arch/arc/include/asm/processor.h

@@ -47,9 +47,6 @@ struct thread_struct {
 /* Forward declaration, a strange C thing */
 struct task_struct;
 
-/* Return saved PC of a blocked thread  */
-unsigned long thread_saved_pc(struct task_struct *t);
-
 #define task_pt_regs(p) \
 	((struct pt_regs *)(THREAD_SIZE + (void *)task_stack_page(p)) - 1)
 
@@ -72,18 +69,21 @@ unsigned long thread_saved_pc(struct task_struct *t);
 #define release_segments(mm)        do { } while (0)
 
 #define KSTK_EIP(tsk)   (task_pt_regs(tsk)->ret)
+#define KSTK_ESP(tsk)   (task_pt_regs(tsk)->sp)
 
 /*
  * Where abouts of Task's sp, fp, blink when it was last seen in kernel mode.
  * Look in process.c for details of kernel stack layout
  */
-#define KSTK_ESP(tsk)   (tsk->thread.ksp)
+#define TSK_K_ESP(tsk)		(tsk->thread.ksp)
 
-#define KSTK_REG(tsk, off)	(*((unsigned int *)(KSTK_ESP(tsk) + \
+#define TSK_K_REG(tsk, off)	(*((unsigned int *)(TSK_K_ESP(tsk) + \
 					sizeof(struct callee_regs) + off)))
 
-#define KSTK_BLINK(tsk) KSTK_REG(tsk, 4)
-#define KSTK_FP(tsk)    KSTK_REG(tsk, 0)
+#define TSK_K_BLINK(tsk)	TSK_K_REG(tsk, 4)
+#define TSK_K_FP(tsk)		TSK_K_REG(tsk, 0)
+
+#define thread_saved_pc(tsk)	TSK_K_BLINK(tsk)
 
 extern void start_thread(struct pt_regs * regs, unsigned long pc,
 			 unsigned long usp);

arch/arc/include/asm/stacktrace.h

@@ -0,0 +1,37 @@
+/*
+ * Copyright (C) 2014-15 Synopsys, Inc. (www.synopsys.com)
+ * Copyright (C) 2007-2010, 2011-2012 Synopsys, Inc. (www.synopsys.com)
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef __ASM_STACKTRACE_H
+#define __ASM_STACKTRACE_H
+
+#include <linux/sched.h>
+
+/**
+ * arc_unwind_core - Unwind the kernel mode stack for an execution context
+ * @tsk:		NULL for current task, specific task otherwise
+ * @regs:		pt_regs used to seed the unwinder {SP, FP, BLINK, PC}
+ * 			If NULL, use pt_regs of @tsk (if !NULL) otherwise
+ * 			use the current values of {SP, FP, BLINK, PC}
+ * @consumer_fn:	Callback invoked for each frame unwound
+ * 			Returns 0 to continue unwinding, -1 to stop
+ * @arg:		Arg to callback
+ *
+ * Returns the address of first function in stack
+ *
+ * Semantics:
+ *  - synchronous unwinding (e.g. dump_stack): @tsk  NULL, @regs  NULL
+ *  - Asynchronous unwinding of sleeping task: @tsk !NULL, @regs  NULL
+ *  - Asynchronous unwinding of intr/excp etc: @tsk !NULL, @regs !NULL
+ */
+notrace noinline unsigned int arc_unwind_core(
+	struct task_struct *tsk, struct pt_regs *regs,
+	int (*consumer_fn) (unsigned int, void *),
+	void *arg);
+
+#endif /* __ASM_STACKTRACE_H */

arch/arc/kernel/process.c

@@ -192,29 +192,6 @@ int dump_fpu(struct pt_regs *regs, elf_fpregset_t *fpu)
 	return 0;
 }
 
-/*
- * API: expected by schedular Code: If thread is sleeping where is that.
- * What is this good for? it will be always the scheduler or ret_from_fork.
- * So we hard code that anyways.
- */
-unsigned long thread_saved_pc(struct task_struct *t)
-{
-	struct pt_regs *regs = task_pt_regs(t);
-	unsigned long blink = 0;
-
-	/*
-	 * If the thread being queried for in not itself calling this, then it
-	 * implies it is not executing, which in turn implies it is sleeping,
-	 * which in turn implies it got switched OUT by the schedular.
-	 * In that case, it's kernel mode blink can reliably retrieved as per
-	 * the picture above (right above pt_regs).
-	 */
-	if (t != current && t->state != TASK_RUNNING)
-		blink = *((unsigned int *)regs - 1);
-
-	return blink;
-}
-
 int elf_check_arch(const struct elf32_hdr *x)
 {
 	unsigned int eflags;

arch/arc/kernel/stacktrace.c

@@ -43,6 +43,10 @@ static void seed_unwind_frame_info(struct task_struct *tsk,
 				   struct pt_regs *regs,
 				   struct unwind_frame_info *frame_info)
 {
+	/*
+	 * synchronous unwinding (e.g. dump_stack)
+	 *  - uses current values of SP and friends
+	 */
 	if (tsk == NULL && regs == NULL) {
 		unsigned long fp, sp, blink, ret;
 		frame_info->task = current;
@@ -61,12 +65,17 @@ static void seed_unwind_frame_info(struct task_struct *tsk,
 		frame_info->regs.r63 = ret;
 		frame_info->call_frame = 0;
 	} else if (regs == NULL) {
+		/*
+		 * Asynchronous unwinding of sleeping task
+		 *  - Gets SP etc from task's pt_regs (saved bottom of kernel
+		 *    mode stack of task)
+		 */
 
 		frame_info->task = tsk;
 
-		frame_info->regs.r27 = KSTK_FP(tsk);
-		frame_info->regs.r28 = KSTK_ESP(tsk);
-		frame_info->regs.r31 = KSTK_BLINK(tsk);
+		frame_info->regs.r27 = TSK_K_FP(tsk);
+		frame_info->regs.r28 = TSK_K_ESP(tsk);
+		frame_info->regs.r31 = TSK_K_BLINK(tsk);
 		frame_info->regs.r63 = (unsigned int)__switch_to;
 
 		/* In the prologue of __switch_to, first FP is saved on stack
@@ -83,6 +92,10 @@ static void seed_unwind_frame_info(struct task_struct *tsk,
 		frame_info->call_frame = 0;
 
 	} else {
+		/*
+		 * Asynchronous unwinding of intr/exception
+		 *  - Just uses the pt_regs passed
+		 */
 		frame_info->task = tsk;
 
 		frame_info->regs.r27 = regs->fp;
@@ -95,7 +108,7 @@ static void seed_unwind_frame_info(struct task_struct *tsk,
 
 #endif
 
-static noinline unsigned int
+notrace noinline unsigned int
 arc_unwind_core(struct task_struct *tsk, struct pt_regs *regs,
 		int (*consumer_fn) (unsigned int, void *), void *arg)
 {

arch/arc/kernel/unaligned.c

@@ -12,6 +12,7 @@
  */
 
 #include <linux/types.h>
+#include <linux/perf_event.h>
 #include <linux/ptrace.h>
 #include <linux/uaccess.h>
 #include <asm/disasm.h>
@@ -253,6 +254,7 @@ int misaligned_fixup(unsigned long address, struct pt_regs *regs,
 		}
 	}
 
+	perf_sw_event(PERF_COUNT_SW_ALIGNMENT_FAULTS, 1, regs, address);
 	return 0;
 
 fault:

arch/arc/mm/fault.c

@@ -14,6 +14,7 @@
 #include <linux/ptrace.h>
 #include <linux/uaccess.h>
 #include <linux/kdebug.h>
+#include <linux/perf_event.h>
 #include <asm/pgalloc.h>
 #include <asm/mmu.h>
 
@@ -139,13 +140,20 @@ good_area:
 			return;
 	}
 
+	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
+
 	if (likely(!(fault & VM_FAULT_ERROR))) {
 		if (flags & FAULT_FLAG_ALLOW_RETRY) {
 			/* To avoid updating stats twice for retry case */
-			if (fault & VM_FAULT_MAJOR)
+			if (fault & VM_FAULT_MAJOR) {
 				tsk->maj_flt++;
-			else
+				perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
+					      regs, address);
+			} else {
 				tsk->min_flt++;
+				perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
+					      regs, address);
+			}
 
 			if (fault & VM_FAULT_RETRY) {
 				flags &= ~FAULT_FLAG_ALLOW_RETRY;