Re: [PATCH 04/20] arm-nommu: use generic dma_noncoherent_ops
On Fri, May 11, 2018 at 10:11:15AM +0100, Russell King - ARM Linux wrote: > > +void arch_sync_dma_for_device(struct device *dev, phys_addr_t paddr, > > + size_t size, enum dma_data_direction dir) > > Please no. There is a lot of history of these (__dma_page_cpu_to_dev etc) > functions being abused by out of tree drivers, because they think they > know better. This is stopped by making them static and ensuring that > drivers have no access to these functions. > > Please do not re-expose these to the global kernel. > > While it may make things easier for a cross-architecture point of view, > it makes it a lot easier for people to abuse these private APIs. The point of this series, which isn't fully archived yet, is to consolidate the direct mapping dma code, that is all dma_map_ops instances except for iommus. It is in fact in many ways modelled after the ARM code. For that we need the architectures to supply the cache maintainance routines. However, even if they now appear in dma-noncoherent.h they are NOT and will NOT be exported, so using them directly from drivers won't easily be possible. I'll drop the arm-nommu patch for now, but I will pester your with this again once arm is the last architecture not sharing the common code. > > -- > RMK's Patch system: http://www.armlinux.org.uk/developer/patches/ > FTTC broadband for 0.8mile line in suburbia: sync at 8.8Mbps down 630kbps up > According to speedtest.net: 8.21Mbps down 510kbps up ---end quoted text--- ___ linux-snps-arc mailing list linux-snps-arc@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 01/25] hexagon: remove the sync_single_for_cpu DMA operation
hexagon does all the required cache maintainance at dma map time, and none
at unmap time. It thus has to implement sync_single_for_device to match
the map cace for buffer reuse, but there is no point in doing another
invalidation in the sync_single_cpu_case, which in terms of cache
maintainance is equivalent to the unmap case.
Signed-off-by: Christoph Hellwig
---
arch/hexagon/kernel/dma.c | 8
1 file changed, 8 deletions(-)
diff --git a/arch/hexagon/kernel/dma.c b/arch/hexagon/kernel/dma.c
index 77459df34e2e..d2b717f352f4 100644
--- a/arch/hexagon/kernel/dma.c
+++ b/arch/hexagon/kernel/dma.c
@@ -181,13 +181,6 @@ static dma_addr_t hexagon_map_page(struct device *dev,
struct page *page,
return bus;
}
-static void hexagon_sync_single_for_cpu(struct device *dev,
- dma_addr_t dma_handle, size_t size,
- enum dma_data_direction dir)
-{
- dma_sync(dma_addr_to_virt(dma_handle), size, dir);
-}
-
static void hexagon_sync_single_for_device(struct device *dev,
dma_addr_t dma_handle, size_t size,
enum dma_data_direction dir)
@@ -205,7 +198,6 @@ const struct dma_map_ops hexagon_dma_ops = {
.free = hexagon_free_coherent,
.map_sg = hexagon_map_sg,
.map_page = hexagon_map_page,
- .sync_single_for_cpu = hexagon_sync_single_for_cpu,
.sync_single_for_device = hexagon_sync_single_for_device,
.mapping_error = hexagon_mapping_error,
};
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
common non-cache coherent direct dma mapping ops v2
Hi all, this series continues consolidating the dma-mapping code, with a focus on architectures that do not (always) provide cache coherence for DMA. Three architectures (arm, mips and powerpc) are still left to be converted later due to complexity of their dma ops selection. The dma-noncoherent ops calls the dma-direct ops for the actual translation of streaming mappins and allow the architecture to provide any cache flushing required for cpu to device and/or device to cpu ownership transfers. The dma coherent allocator is for now still left entirely to architecture supplied implementations due the amount of variations. Hopefully we can do some consolidation for them later on as well. A lot of architectures are currently doing very questionable things in their dma mapping routines, which are documented in the changelogs for each patch. Please review them very careful and correct me on incorrect assumptions. Because this series sits on top of three previously submitted series a git tree might be useful to actually test it. It is provided here: git://git.infradead.org/users/hch/misc.git generic-dma-noncoherent.2 Gitweb: http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/generic-dma-noncoherent.2 Changes since v1: - dropped the already merged generic, arc and c6x patches (with the map_single offset fix from Alexey Brodkin) - split various changes into smaller patches - dropped the arm-nommu changes for now - fixed a typo Changes since RFC: - fix a typo accidentally disabling the device to cpu transfer sync - fixed a few compile failures ___ linux-snps-arc mailing list linux-snps-arc@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 05/25] microblaze: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
This removes the direction-based optimizations in
sync_{single,sg}_for_{cpu,device} which were marked untestested and
do not match the usually very well tested {un,}map_{single,sg}
implementations.
Signed-off-by: Christoph Hellwig
---
arch/microblaze/Kconfig | 4 +
arch/microblaze/include/asm/Kbuild| 1 +
arch/microblaze/include/asm/dma-mapping.h | 28 -
arch/microblaze/include/asm/pgtable.h | 2 -
arch/microblaze/kernel/dma.c | 144 ++
arch/microblaze/mm/consistent.c | 9 +-
6 files changed, 22 insertions(+), 166 deletions(-)
delete mode 100644 arch/microblaze/include/asm/dma-mapping.h
diff --git a/arch/microblaze/Kconfig b/arch/microblaze/Kconfig
index d14782100088..848e31a86ba5 100644
--- a/arch/microblaze/Kconfig
+++ b/arch/microblaze/Kconfig
@@ -1,6 +1,8 @@
config MICROBLAZE
def_bool y
select ARCH_HAS_GCOV_PROFILE_ALL
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_NO_COHERENT_DMA_MMAP if !MMU
select ARCH_WANT_IPC_PARSE_VERSION
@@ -8,6 +10,8 @@ config MICROBLAZE
select TIMER_OF
select CLONE_BACKWARDS3
select COMMON_CLK
+ select DMA_NONCOHERENT_OPS
+ select DMA_NONCOHERENT_MMAP
select GENERIC_ATOMIC64
select GENERIC_CLOCKEVENTS
select GENERIC_CPU_DEVICES
diff --git a/arch/microblaze/include/asm/Kbuild
b/arch/microblaze/include/asm/Kbuild
index 3c80a5a308ed..8d3e71f43a3e 100644
--- a/arch/microblaze/include/asm/Kbuild
+++ b/arch/microblaze/include/asm/Kbuild
@@ -4,6 +4,7 @@ generic-y += bug.h
generic-y += bugs.h
generic-y += device.h
generic-y += div64.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += exec.h
generic-y += extable.h
diff --git a/arch/microblaze/include/asm/dma-mapping.h
b/arch/microblaze/include/asm/dma-mapping.h
deleted file mode 100644
index add50c1373bf..
--- a/arch/microblaze/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,28 +0,0 @@
-/*
- * Implements the generic device dma API for microblaze and the pci
- *
- * Copyright (C) 2009-2010 Michal Simek
- * Copyright (C) 2009-2010 PetaLogix
- *
- * 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.
- *
- * This file is base on powerpc and x86 dma-mapping.h versions
- * Copyright (C) 2004 IBM
- */
-
-#ifndef _ASM_MICROBLAZE_DMA_MAPPING_H
-#define _ASM_MICROBLAZE_DMA_MAPPING_H
-
-/*
- * Available generic sets of operations
- */
-extern const struct dma_map_ops dma_nommu_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
- return &dma_nommu_ops;
-}
-
-#endif /* _ASM_MICROBLAZE_DMA_MAPPING_H */
diff --git a/arch/microblaze/include/asm/pgtable.h
b/arch/microblaze/include/asm/pgtable.h
index db8b1fa83452..8a2e654b709f 100644
--- a/arch/microblaze/include/asm/pgtable.h
+++ b/arch/microblaze/include/asm/pgtable.h
@@ -553,8 +553,6 @@ void __init *early_get_page(void);
extern unsigned long ioremap_bot, ioremap_base;
-void *consistent_alloc(gfp_t gfp, size_t size, dma_addr_t *dma_handle);
-void consistent_free(size_t size, void *vaddr);
void consistent_sync(void *vaddr, size_t size, int direction);
void consistent_sync_page(struct page *page, unsigned long offset,
size_t size, int direction);
diff --git a/arch/microblaze/kernel/dma.c b/arch/microblaze/kernel/dma.c
index 3145e7dc8ab1..71032cf64669 100644
--- a/arch/microblaze/kernel/dma.c
+++ b/arch/microblaze/kernel/dma.c
@@ -8,29 +8,15 @@
*/
#include
-#include
+#include
#include
#include
#include
#include
#include
-static void *dma_nommu_alloc_coherent(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t flag,
- unsigned long attrs)
-{
- return consistent_alloc(flag, size, dma_handle);
-}
-
-static void dma_nommu_free_coherent(struct device *dev, size_t size,
-void *vaddr, dma_addr_t dma_handle,
-unsigned long attrs)
-{
- consistent_free(size, vaddr);
-}
-
-static inline void __dma_sync(unsigned long paddr,
- size_t size, enum dma_data_direction direction)
+static void __dma_sync(struct device *dev, phys_addr_t paddr, size_t size,
+ enum dma_data_direction direction)
{
switch (direction) {
case DMA_TO_DEVICE:
@@ -45,113 +31,21 @@ static inline void __dma_sync(unsigned long paddr,
}
}
-static int dma_nommu_map_sg(struct device *dev, struct scatterlist *sgl,
-int nents, enum dma_data_direction direction,
-unsigned long attrs)
+void arch
[PATCH 04/25] m68k: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/m68k/Kconfig | 2 +
arch/m68k/include/asm/Kbuild| 1 +
arch/m68k/include/asm/dma-mapping.h | 12 -
arch/m68k/kernel/dma.c | 68 -
4 files changed, 11 insertions(+), 72 deletions(-)
delete mode 100644 arch/m68k/include/asm/dma-mapping.h
diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig
index 785612b576f7..3f61327da2d5 100644
--- a/arch/m68k/Kconfig
+++ b/arch/m68k/Kconfig
@@ -2,6 +2,7 @@
config M68K
bool
default y
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE if HAS_DMA
select ARCH_MIGHT_HAVE_PC_PARPORT if ISA
select ARCH_NO_COHERENT_DMA_MMAP if !MMU
select HAVE_IDE
@@ -24,6 +25,7 @@ config M68K
select MODULES_USE_ELF_RELA
select OLD_SIGSUSPEND3
select OLD_SIGACTION
+ select DMA_NONCOHERENT_OPS if HAS_DMA
config CPU_BIG_ENDIAN
def_bool y
diff --git a/arch/m68k/include/asm/Kbuild b/arch/m68k/include/asm/Kbuild
index 88a9d27df1ac..a853c00f1374 100644
--- a/arch/m68k/include/asm/Kbuild
+++ b/arch/m68k/include/asm/Kbuild
@@ -1,5 +1,6 @@
generic-y += barrier.h
generic-y += device.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += exec.h
generic-y += extable.h
diff --git a/arch/m68k/include/asm/dma-mapping.h
b/arch/m68k/include/asm/dma-mapping.h
deleted file mode 100644
index e3722ed04fbb..
--- a/arch/m68k/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,12 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _M68K_DMA_MAPPING_H
-#define _M68K_DMA_MAPPING_H
-
-extern const struct dma_map_ops m68k_dma_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
-return &m68k_dma_ops;
-}
-
-#endif /* _M68K_DMA_MAPPING_H */
diff --git a/arch/m68k/kernel/dma.c b/arch/m68k/kernel/dma.c
index c01b9b8f97bf..3d561c577d35 100644
--- a/arch/m68k/kernel/dma.c
+++ b/arch/m68k/kernel/dma.c
@@ -6,7 +6,7 @@
#undef DEBUG
-#include
+#include
#include
#include
#include
@@ -18,7 +18,7 @@
#if defined(CONFIG_MMU) && !defined(CONFIG_COLDFIRE)
-static void *m68k_dma_alloc(struct device *dev, size_t size, dma_addr_t
*handle,
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle,
gfp_t flag, unsigned long attrs)
{
struct page *page, **map;
@@ -61,7 +61,7 @@ static void *m68k_dma_alloc(struct device *dev, size_t size,
dma_addr_t *handle,
return addr;
}
-static void m68k_dma_free(struct device *dev, size_t size, void *addr,
+void arch_dma_free(struct device *dev, size_t size, void *addr,
dma_addr_t handle, unsigned long attrs)
{
pr_debug("dma_free_coherent: %p, %x\n", addr, handle);
@@ -72,8 +72,8 @@ static void m68k_dma_free(struct device *dev, size_t size,
void *addr,
#include
-static void *m68k_dma_alloc(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t gfp, unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_handle,
+ gfp_t gfp, unsigned long attrs)
{
void *ret;
@@ -88,7 +88,7 @@ static void *m68k_dma_alloc(struct device *dev, size_t size,
return ret;
}
-static void m68k_dma_free(struct device *dev, size_t size, void *vaddr,
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
dma_addr_t dma_handle, unsigned long attrs)
{
free_pages((unsigned long)vaddr, get_order(size));
@@ -96,8 +96,8 @@ static void m68k_dma_free(struct device *dev, size_t size,
void *vaddr,
#endif /* CONFIG_MMU && !CONFIG_COLDFIRE */
-static void m68k_dma_sync_single_for_device(struct device *dev,
- dma_addr_t handle, size_t size, enum dma_data_direction dir)
+void arch_sync_dma_for_device(struct device *dev, phys_addr_t handle,
+ size_t size, enum dma_data_direction dir)
{
switch (dir) {
case DMA_BIDIRECTIONAL:
@@ -113,55 +113,3 @@ static void m68k_dma_sync_single_for_device(struct device
*dev,
break;
}
}
-
-static void m68k_dma_sync_sg_for_device(struct device *dev,
- struct scatterlist *sglist, int nents, enum dma_data_direction
dir)
-{
- int i;
- struct scatterlist *sg;
-
- for_each_sg(sglist, sg, nents, i) {
- dma_sync_single_for_device(dev, sg->dma_address, sg->length,
- dir);
- }
-}
-
-static dma_addr_t m68k_dma_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 handle = page_to_phys(page) + offset;
-
- if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- dma_sync_single_for_device(dev, handle, size, dir);
-
- return handle;
-}
-
-static int m68k_dma_map_sg
[PATCH 06/25] microblaze: remove the consistent_sync and consistent_sync_page
Both unused.
Signed-off-by: Christoph Hellwig
---
arch/microblaze/include/asm/pgtable.h | 3 --
arch/microblaze/mm/consistent.c | 45 ---
2 files changed, 48 deletions(-)
diff --git a/arch/microblaze/include/asm/pgtable.h
b/arch/microblaze/include/asm/pgtable.h
index 8a2e654b709f..7b650ab14fa0 100644
--- a/arch/microblaze/include/asm/pgtable.h
+++ b/arch/microblaze/include/asm/pgtable.h
@@ -553,9 +553,6 @@ void __init *early_get_page(void);
extern unsigned long ioremap_bot, ioremap_base;
-void consistent_sync(void *vaddr, size_t size, int direction);
-void consistent_sync_page(struct page *page, unsigned long offset,
- size_t size, int direction);
unsigned long consistent_virt_to_pfn(void *vaddr);
void setup_memory(void);
diff --git a/arch/microblaze/mm/consistent.c b/arch/microblaze/mm/consistent.c
index b9a9c8c3397b..c9a278ac795a 100644
--- a/arch/microblaze/mm/consistent.c
+++ b/arch/microblaze/mm/consistent.c
@@ -220,48 +220,3 @@ void arch_dma_free(struct device *dev, size_t size, void
*vaddr,
flush_tlb_all();
#endif
}
-
-/*
- * make an area consistent.
- */
-void consistent_sync(void *vaddr, size_t size, int direction)
-{
- unsigned long start;
- unsigned long end;
-
- start = (unsigned long)vaddr;
-
- /* Convert start address back down to unshadowed memory region */
-#ifdef CONFIG_XILINX_UNCACHED_SHADOW
- start &= ~UNCACHED_SHADOW_MASK;
-#endif
- end = start + size;
-
- switch (direction) {
- case PCI_DMA_NONE:
- BUG();
- case PCI_DMA_FROMDEVICE:/* invalidate only */
- invalidate_dcache_range(start, end);
- break;
- case PCI_DMA_TODEVICE: /* writeback only */
- flush_dcache_range(start, end);
- break;
- case PCI_DMA_BIDIRECTIONAL: /* writeback and invalidate */
- flush_dcache_range(start, end);
- break;
- }
-}
-EXPORT_SYMBOL(consistent_sync);
-
-/*
- * consistent_sync_page makes memory consistent. identical
- * to consistent_sync, but takes a struct page instead of a
- * virtual address
- */
-void consistent_sync_page(struct page *page, unsigned long offset,
- size_t size, int direction)
-{
- unsigned long start = (unsigned long)page_address(page) + offset;
- consistent_sync((void *)start, size, direction);
-}
-EXPORT_SYMBOL(consistent_sync_page);
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 02/25] hexagon: implement the sync_sg_for_device DMA operation
This methods needs to provide the equivalent of sync_single_for_device
for each S/G list element, but was missing.
Signed-off-by: Christoph Hellwig
---
arch/hexagon/kernel/dma.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/arch/hexagon/kernel/dma.c b/arch/hexagon/kernel/dma.c
index d2b717f352f4..9e46556a227d 100644
--- a/arch/hexagon/kernel/dma.c
+++ b/arch/hexagon/kernel/dma.c
@@ -188,6 +188,18 @@ static void hexagon_sync_single_for_device(struct device
*dev,
dma_sync(dma_addr_to_virt(dma_handle), size, dir);
}
+static void hexagon_sync_sg_for_device(struct device *dev,
+ struct scatterlist *sgl, int nents, enum dma_data_direction dir)
+{
+ struct scatterlist *sg;
+ int i;
+
+ for_each_sg(sgl, sg, nents, i)
+ hexagon_sync_single_for_device(dev, sg_dma_address(sg),
+ sg->length, dir);
+}
+
+
static int hexagon_mapping_error(struct device *dev, dma_addr_t dma_addr)
{
return dma_addr == HEXAGON_MAPPING_ERROR;
@@ -199,6 +211,7 @@ const struct dma_map_ops hexagon_dma_ops = {
.map_sg = hexagon_map_sg,
.map_page = hexagon_map_page,
.sync_single_for_device = hexagon_sync_single_for_device,
+ .sync_sg_for_device = hexagon_sync_sg_for_device,
.mapping_error = hexagon_mapping_error,
};
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 03/25] hexagon: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/hexagon/Kconfig | 2 +
arch/hexagon/include/asm/Kbuild| 1 +
arch/hexagon/include/asm/dma-mapping.h | 40 ---
arch/hexagon/kernel/dma.c | 148 ++---
4 files changed, 11 insertions(+), 180 deletions(-)
delete mode 100644 arch/hexagon/include/asm/dma-mapping.h
diff --git a/arch/hexagon/Kconfig b/arch/hexagon/Kconfig
index 37adb2003033..bcbdcb32935c 100644
--- a/arch/hexagon/Kconfig
+++ b/arch/hexagon/Kconfig
@@ -4,6 +4,7 @@ comment "Linux Kernel Configuration for Hexagon"
config HEXAGON
def_bool y
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select HAVE_OPROFILE
# Other pending projects/to-do items.
# select HAVE_REGS_AND_STACK_ACCESS_API
@@ -28,6 +29,7 @@ config HEXAGON
select GENERIC_CLOCKEVENTS_BROADCAST
select MODULES_USE_ELF_RELA
select GENERIC_CPU_DEVICES
+ select DMA_NONCOHERENT_OPS
---help---
Qualcomm Hexagon is a processor architecture designed for high
performance and low power across a wide variety of applications.
diff --git a/arch/hexagon/include/asm/Kbuild b/arch/hexagon/include/asm/Kbuild
index e9743f689fb8..843a8086e980 100644
--- a/arch/hexagon/include/asm/Kbuild
+++ b/arch/hexagon/include/asm/Kbuild
@@ -5,6 +5,7 @@ generic-y += bugs.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += extable.h
generic-y += fb.h
diff --git a/arch/hexagon/include/asm/dma-mapping.h
b/arch/hexagon/include/asm/dma-mapping.h
deleted file mode 100644
index 263f6acbfb0f..
--- a/arch/hexagon/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,40 +0,0 @@
-/*
- * DMA operations for the Hexagon architecture
- *
- * Copyright (c) 2010-2011, The Linux Foundation. All rights reserved.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 and
- * only version 2 as published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
- * 02110-1301, USA.
- */
-
-#ifndef _ASM_DMA_MAPPING_H
-#define _ASM_DMA_MAPPING_H
-
-#include
-#include
-#include
-#include
-#include
-#include
-
-struct device;
-
-extern const struct dma_map_ops *dma_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
- return dma_ops;
-}
-
-#endif
diff --git a/arch/hexagon/kernel/dma.c b/arch/hexagon/kernel/dma.c
index 9e46556a227d..ffc4ae8e126f 100644
--- a/arch/hexagon/kernel/dma.c
+++ b/arch/hexagon/kernel/dma.c
@@ -18,32 +18,19 @@
* 02110-1301, USA.
*/
-#include
-#include
+#include
#include
#include
-#include
#include
#include
-#define HEXAGON_MAPPING_ERROR 0
-
-const struct dma_map_ops *dma_ops;
-EXPORT_SYMBOL(dma_ops);
-
-static inline void *dma_addr_to_virt(dma_addr_t dma_addr)
-{
- return phys_to_virt((unsigned long) dma_addr);
-}
-
static struct gen_pool *coherent_pool;
/* Allocates from a pool of uncached memory that was reserved at boot time */
-static void *hexagon_dma_alloc_coherent(struct device *dev, size_t size,
-dma_addr_t *dma_addr, gfp_t flag,
-unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_addr,
+ gfp_t flag, unsigned long attrs)
{
void *ret;
@@ -75,58 +62,17 @@ static void *hexagon_dma_alloc_coherent(struct device *dev,
size_t size,
return ret;
}
-static void hexagon_free_coherent(struct device *dev, size_t size, void *vaddr,
- dma_addr_t dma_addr, unsigned long attrs)
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
+ dma_addr_t dma_addr, unsigned long attrs)
{
gen_pool_free(coherent_pool, (unsigned long) vaddr, size);
}
-static int check_addr(const char *name, struct device *hwdev,
- dma_addr_t bus, size_t size)
-{
- if (hwdev && hwdev->dma_mask && !dma_capable(hwdev, bus, size)) {
- if (*hwdev->dma_mask >= DMA_BIT_MASK(32))
- printk(KERN_ERR
- "%s: overflow %Lx+%zu of device mask %Lx\n",
- name, (long long)bus, size,
- (long long)*hwdev->dma_mask);
- return 0;
- }
- return 1;
-}
[PATCH 07/25] nds32: remove the broken kmap code in nds32_dma_map_sg
nds32_dma_map_sg is the only of the various DMA operations that tries
to deal with highmem (the single page variants and SG sync routines are
missing, SG unmap is entirely unimplemented), and it does so without
taking into account S/G list items that are bigger than a page, which
are legal and can happen frequently.
Remove this code for now - if highmem support on nds32 becomes a real
thing it needs to be added back as a loop over pages in the newly
consolidated code that deals with all operations.
Signed-off-by: Christoph Hellwig
---
arch/nds32/kernel/dma.c | 21 +++--
1 file changed, 3 insertions(+), 18 deletions(-)
diff --git a/arch/nds32/kernel/dma.c b/arch/nds32/kernel/dma.c
index d291800fc621..e1bf7206e015 100644
--- a/arch/nds32/kernel/dma.c
+++ b/arch/nds32/kernel/dma.c
@@ -393,24 +393,9 @@ static int nds32_dma_map_sg(struct device *dev, struct
scatterlist *sg,
int i;
for (i = 0; i < nents; i++, sg++) {
- void *virt;
- unsigned long pfn;
- struct page *page = sg_page(sg);
-
- sg->dma_address = sg_phys(sg);
- pfn = page_to_pfn(page) + sg->offset / PAGE_SIZE;
- page = pfn_to_page(pfn);
- if (PageHighMem(page)) {
- virt = kmap_atomic(page);
- consistent_sync(virt, sg->length, dir, FOR_CPU);
- kunmap_atomic(virt);
- } else {
- if (sg->offset > PAGE_SIZE)
- panic("sg->offset:%08x > PAGE_SIZE\n",
- sg->offset);
- virt = page_address(page) + sg->offset;
- consistent_sync(virt, sg->length, dir, FOR_CPU);
- }
+ char *virt =
+ page_address((struct page *)sg->page_link) + sg->offset;
+ consistent_sync(virt, sg->length, dir, FOR_CPU);
}
return nents;
}
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 08/25] nds32: consolidate DMA cache maintainance routines
Make sure all other DMA methods call nds32_dma_sync_single_for_{device,cpu}
to perform cache maintaince, and remove the consisteny_sync helper that
implemented both with entirely separate code based off an argument.
Signed-off-by: Christoph Hellwig
---
arch/nds32/kernel/dma.c | 140 +---
1 file changed, 61 insertions(+), 79 deletions(-)
diff --git a/arch/nds32/kernel/dma.c b/arch/nds32/kernel/dma.c
index e1bf7206e015..4e6fb4ffd3f7 100644
--- a/arch/nds32/kernel/dma.c
+++ b/arch/nds32/kernel/dma.c
@@ -22,11 +22,6 @@
static pte_t *consistent_pte;
static DEFINE_RAW_SPINLOCK(consistent_lock);
-enum master_type {
- FOR_CPU = 0,
- FOR_DEVICE = 1,
-};
-
/*
* VM region handling support.
*
@@ -333,15 +328,53 @@ static int __init consistent_init(void)
}
core_initcall(consistent_init);
-static void consistent_sync(void *vaddr, size_t size, int direction, int
master_type);
+
+static void
+nds32_dma_sync_single_for_device(struct device *dev, dma_addr_t handle,
+size_t size, enum dma_data_direction dir)
+{
+ unsigned long start = (unsigned long)phys_to_virt(handle);
+
+ switch (dir) {
+ case DMA_FROM_DEVICE:
+ break;
+ case DMA_TO_DEVICE:
+ case DMA_BIDIRECTIONAL:
+ cpu_dma_wb_range(start, start + size);
+ break;
+ default:
+ BUG();
+ }
+}
+
+static void
+nds32_dma_sync_single_for_cpu(struct device *dev, dma_addr_t handle,
+ size_t size, enum dma_data_direction dir)
+{
+ unsigned long start = (unsigned long)phys_to_virt(handle);
+
+ switch (dir) {
+ case DMA_TO_DEVICE:
+ break;
+ case DMA_FROM_DEVICE:
+ case DMA_BIDIRECTIONAL:
+ cpu_dma_inval_range(start, start + size);
+ break;
+ default:
+ BUG();
+ }
+}
+
static dma_addr_t nds32_dma_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_addr = page_to_phys(page) + offset;
+
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- consistent_sync((void *)(page_address(page) + offset), size,
dir, FOR_DEVICE);
- return page_to_phys(page) + offset;
+ nds32_dma_sync_single_for_device(dev, handle, size, dir);
+ return dma_addr;
}
static void nds32_dma_unmap_page(struct device *dev, dma_addr_t handle,
@@ -349,75 +382,19 @@ static void nds32_dma_unmap_page(struct device *dev,
dma_addr_t handle,
unsigned long attrs)
{
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- consistent_sync(phys_to_virt(handle), size, dir, FOR_CPU);
+ nds32_dma_sync_single_for_cpu(dev, handle, size, dir);
}
-/*
- * Make an area consistent for devices.
- */
-static void consistent_sync(void *vaddr, size_t size, int direction, int
master_type)
-{
- unsigned long start = (unsigned long)vaddr;
- unsigned long end = start + size;
-
- if (master_type == FOR_CPU) {
- switch (direction) {
- case DMA_TO_DEVICE:
- break;
- case DMA_FROM_DEVICE:
- case DMA_BIDIRECTIONAL:
- cpu_dma_inval_range(start, end);
- break;
- default:
- BUG();
- }
- } else {
- /* FOR_DEVICE */
- switch (direction) {
- case DMA_FROM_DEVICE:
- break;
- case DMA_TO_DEVICE:
- case DMA_BIDIRECTIONAL:
- cpu_dma_wb_range(start, end);
- break;
- default:
- BUG();
- }
- }
-}
-
-static int nds32_dma_map_sg(struct device *dev, struct scatterlist *sg,
- int nents, enum dma_data_direction dir,
- unsigned long attrs)
+static void
+nds32_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg,
+int nents, enum dma_data_direction dir)
{
int i;
for (i = 0; i < nents; i++, sg++) {
- char *virt =
- page_address((struct page *)sg->page_link) + sg->offset;
- consistent_sync(virt, sg->length, dir, FOR_CPU);
+ nds32_dma_sync_single_for_device(dev, sg_dma_address(sg),
+ sg->length, dir);
}
- return nents;
-}
-
-static void nds32_dma_unmap_sg(struct device *dev, struct scatterlist *sg,
- int nhwentries, enum dma_data_direction dir,
- unsigned long attrs)
-{
-}
-
-static void
-nds32_dma_sy
[PATCH 19/25] sh: split arch/sh/mm/consistent.c
Half of the file just contains platform device memory setup code which
is required for all builds, and half contains helpers for dma coherent
allocation, which is only needed if CONFIG_DMA_NONCOHERENT is enabled.
Signed-off-by: Christoph Hellwig
---
arch/sh/kernel/Makefile | 2 +-
arch/sh/kernel/dma-coherent.c | 80 +++
arch/sh/mm/consistent.c | 75
3 files changed, 81 insertions(+), 76 deletions(-)
create mode 100644 arch/sh/kernel/dma-coherent.c
diff --git a/arch/sh/kernel/Makefile b/arch/sh/kernel/Makefile
index cb5f1bfb52de..d5ddb64bfffe 100644
--- a/arch/sh/kernel/Makefile
+++ b/arch/sh/kernel/Makefile
@@ -45,7 +45,7 @@ obj-$(CONFIG_DUMP_CODE) += disassemble.o
obj-$(CONFIG_HIBERNATION) += swsusp.o
obj-$(CONFIG_DWARF_UNWINDER) += dwarf.o
obj-$(CONFIG_PERF_EVENTS) += perf_event.o perf_callchain.o
-obj-$(CONFIG_DMA_NONCOHERENT) += dma-nommu.o
+obj-$(CONFIG_DMA_NONCOHERENT) += dma-nommu.o dma-coherent.o
obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o
ccflags-y := -Werror
diff --git a/arch/sh/kernel/dma-coherent.c b/arch/sh/kernel/dma-coherent.c
new file mode 100644
index ..4f41e5cd5207
--- /dev/null
+++ b/arch/sh/kernel/dma-coherent.c
@@ -0,0 +1,80 @@
+/*
+ * Copyright (C) 2004 - 2007 Paul Mundt
+ *
+ * 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.
+ */
+#include
+#include
+#include
+#include
+#include
+#include
+
+void *dma_generic_alloc_coherent(struct device *dev, size_t size,
+dma_addr_t *dma_handle, gfp_t gfp,
+unsigned long attrs)
+{
+ void *ret, *ret_nocache;
+ int order = get_order(size);
+
+ gfp |= __GFP_ZERO;
+
+ ret = (void *)__get_free_pages(gfp, order);
+ if (!ret)
+ return NULL;
+
+ /*
+* Pages from the page allocator may have data present in
+* cache. So flush the cache before using uncached memory.
+*/
+ sh_sync_dma_for_device(ret, size, DMA_BIDIRECTIONAL);
+
+ ret_nocache = (void __force *)ioremap_nocache(virt_to_phys(ret), size);
+ if (!ret_nocache) {
+ free_pages((unsigned long)ret, order);
+ return NULL;
+ }
+
+ split_page(pfn_to_page(virt_to_phys(ret) >> PAGE_SHIFT), order);
+
+ *dma_handle = virt_to_phys(ret) - PFN_PHYS(dev->dma_pfn_offset);
+
+ return ret_nocache;
+}
+
+void dma_generic_free_coherent(struct device *dev, size_t size,
+ void *vaddr, dma_addr_t dma_handle,
+ unsigned long attrs)
+{
+ int order = get_order(size);
+ unsigned long pfn = (dma_handle >> PAGE_SHIFT) + dev->dma_pfn_offset;
+ int k;
+
+ for (k = 0; k < (1 << order); k++)
+ __free_pages(pfn_to_page(pfn + k), 0);
+
+ iounmap(vaddr);
+}
+
+void sh_sync_dma_for_device(void *vaddr, size_t size,
+ enum dma_data_direction direction)
+{
+ void *addr = sh_cacheop_vaddr(vaddr);
+
+ switch (direction) {
+ case DMA_FROM_DEVICE: /* invalidate only */
+ __flush_invalidate_region(addr, size);
+ break;
+ case DMA_TO_DEVICE: /* writeback only */
+ __flush_wback_region(addr, size);
+ break;
+ case DMA_BIDIRECTIONAL: /* writeback and invalidate */
+ __flush_purge_region(addr, size);
+ break;
+ default:
+ BUG();
+ }
+}
+EXPORT_SYMBOL(sh_sync_dma_for_device);
diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index 5f86ae24025b..5da5be74fb6b 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -1,10 +1,6 @@
/*
- * arch/sh/mm/consistent.c
- *
* Copyright (C) 2004 - 2007 Paul Mundt
*
- * Declared coherent memory functions based on arch/x86/kernel/pci-dma_32.c
- *
* 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.
@@ -13,78 +9,7 @@
#include
#include
#include
-#include
#include
-#include
-#include
-#include
-#include
-
-void *dma_generic_alloc_coherent(struct device *dev, size_t size,
-dma_addr_t *dma_handle, gfp_t gfp,
-unsigned long attrs)
-{
- void *ret, *ret_nocache;
- int order = get_order(size);
-
- gfp |= __GFP_ZERO;
-
- ret = (void *)__get_free_pages(gfp, order);
- if (!ret)
- return NULL;
-
- /*
-* Pages from the page allocator may have data present in
-* cache. So flush the cache before using uncached memory.
-*/
- sh_sync_dma_for_device(ret, size, D
[PATCH 20/25] sh: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/sh/Kconfig | 3 +-
arch/sh/include/asm/Kbuild| 1 +
arch/sh/include/asm/dma-mapping.h | 26 ---
arch/sh/kernel/Makefile | 2 +-
arch/sh/kernel/dma-coherent.c | 23 +
arch/sh/kernel/dma-nommu.c| 78 ---
6 files changed, 15 insertions(+), 118 deletions(-)
delete mode 100644 arch/sh/include/asm/dma-mapping.h
delete mode 100644 arch/sh/kernel/dma-nommu.c
diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index d0b095323d62..9809e0604af3 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -49,7 +49,6 @@ config SUPERH
select HAVE_ARCH_AUDITSYSCALL
select HAVE_FUTEX_CMPXCHG if FUTEX
select HAVE_NMI
- select NEED_DMA_MAP_STATE
select NEED_SG_DMA_LENGTH
help
@@ -162,6 +161,8 @@ config DMA_COHERENT
config DMA_NONCOHERENT
def_bool !DMA_COHERENT
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select DMA_NONCOHERENT_OPS
config PGTABLE_LEVELS
default 3 if X2TLB
diff --git a/arch/sh/include/asm/Kbuild b/arch/sh/include/asm/Kbuild
index 1efcce74997b..50f7e878ea1b 100644
--- a/arch/sh/include/asm/Kbuild
+++ b/arch/sh/include/asm/Kbuild
@@ -1,6 +1,7 @@
generic-y += current.h
generic-y += delay.h
generic-y += div64.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += exec.h
generic-y += irq_regs.h
diff --git a/arch/sh/include/asm/dma-mapping.h
b/arch/sh/include/asm/dma-mapping.h
deleted file mode 100644
index 1ebc6a4eb1c5..
--- a/arch/sh/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,26 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __ASM_SH_DMA_MAPPING_H
-#define __ASM_SH_DMA_MAPPING_H
-
-extern const struct dma_map_ops nommu_dma_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
-#ifdef CONFIG_DMA_NONCOHERENT
- return &nommu_dma_ops;
-#else
- return &dma_direct_ops;
-#endif
-}
-
-extern void *dma_generic_alloc_coherent(struct device *dev, size_t size,
- dma_addr_t *dma_addr, gfp_t flag,
- unsigned long attrs);
-extern void dma_generic_free_coherent(struct device *dev, size_t size,
- void *vaddr, dma_addr_t dma_handle,
- unsigned long attrs);
-
-void sh_sync_dma_for_device(void *vaddr, size_t size,
- enum dma_data_direction dir);
-
-#endif /* __ASM_SH_DMA_MAPPING_H */
diff --git a/arch/sh/kernel/Makefile b/arch/sh/kernel/Makefile
index d5ddb64bfffe..59673f8a3379 100644
--- a/arch/sh/kernel/Makefile
+++ b/arch/sh/kernel/Makefile
@@ -45,7 +45,7 @@ obj-$(CONFIG_DUMP_CODE) += disassemble.o
obj-$(CONFIG_HIBERNATION) += swsusp.o
obj-$(CONFIG_DWARF_UNWINDER) += dwarf.o
obj-$(CONFIG_PERF_EVENTS) += perf_event.o perf_callchain.o
-obj-$(CONFIG_DMA_NONCOHERENT) += dma-nommu.o dma-coherent.o
+obj-$(CONFIG_DMA_NONCOHERENT) += dma-coherent.o
obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o
ccflags-y := -Werror
diff --git a/arch/sh/kernel/dma-coherent.c b/arch/sh/kernel/dma-coherent.c
index 4f41e5cd5207..67b9c9e6e43a 100644
--- a/arch/sh/kernel/dma-coherent.c
+++ b/arch/sh/kernel/dma-coherent.c
@@ -7,14 +7,13 @@
*/
#include
#include
-#include
+#include
#include
#include
#include
-void *dma_generic_alloc_coherent(struct device *dev, size_t size,
-dma_addr_t *dma_handle, gfp_t gfp,
-unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_handle,
+ gfp_t gfp, unsigned long attrs)
{
void *ret, *ret_nocache;
int order = get_order(size);
@@ -29,7 +28,8 @@ void *dma_generic_alloc_coherent(struct device *dev, size_t
size,
* Pages from the page allocator may have data present in
* cache. So flush the cache before using uncached memory.
*/
- sh_sync_dma_for_device(ret, size, DMA_BIDIRECTIONAL);
+ arch_sync_dma_for_device(dev, virt_to_phys(ret), size,
+ DMA_BIDIRECTIONAL);
ret_nocache = (void __force *)ioremap_nocache(virt_to_phys(ret), size);
if (!ret_nocache) {
@@ -44,9 +44,8 @@ void *dma_generic_alloc_coherent(struct device *dev, size_t
size,
return ret_nocache;
}
-void dma_generic_free_coherent(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)
{
int order = get_order(size);
unsigned long pfn = (dma_handle >> PAGE_SHIFT) + dev->dma_pfn_offset;
@@ -58,12 +57,12 @@ void dma_generic_free_cohere
[PATCH 21/25] xtensa: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/xtensa/Kconfig | 3 +
arch/xtensa/include/asm/Kbuild| 1 +
arch/xtensa/include/asm/dma-mapping.h | 26 --
arch/xtensa/kernel/pci-dma.c | 130 +++---
4 files changed, 19 insertions(+), 141 deletions(-)
delete mode 100644 arch/xtensa/include/asm/dma-mapping.h
diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig
index 17df332269b2..ef114648e954 100644
--- a/arch/xtensa/Kconfig
+++ b/arch/xtensa/Kconfig
@@ -5,11 +5,14 @@ config ZONE_DMA
config XTENSA
def_bool y
select ARCH_NO_COHERENT_DMA_MMAP if !MMU
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
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
diff --git a/arch/xtensa/include/asm/Kbuild b/arch/xtensa/include/asm/Kbuild
index 436b20337168..a8d6cd3bee4b 100644
--- a/arch/xtensa/include/asm/Kbuild
+++ b/arch/xtensa/include/asm/Kbuild
@@ -2,6 +2,7 @@ generic-y += bug.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
diff --git a/arch/xtensa/include/asm/dma-mapping.h
b/arch/xtensa/include/asm/dma-mapping.h
deleted file mode 100644
index 44098800dad7..
--- a/arch/xtensa/include/asm/dma-mapping.h
+++ /dev/null
@@ -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
-#include
-
-#include
-#include
-
-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 */
diff --git a/arch/xtensa/kernel/pci-dma.c b/arch/xtensa/kernel/pci-dma.c
index 392b4a80ebc2..a83d60e92908 100644
--- a/arch/xtensa/kernel/pci-dma.c
+++ b/arch/xtensa/kernel/pci-dma.c
@@ -16,26 +16,24 @@
*/
#include
+#include
#include
#include
#include
#include
-#include
-#include
-#include
#include
#include
#include
-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 +47,13 @@ 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 +65,14 @@ 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:
@@ -88,40 +84,13 @@ static void xtensa_sync_single_for_device(struct device
*dev,
}
}
-static void xten
[PATCH 24/25] parisc: always use flush_kernel_dcache_range for DMA cache maintainance
Current the S/G list based DMA ops use flush_kernel_vmap_range which contains a few UP optimizations, while the rest of the DMA operations uses flush_kernel_dcache_range. The single vs sg operations are supposed to have the same effect, so they should use the same routines. Use the more conservation version for now, but if people more familiar with parisc think the vmap version is generally fine for DMA we should switch all interfaces over to it. Signed-off-by: Christoph Hellwig --- arch/parisc/kernel/pci-dma.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c index 52304cb290f9..4d7506336918 100644 --- a/arch/parisc/kernel/pci-dma.c +++ b/arch/parisc/kernel/pci-dma.c @@ -550,7 +550,7 @@ static void pa11_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, /* once we do combining we'll need to use phys_to_virt(sg_dma_address(sglist)) */ for_each_sg(sglist, sg, nents, i) - flush_kernel_vmap_range(sg_virt(sg), sg->length); + flush_kernel_dcache_range(sg_virt(sg), sg->length); } static void pa11_dma_sync_single_for_cpu(struct device *dev, @@ -581,7 +581,7 @@ static void pa11_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sgl /* once we do combining we'll need to use phys_to_virt(sg_dma_address(sglist)) */ for_each_sg(sglist, sg, nents, i) - flush_kernel_vmap_range(sg_virt(sg), sg->length); + flush_kernel_dcache_range(sg_virt(sg), sg->length); } static void pa11_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sglist, int nents, enum dma_data_direction direction) @@ -592,7 +592,7 @@ static void pa11_dma_sync_sg_for_device(struct device *dev, struct scatterlist * /* once we do combining we'll need to use phys_to_virt(sg_dma_address(sglist)) */ for_each_sg(sglist, sg, nents, i) - flush_kernel_vmap_range(sg_virt(sg), sg->length); + flush_kernel_dcache_range(sg_virt(sg), sg->length); } static void pa11_dma_cache_sync(struct device *dev, void *vaddr, size_t size, -- 2.17.0 ___ linux-snps-arc mailing list linux-snps-arc@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 23/25] parisc: merge pcx_dma_ops and pcxl_dma_ops
The only difference is that pcxl supports dma coherent allocations, while
pcx only supports non-consistent allocations and otherwise fails.
But dma_alloc* is not in the fast path, and merging these two allows an
easy migration path to the generic dma-noncoherent implementation, so
do it.
Signed-off-by: Christoph Hellwig
---
arch/parisc/include/asm/dma-mapping.h | 3 +-
arch/parisc/kernel/pci-dma.c | 80 ---
arch/parisc/kernel/setup.c| 8 +--
arch/parisc/mm/init.c | 11 +---
4 files changed, 43 insertions(+), 59 deletions(-)
diff --git a/arch/parisc/include/asm/dma-mapping.h
b/arch/parisc/include/asm/dma-mapping.h
index 01e1fc057c83..eeec8dd18e74 100644
--- a/arch/parisc/include/asm/dma-mapping.h
+++ b/arch/parisc/include/asm/dma-mapping.h
@@ -22,8 +22,7 @@
*/
#ifdef CONFIG_PA11
-extern const struct dma_map_ops pcxl_dma_ops;
-extern const struct dma_map_ops pcx_dma_ops;
+extern const struct dma_map_ops pa11_dma_ops;
#endif
extern const struct dma_map_ops *hppa_dma_ops;
diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c
index 91bc0cac03a1..52304cb290f9 100644
--- a/arch/parisc/kernel/pci-dma.c
+++ b/arch/parisc/kernel/pci-dma.c
@@ -408,7 +408,7 @@ pcxl_dma_init(void)
__initcall(pcxl_dma_init);
-static void *pa11_dma_alloc(struct device *dev, size_t size,
+static void *pcxl_dma_alloc(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t flag, unsigned long attrs)
{
unsigned long vaddr;
@@ -435,16 +435,44 @@ static void *pa11_dma_alloc(struct device *dev, size_t
size,
return (void *)vaddr;
}
+static void *pcx_dma_alloc(struct device *dev, size_t size,
+ dma_addr_t *dma_handle, gfp_t flag, unsigned long attrs)
+{
+ void *addr;
+
+ if ((attrs & DMA_ATTR_NON_CONSISTENT) == 0)
+ return NULL;
+
+ addr = (void *)__get_free_pages(flag, get_order(size));
+ if (addr)
+ *dma_handle = (dma_addr_t)virt_to_phys(addr);
+
+ return addr;
+}
+
+static void *pa11_dma_alloc(struct device *dev, size_t size,
+ dma_addr_t *dma_handle, gfp_t gfp, unsigned long attrs)
+{
+
+ if (boot_cpu_data.cpu_type == pcxl2 || boot_cpu_data.cpu_type == pcxl)
+ return pcxl_dma_alloc(dev, size, dma_handle, gfp, attrs);
+ else
+ return pcx_dma_alloc(dev, size, dma_handle, gfp, attrs);
+}
+
static void pa11_dma_free(struct device *dev, size_t size, void *vaddr,
dma_addr_t dma_handle, unsigned long attrs)
{
- int order;
+ int order = get_order(size);
- order = get_order(size);
- size = 1 << (order + PAGE_SHIFT);
- unmap_uncached_pages((unsigned long)vaddr, size);
- pcxl_free_range((unsigned long)vaddr, size);
- free_pages((unsigned long)__va(dma_handle), order);
+ if (boot_cpu_data.cpu_type == pcxl2 || boot_cpu_data.cpu_type == pcxl) {
+ size = 1 << (order + PAGE_SHIFT);
+ unmap_uncached_pages((unsigned long)vaddr, size);
+ pcxl_free_range((unsigned long)vaddr, size);
+
+ vaddr = __va(dma_handle);
+ }
+ free_pages((unsigned long)vaddr, get_order(size));
}
static dma_addr_t pa11_dma_map_page(struct device *dev, struct page *page,
@@ -573,7 +601,7 @@ static void pa11_dma_cache_sync(struct device *dev, void
*vaddr, size_t size,
flush_kernel_dcache_range((unsigned long)vaddr, size);
}
-const struct dma_map_ops pcxl_dma_ops = {
+const struct dma_map_ops pa11_dma_ops = {
.alloc =pa11_dma_alloc,
.free = pa11_dma_free,
.map_page = pa11_dma_map_page,
@@ -586,39 +614,3 @@ const struct dma_map_ops pcxl_dma_ops = {
.sync_sg_for_device = pa11_dma_sync_sg_for_device,
.cache_sync = pa11_dma_cache_sync,
};
-
-static void *pcx_dma_alloc(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t flag, unsigned long attrs)
-{
- void *addr;
-
- if ((attrs & DMA_ATTR_NON_CONSISTENT) == 0)
- return NULL;
-
- addr = (void *)__get_free_pages(flag, get_order(size));
- if (addr)
- *dma_handle = (dma_addr_t)virt_to_phys(addr);
-
- return addr;
-}
-
-static void pcx_dma_free(struct device *dev, size_t size, void *vaddr,
- dma_addr_t iova, unsigned long attrs)
-{
- free_pages((unsigned long)vaddr, get_order(size));
- return;
-}
-
-const struct dma_map_ops pcx_dma_ops = {
- .alloc =pcx_dma_alloc,
- .free = pcx_dma_free,
- .map_page = pa11_dma_map_page,
- .unmap_page = pa11_dma_unmap_page,
- .map_sg = pa11_dma_map_sg,
- .unmap_sg = pa11_dma_unmap_sg,
- .sync_single_for_cpu = pa11_dma_sync_single_for_cpu,
- .sync_single_for_device = pa11_dm
[PATCH 25/25] parisc: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Fix sync_single_for_cpu to do skip the cache flush unless the transfer
is to the device to match the more tested unmap_single path which should
have the same cache coherency implications.
Signed-off-by: Christoph Hellwig
---
arch/parisc/Kconfig | 4 +
arch/parisc/include/asm/dma-mapping.h | 4 -
arch/parisc/kernel/pci-dma.c | 145 ++
arch/parisc/kernel/setup.c| 2 +-
4 files changed, 16 insertions(+), 139 deletions(-)
diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig
index 4d8f64d48597..4993c6dc8358 100644
--- a/arch/parisc/Kconfig
+++ b/arch/parisc/Kconfig
@@ -188,6 +188,10 @@ config PA20
config PA11
def_bool y
depends on PA7000 || PA7100LC || PA7200 || PA7300LC
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select DMA_NONCOHERENT_OPS
+ select DMA_NONCOHERENT_CACHE_SYNC
config PREFETCH
def_bool y
diff --git a/arch/parisc/include/asm/dma-mapping.h
b/arch/parisc/include/asm/dma-mapping.h
index eeec8dd18e74..44a9f97194aa 100644
--- a/arch/parisc/include/asm/dma-mapping.h
+++ b/arch/parisc/include/asm/dma-mapping.h
@@ -21,10 +21,6 @@
** flush/purge and allocate "regular" cacheable pages for everything.
*/
-#ifdef CONFIG_PA11
-extern const struct dma_map_ops pa11_dma_ops;
-#endif
-
extern const struct dma_map_ops *hppa_dma_ops;
static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c
index 4d7506336918..87a7926cda6f 100644
--- a/arch/parisc/kernel/pci-dma.c
+++ b/arch/parisc/kernel/pci-dma.c
@@ -21,13 +21,12 @@
#include
#include
#include
-#include
#include
#include
#include
#include
-#include
-#include
+#include
+#include
#include
#include /* for DMA_CHUNK_SIZE */
@@ -450,7 +449,7 @@ static void *pcx_dma_alloc(struct device *dev, size_t size,
return addr;
}
-static void *pa11_dma_alloc(struct device *dev, size_t size,
+void *arch_dma_alloc(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t gfp, unsigned long attrs)
{
@@ -460,7 +459,7 @@ static void *pa11_dma_alloc(struct device *dev, size_t size,
return pcx_dma_alloc(dev, size, dma_handle, gfp, attrs);
}
-static void pa11_dma_free(struct device *dev, size_t size, void *vaddr,
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
dma_addr_t dma_handle, unsigned long attrs)
{
int order = get_order(size);
@@ -475,142 +474,20 @@ static void pa11_dma_free(struct device *dev, size_t
size, void *vaddr,
free_pages((unsigned long)vaddr, get_order(size));
}
-static dma_addr_t pa11_dma_map_page(struct device *dev, struct page *page,
- unsigned long offset, size_t size,
- enum dma_data_direction direction, unsigned long attrs)
+void arch_sync_dma_for_device(struct device *dev, phys_addr_t paddr,
+ size_t size, enum dma_data_direction dir)
{
- void *addr = page_address(page) + offset;
- BUG_ON(direction == DMA_NONE);
-
- if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- flush_kernel_dcache_range((unsigned long) addr, size);
-
- return virt_to_phys(addr);
-}
-
-static void pa11_dma_unmap_page(struct device *dev, dma_addr_t dma_handle,
- size_t size, enum dma_data_direction direction,
- unsigned long attrs)
-{
- BUG_ON(direction == DMA_NONE);
-
- if (attrs & DMA_ATTR_SKIP_CPU_SYNC)
- return;
-
- if (direction == DMA_TO_DEVICE)
- return;
-
- /*
-* For PCI_DMA_FROMDEVICE this flush is not necessary for the
-* simple map/unmap case. However, it IS necessary if if
-* pci_dma_sync_single_* has been called and the buffer reused.
-*/
-
- flush_kernel_dcache_range((unsigned long) phys_to_virt(dma_handle),
size);
-}
-
-static int pa11_dma_map_sg(struct device *dev, struct scatterlist *sglist,
- int nents, enum dma_data_direction direction,
- unsigned long attrs)
-{
- int i;
- struct scatterlist *sg;
-
- BUG_ON(direction == DMA_NONE);
-
- for_each_sg(sglist, sg, nents, i) {
- unsigned long vaddr = (unsigned long)sg_virt(sg);
-
- sg_dma_address(sg) = (dma_addr_t) virt_to_phys(vaddr);
- sg_dma_len(sg) = sg->length;
-
- if (attrs & DMA_ATTR_SKIP_CPU_SYNC)
- continue;
-
- flush_kernel_dcache_range(vaddr, sg->length);
- }
- return nents;
-}
-
-static void pa11_dma_unmap_sg(struct device *dev, struct scatterlist *sglist,
- int nents, enum dma_data_direction direction,
- unsigned long attrs)
-{
- int i;
- struct scatterlist *sg;
-
- BUG_ON(direction
[PATCH 09/25] nds32: implement the unmap_sg DMA operation
This matches the implementation of the more commonly used unmap_single
routines and the sync_sg_for_cpu method which should provide equivalent
cache maintainance.
Signed-off-by: Christoph Hellwig
---
arch/nds32/kernel/dma.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/arch/nds32/kernel/dma.c b/arch/nds32/kernel/dma.c
index 4e6fb4ffd3f7..43d7fd432bb6 100644
--- a/arch/nds32/kernel/dma.c
+++ b/arch/nds32/kernel/dma.c
@@ -426,6 +426,12 @@ static void nds32_dma_unmap_sg(struct device *dev, struct
scatterlist *sg,
int nhwentries, enum dma_data_direction dir,
unsigned long attrs)
{
+ int i;
+
+ for (i = 0; i < nhwentries; i++, sg++) {
+ nds32_dma_sync_single_for_cpu(dev, sg_dma_address(sg),
+ sg->length, dir);
+ }
}
struct dma_map_ops nds32_dma_ops = {
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 10/25] nds32: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/nds32/Kconfig | 3 +
arch/nds32/include/asm/Kbuild| 1 +
arch/nds32/include/asm/dma-mapping.h | 14
arch/nds32/kernel/dma.c | 113 +++
4 files changed, 15 insertions(+), 116 deletions(-)
delete mode 100644 arch/nds32/include/asm/dma-mapping.h
diff --git a/arch/nds32/Kconfig b/arch/nds32/Kconfig
index 249f38d3388f..67d0ac0a989c 100644
--- a/arch/nds32/Kconfig
+++ b/arch/nds32/Kconfig
@@ -5,10 +5,13 @@
config NDS32
def_bool y
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select ARCH_WANT_FRAME_POINTERS if FTRACE
select CLKSRC_MMIO
select CLONE_BACKWARDS
select COMMON_CLK
+ select DMA_NONCOHERENT_OPS
select GENERIC_ATOMIC64
select GENERIC_CPU_DEVICES
select GENERIC_CLOCKEVENTS
diff --git a/arch/nds32/include/asm/Kbuild b/arch/nds32/include/asm/Kbuild
index 06bdf8167f5a..b3e951f805f8 100644
--- a/arch/nds32/include/asm/Kbuild
+++ b/arch/nds32/include/asm/Kbuild
@@ -13,6 +13,7 @@ generic-y += cputime.h
generic-y += device.h
generic-y += div64.h
generic-y += dma.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += errno.h
generic-y += exec.h
diff --git a/arch/nds32/include/asm/dma-mapping.h
b/arch/nds32/include/asm/dma-mapping.h
deleted file mode 100644
index 2dd47d245c25..
--- a/arch/nds32/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,14 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0
-// Copyright (C) 2005-2017 Andes Technology Corporation
-
-#ifndef ASMNDS32_DMA_MAPPING_H
-#define ASMNDS32_DMA_MAPPING_H
-
-extern struct dma_map_ops nds32_dma_ops;
-
-static inline struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
- return &nds32_dma_ops;
-}
-
-#endif
diff --git a/arch/nds32/kernel/dma.c b/arch/nds32/kernel/dma.c
index 43d7fd432bb6..2b3167a1f259 100644
--- a/arch/nds32/kernel/dma.c
+++ b/arch/nds32/kernel/dma.c
@@ -3,17 +3,14 @@
#include
#include
-#include
#include
-#include
-#include
+#include
#include
#include
#include
#include
#include
#include
-#include
#include
/*
@@ -119,10 +116,8 @@ static struct arch_vm_region *vm_region_find(struct
arch_vm_region *head,
return c;
}
-/* FIXME: attrs is not used. */
-static void *nds32_dma_alloc_coherent(struct device *dev, size_t size,
- dma_addr_t * handle, gfp_t gfp,
- unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *handle,
+ gfp_t gfp, unsigned long attrs)
{
struct page *page;
struct arch_vm_region *c;
@@ -227,8 +222,8 @@ static void *nds32_dma_alloc_coherent(struct device *dev,
size_t size,
return NULL;
}
-static void nds32_dma_free(struct device *dev, size_t size, void *cpu_addr,
- dma_addr_t handle, unsigned long attrs)
+void arch_dma_free(struct device *dev, size_t size, void *cpu_addr,
+ dma_addr_t handle, unsigned long attrs)
{
struct arch_vm_region *c;
unsigned long flags, addr;
@@ -329,11 +324,10 @@ static int __init consistent_init(void)
core_initcall(consistent_init);
-static void
-nds32_dma_sync_single_for_device(struct device *dev, dma_addr_t 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)
{
- unsigned long start = (unsigned long)phys_to_virt(handle);
+ unsigned long start = (unsigned long)phys_to_virt(paddr);
switch (dir) {
case DMA_FROM_DEVICE:
@@ -347,11 +341,10 @@ nds32_dma_sync_single_for_device(struct device *dev,
dma_addr_t handle,
}
}
-static void
-nds32_dma_sync_single_for_cpu(struct device *dev, dma_addr_t 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)
{
- unsigned long start = (unsigned long)phys_to_virt(handle);
+ unsigned long start = (unsigned long)phys_to_virt(paddr);
switch (dir) {
case DMA_TO_DEVICE:
@@ -364,87 +357,3 @@ nds32_dma_sync_single_for_cpu(struct device *dev,
dma_addr_t handle,
BUG();
}
}
-
-static dma_addr_t nds32_dma_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_addr = page_to_phys(page) + offset;
-
- if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- nds32_dma_sync_single_for_devic
[PATCH 14/25] openrisc: fix cache maintainance the the sync_single_for_device DMA operation
The cache maintaince in the sync_single_for_device operation should be
equivalent to the map_page operation to facilitate reusing buffers. Fix the
openrisc implementation by moving the cache maintaince performed in map_page
into the sync_single method, and calling that from map_page.
Signed-off-by: Christoph Hellwig
---
arch/openrisc/kernel/dma.c | 42 +-
1 file changed, 19 insertions(+), 23 deletions(-)
diff --git a/arch/openrisc/kernel/dma.c b/arch/openrisc/kernel/dma.c
index 7cadff93d179..d6a0bf1fa713 100644
--- a/arch/openrisc/kernel/dma.c
+++ b/arch/openrisc/kernel/dma.c
@@ -133,19 +133,15 @@ or1k_dma_free(struct device *dev, size_t size, void
*vaddr,
free_pages_exact(vaddr, size);
}
-static dma_addr_t
-or1k_map_page(struct device *dev, struct page *page,
- unsigned long offset, size_t size,
- enum dma_data_direction dir,
- unsigned long attrs)
+static void
+or1k_sync_single_for_device(struct device *dev,
+ dma_addr_t dma_handle, size_t size,
+ enum dma_data_direction dir)
{
unsigned long cl;
- dma_addr_t addr = page_to_phys(page) + offset;
+ dma_addr_t addr = dma_handle;
struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
- if (attrs & DMA_ATTR_SKIP_CPU_SYNC)
- return addr;
-
switch (dir) {
case DMA_TO_DEVICE:
/* Flush the dcache for the requested range */
@@ -168,6 +164,20 @@ or1k_map_page(struct device *dev, struct page *page,
break;
}
+}
+
+static dma_addr_t
+or1k_map_page(struct device *dev, struct page *page,
+ unsigned long offset, size_t size,
+ enum dma_data_direction dir,
+ unsigned long attrs)
+{
+ unsigned long cl;
+ dma_addr_t addr = page_to_phys(page) + offset;
+ struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
+
+ if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
+ or1k_sync_single_for_device(dev, addr, size, dir);
return addr;
}
@@ -187,20 +197,6 @@ or1k_map_sg(struct device *dev, struct scatterlist *sg,
return nents;
}
-static void
-or1k_sync_single_for_device(struct device *dev,
- dma_addr_t dma_handle, size_t size,
- enum dma_data_direction dir)
-{
- unsigned long cl;
- dma_addr_t addr = dma_handle;
- struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
-
- /* Flush the dcache for the requested range */
- for (cl = addr; cl < addr + size; cl += cpuinfo->dcache_block_size)
- mtspr(SPR_DCBFR, cl);
-}
-
const struct dma_map_ops or1k_dma_map_ops = {
.alloc = or1k_dma_alloc,
.free = or1k_dma_free,
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 12/25] openrisc: remove the sync_single_for_cpu DMA operation
openrisc does all the required cache maintainance at dma map time, and none
at unmap time. It thus has to implement sync_single_for_device to match
the map cace for buffer reuse, but there is no point in doing another
invalidation in the sync_single_cpu_case, which in terms of cache
maintainance is equivalent to the unmap case.
Signed-off-by: Christoph Hellwig
---
arch/openrisc/kernel/dma.c | 15 ---
1 file changed, 15 deletions(-)
diff --git a/arch/openrisc/kernel/dma.c b/arch/openrisc/kernel/dma.c
index ec7fd45704d2..47601274abf7 100644
--- a/arch/openrisc/kernel/dma.c
+++ b/arch/openrisc/kernel/dma.c
@@ -208,20 +208,6 @@ or1k_unmap_sg(struct device *dev, struct scatterlist *sg,
}
}
-static void
-or1k_sync_single_for_cpu(struct device *dev,
-dma_addr_t dma_handle, size_t size,
-enum dma_data_direction dir)
-{
- unsigned long cl;
- dma_addr_t addr = dma_handle;
- struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
-
- /* Invalidate the dcache for the requested range */
- for (cl = addr; cl < addr + size; cl += cpuinfo->dcache_block_size)
- mtspr(SPR_DCBIR, cl);
-}
-
static void
or1k_sync_single_for_device(struct device *dev,
dma_addr_t dma_handle, size_t size,
@@ -243,7 +229,6 @@ const struct dma_map_ops or1k_dma_map_ops = {
.unmap_page = or1k_unmap_page,
.map_sg = or1k_map_sg,
.unmap_sg = or1k_unmap_sg,
- .sync_single_for_cpu = or1k_sync_single_for_cpu,
.sync_single_for_device = or1k_sync_single_for_device,
};
EXPORT_SYMBOL(or1k_dma_map_ops);
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 13/25] openrisc: remove the no-op unmap_page and unmap_sg DMA operations
Signed-off-by: Christoph Hellwig
---
arch/openrisc/kernel/dma.c | 23 ---
1 file changed, 23 deletions(-)
diff --git a/arch/openrisc/kernel/dma.c b/arch/openrisc/kernel/dma.c
index 47601274abf7..7cadff93d179 100644
--- a/arch/openrisc/kernel/dma.c
+++ b/arch/openrisc/kernel/dma.c
@@ -171,14 +171,6 @@ or1k_map_page(struct device *dev, struct page *page,
return addr;
}
-static void
-or1k_unmap_page(struct device *dev, dma_addr_t dma_handle,
- size_t size, enum dma_data_direction dir,
- unsigned long attrs)
-{
- /* Nothing special to do here... */
-}
-
static int
or1k_map_sg(struct device *dev, struct scatterlist *sg,
int nents, enum dma_data_direction dir,
@@ -195,19 +187,6 @@ or1k_map_sg(struct device *dev, struct scatterlist *sg,
return nents;
}
-static void
-or1k_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) {
- or1k_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir, 0);
- }
-}
-
static void
or1k_sync_single_for_device(struct device *dev,
dma_addr_t dma_handle, size_t size,
@@ -226,9 +205,7 @@ const struct dma_map_ops or1k_dma_map_ops = {
.alloc = or1k_dma_alloc,
.free = or1k_dma_free,
.map_page = or1k_map_page,
- .unmap_page = or1k_unmap_page,
.map_sg = or1k_map_sg,
- .unmap_sg = or1k_unmap_sg,
.sync_single_for_device = or1k_sync_single_for_device,
};
EXPORT_SYMBOL(or1k_dma_map_ops);
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 15/25] openrisc: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/openrisc/Kconfig | 2 +
arch/openrisc/include/asm/Kbuild| 1 +
arch/openrisc/include/asm/dma-mapping.h | 35 -
arch/openrisc/kernel/dma.c | 65 -
4 files changed, 12 insertions(+), 91 deletions(-)
delete mode 100644 arch/openrisc/include/asm/dma-mapping.h
diff --git a/arch/openrisc/Kconfig b/arch/openrisc/Kconfig
index 9ecad05bfc73..65e3c574c9d3 100644
--- a/arch/openrisc/Kconfig
+++ b/arch/openrisc/Kconfig
@@ -6,6 +6,8 @@
config OPENRISC
def_bool y
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select DMA_NONCOHERENT_OPS
select OF
select OF_EARLY_FLATTREE
select IRQ_DOMAIN
diff --git a/arch/openrisc/include/asm/Kbuild b/arch/openrisc/include/asm/Kbuild
index f05c722a21f8..e663a996b612 100644
--- a/arch/openrisc/include/asm/Kbuild
+++ b/arch/openrisc/include/asm/Kbuild
@@ -6,6 +6,7 @@ generic-y += current.h
generic-y += device.h
generic-y += div64.h
generic-y += dma.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += exec.h
generic-y += extable.h
diff --git a/arch/openrisc/include/asm/dma-mapping.h
b/arch/openrisc/include/asm/dma-mapping.h
deleted file mode 100644
index e212a1f0b6d2..
--- a/arch/openrisc/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,35 +0,0 @@
-/*
- * OpenRISC Linux
- *
- * Linux architectural port borrowing liberally from similar works of
- * others. All original copyrights apply as per the original source
- * declaration.
- *
- * OpenRISC implementation:
- * Copyright (C) 2010-2011 Jonas Bonn
- *
- * 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.
- */
-
-#ifndef __ASM_OPENRISC_DMA_MAPPING_H
-#define __ASM_OPENRISC_DMA_MAPPING_H
-
-/*
- * See Documentation/DMA-API-HOWTO.txt and
- * Documentation/DMA-API.txt for documentation.
- */
-
-#include
-#include
-
-extern const struct dma_map_ops or1k_dma_map_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
- return &or1k_dma_map_ops;
-}
-
-#endif /* __ASM_OPENRISC_DMA_MAPPING_H */
diff --git a/arch/openrisc/kernel/dma.c b/arch/openrisc/kernel/dma.c
index d6a0bf1fa713..159336adfa2f 100644
--- a/arch/openrisc/kernel/dma.c
+++ b/arch/openrisc/kernel/dma.c
@@ -19,9 +19,7 @@
* the only thing implemented properly. The rest need looking into...
*/
-#include
-#include
-#include
+#include
#include
#include
@@ -80,10 +78,9 @@ page_clear_nocache(pte_t *pte, unsigned long addr,
* is being ignored for now; uncached but write-combined memory is a
* missing feature of the OR1K.
*/
-static void *
-or1k_dma_alloc(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t gfp,
- unsigned long attrs)
+void *
+arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_handle,
+ gfp_t gfp, unsigned long attrs)
{
unsigned long va;
void *page;
@@ -115,9 +112,9 @@ or1k_dma_alloc(struct device *dev, size_t size,
return (void *)va;
}
-static void
-or1k_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 va = (unsigned long)vaddr;
struct mm_walk walk = {
@@ -133,13 +130,10 @@ or1k_dma_free(struct device *dev, size_t size, void
*vaddr,
free_pages_exact(vaddr, size);
}
-static void
-or1k_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 addr, size_t
size,
+ enum dma_data_direction dir)
{
unsigned long cl;
- dma_addr_t addr = dma_handle;
struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
switch (dir) {
@@ -163,45 +157,4 @@ or1k_sync_single_for_device(struct device *dev,
*/
break;
}
-
-}
-
-static dma_addr_t
-or1k_map_page(struct device *dev, struct page *page,
- unsigned long offset, size_t size,
- enum dma_data_direction dir,
- unsigned long attrs)
-{
- unsigned long cl;
- dma_addr_t addr = page_to_phys(page) + offset;
- struct cpuinfo_or1k *cpuinfo = &cpuinfo_or1k[smp_processor_id()];
-
- if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- or1k_sync_single_for_device(dev, addr, size, dir);
- return addr;
}
-
-static int
-or1k_map_sg(struct device *dev, struct scatterlist *sg,
-
[PATCH 16/25] sh: simplify get_arch_dma_ops
Remove the indirection through the dma_ops variable, and just return
nommu_dma_ops directly from get_arch_dma_ops.
Signed-off-by: Christoph Hellwig
---
arch/sh/include/asm/dma-mapping.h | 5 ++---
arch/sh/kernel/dma-nommu.c| 8 +---
arch/sh/mm/consistent.c | 3 ---
arch/sh/mm/init.c | 10 --
4 files changed, 3 insertions(+), 23 deletions(-)
diff --git a/arch/sh/include/asm/dma-mapping.h
b/arch/sh/include/asm/dma-mapping.h
index 41167931e5d9..149e71f95be7 100644
--- a/arch/sh/include/asm/dma-mapping.h
+++ b/arch/sh/include/asm/dma-mapping.h
@@ -2,12 +2,11 @@
#ifndef __ASM_SH_DMA_MAPPING_H
#define __ASM_SH_DMA_MAPPING_H
-extern const struct dma_map_ops *dma_ops;
-extern void no_iommu_init(void);
+extern const struct dma_map_ops nommu_dma_ops;
static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
{
- return dma_ops;
+ return &nommu_dma_ops;
}
extern void *dma_generic_alloc_coherent(struct device *dev, size_t size,
diff --git a/arch/sh/kernel/dma-nommu.c b/arch/sh/kernel/dma-nommu.c
index 3e3a32fc676e..79a9edafa5b0 100644
--- a/arch/sh/kernel/dma-nommu.c
+++ b/arch/sh/kernel/dma-nommu.c
@@ -79,10 +79,4 @@ const struct dma_map_ops nommu_dma_ops = {
.sync_sg_for_device = nommu_sync_sg_for_device,
#endif
};
-
-void __init no_iommu_init(void)
-{
- if (dma_ops)
- return;
- dma_ops = &nommu_dma_ops;
-}
+EXPORT_SYMBOL(nommu_dma_ops);
diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index 35ea3099a3b6..221832eec33b 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -20,9 +20,6 @@
#include
#include
-const struct dma_map_ops *dma_ops;
-EXPORT_SYMBOL(dma_ops);
-
void *dma_generic_alloc_coherent(struct device *dev, size_t size,
dma_addr_t *dma_handle, gfp_t gfp,
unsigned long attrs)
diff --git a/arch/sh/mm/init.c b/arch/sh/mm/init.c
index ce0bbaa7e404..32e09f03e6bf 100644
--- a/arch/sh/mm/init.c
+++ b/arch/sh/mm/init.c
@@ -395,22 +395,12 @@ void __init paging_init(void)
free_area_init_nodes(max_zone_pfns);
}
-/*
- * Early initialization for any I/O MMUs we might have.
- */
-static void __init iommu_init(void)
-{
- no_iommu_init();
-}
-
unsigned int mem_init_done = 0;
void __init mem_init(void)
{
pg_data_t *pgdat;
- iommu_init();
-
high_memory = NULL;
for_each_online_pgdat(pgdat)
high_memory = max_t(void *, high_memory,
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 18/25] sh: use dma_direct_ops for the CONFIG_DMA_COHERENT case
This is a slight change in behavior as we avoid the detour through the
virtual mapping for the coherent allocator, but if this CPU really is
coherent that should be the right thing to do.
Signed-off-by: Christoph Hellwig
---
arch/sh/Kconfig | 1 +
arch/sh/include/asm/dma-mapping.h | 4
arch/sh/kernel/Makefile | 4 ++--
arch/sh/kernel/dma-nommu.c| 4
4 files changed, 7 insertions(+), 6 deletions(-)
diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index 7d521926041e..d0b095323d62 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -157,6 +157,7 @@ config SWAP_IO_SPACE
bool
config DMA_COHERENT
+ select DMA_DIRECT_OPS
bool
config DMA_NONCOHERENT
diff --git a/arch/sh/include/asm/dma-mapping.h
b/arch/sh/include/asm/dma-mapping.h
index 149e71f95be7..1ebc6a4eb1c5 100644
--- a/arch/sh/include/asm/dma-mapping.h
+++ b/arch/sh/include/asm/dma-mapping.h
@@ -6,7 +6,11 @@ extern const struct dma_map_ops nommu_dma_ops;
static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
{
+#ifdef CONFIG_DMA_NONCOHERENT
return &nommu_dma_ops;
+#else
+ return &dma_direct_ops;
+#endif
}
extern void *dma_generic_alloc_coherent(struct device *dev, size_t size,
diff --git a/arch/sh/kernel/Makefile b/arch/sh/kernel/Makefile
index dc80041f7363..cb5f1bfb52de 100644
--- a/arch/sh/kernel/Makefile
+++ b/arch/sh/kernel/Makefile
@@ -12,7 +12,7 @@ endif
CFLAGS_REMOVE_return_address.o = -pg
-obj-y := debugtraps.o dma-nommu.o dumpstack.o \
+obj-y := debugtraps.o dumpstack.o \
idle.o io.o irq.o irq_$(BITS).o kdebugfs.o \
machvec.o nmi_debug.o process.o \
process_$(BITS).o ptrace.o ptrace_$(BITS).o \
@@ -45,7 +45,7 @@ obj-$(CONFIG_DUMP_CODE) += disassemble.o
obj-$(CONFIG_HIBERNATION) += swsusp.o
obj-$(CONFIG_DWARF_UNWINDER) += dwarf.o
obj-$(CONFIG_PERF_EVENTS) += perf_event.o perf_callchain.o
-
+obj-$(CONFIG_DMA_NONCOHERENT) += dma-nommu.o
obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o
ccflags-y := -Werror
diff --git a/arch/sh/kernel/dma-nommu.c b/arch/sh/kernel/dma-nommu.c
index 79a9edafa5b0..d8689b1cb743 100644
--- a/arch/sh/kernel/dma-nommu.c
+++ b/arch/sh/kernel/dma-nommu.c
@@ -51,7 +51,6 @@ static int nommu_map_sg(struct device *dev, struct
scatterlist *sg,
return nents;
}
-#ifdef CONFIG_DMA_NONCOHERENT
static void nommu_sync_single_for_device(struct device *dev, dma_addr_t addr,
size_t size, enum dma_data_direction dir)
{
@@ -67,16 +66,13 @@ static void nommu_sync_sg_for_device(struct device *dev,
struct scatterlist *sg,
for_each_sg(sg, s, nelems, i)
sh_sync_dma_for_device(sg_virt(s), s->length, dir);
}
-#endif
const struct dma_map_ops nommu_dma_ops = {
.alloc = dma_generic_alloc_coherent,
.free = dma_generic_free_coherent,
.map_page = nommu_map_page,
.map_sg = nommu_map_sg,
-#ifdef CONFIG_DMA_NONCOHERENT
.sync_single_for_device = nommu_sync_single_for_device,
.sync_sg_for_device = nommu_sync_sg_for_device,
-#endif
};
EXPORT_SYMBOL(nommu_dma_ops);
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 17/25] sh: introduce a sh_cacheop_vaddr helper
And use it in the maple bus code to avoid a dma API dependency.
Signed-off-by: Christoph Hellwig
---
arch/sh/include/asm/cacheflush.h | 7 +++
arch/sh/mm/consistent.c | 6 +-
drivers/sh/maple/maple.c | 7 ---
3 files changed, 12 insertions(+), 8 deletions(-)
diff --git a/arch/sh/include/asm/cacheflush.h b/arch/sh/include/asm/cacheflush.h
index d103ab5a4e4b..b932e42ef028 100644
--- a/arch/sh/include/asm/cacheflush.h
+++ b/arch/sh/include/asm/cacheflush.h
@@ -101,5 +101,12 @@ void kunmap_coherent(void *kvaddr);
void cpu_cache_init(void);
+static inline void *sh_cacheop_vaddr(void *vaddr)
+{
+ if (__in_29bit_mode())
+ vaddr = (void *)CAC_ADDR((unsigned long)vaddr);
+ return vaddr;
+}
+
#endif /* __KERNEL__ */
#endif /* __ASM_SH_CACHEFLUSH_H */
diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index 221832eec33b..5f86ae24025b 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -69,10 +69,7 @@ void dma_generic_free_coherent(struct device *dev, size_t
size,
void sh_sync_dma_for_device(void *vaddr, size_t size,
enum dma_data_direction direction)
{
- void *addr;
-
- addr = __in_29bit_mode() ?
- (void *)CAC_ADDR((unsigned long)vaddr) : vaddr;
+ void *addr = sh_cacheop_vaddr(vaddr);
switch (direction) {
case DMA_FROM_DEVICE: /* invalidate only */
@@ -88,7 +85,6 @@ void sh_sync_dma_for_device(void *vaddr, size_t size,
BUG();
}
}
-EXPORT_SYMBOL(sh_sync_dma_for_device);
static int __init memchunk_setup(char *str)
{
diff --git a/drivers/sh/maple/maple.c b/drivers/sh/maple/maple.c
index 7525039d812c..c9c354bd713a 100644
--- a/drivers/sh/maple/maple.c
+++ b/drivers/sh/maple/maple.c
@@ -300,8 +300,8 @@ static void maple_send(void)
mutex_unlock(&maple_wlist_lock);
if (maple_packets > 0) {
for (i = 0; i < (1 << MAPLE_DMA_PAGES); i++)
- sh_sync_dma_for_device(maple_sendbuf + i * PAGE_SIZE,
- PAGE_SIZE, DMA_BIDIRECTIONAL);
+ __flush_purge_region(maple_sendbuf + i * PAGE_SIZE,
+ PAGE_SIZE);
}
finish:
@@ -642,7 +642,8 @@ static void maple_dma_handler(struct work_struct *work)
list_for_each_entry_safe(mq, nmq, &maple_sentq, list) {
mdev = mq->dev;
recvbuf = mq->recvbuf->buf;
- sh_sync_dma_for_device(recvbuf, 0x400, DMA_FROM_DEVICE);
+ __flush_invalidate_region(sh_cacheop_vaddr(recvbuf),
+ 0x400);
code = recvbuf[0];
kfree(mq->sendbuf);
list_del_init(&mq->list);
--
2.17.0
___
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc
[PATCH 11/25] nios2: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
Signed-off-by: Christoph Hellwig
---
arch/nios2/Kconfig | 3 +
arch/nios2/include/asm/Kbuild| 1 +
arch/nios2/include/asm/dma-mapping.h | 20
arch/nios2/mm/dma-mapping.c | 139 +++
4 files changed, 17 insertions(+), 146 deletions(-)
delete mode 100644 arch/nios2/include/asm/dma-mapping.h
diff --git a/arch/nios2/Kconfig b/arch/nios2/Kconfig
index 3d4ec88f1db1..92035042cf62 100644
--- a/arch/nios2/Kconfig
+++ b/arch/nios2/Kconfig
@@ -1,6 +1,9 @@
# SPDX-License-Identifier: GPL-2.0
config NIOS2
def_bool y
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select DMA_NONCOHERENT_OPS
select TIMER_OF
select GENERIC_ATOMIC64
select GENERIC_CLOCKEVENTS
diff --git a/arch/nios2/include/asm/Kbuild b/arch/nios2/include/asm/Kbuild
index d232da2cbb38..24f6ee1ee69b 100644
--- a/arch/nios2/include/asm/Kbuild
+++ b/arch/nios2/include/asm/Kbuild
@@ -8,6 +8,7 @@ generic-y += current.h
generic-y += device.h
generic-y += div64.h
generic-y += dma.h
+generic-y += dma-mapping.h
generic-y += emergency-restart.h
generic-y += exec.h
generic-y += extable.h
diff --git a/arch/nios2/include/asm/dma-mapping.h
b/arch/nios2/include/asm/dma-mapping.h
deleted file mode 100644
index 6ceb92251da0..
--- a/arch/nios2/include/asm/dma-mapping.h
+++ /dev/null
@@ -1,20 +0,0 @@
-/*
- * Copyright (C) 2011 Tobias Klauser
- * Copyright (C) 2009 Wind River Systems Inc
- *
- * 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.
- */
-
-#ifndef _ASM_NIOS2_DMA_MAPPING_H
-#define _ASM_NIOS2_DMA_MAPPING_H
-
-extern const struct dma_map_ops nios2_dma_ops;
-
-static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus)
-{
- return &nios2_dma_ops;
-}
-
-#endif /* _ASM_NIOS2_DMA_MAPPING_H */
diff --git a/arch/nios2/mm/dma-mapping.c b/arch/nios2/mm/dma-mapping.c
index 4be815519dd4..4af9e5b5ba1c 100644
--- a/arch/nios2/mm/dma-mapping.c
+++ b/arch/nios2/mm/dma-mapping.c
@@ -12,18 +12,18 @@
#include
#include
-#include
#include
-#include
#include
#include
#include
#include
-static inline void __dma_sync_for_device(void *vaddr, size_t size,
- enum dma_data_direction direction)
+void arch_sync_dma_for_device(struct device *dev, phys_addr_t paddr,
+ size_t size, enum dma_data_direction dir)
{
- switch (direction) {
+ void *vaddr = phys_to_virt(paddr);
+
+ switch (dir) {
case DMA_FROM_DEVICE:
invalidate_dcache_range((unsigned long)vaddr,
(unsigned long)(vaddr + size));
@@ -42,10 +42,12 @@ static inline void __dma_sync_for_device(void *vaddr,
size_t size,
}
}
-static inline void __dma_sync_for_cpu(void *vaddr, size_t size,
- enum dma_data_direction direction)
+void arch_sync_dma_for_cpu(struct device *dev, phys_addr_t paddr,
+ size_t size, enum dma_data_direction dir)
{
- switch (direction) {
+ void *vaddr = phys_to_virt(paddr);
+
+ switch (dir) {
case DMA_BIDIRECTIONAL:
case DMA_FROM_DEVICE:
invalidate_dcache_range((unsigned long)vaddr,
@@ -58,8 +60,8 @@ static inline void __dma_sync_for_cpu(void *vaddr, size_t
size,
}
}
-static void *nios2_dma_alloc(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t gfp, unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_handle,
+ gfp_t gfp, unsigned long attrs)
{
void *ret;
@@ -80,125 +82,10 @@ static void *nios2_dma_alloc(struct device *dev, size_t
size,
return ret;
}
-static void nios2_dma_free(struct device *dev, size_t size, void *vaddr,
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
dma_addr_t dma_handle, unsigned long attrs)
{
unsigned long addr = (unsigned long) CAC_ADDR((unsigned long) vaddr);
free_pages(addr, get_order(size));
}
-
-static int nios2_dma_map_sg(struct device *dev, struct scatterlist *sg,
- int nents, enum dma_data_direction direction,
- unsigned long attrs)
-{
- int i;
-
- for_each_sg(sg, sg, nents, i) {
- void *addr = sg_virt(sg);
-
- if (!addr)
- continue;
-
- sg->dma_address = sg_phys(sg);
-
- if (attrs & DMA_ATTR_SKIP_CPU_SYNC)
- continue;
-
- __dma_sync_for_device(addr, sg->length, direction);
- }
-
- return nents;
-}
-
-static dma_addr_t nios2_dma_map_page(struct device *dev, struct page *page,
- unsigned long offset, size_t size,
-
[PATCH 22/25] sparc: use generic dma_noncoherent_ops
Switch to the generic noncoherent direct mapping implementation.
This removes the previous sync_single_for_device implementation, which
looks bogus given that no syncing is happening in the similar but more
important map_single case.
Signed-off-by: Christoph Hellwig
---
arch/sparc/Kconfig | 2 +
arch/sparc/include/asm/dma-mapping.h | 5 +-
arch/sparc/kernel/ioport.c | 151 ++-
3 files changed, 14 insertions(+), 144 deletions(-)
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index 435dbc033afe..0889b4eabf8b 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -48,6 +48,8 @@ config SPARC
config SPARC32
def_bool !64BIT
+ select ARCH_HAS_SYNC_DMA_FOR_CPU
+ select DMA_NONCOHERENT_OPS
select GENERIC_ATOMIC64
select CLZ_TAB
select HAVE_UID16
diff --git a/arch/sparc/include/asm/dma-mapping.h
b/arch/sparc/include/asm/dma-mapping.h
index 12ae33daf52f..e17566376934 100644
--- a/arch/sparc/include/asm/dma-mapping.h
+++ b/arch/sparc/include/asm/dma-mapping.h
@@ -7,7 +7,6 @@
#include
extern const struct dma_map_ops *dma_ops;
-extern const struct dma_map_ops pci32_dma_ops;
extern struct bus_type pci_bus_type;
@@ -15,11 +14,11 @@ static inline const struct dma_map_ops
*get_arch_dma_ops(struct bus_type *bus)
{
#ifdef CONFIG_SPARC_LEON
if (sparc_cpu_model == sparc_leon)
- return &pci32_dma_ops;
+ return &dma_noncoherent_ops;
#endif
#if defined(CONFIG_SPARC32) && defined(CONFIG_PCI)
if (bus == &pci_bus_type)
- return &pci32_dma_ops;
+ return &dma_noncoherent_ops;
#endif
return dma_ops;
}
diff --git a/arch/sparc/kernel/ioport.c b/arch/sparc/kernel/ioport.c
index 3bcef9ce74df..7954512c42e7 100644
--- a/arch/sparc/kernel/ioport.c
+++ b/arch/sparc/kernel/ioport.c
@@ -38,6 +38,7 @@
#include
#include
#include
+#include
#include
#include
@@ -434,9 +435,8 @@ arch_initcall(sparc_register_ioport);
/* Allocate and map kernel buffer using consistent mode DMA for a device.
* hwdev should be valid struct pci_dev pointer for PCI devices.
*/
-static void *pci32_alloc_coherent(struct device *dev, size_t len,
- dma_addr_t *pba, gfp_t gfp,
- unsigned long attrs)
+void *arch_dma_alloc(struct device *dev, size_t len, dma_addr_t *pba, gfp_t
gfp,
+ unsigned long attrs)
{
unsigned long len_total = PAGE_ALIGN(len);
void *va;
@@ -488,8 +488,8 @@ static void *pci32_alloc_coherent(struct device *dev,
size_t len,
* References to the memory and mappings associated with cpu_addr/dma_addr
* past this call are illegal.
*/
-static void pci32_free_coherent(struct device *dev, size_t n, void *p,
- dma_addr_t ba, unsigned long attrs)
+void arch_dma_free(struct device *dev, size_t n, void *p, dma_addr_t ba,
+ unsigned long attrs)
{
struct resource *res;
@@ -519,146 +519,15 @@ static void pci32_free_coherent(struct device *dev,
size_t n, void *p,
free_pages((unsigned long)phys_to_virt(ba), get_order(n));
}
-/*
- * Same as pci_map_single, but with pages.
- */
-static dma_addr_t pci32_map_page(struct device *dev, struct page *page,
-unsigned long offset, size_t size,
-enum dma_data_direction dir,
-unsigned long attrs)
-{
- /* IIep is write-through, not flushing. */
- return page_to_phys(page) + offset;
-}
-
-static void pci32_unmap_page(struct device *dev, dma_addr_t ba, size_t size,
-enum dma_data_direction dir, unsigned long attrs)
-{
- if (dir != PCI_DMA_TODEVICE && !(attrs & DMA_ATTR_SKIP_CPU_SYNC))
- dma_make_coherent(ba, PAGE_ALIGN(size));
-}
-
-/* Map a set of buffers described by scatterlist in streaming
- * mode for DMA. This is the scatter-gather version of the
- * above pci_map_single interface. Here the scatter gather list
- * elements are each tagged with the appropriate dma address
- * and length. They are obtained via sg_dma_{address,length}(SG).
- *
- * NOTE: An implementation may be able to use a smaller number of
- * DMA address/length pairs than there are SG table elements.
- * (for example via virtual mapping capabilities)
- * The routine returns the number of addr/length pairs actually
- * used, at most nents.
- *
- * Device ownership issues as mentioned above for pci_map_single are
- * the same here.
- */
-static int pci32_map_sg(struct device *device, struct scatterlist *sgl,
- int nents, enum dma_data_direction dir,
- unsigned long attrs)
-{
- struct scatterlist *sg;
- int n;
-
- /* IIep is write-through, not flushing. */
- for_each_sg(sgl, sg, nents, n) {
- sg->dma_address = sg_phy
RE: ARC compact700 NPS platform - EZ_MachineCheck exception handler
There are two cases to consider for this exception: > but others can't so continuing despite it is recipe for disaster. Perhaps > your chip > has some spurious Machine check exceptions ? 1. Except for core 0, which is running the linux os, all other cores are running packet processing code in ZOL isolation mode. If any of these cores hit the compact 700 0x20 exception it is logical to assume all other cores will hit it too. It seems that eventually in any case, will have to reset HW and reboot the system. It might be beneficial for user to try collect more info for debugging the issue even if it’s a disaster for the system. > Hmm, but you have to explain why those machine checks are fine ! 2. The ARC compact700 instruction set was extended to support fast DMA operations to various added HW accelerators and new asm ops to support network packet Processing. In case of an error, some of these instructions are wired to the 0x20 exception. There is an HW mechanism to partition the DDR between linux os and the various accelerators This mechanism unaware of the mmu or virtual memory handling. In a cases where an accelerator access out of its memory bounds this exception is hit but there is no risk to system stability. User signal handler can catch it allowing easier debugging. This is one example. > > 1: > > FAKE_RET_FROM_EXCPN > > You don't need this. When removing FAKE_RET_FROM_EXCPN, first EV_MachineCheck exception Is causing the core running that thread to stall. If not removed multiple exceptions are handled and system seems healthy. Please note that exception is generated by accessing one of the NPS accelerators address which is out its memory space, so no harm is expected to system > Next time please send a real patch so I know right away what was changed. My apologies, here is the patch based on linux-4.16.10 diff -uprN linux-4.16.10/arch/arc/kernel/entry.S linux/arch/arc/kernel/entry.S --- linux-4.16.10/arch/arc/kernel/entry.S 2018-05-19 11:19:37.0 +0300 +++ linux/arch/arc/kernel/entry.S 2018-05-22 14:12:18.065103918 +0300 @@ -106,13 +106,9 @@ ENTRY(EV_MachineCheck) b ret_from_exception 1: - ; DEAD END: can't do much, display Regs and HALT - SAVE_CALLEE_SAVED_USER - - GET_CURR_TASK_FIELD_PTR TASK_THREAD, r10 - st sp, [r10, THREAD_CALLEE_REG] - - j do_machine_check_fault + FAKE_RET_FROM_EXCPN + bl do_machine_check + b ret_from_exception END(EV_MachineCheck) diff -uprN linux-4.16.10/arch/arc/kernel/traps.c linux/arch/arc/kernel/traps.c --- linux-4.16.10/arch/arc/kernel/traps.c 2018-05-19 11:19:37.0 +0300 +++ linux/arch/arc/kernel/traps.c 2018-05-22 14:13:25.162748373 +0300 @@ -86,6 +86,7 @@ DO_ERROR_INFO(SIGBUS, "Invalid Mem Acces DO_ERROR_INFO(SIGTRAP, "Breakpoint Set", trap_is_brkpt, TRAP_BRKPT) DO_ERROR_INFO(SIGBUS, "Misaligned Access", do_misaligned_error, BUS_ADRALN) DO_ERROR_INFO(SIGSEGV, "gcc generated __builtin_trap", do_trap5_error, 0) +DO_ERROR_INFO(SIGBUS, "Machine Check", do_machine_check, BUS_MCEERR_AR ) /* * Entry Point for Misaligned Data access Exception, for emulating in software > -Original Message- > From: Vineet Gupta [mailto:vineet.gup...@synopsys.com] > Sent: Monday, May 21, 2018 19:59 > To: Ofer Levi(SW) > Cc: linux-ker...@vger.kernel.org; Meir Lichtinger ; > arcml > Subject: Re: ARC compact700 NPS platform - EZ_MachineCheck exception > handler > > On 05/21/2018 07:14 AM, Ofer Levi(SW) wrote: > > Resending, due to typo in LKML mail address. > > Also please CC linux-snps-arc@lists.infradead.org for any ARC Linux related > posts. > > > > > The EV_MachineCheck exception handler is halting the core for > exceptions > > which are not tlb_overlap_fault. > > Since for the NPS platform each core is running a single thread in ZOL > > (Zero > > Overhead Linux) isolation mode, we feel that most of the time it is safe > > to > > resume execution instead of halting the core. > > Most of the time is not good enough when dealing with OS code :-( A > Machine check excepting implies something went terribly wrong. Some of > those cases can be handled gracefully (such as duplicate TLB entry), but > others can't so continuing despite it is recipe for disaster. Perhaps your > chip > has some spurious Machine check exceptions ? > > > I would appreciate it if you could review the change below > > Next time please send a real patch so I know right away what was changed. > > > and let me know > > what you think, if this change is valid or if we missed or overlooked > > something. > > We are not looking to push this change upstream, but will be used on > some > > systems. > > Hmm, but you have to explain why those machine checks are fine ! > > > > > Please see below our implementation after label 1. > > > > Thanks > > Ofer > > > > ENTRY(EV_MachineCheck) > > > > EXCEPTIO
