Merge tag 'xtensa-20150830' of git://github.com/czankel/xtensa-linux

Pull xtensa updates from Chris Zankel:
 "Xtensa fixes and improvements for 4.3:

   - reimplement DMA API using common helpers
   - implement counting and sampling perf events using hardware perf
     counters
   - add fake NMI support for hardware perf counters
   - fix THREADPTR register reloading on return to userspace
   - keep exception/interrupt stack continuous for debugger
   - improve vmlinux.lds.S post-processing"

* tag 'xtensa-20150830' of git://github.com/czankel/xtensa-linux:
  xtensa: improve vmlinux.lds.S sed post-processing
  xtensa: drop unused irq_err_count
  xtensa: implement fake NMI
  xtensa: don't touch EXC_TABLE_FIXUP in _switch_to
  xtensa: fix kernel register spilling
  xtensa: reorganize irq flags tracing
  perf tools: xtensa: add DWARF register names
  xtensa: implement counting and sampling perf events
  xtensa: count software page fault perf events
  xtensa: add profiling IRQ type to xtensa_irq_map
  xtensa: select PERF_USE_VMALLOC for cache-aliasing configurations
  xtensa: move oprofile stack tracing to stacktrace.c
  xtensa: keep exception/interrupt stack continuous
  xtensa: clean up Kconfig dependencies for custom cores
  xtensa: reimplement DMA API using common helpers
  xtensa: fix threadptr reload on return to userspace
  xtensa: ISS: add missing va_end into split_if_spec

9 files changed, 142 insertions, 168 deletions

diff --git a/arch/xtensa/include/asm/Kbuild b/arch/xtensa/include/asm/Kbuild
index 5b478accd5fc..63c223dff5f1 100644
--- a/arch/xtensa/include/asm/Kbuild
+++ b/arch/xtensa/include/asm/Kbuild
@@ -2,7 +2,6 @@ generic-y += bitsperlong.h
 generic-y += bug.h
 generic-y += clkdev.h
 generic-y += cputime.h
-generic-y += device.h
 generic-y += div64.h
 generic-y += emergency-restart.h
 generic-y += errno.h
diff --git a/arch/xtensa/include/asm/atomic.h b/arch/xtensa/include/asm/atomic.h
index 00b7d46b35b8..ebcd1f6fc8cb 100644
--- a/arch/xtensa/include/asm/atomic.h
+++ b/arch/xtensa/include/asm/atomic.h
@@ -29,7 +29,7 @@
  *
  * Locking interrupts looks like this:
  *
- *    rsil a15, LOCKLEVEL
+ *    rsil a15, TOPLEVEL
  *    <code>
  *    wsr  a15, PS
  *    rsync
@@ -106,7 +106,7 @@ static inline void atomic_##op(int i, atomic_t * v)			\
 	unsigned int vval;						\
 									\
 	__asm__ __volatile__(						\
-			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"\
+			"       rsil    a15, "__stringify(TOPLEVEL)"\n"\
 			"       l32i    %0, %2, 0\n"			\
 			"       " #op " %0, %0, %1\n"			\
 			"       s32i    %0, %2, 0\n"			\
@@ -124,7 +124,7 @@ static inline int atomic_##op##_return(int i, atomic_t * v)		\
 	unsigned int vval;						\
 									\
 	__asm__ __volatile__(						\
-			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"	\
+			"       rsil    a15,"__stringify(TOPLEVEL)"\n"	\
 			"       l32i    %0, %2, 0\n"			\
 			"       " #op " %0, %0, %1\n"			\
 			"       s32i    %0, %2, 0\n"			\
@@ -272,7 +272,7 @@ static inline void atomic_clear_mask(unsigned int mask, atomic_t *v)
 	unsigned int vval;
 
 	__asm__ __volatile__(
-			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       rsil    a15,"__stringify(TOPLEVEL)"\n"
 			"       l32i    %0, %2, 0\n"
 			"       xor     %1, %4, %3\n"
 			"       and     %0, %0, %4\n"
@@ -306,7 +306,7 @@ static inline void atomic_set_mask(unsigned int mask, atomic_t *v)
 	unsigned int vval;
 
 	__asm__ __volatile__(
-			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       rsil    a15,"__stringify(TOPLEVEL)"\n"
 			"       l32i    %0, %2, 0\n"
 			"       or      %0, %0, %1\n"
 			"       s32i    %0, %2, 0\n"
diff --git a/arch/xtensa/include/asm/cmpxchg.h b/arch/xtensa/include/asm/cmpxchg.h
index 370b26f38414..201e9009efd8 100644
--- a/arch/xtensa/include/asm/cmpxchg.h
+++ b/arch/xtensa/include/asm/cmpxchg.h
@@ -34,7 +34,7 @@ __cmpxchg_u32(volatile int *p, int old, int new)
 	return new;
 #else
 	__asm__ __volatile__(
-			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       rsil    a15, "__stringify(TOPLEVEL)"\n"
 			"       l32i    %0, %1, 0\n"
 			"       bne     %0, %2, 1f\n"
 			"       s32i    %3, %1, 0\n"
@@ -123,7 +123,7 @@ static inline unsigned long xchg_u32(volatile int * m, unsigned long val)
 #else
 	unsigned long tmp;
 	__asm__ __volatile__(
-			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       rsil    a15, "__stringify(TOPLEVEL)"\n"
 			"       l32i    %0, %1, 0\n"
 			"       s32i    %2, %1, 0\n"
 			"       wsr     a15, ps\n"
diff --git a/arch/xtensa/include/asm/device.h b/arch/xtensa/include/asm/device.h
new file mode 100644
index 000000000000..fe1f5c878493
--- /dev/null
+++ b/arch/xtensa/include/asm/device.h
@@ -0,0 +1,19 @@
+/*
+ * Arch specific extensions to struct device
+ *
+ * This file is released under the GPLv2
+ */
+#ifndef _ASM_XTENSA_DEVICE_H
+#define _ASM_XTENSA_DEVICE_H
+
+struct dma_map_ops;
+
+struct dev_archdata {
+	/* DMA operations on that device */
+	struct dma_map_ops *dma_ops;
+};
+
+struct pdev_archdata {
+};
+
+#endif /* _ASM_XTENSA_DEVICE_H */
diff --git a/arch/xtensa/include/asm/dma-mapping.h b/arch/xtensa/include/asm/dma-mapping.h
index 1f5f6dc09736..f01cb3044e50 100644
--- a/arch/xtensa/include/asm/dma-mapping.h
+++ b/arch/xtensa/include/asm/dma-mapping.h
@@ -1,11 +1,10 @@
 /*
- * include/asm-xtensa/dma-mapping.h
- *
  * This file is subject to the terms and conditions of the GNU General Public
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
  * Copyright (C) 2003 - 2005 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
  */
 
 #ifndef _XTENSA_DMA_MAPPING_H
@@ -13,142 +12,67 @@
 
 #include <asm/cache.h>
 #include <asm/io.h>
+
+#include <asm-generic/dma-coherent.h>
+
 #include <linux/mm.h>
 #include <linux/scatterlist.h>
 
 #define DMA_ERROR_CODE		(~(dma_addr_t)0x0)
 
-/*
- * DMA-consistent mapping functions.
- */
-
-extern void *consistent_alloc(int, size_t, dma_addr_t, unsigned long);
-extern void consistent_free(void*, size_t, dma_addr_t);
-extern void consistent_sync(void*, size_t, int);
-
-#define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f)
-#define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h)
-
-void *dma_alloc_coherent(struct device *dev, size_t size,
-			   dma_addr_t *dma_handle, gfp_t flag);
+extern struct dma_map_ops xtensa_dma_map_ops;
 
-void dma_free_coherent(struct device *dev, size_t size,
-			 void *vaddr, dma_addr_t dma_handle);
-
-static inline dma_addr_t
-dma_map_single(struct device *dev, void *ptr, size_t size,
-	       enum dma_data_direction direction)
-{
-	BUG_ON(direction == DMA_NONE);
-	consistent_sync(ptr, size, direction);
-	return virt_to_phys(ptr);
-}
-
-static inline void
-dma_unmap_single(struct device *dev, dma_addr_t dma_addr, size_t size,
-		 enum dma_data_direction direction)
+static inline struct dma_map_ops *get_dma_ops(struct device *dev)
 {
-	BUG_ON(direction == DMA_NONE);
+	if (dev && dev->archdata.dma_ops)
+		return dev->archdata.dma_ops;
+	else
+		return &xtensa_dma_map_ops;
 }
 
-static inline int
-dma_map_sg(struct device *dev, struct scatterlist *sglist, int nents,
-	   enum dma_data_direction direction)
-{
-	int i;
-	struct scatterlist *sg;
-
-	BUG_ON(direction == DMA_NONE);
-
-	for_each_sg(sglist, sg, nents, i) {
-		BUG_ON(!sg_page(sg));
+#include <asm-generic/dma-mapping-common.h>
 
-		sg->dma_address = sg_phys(sg);
-		consistent_sync(sg_virt(sg), sg->length, direction);
-	}
+#define dma_alloc_noncoherent(d, s, h, f) dma_alloc_attrs(d, s, h, f, NULL)
+#define dma_free_noncoherent(d, s, v, h) dma_free_attrs(d, s, v, h, NULL)
+#define dma_alloc_coherent(d, s, h, f) dma_alloc_attrs(d, s, h, f, NULL)
+#define dma_free_coherent(d, s, c, h) dma_free_attrs(d, s, c, h, NULL)
 
-	return nents;
-}
-
-static inline dma_addr_t
-dma_map_page(struct device *dev, struct page *page, unsigned long offset,
-	     size_t size, enum dma_data_direction direction)
-{
-	BUG_ON(direction == DMA_NONE);
-	return (dma_addr_t)(page_to_pfn(page)) * PAGE_SIZE + offset;
-}
-
-static inline void
-dma_unmap_page(struct device *dev, dma_addr_t dma_address, size_t size,
-	       enum dma_data_direction direction)
+static inline void *dma_alloc_attrs(struct device *dev, size_t size,
+				    dma_addr_t *dma_handle, gfp_t gfp,
+				    struct dma_attrs *attrs)
 {
-	BUG_ON(direction == DMA_NONE);
-}
+	void *ret;
+	struct dma_map_ops *ops = get_dma_ops(dev);
 
+	if (dma_alloc_from_coherent(dev, size, dma_handle, &ret))
+		return ret;
 
-static inline void
-dma_unmap_sg(struct device *dev, struct scatterlist *sg, int nhwentries,
-	     enum dma_data_direction direction)
-{
-	BUG_ON(direction == DMA_NONE);
-}
-
-static inline void
-dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
-		enum dma_data_direction direction)
-{
-	consistent_sync((void *)bus_to_virt(dma_handle), size, direction);
-}
+	ret = ops->alloc(dev, size, dma_handle, gfp, attrs);
+	debug_dma_alloc_coherent(dev, size, *dma_handle, ret);
 
-static inline void
-dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
-		           size_t size, enum dma_data_direction direction)
-{
-	consistent_sync((void *)bus_to_virt(dma_handle), size, direction);
+	return ret;
 }
 
-static inline void
-dma_sync_single_range_for_cpu(struct device *dev, dma_addr_t dma_handle,
-		      unsigned long offset, size_t size,
-		      enum dma_data_direction direction)
-{
-
-	consistent_sync((void *)bus_to_virt(dma_handle)+offset,size,direction);
-}
-
-static inline void
-dma_sync_single_range_for_device(struct device *dev, dma_addr_t dma_handle,
-		      unsigned long offset, size_t size,
-		      enum dma_data_direction direction)
+static inline void dma_free_attrs(struct device *dev, size_t size,
+				  void *vaddr, dma_addr_t dma_handle,
+				  struct dma_attrs *attrs)
 {
+	struct dma_map_ops *ops = get_dma_ops(dev);
 
-	consistent_sync((void *)bus_to_virt(dma_handle)+offset,size,direction);
-}
-static inline void
-dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sglist, int nelems,
-		 enum dma_data_direction dir)
-{
-	int i;
-	struct scatterlist *sg;
+	if (dma_release_from_coherent(dev, get_order(size), vaddr))
+		return;
 
-	for_each_sg(sglist, sg, nelems, i)
-		consistent_sync(sg_virt(sg), sg->length, dir);
+	ops->free(dev, size, vaddr, dma_handle, attrs);
+	debug_dma_free_coherent(dev, size, vaddr, dma_handle);
 }
 
-static inline void
-dma_sync_sg_for_device(struct device *dev, struct scatterlist *sglist,
-		       int nelems, enum dma_data_direction dir)
-{
-	int i;
-	struct scatterlist *sg;
-
-	for_each_sg(sglist, sg, nelems, i)
-		consistent_sync(sg_virt(sg), sg->length, dir);
-}
 static inline int
 dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
 {
-	return 0;
+	struct dma_map_ops *ops = get_dma_ops(dev);
+
+	debug_dma_mapping_error(dev, dma_addr);
+	return ops->mapping_error(dev, dma_addr);
 }
 
 static inline int
@@ -168,39 +92,7 @@ dma_set_mask(struct device *dev, u64 mask)
 	return 0;
 }
 
-static inline void
-dma_cache_sync(struct device *dev, void *vaddr, size_t size,
-	       enum dma_data_direction direction)
-{
-	consistent_sync(vaddr, size, direction);
-}
-
-/* Not supported for now */
-static inline int dma_mmap_coherent(struct device *dev,
-				    struct vm_area_struct *vma, void *cpu_addr,
-				    dma_addr_t dma_addr, size_t size)
-{
-	return -EINVAL;
-}
-
-static inline int dma_get_sgtable(struct device *dev, struct sg_table *sgt,
-				  void *cpu_addr, dma_addr_t dma_addr,
-				  size_t size)
-{
-	return -EINVAL;
-}
-
-static inline void *dma_alloc_attrs(struct device *dev, size_t size,
-				    dma_addr_t *dma_handle, gfp_t flag,
-				    struct dma_attrs *attrs)
-{
-	return NULL;
-}
-
-static inline void dma_free_attrs(struct device *dev, size_t size,
-				  void *vaddr, dma_addr_t dma_handle,
-				  struct dma_attrs *attrs)
-{
-}
+void dma_cache_sync(struct device *dev, void *vaddr, size_t size,
+		    enum dma_data_direction direction);
 
 #endif	/* _XTENSA_DMA_MAPPING_H */
diff --git a/arch/xtensa/include/asm/irqflags.h b/arch/xtensa/include/asm/irqflags.h
index ea36674c6ec5..8e090c709046 100644
--- a/arch/xtensa/include/asm/irqflags.h
+++ b/arch/xtensa/include/asm/irqflags.h
@@ -6,6 +6,7 @@
  * for more details.
  *
  * Copyright (C) 2001 - 2005 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
  */
 
 #ifndef _XTENSA_IRQFLAGS_H
@@ -23,8 +24,27 @@ static inline unsigned long arch_local_save_flags(void)
 static inline unsigned long arch_local_irq_save(void)
 {
 	unsigned long flags;
-	asm volatile("rsil %0, "__stringify(LOCKLEVEL)
+#if XTENSA_FAKE_NMI
+#if defined(CONFIG_DEBUG_KERNEL) && (LOCKLEVEL | TOPLEVEL) >= XCHAL_DEBUGLEVEL
+	unsigned long tmp;
+
+	asm volatile("rsr	%0, ps\t\n"
+		     "extui	%1, %0, 0, 4\t\n"
+		     "bgei	%1, "__stringify(LOCKLEVEL)", 1f\t\n"
+		     "rsil	%0, "__stringify(LOCKLEVEL)"\n"
+		     "1:"
+		     : "=a" (flags), "=a" (tmp) :: "memory");
+#else
+	asm volatile("rsr	%0, ps\t\n"
+		     "or	%0, %0, %1\t\n"
+		     "xsr	%0, ps\t\n"
+		     "rsync"
+		     : "=&a" (flags) : "a" (LOCKLEVEL) : "memory");
+#endif
+#else
+	asm volatile("rsil	%0, "__stringify(LOCKLEVEL)
 		     : "=a" (flags) :: "memory");
+#endif
 	return flags;
 }
 
diff --git a/arch/xtensa/include/asm/processor.h b/arch/xtensa/include/asm/processor.h
index b61bdf0eea25..83e2e4bc01ba 100644
--- a/arch/xtensa/include/asm/processor.h
+++ b/arch/xtensa/include/asm/processor.h
@@ -1,11 +1,10 @@
 /*
- * include/asm-xtensa/processor.h
- *
  * This file is subject to the terms and conditions of the GNU General Public
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
  * Copyright (C) 2001 - 2008 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
  */
 
 #ifndef _XTENSA_PROCESSOR_H
@@ -45,6 +44,14 @@
 #define STACK_TOP_MAX	STACK_TOP
 
 /*
+ * General exception cause assigned to fake NMI. Fake NMI needs to be handled
+ * differently from other interrupts, but it uses common kernel entry/exit
+ * code.
+ */
+
+#define EXCCAUSE_MAPPED_NMI	62
+
+/*
  * General exception cause assigned to debug exceptions. Debug exceptions go
  * to their own vector, rather than the general exception vectors (user,
  * kernel, double); and their specific causes are reported via DEBUGCAUSE
@@ -65,10 +72,30 @@
 
 #define VALID_DOUBLE_EXCEPTION_ADDRESS	64
 
+#define XTENSA_INT_LEVEL(intno) _XTENSA_INT_LEVEL(intno)
+#define _XTENSA_INT_LEVEL(intno) XCHAL_INT##intno##_LEVEL
+
+#define XTENSA_INTLEVEL_MASK(level) _XTENSA_INTLEVEL_MASK(level)
+#define _XTENSA_INTLEVEL_MASK(level) (XCHAL_INTLEVEL##level##_MASK)
+
+#define IS_POW2(v) (((v) & ((v) - 1)) == 0)
+
+#define PROFILING_INTLEVEL XTENSA_INT_LEVEL(XCHAL_PROFILING_INTERRUPT)
+
 /* LOCKLEVEL defines the interrupt level that masks all
  * general-purpose interrupts.
  */
+#if defined(CONFIG_XTENSA_VARIANT_HAVE_PERF_EVENTS) && \
+	defined(XCHAL_PROFILING_INTERRUPT) && \
+	PROFILING_INTLEVEL == XCHAL_EXCM_LEVEL && \
+	XCHAL_EXCM_LEVEL > 1 && \
+	IS_POW2(XTENSA_INTLEVEL_MASK(PROFILING_INTLEVEL))
+#define LOCKLEVEL (XCHAL_EXCM_LEVEL - 1)
+#else
 #define LOCKLEVEL XCHAL_EXCM_LEVEL
+#endif
+#define TOPLEVEL XCHAL_EXCM_LEVEL
+#define XTENSA_FAKE_NMI (LOCKLEVEL < TOPLEVEL)
 
 /* WSBITS and WBBITS are the width of the WINDOWSTART and WINDOWBASE
  * registers
diff --git a/arch/xtensa/include/asm/stacktrace.h b/arch/xtensa/include/asm/stacktrace.h
index 6a05fcb0a20d..fe06e8ed162b 100644
--- a/arch/xtensa/include/asm/stacktrace.h
+++ b/arch/xtensa/include/asm/stacktrace.h
@@ -33,4 +33,12 @@ void walk_stackframe(unsigned long *sp,
 		int (*fn)(struct stackframe *frame, void *data),
 		void *data);
 
+void xtensa_backtrace_kernel(struct pt_regs *regs, unsigned int depth,
+			     int (*kfn)(struct stackframe *frame, void *data),
+			     int (*ufn)(struct stackframe *frame, void *data),
+			     void *data);
+void xtensa_backtrace_user(struct pt_regs *regs, unsigned int depth,
+			   int (*ufn)(struct stackframe *frame, void *data),
+			   void *data);
+
 #endif /* _XTENSA_STACKTRACE_H */
diff --git a/arch/xtensa/include/asm/traps.h b/arch/xtensa/include/asm/traps.h
index 677bfcf4ee5d..28f33a8b7f5f 100644
--- a/arch/xtensa/include/asm/traps.h
+++ b/arch/xtensa/include/asm/traps.h
@@ -25,30 +25,39 @@ static inline void spill_registers(void)
 {
 #if XCHAL_NUM_AREGS > 16
 	__asm__ __volatile__ (
-		"	call12	1f\n"
+		"	call8	1f\n"
 		"	_j	2f\n"
 		"	retw\n"
 		"	.align	4\n"
 		"1:\n"
+#if XCHAL_NUM_AREGS == 32
+		"	_entry	a1, 32\n"
+		"	addi	a8, a0, 3\n"
+		"	_entry	a1, 16\n"
+		"	mov	a12, a12\n"
+		"	retw\n"
+#else
 		"	_entry	a1, 48\n"
-		"	addi	a12, a0, 3\n"
-#if XCHAL_NUM_AREGS > 32
-		"	.rept	(" __stringify(XCHAL_NUM_AREGS) " - 32) / 12\n"
+		"	call12	1f\n"
+		"	retw\n"
+		"	.align	4\n"
+		"1:\n"
+		"	.rept	(" __stringify(XCHAL_NUM_AREGS) " - 16) / 12\n"
 		"	_entry	a1, 48\n"
 		"	mov	a12, a0\n"
 		"	.endr\n"
-#endif
-		"	_entry	a1, 48\n"
+		"	_entry	a1, 16\n"
 #if XCHAL_NUM_AREGS % 12 == 0
-		"	mov	a8, a8\n"
-#elif XCHAL_NUM_AREGS % 12 == 4
 		"	mov	a12, a12\n"
-#elif XCHAL_NUM_AREGS % 12 == 8
+#elif XCHAL_NUM_AREGS % 12 == 4
 		"	mov	a4, a4\n"
+#elif XCHAL_NUM_AREGS % 12 == 8
+		"	mov	a8, a8\n"
 #endif
 		"	retw\n"
+#endif
 		"2:\n"
-		: : : "a12", "a13", "memory");
+		: : : "a8", "a9", "memory");
 #else
 	__asm__ __volatile__ (
 		"	mov	a12, a12\n"