Merge branch 'xtensa-dma-fixes' (early part) into xtensa-fixes

This switches xtensa arch to the generic noncoherent direct mapping
operations, adds support for DMA_ATTR_NO_KERNEL_MAPPING attribute and
allows for platform-specific handling of coherent memory.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>

arch/xtensa/Kconfig

@@ -4,12 +4,15 @@ config ZONE_DMA
 
 config XTENSA
 	def_bool y
+	select ARCH_HAS_SYNC_DMA_FOR_CPU
+	select ARCH_HAS_SYNC_DMA_FOR_DEVICE
 	select ARCH_NO_COHERENT_DMA_MMAP if !MMU
 	select ARCH_WANT_FRAME_POINTERS
 	select ARCH_WANT_IPC_PARSE_VERSION
 	select BUILDTIME_EXTABLE_SORT
 	select CLONE_BACKWARDS
 	select COMMON_CLK
+	select DMA_NONCOHERENT_OPS
 	select GENERIC_ATOMIC64
 	select GENERIC_CLOCKEVENTS
 	select GENERIC_IRQ_SHOW

arch/xtensa/include/asm/Kbuild

@@ -3,6 +3,7 @@ generic-y += compat.h
 generic-y += device.h
 generic-y += div64.h
 generic-y += dma-contiguous.h
+generic-y += dma-mapping.h
 generic-y += emergency-restart.h
 generic-y += exec.h
 generic-y += extable.h

arch/xtensa/include/asm/dma-mapping.h

@@ -1,26 +0,0 @@
-/*
- * 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
-#define _XTENSA_DMA_MAPPING_H
-
-#include <asm/cache.h>
-#include <asm/io.h>
-
-#include <linux/mm.h>
-#include <linux/scatterlist.h>
-
-extern const struct dma_map_ops xtensa_dma_map_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
-	return &xtensa_dma_map_ops;
-}
-
-#endif	/* _XTENSA_DMA_MAPPING_H */

arch/xtensa/include/asm/kmem_layout.h

@@ -63,12 +63,6 @@
 #error XCHAL_KSEG_PADDR is not properly aligned to XCHAL_KSEG_ALIGNMENT
 #endif
 
-#else
-
-#define XCHAL_KSEG_CACHED_VADDR	__XTENSA_UL_CONST(0xd0000000)
-#define XCHAL_KSEG_BYPASS_VADDR	__XTENSA_UL_CONST(0xd8000000)
-#define XCHAL_KSEG_SIZE		__XTENSA_UL_CONST(0x08000000)
-
 #endif
 
 #ifndef CONFIG_KASAN

arch/xtensa/include/asm/pgtable.h

@@ -66,6 +66,7 @@
 #define FIRST_USER_ADDRESS	0UL
 #define FIRST_USER_PGD_NR	(FIRST_USER_ADDRESS >> PGDIR_SHIFT)
 
+#ifdef CONFIG_MMU
 /*
  * Virtual memory area. We keep a distance to other memory regions to be
  * on the safe side. We also use this area for cache aliasing.
@@ -80,6 +81,13 @@
 #define TLBTEMP_SIZE		ICACHE_WAY_SIZE
 #endif
 
+#else
+
+#define VMALLOC_START		__XTENSA_UL_CONST(0)
+#define VMALLOC_END		__XTENSA_UL_CONST(0xffffffff)
+
+#endif
+
 /*
  * For the Xtensa architecture, the PTE layout is as follows:
  *

arch/xtensa/include/asm/platform.h

@@ -75,4 +75,31 @@ extern void platform_calibrate_ccount (void);
  */
 void cpu_reset(void) __attribute__((noreturn));
 
+/*
+ * Memory caching is platform-dependent in noMMU xtensa configurations.
+ * The following set of functions should be implemented in platform code
+ * in order to enable coherent DMA memory operations when CONFIG_MMU is not
+ * enabled. Default implementations do nothing and issue a warning.
+ */
+
+/*
+ * Check whether p points to a cached memory.
+ */
+bool platform_vaddr_cached(const void *p);
+
+/*
+ * Check whether p points to an uncached memory.
+ */
+bool platform_vaddr_uncached(const void *p);
+
+/*
+ * Return pointer to an uncached view of the cached sddress p.
+ */
+void *platform_vaddr_to_uncached(void *p);
+
+/*
+ * Return pointer to a cached view of the uncached sddress p.
+ */
+void *platform_vaddr_to_cached(void *p);
+
 #endif	/* _XTENSA_PLATFORM_H */

arch/xtensa/kernel/pci-dma.c

@@ -16,26 +16,25 @@
  */
 
 #include <linux/dma-contiguous.h>
+#include <linux/dma-noncoherent.h>
 #include <linux/dma-direct.h>
 #include <linux/gfp.h>
 #include <linux/highmem.h>
 #include <linux/mm.h>
-#include <linux/module.h>
-#include <linux/pci.h>
-#include <linux/string.h>
 #include <linux/types.h>
 #include <asm/cacheflush.h>
 #include <asm/io.h>
+#include <asm/platform.h>
 
-static void do_cache_op(dma_addr_t dma_handle, size_t size,
+static void do_cache_op(phys_addr_t paddr, size_t size,
 			void (*fn)(unsigned long, unsigned long))
 {
-	unsigned long off = dma_handle & (PAGE_SIZE - 1);
-	unsigned long pfn = PFN_DOWN(dma_handle);
+	unsigned long off = paddr & (PAGE_SIZE - 1);
+	unsigned long pfn = PFN_DOWN(paddr);
 	struct page *page = pfn_to_page(pfn);
 
 	if (!PageHighMem(page))
-		fn((unsigned long)bus_to_virt(dma_handle), size);
+		fn((unsigned long)phys_to_virt(paddr), size);
 	else
 		while (size > 0) {
 			size_t sz = min_t(size_t, size, PAGE_SIZE - off);
@@ -49,14 +48,13 @@ static void do_cache_op(dma_addr_t dma_handle, size_t size,
 		}
 }
 
-static void xtensa_sync_single_for_cpu(struct device *dev,
-				       dma_addr_t dma_handle, size_t size,
-				       enum dma_data_direction dir)
+void arch_sync_dma_for_cpu(struct device *dev, phys_addr_t paddr,
+		size_t size, enum dma_data_direction dir)
 {
 	switch (dir) {
 	case DMA_BIDIRECTIONAL:
 	case DMA_FROM_DEVICE:
-		do_cache_op(dma_handle, size, __invalidate_dcache_range);
+		do_cache_op(paddr, size, __invalidate_dcache_range);
 		break;
 
 	case DMA_NONE:
@@ -68,15 +66,14 @@ static void xtensa_sync_single_for_cpu(struct device *dev,
 	}
 }
 
-static void xtensa_sync_single_for_device(struct device *dev,
-					  dma_addr_t dma_handle, size_t size,
-					  enum dma_data_direction dir)
+void arch_sync_dma_for_device(struct device *dev, phys_addr_t paddr,
+		size_t size, enum dma_data_direction dir)
 {
 	switch (dir) {
 	case DMA_BIDIRECTIONAL:
 	case DMA_TO_DEVICE:
 		if (XCHAL_DCACHE_IS_WRITEBACK)
-			do_cache_op(dma_handle, size, __flush_dcache_range);
+			do_cache_op(paddr, size, __flush_dcache_range);
 		break;
 
 	case DMA_NONE:
@@ -88,43 +85,66 @@ static void xtensa_sync_single_for_device(struct device *dev,
 	}
 }
 
-static void xtensa_sync_sg_for_cpu(struct device *dev,
-				   struct scatterlist *sg, int nents,
-				   enum dma_data_direction dir)
+#ifdef CONFIG_MMU
+bool platform_vaddr_cached(const void *p)
 {
-	struct scatterlist *s;
-	int i;
+	unsigned long addr = (unsigned long)p;
 
-	for_each_sg(sg, s, nents, i) {
-		xtensa_sync_single_for_cpu(dev, sg_dma_address(s),
-					   sg_dma_len(s), dir);
-	}
+	return addr >= XCHAL_KSEG_CACHED_VADDR &&
+	       addr - XCHAL_KSEG_CACHED_VADDR < XCHAL_KSEG_SIZE;
 }
 
-static void xtensa_sync_sg_for_device(struct device *dev,
-				      struct scatterlist *sg, int nents,
-				      enum dma_data_direction dir)
+bool platform_vaddr_uncached(const void *p)
 {
-	struct scatterlist *s;
-	int i;
+	unsigned long addr = (unsigned long)p;
 
-	for_each_sg(sg, s, nents, i) {
-		xtensa_sync_single_for_device(dev, sg_dma_address(s),
-					      sg_dma_len(s), dir);
-	}
+	return addr >= XCHAL_KSEG_BYPASS_VADDR &&
+	       addr - XCHAL_KSEG_BYPASS_VADDR < XCHAL_KSEG_SIZE;
+}
+
+void *platform_vaddr_to_uncached(void *p)
+{
+	return p + XCHAL_KSEG_BYPASS_VADDR - XCHAL_KSEG_CACHED_VADDR;
+}
+
+void *platform_vaddr_to_cached(void *p)
+{
+	return p + XCHAL_KSEG_CACHED_VADDR - XCHAL_KSEG_BYPASS_VADDR;
+}
+#else
+bool __attribute__((weak)) platform_vaddr_cached(const void *p)
+{
+	WARN_ONCE(1, "Default %s implementation is used\n", __func__);
+	return true;
+}
+
+bool __attribute__((weak)) platform_vaddr_uncached(const void *p)
+{
+	WARN_ONCE(1, "Default %s implementation is used\n", __func__);
+	return false;
+}
+
+void __attribute__((weak)) *platform_vaddr_to_uncached(void *p)
+{
+	WARN_ONCE(1, "Default %s implementation is used\n", __func__);
+	return p;
+}
+
+void __attribute__((weak)) *platform_vaddr_to_cached(void *p)
+{
+	WARN_ONCE(1, "Default %s implementation is used\n", __func__);
+	return p;
 }
+#endif
 
 /*
  * Note: We assume that the full memory space is always mapped to 'kseg'
  *	 Otherwise we have to use page attributes (not implemented).
  */
 
-static void *xtensa_dma_alloc(struct device *dev, size_t size,
-			      dma_addr_t *handle, gfp_t flag,
-			      unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle,
+		gfp_t flag, unsigned long attrs)
 {
-	unsigned long ret;
-	unsigned long uncached;
 	unsigned long count = PAGE_ALIGN(size) >> PAGE_SHIFT;
 	struct page *page = NULL;
 
@@ -147,6 +167,10 @@ static void *xtensa_dma_alloc(struct device *dev, size_t size,
 
 	*handle = phys_to_dma(dev, page_to_phys(page));
 
+	if (attrs & DMA_ATTR_NO_KERNEL_MAPPING) {
+		return page;
+	}
+
 #ifdef CONFIG_MMU
 	if (PageHighMem(page)) {
 		void *p;
@@ -161,27 +185,21 @@ static void *xtensa_dma_alloc(struct device *dev, size_t size,
 		return p;
 	}
 #endif
-	ret = (unsigned long)page_address(page);
-	BUG_ON(ret < XCHAL_KSEG_CACHED_VADDR ||
-	       ret > XCHAL_KSEG_CACHED_VADDR + XCHAL_KSEG_SIZE - 1);
-
-	uncached = ret + XCHAL_KSEG_BYPASS_VADDR - XCHAL_KSEG_CACHED_VADDR;
-	__invalidate_dcache_range(ret, size);
-
-	return (void *)uncached;
+	BUG_ON(!platform_vaddr_cached(page_address(page)));
+	__invalidate_dcache_range((unsigned long)page_address(page), size);
+	return platform_vaddr_to_uncached(page_address(page));
 }
 
-static void xtensa_dma_free(struct device *dev, size_t size, void *vaddr,
-			    dma_addr_t dma_handle, unsigned long attrs)
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
+		dma_addr_t dma_handle, unsigned long attrs)
 {
 	unsigned long count = PAGE_ALIGN(size) >> PAGE_SHIFT;
-	unsigned long addr = (unsigned long)vaddr;
 	struct page *page;
 
-	if (addr >= XCHAL_KSEG_BYPASS_VADDR &&
-	    addr - XCHAL_KSEG_BYPASS_VADDR < XCHAL_KSEG_SIZE) {
-		addr += XCHAL_KSEG_CACHED_VADDR - XCHAL_KSEG_BYPASS_VADDR;
-		page = virt_to_page(addr);
+	if (attrs & DMA_ATTR_NO_KERNEL_MAPPING) {
+		page = vaddr;
+	} else if (platform_vaddr_uncached(vaddr)) {
+		page = virt_to_page(platform_vaddr_to_cached(vaddr));
 	} else {
 #ifdef CONFIG_MMU
 		dma_common_free_remap(vaddr, size, VM_MAP);
@@ -192,72 +210,3 @@ static void xtensa_dma_free(struct device *dev, size_t size, void *vaddr,
 	if (!dma_release_from_contiguous(dev, page, count))
 		__free_pages(page, get_order(size));
 }
-
-static dma_addr_t xtensa_map_page(struct device *dev, struct page *page,
-				  unsigned long offset, size_t size,
-				  enum dma_data_direction dir,
-				  unsigned long attrs)
-{
-	dma_addr_t dma_handle = page_to_phys(page) + offset;
-
-	if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
-		xtensa_sync_single_for_device(dev, dma_handle, size, dir);
-
-	return dma_handle;
-}
-
-static void xtensa_unmap_page(struct device *dev, dma_addr_t dma_handle,
-			      size_t size, enum dma_data_direction dir,
-			      unsigned long attrs)
-{
-	if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
-		xtensa_sync_single_for_cpu(dev, dma_handle, size, dir);
-}
-
-static int xtensa_map_sg(struct device *dev, struct scatterlist *sg,
-			 int nents, enum dma_data_direction dir,
-			 unsigned long attrs)
-{
-	struct scatterlist *s;
-	int i;
-
-	for_each_sg(sg, s, nents, i) {
-		s->dma_address = xtensa_map_page(dev, sg_page(s), s->offset,
-						 s->length, dir, attrs);
-	}
-	return nents;
-}
-
-static void xtensa_unmap_sg(struct device *dev,
-			    struct scatterlist *sg, int nents,
-			    enum dma_data_direction dir,
-			    unsigned long attrs)
-{
-	struct scatterlist *s;
-	int i;
-
-	for_each_sg(sg, s, nents, i) {
-		xtensa_unmap_page(dev, sg_dma_address(s),
-				  sg_dma_len(s), dir, attrs);
-	}
-}
-
-int xtensa_dma_mapping_error(struct device *dev, dma_addr_t dma_addr)
-{
-	return 0;
-}
-
-const struct dma_map_ops xtensa_dma_map_ops = {
-	.alloc = xtensa_dma_alloc,
-	.free = xtensa_dma_free,
-	.map_page = xtensa_map_page,
-	.unmap_page = xtensa_unmap_page,
-	.map_sg = xtensa_map_sg,
-	.unmap_sg = xtensa_unmap_sg,
-	.sync_single_for_cpu = xtensa_sync_single_for_cpu,
-	.sync_single_for_device = xtensa_sync_single_for_device,
-	.sync_sg_for_cpu = xtensa_sync_sg_for_cpu,
-	.sync_sg_for_device = xtensa_sync_sg_for_device,
-	.mapping_error = xtensa_dma_mapping_error,
-};
-EXPORT_SYMBOL(xtensa_dma_map_ops);