diff options
40 files changed, 1899 insertions, 844 deletions
diff --git a/Documentation/DMA-API.txt b/Documentation/DMA-API.txt index 2a3fcc55e98..d9aa43d78bc 100644 --- a/Documentation/DMA-API.txt +++ b/Documentation/DMA-API.txt @@ -609,3 +609,109 @@ size is the size (and should be a page-sized multiple). The return value will be either a pointer to the processor virtual address of the memory, or an error (via PTR_ERR()) if any part of the region is occupied. + +Part III - Debug drivers use of the DMA-API +------------------------------------------- + +The DMA-API as described above as some constraints. DMA addresses must be +released with the corresponding function with the same size for example. With +the advent of hardware IOMMUs it becomes more and more important that drivers +do not violate those constraints. In the worst case such a violation can +result in data corruption up to destroyed filesystems. + +To debug drivers and find bugs in the usage of the DMA-API checking code can +be compiled into the kernel which will tell the developer about those +violations. If your architecture supports it you can select the "Enable +debugging of DMA-API usage" option in your kernel configuration. Enabling this +option has a performance impact. Do not enable it in production kernels. + +If you boot the resulting kernel will contain code which does some bookkeeping +about what DMA memory was allocated for which device. If this code detects an +error it prints a warning message with some details into your kernel log. An +example warning message may look like this: + +------------[ cut here ]------------ +WARNING: at /data2/repos/linux-2.6-iommu/lib/dma-debug.c:448 + check_unmap+0x203/0x490() +Hardware name: +forcedeth 0000:00:08.0: DMA-API: device driver frees DMA memory with wrong + function [device address=0x00000000640444be] [size=66 bytes] [mapped as +single] [unmapped as page] +Modules linked in: nfsd exportfs bridge stp llc r8169 +Pid: 0, comm: swapper Tainted: G W 2.6.28-dmatest-09289-g8bb99c0 #1 +Call Trace: + <IRQ> [<ffffffff80240b22>] warn_slowpath+0xf2/0x130 + [<ffffffff80647b70>] _spin_unlock+0x10/0x30 + [<ffffffff80537e75>] usb_hcd_link_urb_to_ep+0x75/0xc0 + [<ffffffff80647c22>] _spin_unlock_irqrestore+0x12/0x40 + [<ffffffff8055347f>] ohci_urb_enqueue+0x19f/0x7c0 + [<ffffffff80252f96>] queue_work+0x56/0x60 + [<ffffffff80237e10>] enqueue_task_fair+0x20/0x50 + [<ffffffff80539279>] usb_hcd_submit_urb+0x379/0xbc0 + [<ffffffff803b78c3>] cpumask_next_and+0x23/0x40 + [<ffffffff80235177>] find_busiest_group+0x207/0x8a0 + [<ffffffff8064784f>] _spin_lock_irqsave+0x1f/0x50 + [<ffffffff803c7ea3>] check_unmap+0x203/0x490 + [<ffffffff803c8259>] debug_dma_unmap_page+0x49/0x50 + [<ffffffff80485f26>] nv_tx_done_optimized+0xc6/0x2c0 + [<ffffffff80486c13>] nv_nic_irq_optimized+0x73/0x2b0 + [<ffffffff8026df84>] handle_IRQ_event+0x34/0x70 + [<ffffffff8026ffe9>] handle_edge_irq+0xc9/0x150 + [<ffffffff8020e3ab>] do_IRQ+0xcb/0x1c0 + [<ffffffff8020c093>] ret_from_intr+0x0/0xa + <EOI> <4>---[ end trace f6435a98e2a38c0e ]--- + +The driver developer can find the driver and the device including a stacktrace +of the DMA-API call which caused this warning. + +Per default only the first error will result in a warning message. All other +errors will only silently counted. This limitation exist to prevent the code +from flooding your kernel log. To support debugging a device driver this can +be disabled via debugfs. See the debugfs interface documentation below for +details. + +The debugfs directory for the DMA-API debugging code is called dma-api/. In +this directory the following files can currently be found: + + dma-api/all_errors This file contains a numeric value. If this + value is not equal to zero the debugging code + will print a warning for every error it finds + into the kernel log. Be carefull with this + option. It can easily flood your logs. + + dma-api/disabled This read-only file contains the character 'Y' + if the debugging code is disabled. This can + happen when it runs out of memory or if it was + disabled at boot time + + dma-api/error_count This file is read-only and shows the total + numbers of errors found. + + dma-api/num_errors The number in this file shows how many + warnings will be printed to the kernel log + before it stops. This number is initialized to + one at system boot and be set by writing into + this file + + dma-api/min_free_entries + This read-only file can be read to get the + minimum number of free dma_debug_entries the + allocator has ever seen. If this value goes + down to zero the code will disable itself + because it is not longer reliable. + + dma-api/num_free_entries + The current number of free dma_debug_entries + in the allocator. + +If you have this code compiled into your kernel it will be enabled by default. +If you want to boot without the bookkeeping anyway you can provide +'dma_debug=off' as a boot parameter. This will disable DMA-API debugging. +Notice that you can not enable it again at runtime. You have to reboot to do +so. + +When the code disables itself at runtime this is most likely because it ran +out of dma_debug_entries. These entries are preallocated at boot. The number +of preallocated entries is defined per architecture. If it is too low for you +boot with 'dma_debug_entries=<your_desired_number>' to overwrite the +architectural default. diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index be3bde51b56..aeedb89a307 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -492,6 +492,16 @@ and is between 256 and 4096 characters. It is defined in the file Range: 0 - 8192 Default: 64 + dma_debug=off If the kernel is compiled with DMA_API_DEBUG support + this option disables the debugging code at boot. + + dma_debug_entries=<number> + This option allows to tune the number of preallocated + entries for DMA-API debugging code. One entry is + required per DMA-API allocation. Use this if the + DMA-API debugging code disables itself because the + architectural default is too low. + hpet= [X86-32,HPET] option to control HPET usage Format: { enable (default) | disable | force | verbose } diff --git a/arch/Kconfig b/arch/Kconfig index 550dab22daa..830c16a2b80 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -106,3 +106,5 @@ config HAVE_CLK The <linux/clk.h> calls support software clock gating and thus are a key power management tool on many systems. +config HAVE_DMA_API_DEBUG + bool diff --git a/arch/ia64/dig/Makefile b/arch/ia64/dig/Makefile index 5c0283830bd..2f7caddf093 100644 --- a/arch/ia64/dig/Makefile +++ b/arch/ia64/dig/Makefile @@ -7,8 +7,8 @@ obj-y := setup.o ifeq ($(CONFIG_DMAR), y) -obj-$(CONFIG_IA64_GENERIC) += machvec.o machvec_vtd.o dig_vtd_iommu.o +obj-$(CONFIG_IA64_GENERIC) += machvec.o machvec_vtd.o else obj-$(CONFIG_IA64_GENERIC) += machvec.o endif -obj-$(CONFIG_IA64_DIG_VTD) += dig_vtd_iommu.o + diff --git a/arch/ia64/dig/dig_vtd_iommu.c b/arch/ia64/dig/dig_vtd_iommu.c deleted file mode 100644 index 1c8a079017a..00000000000 --- a/arch/ia64/dig/dig_vtd_iommu.c +++ /dev/null @@ -1,59 +0,0 @@ -#include <linux/types.h> -#include <linux/kernel.h> -#include <linux/module.h> -#include <linux/intel-iommu.h> - -void * -vtd_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, - gfp_t flags) -{ - return intel_alloc_coherent(dev, size, dma_handle, flags); -} -EXPORT_SYMBOL_GPL(vtd_alloc_coherent); - -void -vtd_free_coherent(struct device *dev, size_t size, void *vaddr, - dma_addr_t dma_handle) -{ - intel_free_coherent(dev, size, vaddr, dma_handle); -} -EXPORT_SYMBOL_GPL(vtd_free_coherent); - -dma_addr_t -vtd_map_single_attrs(struct device *dev, void *addr, size_t size, - int dir, struct dma_attrs *attrs) -{ - return intel_map_single(dev, (phys_addr_t)addr, size, dir); -} -EXPORT_SYMBOL_GPL(vtd_map_single_attrs); - -void -vtd_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size, - int dir, struct dma_attrs *attrs) -{ - intel_unmap_single(dev, iova, size, dir); -} -EXPORT_SYMBOL_GPL(vtd_unmap_single_attrs); - -int -vtd_map_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents, - int dir, struct dma_attrs *attrs) -{ - return intel_map_sg(dev, sglist, nents, dir); -} -EXPORT_SYMBOL_GPL(vtd_map_sg_attrs); - -void -vtd_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist, - int nents, int dir, struct dma_attrs *attrs) -{ - intel_unmap_sg(dev, sglist, nents, dir); -} -EXPORT_SYMBOL_GPL(vtd_unmap_sg_attrs); - -int -vtd_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) -{ - return 0; -} -EXPORT_SYMBOL_GPL(vtd_dma_mapping_error); diff --git a/arch/ia64/hp/common/hwsw_iommu.c b/arch/ia64/hp/common/hwsw_iommu.c index 2769dbfd03b..e4a80d82e3d 100644 --- a/arch/ia64/hp/common/hwsw_iommu.c +++ b/arch/ia64/hp/common/hwsw_iommu.c @@ -13,49 +13,34 @@ */ #include <linux/device.h> +#include <linux/dma-mapping.h> #include <linux/swiotlb.h> - #include <asm/machvec.h> +extern struct dma_map_ops sba_dma_ops, swiotlb_dma_ops; + /* swiotlb declarations & definitions: */ extern int swiotlb_late_init_with_default_size (size_t size); -/* hwiommu declarations & definitions: */ - -extern ia64_mv_dma_alloc_coherent sba_alloc_coherent; -extern ia64_mv_dma_free_coherent sba_free_coherent; -extern ia64_mv_dma_map_single_attrs sba_map_single_attrs; -extern ia64_mv_dma_unmap_single_attrs sba_unmap_single_attrs; -extern ia64_mv_dma_map_sg_attrs sba_map_sg_attrs; -extern ia64_mv_dma_unmap_sg_attrs sba_unmap_sg_attrs; -extern ia64_mv_dma_supported sba_dma_supported; -extern ia64_mv_dma_mapping_error sba_dma_mapping_error; - -#define hwiommu_alloc_coherent sba_alloc_coherent -#define hwiommu_free_coherent sba_free_coherent -#define hwiommu_map_single_attrs sba_map_single_attrs -#define hwiommu_unmap_single_attrs sba_unmap_single_attrs -#define hwiommu_map_sg_attrs sba_map_sg_attrs -#define hwiommu_unmap_sg_attrs sba_unmap_sg_attrs -#define hwiommu_dma_supported sba_dma_supported -#define hwiommu_dma_mapping_error sba_dma_mapping_error -#define hwiommu_sync_single_for_cpu machvec_dma_sync_single -#define hwiommu_sync_sg_for_cpu machvec_dma_sync_sg -#define hwiommu_sync_single_for_device machvec_dma_sync_single -#define hwiommu_sync_sg_for_device machvec_dma_sync_sg - - /* * Note: we need to make the determination of whether or not to use * the sw I/O TLB based purely on the device structure. Anything else * would be unreliable or would be too intrusive. */ -static inline int -use_swiotlb (struct device *dev) +static inline int use_swiotlb(struct device *dev) { - return dev && dev->dma_mask && !hwiommu_dma_supported(dev, *dev->dma_mask); + return dev && dev->dma_mask && + !sba_dma_ops.dma_supported(dev, *dev->dma_mask); } +struct dma_map_ops *hwsw_dma_get_ops(struct device *dev) +{ + if (use_swiotlb(dev)) + return &swiotlb_dma_ops; + return &sba_dma_ops; +} +EXPORT_SYMBOL(hwsw_dma_get_ops); + void __init hwsw_init (void) { @@ -71,125 +56,3 @@ hwsw_init (void) #endif } } - -void * -hwsw_alloc_coherent (struct device *dev, size_t size, dma_addr_t *dma_handle, gfp_t flags) -{ - if (use_swiotlb(dev)) - return swiotlb_alloc_coherent(dev, size, dma_handle, flags); - else - return hwiommu_alloc_coherent(dev, size, dma_handle, flags); -} - -void -hwsw_free_coherent (struct device *dev, size_t size, void *vaddr, dma_addr_t dma_handle) -{ - if (use_swiotlb(dev)) - swiotlb_free_coherent(dev, size, vaddr, dma_handle); - else - hwiommu_free_coherent(dev, size, vaddr, dma_handle); -} - -dma_addr_t -hwsw_map_single_attrs(struct device *dev, void *addr, size_t size, int dir, - struct dma_attrs *attrs) -{ - if (use_swiotlb(dev)) - return swiotlb_map_single_attrs(dev, addr, size, dir, attrs); - else - return hwiommu_map_single_attrs(dev, addr, size, dir, attrs); -} -EXPORT_SYMBOL(hwsw_map_single_attrs); - -void -hwsw_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size, - int dir, struct dma_attrs *attrs) -{ - if (use_swiotlb(dev)) - return swiotlb_unmap_single_attrs(dev, iova, size, dir, attrs); - else - return hwiommu_unmap_single_attrs(dev, iova, size, dir, attrs); -} -EXPORT_SYMBOL(hwsw_unmap_single_attrs); - -int -hwsw_map_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents, - int dir, struct dma_attrs *attrs) -{ - if (use_swiotlb(dev)) - return swiotlb_map_sg_attrs(dev, sglist, nents, dir, attrs); - else - return hwiommu_map_sg_attrs(dev, sglist, nents, dir, attrs); -} -EXPORT_SYMBOL(hwsw_map_sg_attrs); - -void -hwsw_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents, - int dir, struct dma_attrs *attrs) -{ - if (use_swiotlb(dev)) - return swiotlb_unmap_sg_attrs(dev, sglist, nents, dir, attrs); - else - return hwiommu_unmap_sg_attrs(dev, sglist, nents, dir, attrs); -} -EXPORT_SYMBOL(hwsw_unmap_sg_attrs); - -void -hwsw_sync_single_for_cpu (struct device *dev, dma_addr_t addr, size_t size, int dir) -{ - if (use_swiotlb(dev)) - swiotlb_sync_single_for_cpu(dev, addr, size, dir); - else - hwiommu_sync_single_for_cpu(dev, addr, size, dir); -} - -void -hwsw_sync_sg_for_cpu (struct device *dev, struct scatterlist *sg, int nelems, int dir) -{ - if (use_swiotlb(dev)) - swiotlb_sync_sg_for_cpu(dev, sg, nelems, dir); - else - hwiommu_sync_sg_for_cpu(dev, sg, nelems, dir); -} - -void -hwsw_sync_single_for_device (struct device *dev, dma_addr_t addr, size_t size, int dir) -{ - if (use_swiotlb(dev)) - swiotlb_sync_single_for_device(dev, addr, size, dir); - else - hwiommu_sync_single_for_device(dev, addr, size, dir); -} - -void -hwsw_sync_sg_for_device (struct device *dev, struct scatterlist *sg, int nelems, int dir) -{ - if (use_swiotlb(dev)) - swiotlb_sync_sg_for_device(dev, sg, nelems, dir); - else - hwiommu_sync_sg_for_device(dev, sg, nelems, dir); -} - -int -hwsw_dma_supported (struct device *dev, u64 mask) -{ - if (hwiommu_dma_supported(dev, mask)) - return 1; - return swiotlb_dma_supported(dev, mask); -} - -int -hwsw_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) -{ - return hwiommu_dma_mapping_error(dev, dma_addr) || - swiotlb_dma_mapping_error(dev, dma_addr); -} - -EXPORT_SYMBOL(hwsw_dma_mapping_error); -EXPORT_SYMBOL(hwsw_dma_supported); -EXPORT_SYMBOL(hwsw_alloc_coherent); -EXPORT_SYMBOL(hwsw_free_coherent); -EXPORT_SYMBOL(hwsw_sync_single_for_cpu); -EXPORT_SYMBOL(hwsw_sync_single_for_device); -EXPORT_SYMBOL(hwsw_sync_sg_for_cpu); -EXPORT_SYMBOL(hwsw_sync_sg_for_device); diff --git a/arch/ia64/hp/common/sba_iommu.c b/arch/ia64/hp/common/sba_iommu.c index 6d5e6c5630e..56ceb68eb99 100644 --- a/arch/ia64/hp/common/sba_iommu.c +++ b/arch/ia64/hp/common/sba_iommu.c @@ -36,6 +36,7 @@ #include <linux/bitops.h> /* hweight64() */ #include <linux/crash_dump.h> #include <linux/iommu-helper.h> +#include <linux/dma-mapping.h> #include <asm/delay.h> /* ia64_get_itc() */ #include <asm/io.h> @@ -908,11 +909,13 @@ sba_mark_invalid(struct ioc *ioc, dma_addr_t iova, size_t byte_cnt) * * See Documentation/PCI/PCI-DMA-mapping.txt */ -dma_addr_t -sba_map_single_attrs(struct device *dev, void *addr, size_t size, int dir, - struct dma_attrs *attrs) +static dma_addr_t sba_map_page(struct device *dev, struct page *page, + unsigned long poff, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { struct ioc *ioc; + void *addr = page_address(page) + poff; dma_addr_t iovp; dma_addr_t offset; u64 *pdir_start; @@ -990,7 +993,14 @@ sba_map_single_attrs(struct device *dev, void *addr, size_t size, int dir, #endif return SBA_IOVA(ioc, iovp, offset); } -EXPORT_SYMBOL(sba_map_single_attrs); + +static dma_addr_t sba_map_single_attrs(struct device *dev, void *addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + return sba_map_page(dev, virt_to_page(addr), + (unsigned long)addr & ~PAGE_MASK, size, dir, attrs); +} #ifdef ENABLE_MARK_CLEAN static SBA_INLINE void @@ -1026,8 +1036,8 @@ sba_mark_clean(struct ioc *ioc, dma_addr_t iova, size_t size) * * See Documentation/PCI/PCI-DMA-mapping.txt */ -void sba_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size, - int dir, struct dma_attrs *attrs) +static void sba_unmap_page(struct device *dev, dma_addr_t iova, size_t size, + enum dma_data_direction dir, struct dma_attrs *attrs) { struct ioc *ioc; #if DELAYED_RESOURCE_CNT > 0 @@ -1094,7 +1104,12 @@ void sba_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size, spin_unlock_irqrestore(&ioc->res_lock, flags); #endif /* DELAYED_RESOURCE_CNT == 0 */ } -EXPORT_SYMBOL(sba_unmap_single_attrs); + +void sba_unmap_single_attrs(struct device *dev, dma_addr_t iova, size_t size, + enum dma_data_direction dir, struct dma_attrs *attrs) +{ + sba_unmap_page(dev, iova, size, dir, attrs); +} /** * sba_alloc_coherent - allocate/map shared mem for DMA @@ -1104,7 +1119,7 @@ EXPORT_SYMBOL(sba_unmap_single_attrs); * * See Documentation/PCI/PCI-DMA-mapping.txt */ -void * +static void * sba_alloc_coherent (struct device *dev, size_t size, dma_addr_t *dma_handle, gfp_t flags) { struct ioc *ioc; @@ -1167,7 +1182,8 @@ sba_alloc_coherent (struct device *dev, size_t size, dma_addr_t *dma_handle, gfp * * See Documentation/PCI/PCI-DMA-mapping.txt */ -void sba_free_coherent (struct device *dev, size_t size, void *vaddr, dma_addr_t dma_handle) +static void sba_free_coherent (struct device *dev, size_t size, void *vaddr, + dma_addr_t dma_handle) { sba_unmap_single_attrs(dev, dma_handle, size, 0, NULL); free_pages((unsigned long) vaddr, get_order(size)); @@ -1422,8 +1438,9 @@ sba_coalesce_chunks(struct ioc *ioc, struct device *dev, * * See Documentation/PCI/PCI-DMA-mapping.txt */ -int sba_map_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents, - int dir, struct dma_attrs *attrs) +static int sba_map_sg_attrs(struct device *dev, struct scatterlist *sglist, + int nents, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct ioc *ioc; int coalesced, filled = 0; @@ -1502,7 +1519,6 @@ int sba_map_sg_attrs(struct device *dev, struct scatterlist *sglist, int nents, return filled; } -EXPORT_SYMBOL(sba_map_sg_attrs); /** * sba_unmap_sg_attrs - unmap Scatter/Gather list @@ -1514,8 +1530,9 @@ EXPORT_SYMBOL(sba_map_sg_attrs); * * See Documentation/PCI/PCI-DMA-mapping.txt */ -void sba_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist, - int nents, int dir, struct dma_attrs *attrs) +static void sba_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist, + int nents, enum dma_data_direction dir, + struct dma_attrs *attrs) { #ifdef ASSERT_PDIR_SANITY struct ioc *ioc; @@ -1551,7 +1568,6 @@ void sba_unmap_sg_attrs(struct device *dev, struct scatterlist *sglist, #endif } -EXPORT_SYMBOL(sba_unmap_sg_attrs); /************************************************************** * @@ -2064,6 +2080,8 @@ static struct acpi_driver acpi_sba_ioc_driver = { }, }; +extern struct dma_map_ops swiotlb_dma_ops; + static int __init sba_init(void) { @@ -2077,6 +2095,7 @@ sba_init(void) * a successful kdump kernel boot is to use the swiotlb. */ if (is_kdump_kernel()) { + dma_ops = &swiotlb_dma_ops; if (swiotlb_late_init_with_default_size(64 * (1<<20)) != 0) panic("Unable to initialize software I/O TLB:" " Try machvec=dig boot option"); @@ -2092,6 +2111,7 @@ sba_init(void) * If we didn't find something sba_iommu can claim, we * need to setup the swiotlb and switch to the dig machvec. */ + dma_ops = &swiotlb_dma_ops; if (swiotlb_late_init_with_default_size(64 * (1<<20)) != 0) panic("Unable to find SBA IOMMU or initialize " "software I/O TLB: Try machvec=dig boot option"); @@ -2138,15 +2158,13 @@ nosbagart(char *str) return 1; } -int -sba_dma_supported (struct device *dev, u64 mask) +static int sba_dma_supported (struct device *dev, u64 mask) { /* make sure it's at least 32bit capable */ return ((mask & 0xFFFFFFFFUL) == 0xFFFFFFFFUL); } -int -sba_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) +static int sba_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) { return 0; } @@ -2176,7 +2194,22 @@ sba_page_override(char *str) __setup("sbapagesize=",sba_page_override); -EXPORT_SYMBOL(sba_dma_mapping_error); -EXPORT_SYMBOL(sba_dma_supported); -EXPORT_SYMBOL(sba_alloc_coherent); -EXPORT_SYMBOL(sba_free_coherent); +struct dma_map_ops sba_dma_ops = { + .alloc_coherent = sba_alloc_coherent, + .free_coherent = sba_free_coherent, + .map_page = sba_map_page, + .unmap_page = sba_unmap_page, + .map_sg = sba_map_sg_attrs, + .unmap_sg = sba_unmap_sg_attrs, + .sync_single_for_cpu = machvec_dma_sync_single, + .sync_sg_for_cpu = machvec_dma_sync_sg, + .sync_single_for_device = machvec_dma_sync_single, + .sync_sg_for_device = machvec_dma_sync_sg, + .dma_supported = sba_dma_supported, + .mapping_error = sba_dma_mapping_error, +}; + +void sba_dma_init(void) +{ + dma_ops = &sba_dma_ops; +} diff --git a/arch/ia64/include/asm/dma-mapping.h b/arch/ia64/include/asm/dma-mapping.h index 1f912d92758..36c0009dbec 100644 --- a/arch/ia64/include/asm/dma-mapping.h +++ b/arch/ia64/include/asm/dma-mapping.h @@ -11,99 +11,128 @@ #define ARCH_HAS_DMA_GET_REQUIRED_MASK -struct dma_mapping_ops { - int (*mapping_error)(struct device *dev, - dma_addr_t dma_addr); - void* (*alloc_coherent)(struct device *dev, size_t size, - dma_addr_t *dma_handle, gfp_t gfp); - void (*free_coherent)(struct device *dev, size_t size, - void *vaddr, dma_addr_t dma_handle); - dma_addr_t (*map_single)(struct device *hwdev, unsigned long ptr, - size_t size, int direction); - void (*unmap_single)(struct device *dev, dma_addr_t addr, - size_t size, int direction); - void (*sync_single_for_cpu)(struct device *hwdev, - dma_addr_t dma_handle, size_t size, - int direction); - void (*sync_single_for_device)(struct device *hwdev, - dma_addr_t dma_handle, size_t size, - int direction); - void (*sync_single_range_for_cpu)(struct device *hwdev, - dma_addr_t dma_handle, unsigned long offset, - size_t size, int direction); - void (*sync_single_range_for_device)(struct device *hwdev, - dma_addr_t dma_handle, unsigned long offset, - size_t size, int direction); - void (*sync_sg_for_cpu)(struct device *hwdev, - struct scatterlist *sg, int nelems, - int direction); - void (*sync_sg_for_device)(struct device *hwdev, - struct scatterlist *sg, int nelems, - int direction); - int (*map_sg)(struct device *hwdev, struct scatterlist *sg, - int nents, int direction); - void (*unmap_sg)(struct device *hwdev, - struct scatterlist *sg, int nents, - int direction); - int (*dma_supported_op)(struct device *hwdev, u64 mask); - int is_phys; -}; - -extern struct dma_mapping_ops *dma_ops; +extern struct dma_map_ops *dma_ops; extern struct ia64_machine_vector ia64_mv; extern void set_iommu_machvec(void); -#define dma_alloc_coherent(dev, size, handle, gfp) \ - platform_dma_alloc_coherent(dev, size, handle, (gfp) | GFP_DMA) +extern void machvec_dma_sync_single(struct device *, dma_addr_t, size_t, + enum dma_data_direction); +extern void machvec_dma_sync_sg(struct device *, struct scatterlist *, int, + enum dma_data_direction); -/* coherent mem. is cheap */ -static inline void * -dma_alloc_noncoherent(struct device *dev, size_t size, dma_addr_t *dma_handle, - gfp_t flag) +static inline void *dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t *daddr, gfp_t gfp) { - return dma_alloc_coherent(dev, size, dma_handle, flag); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->alloc_coherent(dev, size, daddr, gfp); } -#define dma_free_coherent platform_dma_free_coherent -static inline void -dma_free_noncoherent(struct device *dev, size_t size, void *cpu_addr, - dma_addr_t dma_handle) + +static inline void dma_free_coherent(struct device *dev, size_t size, + void *caddr, dma_addr_t daddr) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->free_coherent(dev, size, caddr, daddr); +} + +#define dma_alloc_noncoherent(d, s, h, f) dma_alloc_coherent(d, s, h, f) +#define dma_free_noncoherent(d, s, v, h) dma_free_coherent(d, s, v, h) + +static inline dma_addr_t dma_map_single_attrs(struct device *dev, + void *caddr, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->map_page(dev, virt_to_page(caddr), + (unsigned long)caddr & ~PAGE_MASK, size, + dir, attrs); +} + +static inline void dma_unmap_single_attrs(struct device *dev, dma_addr_t daddr, + size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->unmap_page(dev, daddr, size, dir, attrs); +} + +#define dma_map_single(d, a, s, r) dma_map_single_attrs(d, a, s, r, NULL) +#define dma_unmap_single(d, a, s, r) dma_unmap_single_attrs(d, a, s, r, NULL) + +static inline int dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl, + int nents, enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->map_sg(dev, sgl, nents, dir, attrs); +} + +static inline void dma_unmap_sg_attrs(struct device *dev, + struct scatterlist *sgl, int nents, + enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->unmap_sg(dev, sgl, nents, dir, attrs); +} + +#define dma_map_sg(d, s, n, r) dma_map_sg_attrs(d, s, n, r, NULL) +#define dma_unmap_sg(d, s, n, r) dma_unmap_sg_attrs(d, s, n, r, NULL) + +static inline void dma_sync_single_for_cpu(struct device *dev, dma_addr_t daddr, + size_t size, + enum dma_data_direction dir) { - dma_free_coherent(dev, size, cpu_addr, dma_handle); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->sync_single_for_cpu(dev, daddr, size, dir); } -#define dma_map_single_attrs platform_dma_map_single_attrs -static inline dma_addr_t dma_map_single(struct device *dev, void *cpu_addr, - size_t size, int dir) + +static inline void dma_sync_sg_for_cpu(struct device *dev, + struct scatterlist *sgl, + int nents, enum dma_data_direction dir) { - return dma_map_single_attrs(dev, cpu_addr, size, dir, NULL); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->sync_sg_for_cpu(dev, sgl, nents, dir); } -#define dma_map_sg_attrs platform_dma_map_sg_attrs -static inline int dma_map_sg(struct device *dev, struct scatterlist *sgl, - int nents, int dir) + +static inline void dma_sync_single_for_device(struct device *dev, + dma_addr_t daddr, + size_t size, + enum dma_data_direction dir) { - return dma_map_sg_attrs(dev, sgl, nents, dir, NULL); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->sync_single_for_device(dev, daddr, size, dir); } -#define dma_unmap_single_attrs platform_dma_unmap_single_attrs -static inline void dma_unmap_single(struct device *dev, dma_addr_t cpu_addr, - size_t size, int dir) + +static inline void dma_sync_sg_for_device(struct device *dev, + struct scatterlist *sgl, + int nents, + enum dma_data_direction dir) { - return dma_unmap_single_attrs(dev, cpu_addr, size, dir, NULL); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + ops->sync_sg_for_device(dev, sgl, nents, dir); } -#define dma_unmap_sg_attrs platform_dma_unmap_sg_attrs -static inline void dma_unmap_sg(struct device *dev, struct scatterlist *sgl, - int nents, int dir) + +static inline int dma_mapping_error(struct device *dev, dma_addr_t daddr) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->mapping_error(dev, daddr); +} + +static inline dma_addr_t dma_map_page(struct device *dev, struct page *page, + size_t offset, size_t size, + enum dma_data_direction dir) { - return dma_unmap_sg_attrs(dev, sgl, nents, dir, NULL); + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->map_page(dev, page, offset, size, dir, NULL); } -#define dma_sync_single_for_cpu platform_dma_sync_single_for_cpu -#define dma_sync_sg_for_cpu platform_dma_sync_sg_for_cpu -#define dma_sync_single_for_device platform_dma_sync_single_for_device -#define dma_sync_sg_for_device platform_dma_sync_sg_for_device -#define dma_mapping_error platform_dma_mapping_error -#define dma_map_page(dev, pg, off, size, dir) \ - dma_map_single(dev, page_address(pg) + (off), (size), (dir)) -#define dma_unmap_page(dev, dma_addr, size, dir) \ - dma_unmap_single(dev, dma_addr, size, dir) +static inline void dma_unmap_page(struct device *dev, dma_addr_t addr, + size_t size, enum dma_data_direction dir) +{ + dma_unmap_single(dev, addr, size, dir); +} /* * Rest of this file is part of the "Advanced DMA API". Use at your own risk. @@ -115,7 +144,11 @@ static inline void dma_unmap_sg(struct device *dev, struct scatterlist *sgl, #define dma_sync_single_range_for_device(dev, dma_handle, offset, size, dir) \ dma_sync_single_for_device(dev, dma_handle, size, dir) -#define dma_supported platform_dma_supported +static inline int dma_supported(struct device *dev, u64 mask) +{ + struct dma_map_ops *ops = platform_dma_get_ops(dev); + return ops->dma_supported(dev, mask); +} static inline int dma_set_mask (struct device *dev, u64 mask) @@ -141,11 +174,4 @@ dma_cache_sync (struct device *dev, void *vaddr, size_t size, #define dma_is_consistent(d, h) (1) /* all we do is coherent memory... */ -static inline struct dma_mapping_ops *get_dma_ops(struct device *dev) -{ - return dma_ops; -} - - - #endif /* _ASM_IA64_DMA_MAPPING_H */ diff --git a/arch/ia64/include/asm/machvec.h b/arch/ia64/include/asm/machvec.h index fe87b212170..367d299d993 100644 --- a/arch/ia64/include/asm/machvec.h +++ b/arch/ia64/include/asm/machvec.h @@ -11,7 +11,6 @@ #define _ASM_IA64_MACHVEC_H #include <linux/types.h> -#include <linux/swiotlb.h> /* forward declarations: */ struct device; @@ -45,24 +44,8 @@ typedef void ia64_mv_kernel_launch_event_t(void); /* DMA-mapping interface: */ typedef void ia64_mv_dma_init (void); -typedef void *ia64_mv_dma_alloc_coherent (struct device *, size_t, dma_addr_t *, gfp_t); -typedef void ia64_mv_dma_free_coherent (struct device *, size_t, void *, dma_addr_t); -typedef dma_addr_t ia64_mv_dma_map_single (struct device *, void *, size_t, int); -typedef void ia64_mv_dma_unmap_single (struct device *, dma_addr_t, size_t, int); -typedef int ia64_mv_dma_map_sg (struct device *, struct scatterlist *, int, int); -typedef void ia64_mv_dma_unmap_sg (struct device *, struct scatterlist *, int, int); -typedef void ia64_mv_dma_sync_single_for_cpu (struct device *, dma_addr_t, size_t, int); -typedef void ia64_mv_dma_sync_sg_for_cpu (struct device *, struct scatterlist *, int, int); -typedef void ia64_mv_dma_sync_single_for_device (struct device *, dma_addr_t, size_t, int); -typedef void ia64_mv_dma_sync_sg_for_device (struct device *, struct scatterlist *, int, int); -typedef int ia64_mv_dma_mapping_error(struct device *, dma_addr_t dma_addr); -typedef int ia64_mv_dma_supported (struct device *, u64); - -typedef dma_addr_t ia64_mv_dma_map_single_attrs (struct device *, void *, size_t, int, struct dma_attrs *); -typedef void ia64_mv_dma_unmap_single_attrs (struct device *, dma_addr_t, size_t, int, struct dma_attrs *); -typedef int ia64_mv_dma_map_sg_attrs (struct device *, struct scatterlist *, int, int, struct dma_attrs *); -typedef void ia64_mv_dma_unmap_sg_attrs (struct device *, struct scatterlist *, int, int, struct dma_attrs *); typedef u64 ia64_mv_dma_get_required_mask (struct device *); +typedef struct dma_map_ops *ia64_mv_dma_get_ops(struct device *); /* * WARNING: The legacy I/O space is _architected_. Platforms are @@ -114,8 +97,6 @@ machvec_noop_bus (struct pci_bus *bus) extern void machvec_setup (char **); extern void machvec_timer_interrupt (int, void *); -extern void machvec_dma_sync_single (struct device *, dma_addr_t, size_t, int); -extern void machvec_dma_sync_sg (struct device *, struct scatterlist *, int, int); extern void machvec_tlb_migrate_finish (struct mm_struct *); # if defined (CONFIG_IA64_HP_SIM) @@ -148,19 +129,8 @@ extern void machvec_tlb_migrate_finish (struct mm_struct *); # define platform_global_tlb_purge ia64_mv.global_tlb_purge # define platform_tlb_migrate_finish ia64_mv.tlb_migrate_finish # define platform_dma_init ia64_mv.dma_init -# define platform_dma_alloc_coherent ia64_mv.dma_alloc_coherent -# define platform_dma_free_coherent ia64_mv.dma_free_coherent -# define platform_dma_map_single_attrs ia64_mv.dma_map_single_attrs -# define platform_dma_unmap_single_attrs ia64_mv.dma_unmap_single_attrs -# define platform_dma_map_sg_attrs ia64_mv.dma_map_sg_attrs -# define platform_dma_unmap_sg_attrs ia64_mv.dma_unmap_sg_attrs -# define platform_dma_sync_single_for_cpu ia64_mv.dma_sync_single_for_cpu -# define platform_dma_sync_sg_for_cpu ia64_mv.dma_sync_sg_for_cpu -# define platform_dma_sync_single_for_device ia64_mv.dma_sync_single_for_device -# define platform_dma_sync_sg_for_device ia64_mv.dma_sync_sg_for_device -# define platform_dma_mapping_error ia64_mv.dma_mapping_error -# define platform_dma_supported ia64_mv.dma_supported # define platform_dma_get_required_mask ia64_mv.dma_get_required_mask +# define platform_dma_get_ops ia64_mv.dma_get_ops # define platform_irq_to_vector ia64_mv.irq_to_vector # define platform_local_vector_to_irq ia64_mv.local_vector_to_irq # define platform_pci_get_legacy_mem ia64_mv.pci_get_legacy_mem @@ -203,19 +173,8 @@ struct ia64_machine_vector { ia64_mv_global_tlb_purge_t *global_tlb_purge; ia64_mv_tlb_migrate_finish_t *tlb_migrate_finish; ia64_mv_dma_init *dma_init; - ia64_mv_dma_alloc_coherent *dma_alloc_coherent; - ia64_mv_dma_free_coherent *dma_free_coherent; - ia64_mv_dma_map_single_attrs *dma_map_single_attrs; - ia64_mv_dma_unmap_single_attrs *dma_unmap_single_attrs; - ia64_mv_dma_map_sg_attrs *dma_map_sg_attrs; - ia64_mv_dma_unmap_sg_attrs *dma_unmap_sg_attrs; - ia64_mv_dma_sync_single_for_cpu *dma_sync_single_for_cpu; - ia64_mv_dma_sync_sg_for_cpu *dma_sync_sg_for_cpu; - ia64_mv_dma_sync_single_for_device *dma_sync_single_for_device; - ia64_mv_dma_sync_sg_for_device *dma_sync_sg_for_device; - ia64_mv_dma_mapping_error *dma_mapping_error; - ia64_mv_dma_supported *dma_supported; ia64_mv_dma_get_required_mask *dma_get_required_mask; + ia64_mv_dma_get_ops *dma_get_ops; ia64_mv_irq_to_vector *irq_to_vector; ia64_mv_local_vector_to_irq *local_vector_to_irq; ia64_mv_pci_get_legacy_mem_t *pci_get_legacy_mem; @@ -254,19 +213,8 @@ struct ia64_machine_vector { platform_global_tlb_purge, \ platform_tlb_migrate_finish, \ platform_dma_init, \ - platform_dma_alloc_coherent, \ - platform_dma_free_coherent, \ - platform_dma_map_single_attrs, \ - platform_dma_unmap_single_attrs, \ - platform_dma_map_sg_attrs, \ - platform_dma_unmap_sg_attrs, \ - platform_dma_sync_single_for_cpu, \ - platform_dma_sync_sg_for_cpu, \ - platform_dma_sync_single_for_device, \ - platform_dma_sync_sg_for_device, \ - platform_dma_mapping_error, \ - platform_dma_supported, \ platform_dma_get_required_mask, \ + platform_dma_get_ops, \ platform_irq_to_vector, \ platform_local_vector_to_irq, \ platform_pci_get_legacy_mem, \ @@ -302,6 +250,9 @@ extern void machvec_init_from_cmdline(const char *cmdline); # error Unknown configuration. Update arch/ia64/include/asm/machvec.h. # endif /* CONFIG_IA64_GENERIC */ +extern void swiotlb_dma_init(void); +extern struct dma_map_ops *dma_get_ops(struct device *); + /* * Define default versions so we can extend machvec for new platforms without having * to update the machvec files for all existing platforms. @@ -332,43 +283,10 @@ extern void machvec_init_from_cmdline(const char *cmdline); # define platform_kernel_launch_event machvec_noop #endif #ifndef platform_dma_init -# define platform_dma_init swiotlb_init -#endif -#ifndef platform_dma_alloc_coherent -# define platform_dma_alloc_coherent swiotlb_alloc_coherent -#endif -#ifndef platform_dma_free_coherent -# define platform_dma_free_coherent swiotlb_free_coherent -#endif -#ifndef platform_dma_map_single_attrs -# define platform_dma_map_single_attrs swiotlb_map_single_attrs -#endif -#ifndef platform_dma_unmap_single_attrs -# define platform_dma_unmap_single_attrs swiotlb_unmap_single_attrs -#endif -#ifndef platform_dma_map_sg_attrs -# define platform_dma_map_sg_attrs swiotlb_map_sg_attrs -#endif -#ifndef platform_dma_unmap_sg_attrs -# define platform_dma_unmap_sg_attrs swiotlb_unmap_sg_attrs -#endif -#ifndef platform_dma_sync_single_for_cpu -# define platform_dma_sync_single_for_cpu swiotlb_sync_single_for_cpu -#endif -#ifndef platform_dma_sync_sg_for_cpu -# define platform_dma_sync_sg_for_cpu swiotlb_sync_sg_for_cpu -#endif -#ifndef platform_dma_sync_single_for_device -# define platform_dma_sync_single_for_device swiotlb_sync_single_for_device -#endif -#ifndef platform_dma_sync_sg_for_device -# define platform_dma_sync_sg_for_device swiotlb_sync_sg_for_device -#endif -#ifndef platform_dma_mapping_error -# define platform_dma_mapping_error swiotlb_dma_mapping_error +# define platform_dma_init swiotlb_dma_init #endif -#ifndef platform_dma_supported -# define platform_dma_supported swiotlb_dma_supported +#ifndef platform_dma_get_ops +# define platform_dma_get_ops dma_get_ops #endif #ifndef platform_dma_get_required_mask # define platform_dma_get_required_mask ia64_dma_get_required_mask diff --git a/arch/ia64/include/asm/machvec_dig_vtd.h b/arch/ia64/include/asm/machvec_dig_vtd.h index 3400b561e71..6ab1de5c45e 100644 --- a/arch/ia64/include/asm/machvec_dig_vtd.h +++ b/arch/ia64/include/asm/machvec_dig_vtd.h @@ -2,14 +2,6 @@ #define _ASM_IA64_MACHVEC_DIG_VTD_h extern ia64_mv_setup_t dig_setup; -extern ia64_mv_dma_alloc_coherent vtd_alloc_coherent; -extern ia64_mv_dma_free_coherent vtd_free_coherent; -extern ia64_mv_dma_map_single_attrs vtd_map_single_attrs; -extern ia64_mv_dma_unmap_single_attrs vtd_unmap_single_attrs; -extern ia64_mv_dma_map_sg_attrs vtd_map_sg_attrs; -extern ia64_mv_dma_unmap_sg_attrs vtd_unmap_sg_attrs; -extern ia64_mv_dma_supported iommu_dma_supported; -extern ia64_mv_dma_mapping_error vtd_dma_mapping_error; extern ia64_mv_dma_init pci_iommu_alloc; /* @@ -22,17 +14,5 @@ extern ia64_mv_dma_init pci_iommu_alloc; #define platform_name "dig_vtd" #define platform_setup dig_setup #define platform_dma_init pci_iommu_alloc -#define platform_dma_alloc_coherent vtd_alloc_coherent -#define platform_dma_free_coherent vtd_free_coherent -#define platform_dma_map_single_attrs vtd_map_single_attrs -#define platform_dma_unmap_single_attrs vtd_unmap_single_attrs -#define platform_dma_map_sg_attrs vtd_map_sg_attrs -#define platform_dma_unmap_sg_attrs vtd_unmap_sg_attrs -#define platform_dma_sync_single_for_cpu machvec_dma_sync_single -#define platform_dma_sync_sg_for_cpu machvec_dma_sync_sg -#define platform_dma_sync_single_for_device machvec_dma_sync_single -#define platform_dma_sync_sg_for_device machvec_dma_sync_sg -#define platform_dma_supported iommu_dma_supported -#define platform_dma_mapping_error vtd_dma_mapping_error #endif /* _ASM_IA64_MACHVEC_DIG_VTD_h */ diff --git a/arch/ia64/include/asm/machvec_hpzx1.h b/arch/ia64/include/asm/machvec_hpzx1.h index 2f57f5144b9..3bd83d78a41 100644 --- a/arch/ia64/include/asm/machvec_hpzx1.h +++ b/arch/ia64/include/asm/machvec_hpzx1.h @@ -2,14 +2,7 @@ #define _ASM_IA64_MACHVEC_HPZX1_h extern ia64_mv_setup_t dig_setup; -extern ia64_mv_dma_alloc_coherent sba_alloc_coherent; -extern ia64_mv_dma_free_coherent sba_free_coherent; -extern ia64_mv_dma_map_single_attrs sba_map_single_attrs; -extern ia64_mv_dma_unmap_single_attrs sba_unmap_single_attrs; -extern ia64_mv_dma_map_sg_attrs sba_map_sg_attrs; -extern ia64_mv_dma_unmap_sg_attrs sba_unmap_sg_attrs; -extern ia64_mv_dma_supported sba_dma_supported; -extern ia64_mv_dma_mapping_error sba_dma_mapping_error; +extern ia64_mv_dma_init sba_dma_init; /* * This stuff has dual use! @@ -20,18 +13,6 @@ extern ia64_mv_dma_mapping_error sba_dma_mapping_error; */ #define platform_name "hpzx1" #define platform_setup dig_setup -#define platform_dma_init machvec_noop -#define platform_dma_alloc_coherent sba_alloc_coherent -#define platform_dma_free_coherent sba_free_coherent -#define platform_dma_map_single_attrs sba_map_single_attrs -#define platform_dma_unmap_single_attrs sba_unmap_single_attrs -#define platform_dma_map_sg_attrs sba_map_sg_attrs -#define platform_dma_unmap_sg_attrs sba_unmap_sg_attrs -#define platform_dma_sync_single_for_cpu machvec_dma_sync_single -#define platform_dma_sync_sg_for_cpu machvec_dma_sync_sg -#define platform_dma_sync_single_for_device machvec_dma_sync_single -#define platform_dma_sync_sg_for_device machvec_dma_sync_sg -#define platform_dma_supported sba_dma_supported -#define platform_dma_mapping_error sba_dma_mapping_error +#define platform_dma_init sba_dma_init #endif /* _ASM_IA64_MACHVEC_HPZX1_h */ diff --git a/arch/ia64/include/asm/machvec_hpzx1_swiotlb.h b/arch/ia64/include/asm/machvec_hpzx1_swiotlb.h index a842cdda827..1091ac39740 100644 --- a/arch/ia64/include/asm/machvec_hpzx1_swiotlb.h +++ b/arch/ia64/include/asm/machvec_hpzx1_swiotlb.h @@ -2,18 +2,7 @@ #define _ASM_IA64_MACHVEC_HPZX1_SWIOTLB_h extern ia64_mv_setup_t dig_setup; -extern ia64_mv_dma_alloc_coherent hwsw_alloc_coherent; -extern ia64_mv_dma_free_coherent hwsw_free_coherent; -extern ia64_mv_dma_map_single_attrs hwsw_map_single_attrs; -extern ia64_mv_dma_unmap_single_attrs hwsw_unmap_single_attrs; -extern ia64_mv_dma_map_sg_attrs hwsw_map_sg_attrs; -extern ia64_mv_dma_unmap_sg_attrs hwsw_unmap_sg_attrs; -extern ia64_mv_dma_supported hwsw_dma_supported; -extern ia64_mv_dma_mapping_error hwsw_dma_mapping_error; -extern ia64_mv_dma_sync_single_for_cpu hwsw_sync_single_for_cpu; -extern ia64_mv_dma_sync_sg_for_cpu hwsw_sync_sg_for_cpu; -extern ia64_mv_dma_sync_single_for_device hwsw_sync_single_for_device; -extern ia64_mv_dma_sync_sg_for_device hwsw_sync_sg_for_device; +extern ia64_mv_dma_get_ops hwsw_dma_get_ops; /* * This stuff has dual use! @@ -23,20 +12,8 @@ extern ia64_mv_dma_sync_sg_for_device hwsw_sync_sg_for_device; * the macros are used directly. */ #define platform_name "hpzx1_swiotlb" - #define platform_setup dig_setup #define platform_dma_init machvec_noop -#define platform_dma_alloc_coherent hwsw_alloc_coherent -#define platform_dma_free_coherent hwsw_free_coherent -#define platform_dma_map_single_attrs hwsw_map_single_attrs -#define platform_dma_unmap_single_attrs hwsw_unmap_single_attrs -#define platform_dma_map_sg_attrs hwsw_map_sg_attrs -#define platform_dma_unmap_sg_attrs hwsw_unmap_sg_attrs -#define platform_dma_supported hwsw_dma_supported -#define platform_dma_mapping_error hwsw_dma_mapping_error -#define platform_dma_sync_single_for_cpu hwsw_sync_single_for_cpu -#define platform_dma_sync_sg_for_cpu hwsw_sync_sg_for_cpu -#define platform_dma_sync_single_for_device hwsw_sync_single_for_device -#define platform_dma_sync_sg_for_device hwsw_sync_sg_for_device +#define platform_dma_get_ops hwsw_dma_get_ops #endif /* _ASM_IA64_MACHVEC_HPZX1_SWIOTLB_h */ diff --git a/arch/ia64/include/asm/machvec_sn2.h b/arch/ia64/include/asm/machvec_sn2.h index f1a6e0d6dfa..f061a30aac4 100644 --- a/arch/ia64/include/asm/machvec_sn2.h +++ b/arch/ia64/include/asm/machvec_sn2.h @@ -55,19 +55,8 @@ extern ia64_mv_readb_t __sn_readb_relaxed; extern ia64_mv_readw_t __sn_readw_relaxed; extern ia64_mv_readl_t __sn_readl_relaxed; extern ia64_mv_readq_t __sn_readq_relaxed; -extern ia64_mv_dma_alloc_coherent sn_dma_alloc_coherent; -extern ia64_mv_dma_free_coherent sn_dma_free_coherent; -extern ia64_mv_dma_map_single_attrs sn_dma_map_single_attrs; -extern ia64_mv_dma_unmap_single_attrs sn_dma_unmap_single_attrs; -extern ia64_mv_dma_map_sg_attrs sn_dma_map_sg_attrs; -extern ia64_mv_dma_unmap_sg_attrs sn_dma_unmap_sg_attrs; -extern ia64_mv_dma_sync_single_for_cpu sn_dma_sync_single_for_cpu; -extern ia64_mv_dma_sync_sg_for_cpu sn_dma_sync_sg_for_cpu; -extern ia64_mv_dma_sync_single_for_device sn_dma_sync_single_for_device; -extern ia64_mv_dma_sync_sg_for_device sn_dma_sync_sg_for_device; -extern ia64_mv_dma_mapping_error sn_dma_mapping_error; -extern ia64_mv_dma_supported sn_dma_supported; extern ia64_mv_dma_get_required_mask sn_dma_get_required_mask; +extern ia64_mv_dma_init sn_dma_init; extern ia64_mv_migrate_t sn_migrate; extern ia64_mv_kernel_launch_event_t sn_kernel_launch_event; extern ia64_mv_setup_msi_irq_t sn_setup_msi_irq; @@ -111,20 +100,8 @@ extern ia64_mv_pci_fixup_bus_t sn_pci_fixup_bus; #define platform_pci_get_legacy_mem sn_pci_get_legacy_mem #define platform_pci_legacy_read sn_pci_legacy_read #define platform_pci_legacy_write sn_pci_legacy_write -#define platform_dma_init machvec_noop -#define platform_dma_alloc_coherent sn_dma_alloc_coherent -#define platform_dma_free_coherent sn_dma_free_coherent -#define platform_dma_map_single_attrs sn_dma_map_single_attrs -#define platform_dma_unmap_single_attrs sn_dma_unmap_single_attrs -#define platform_dma_map_sg_attrs sn_dma_map_sg_attrs -#define platform_dma_unmap_sg_attrs sn_dma_unmap_sg_attrs -#define platform_dma_sync_single_for_cpu sn_dma_sync_single_for_cpu -#define platform_dma_sync_sg_for_cpu sn_dma_sync_sg_for_cpu -#define platform_dma_sync_single_for_device sn_dma_sync_single_for_device -#define platform_dma_sync_sg_for_device sn_dma_sync_sg_for_device -#define platform_dma_mapping_error sn_dma_mapping_error -#define platform_dma_supported sn_dma_supported #define platform_dma_get_required_mask sn_dma_get_required_mask +#define platform_dma_init sn_dma_init #define platform_migrate sn_migrate #define platform_kernel_launch_event sn_kernel_launch_event #ifdef CONFIG_PCI_MSI diff --git a/arch/ia64/kernel/Makefile b/arch/ia64/kernel/Makefile index c381ea95489..f2778f2c4fd 100644 --- a/arch/ia64/kernel/Makefile +++ b/arch/ia64/kernel/Makefile @@ -7,7 +7,7 @@ extra-y := head.o init_task.o vmlinux.lds obj-y := acpi.o entry.o efi.o efi_stub.o gate-data.o fsys.o ia64_ksyms.o irq.o irq_ia64.o \ irq_lsapic.o ivt.o machvec.o pal.o patch.o process.o perfmon.o ptrace.o sal.o \ salinfo.o setup.o signal.o sys_ia64.o time.o traps.o unaligned.o \ - unwind.o mca.o mca_asm.o topology.o + unwind.o mca.o mca_asm.o topology.o dma-mapping.o obj-$(CONFIG_IA64_BRL_EMU) += brl_emu.o obj-$(CONFIG_IA64_GENERIC) += acpi-ext.o @@ -43,9 +43,7 @@ ifneq ($(CONFIG_IA64_ESI),) obj-y += esi_stub.o # must be in kernel proper endif obj-$(CONFIG_DMAR) += pci-dma.o -ifeq ($(CONFIG_DMAR), y) obj-$(CONFIG_SWIOTLB) += pci-swiotlb.o -endif # The gate DSO image is built using a special linker script. targets += gate.so gate-syms.o diff --git a/arch/ia64/kernel/dma-mapping.c b/arch/ia64/kernel/dma-mapping.c new file mode 100644 index 00000000000..086a2aeb040 --- /dev/null +++ b/arch/ia64/kernel/dma-mapping.c @@ -0,0 +1,13 @@ +#include <linux/dma-mapping.h> + +/* Set this to 1 if there is a HW IOMMU in the system */ +int iommu_detected __read_mostly; + +struct dma_map_ops *dma_ops; +EXPORT_SYMBOL(dma_ops); + +struct dma_map_ops *dma_get_ops(struct device *dev) +{ + return dma_ops; +} +EXPORT_SYMBOL(dma_get_ops); diff --git a/arch/ia64/kernel/machvec.c b/arch/ia64/kernel/machvec.c index 7ccb228ceed..d41a40ef80c 100644 --- a/arch/ia64/kernel/machvec.c +++ b/arch/ia64/kernel/machvec.c @@ -1,5 +1,5 @@ #include <linux/module.h> - +#include <linux/dma-mapping.h> #include <asm/machvec.h> #include <asm/system.h> @@ -75,14 +75,16 @@ machvec_timer_interrupt (int irq, void *dev_id) EXPORT_SYMBOL(machvec_timer_interrupt); void -machvec_dma_sync_single (struct device *hwdev, dma_addr_t dma_handle, size_t size, int dir) +machvec_dma_sync_single(struct device *hwdev, dma_addr_t dma_handle, size_t size, + enum dma_data_direction dir) { mb(); } EXPORT_SYMBOL(machvec_dma_sync_single); void -machvec_dma_sync_sg (struct device *hwdev, struct scatterlist *sg, int n, int dir) +machvec_dma_sync_sg(struct device *hwdev, struct scatterlist *sg, int n, + enum dma_data_direction dir) { mb(); } diff --git a/arch/ia64/kernel/pci-dma.c b/arch/ia64/kernel/pci-dma.c index d0ada067a4a..e4cb443bb98 100644 --- a/arch/ia64/kernel/pci-dma.c +++ b/arch/ia64/kernel/pci-dma.c @@ -32,9 +32,6 @@ int force_iommu __read_mostly = 1; int force_iommu __read_mostly; #endif -/* Set this to 1 if there is a HW IOMMU in the system */ -int iommu_detected __read_mostly; - /* Dummy device used for NULL arguments (normally ISA). Better would be probably a smaller DMA mask, but this is bug-to-bug compatible to i386. */ @@ -44,18 +41,7 @@ struct device fallback_dev = { .dma_mask = &fallback_dev.coherent_dma_mask, }; -void __init pci_iommu_alloc(void) -{ - /* - * The order of these functions is important for - * fall-back/fail-over reasons - */ - detect_intel_iommu(); - -#ifdef CONFIG_SWIOTLB - pci_swiotlb_init(); -#endif -} +extern struct dma_map_ops intel_dma_ops; static int __init pci_iommu_init(void) { @@ -79,15 +65,12 @@ iommu_dma_init(void) return; } -struct dma_mapping_ops *dma_ops; -EXPORT_SYMBOL(dma_ops); - int iommu_dma_supported(struct device *dev, u64 mask) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = platform_dma_get_ops(dev); - if (ops->dma_supported_op) - return ops->dma_supported_op(dev, mask); + if (ops->dma_supported) + return ops->dma_supported(dev, mask); /* Copied from i386. Doesn't make much sense, because it will only work for pci_alloc_coherent. @@ -116,4 +99,25 @@ int iommu_dma_supported(struct device *dev, u64 mask) } EXPORT_SYMBOL(iommu_dma_supported); +void __init pci_iommu_alloc(void) +{ + dma_ops = &intel_dma_ops; + + dma_ops->sync_single_for_cpu = machvec_dma_sync_single; + dma_ops->sync_sg_for_cpu = machvec_dma_sync_sg; + dma_ops->sync_single_for_device = machvec_dma_sync_single; + dma_ops->sync_sg_for_device = machvec_dma_sync_sg; + dma_ops->dma_supported = iommu_dma_supported; + + /* + * The order of these functions is important for + * fall-back/fail-over reasons + */ + detect_intel_iommu(); + +#ifdef CONFIG_SWIOTLB + pci_swiotlb_init(); +#endif +} + #endif diff --git a/arch/ia64/kernel/pci-swiotlb.c b/arch/ia64/kernel/pci-swiotlb.c index 16c50516dbc..573f02c39a0 100644 --- a/arch/ia64/kernel/pci-swiotlb.c +++ b/arch/ia64/kernel/pci-swiotlb.c @@ -13,23 +13,37 @@ int swiotlb __read_mostly; EXPORT_SYMBOL(swiotlb); -struct dma_mapping_ops swiotlb_dma_ops = { - .mapping_error = swiotlb_dma_mapping_error, - .alloc_coherent = swiotlb_alloc_coherent, +static void *ia64_swiotlb_alloc_coherent(struct device *dev, size_t size, + dma_addr_t *dma_handle, gfp_t gfp) +{ + if (dev->coherent_dma_mask != DMA_64BIT_MASK) + gfp |= GFP_DMA; + return swiotlb_alloc_coherent(dev, size, dma_handle, gfp); +} + +struct dma_map_ops swiotlb_dma_ops = { + .alloc_coherent = ia64_swiotlb_alloc_coherent, .free_coherent = swiotlb_free_coherent, - .map_single = swiotlb_map_single, - .unmap_single = swiotlb_unmap_single, + .map_page = swiotlb_map_page, + .unmap_page = swiotlb_unmap_page, + .map_sg = swiotlb_map_sg_attrs, + .unmap_sg = swiotlb_unmap_sg_attrs, .sync_single_for_cpu = swiotlb_sync_single_for_cpu, .sync_single_for_device = swiotlb_sync_single_for_device, .sync_single_range_for_cpu = swiotlb_sync_single_range_for_cpu, .sync_single_range_for_device = swiotlb_sync_single_range_for_device, .sync_sg_for_cpu = swiotlb_sync_sg_for_cpu, .sync_sg_for_device = swiotlb_sync_sg_for_device, - .map_sg = swiotlb_map_sg, - .unmap_sg = swiotlb_unmap_sg, - .dma_supported_op = swiotlb_dma_supported, + .dma_supported = swiotlb_dma_supported, + .mapping_error = swiotlb_dma_mapping_error, }; +void __init swiotlb_dma_init(void) +{ + dma_ops = &swiotlb_dma_ops; + swiotlb_init(); +} + void __init pci_swiotlb_init(void) { if (!iommu_detected) { diff --git a/arch/ia64/sn/pci/pci_dma.c b/arch/ia64/sn/pci/pci_dma.c index 863f5017baa..8c130e8f00e 100644 --- a/arch/ia64/sn/pci/pci_dma.c +++ b/arch/ia64/sn/pci/pci_dma.c @@ -10,7 +10,7 @@ */ #include <linux/module.h> -#include <linux/dma-attrs.h> +#include <linux/dma-mapping.h> #include <asm/dma.h> #include <asm/sn/intr.h> #include <asm/sn/pcibus_provider_defs.h> @@ -31,7 +31,7 @@ * this function. Of course, SN only supports devices that have 32 or more * address bits when using the PMU. */ -int sn_dma_supported(struct device *dev, u64 mask) +static int sn_dma_supported(struct device *dev, u64 mask) { BUG_ON(dev->bus != &pci_bus_type); @@ -39,7 +39,6 @@ int sn_dma_supported(struct device *dev, u64 mask) return 0; return 1; } -EXPORT_SYMBOL(sn_dma_supported); /** * sn_dma_set_mask - set the DMA mask @@ -75,8 +74,8 @@ EXPORT_SYMBOL(sn_dma_set_mask); * queue for a SCSI controller). See Documentation/DMA-API.txt for * more information. */ -void *sn_dma_alloc_coherent(struct device *dev, size_t size, - dma_addr_t * dma_handle, gfp_t flags) +static void *sn_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t * dma_handle, gfp_t flags) { void *cpuaddr; unsigned long phys_addr; @@ -124,7 +123,6 @@ void *sn_dma_alloc_coherent(struct device *dev, size_t size, return cpuaddr; } -EXPORT_SYMBOL(sn_dma_alloc_coherent); /** * sn_pci_free_coherent - free memory associated with coherent DMAable region @@ -136,8 +134,8 @@ EXPORT_SYMBOL(sn_dma_alloc_coherent); * Frees the memory allocated by dma_alloc_coherent(), potentially unmapping * any associated IOMMU mappings. */ -void sn_dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, - dma_addr_t dma_handle) +static void sn_dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, + dma_addr_t dma_handle) { struct pci_dev *pdev = to_pci_dev(dev); struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); @@ -147,7 +145,6 @@ void sn_dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, provider->dma_unmap(pdev, dma_handle, 0); free_pages((unsigned long)cpu_addr, get_order(size)); } -EXPORT_SYMBOL(sn_dma_free_coherent); /** * sn_dma_map_single_attrs - map a single page for DMA @@ -173,10 +170,12 @@ EXPORT_SYMBOL(sn_dma_free_coherent); * TODO: simplify our interface; * figure out how to save dmamap handle so can use two step. */ -dma_addr_t sn_dma_map_single_attrs(struct device *dev, void *cpu_addr, - size_t size, int direction, - struct dma_attrs *attrs) +static dma_addr_t sn_dma_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { + void *cpu_addr = page_address(page) + offset; dma_addr_t dma_addr; unsigned long phys_addr; struct pci_dev *pdev = to_pci_dev(dev); @@ -201,7 +200,6 @@ dma_addr_t sn_dma_map_single_attrs(struct device *dev, void *cpu_addr, } return dma_addr; } -EXPORT_SYMBOL(sn_dma_map_single_attrs); /** * sn_dma_unmap_single_attrs - unamp a DMA mapped page @@ -215,21 +213,20 @@ EXPORT_SYMBOL(sn_dma_map_single_attrs); * by @dma_handle into the coherence domain. On SN, we're always cache * coherent, so we just need to free any ATEs associated with this mapping. */ -void sn_dma_unmap_single_attrs(struct device *dev, dma_addr_t dma_addr, - size_t size, int direction, - struct dma_attrs *attrs) +static void sn_dma_unmap_page(struct device *dev, dma_addr_t dma_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct pci_dev *pdev = to_pci_dev(dev); struct sn_pcibus_provider *provider = SN_PCIDEV_BUSPROVIDER(pdev); BUG_ON(dev->bus != &pci_bus_type); - provider->dma_unmap(pdev, dma_addr, direction); + provider->dma_unmap(pdev, dma_addr, dir); } -EXPORT_SYMBOL(sn_dma_unmap_single_attrs); /** - * sn_dma_unmap_sg_attrs - unmap a DMA scatterlist + * sn_dma_unmap_sg - unmap a DMA scatterlist * @dev: device to unmap * @sg: scatterlist to unmap * @nhwentries: number of scatterlist entries @@ -238,9 +235,9 @@ EXPORT_SYMBOL(sn_dma_unmap_single_attrs); * * Unmap a set of streaming mode DMA translations. */ -void sn_dma_unmap_sg_attrs(struct device *dev, struct scatterlist *sgl, - int nhwentries, int direction, - struct dma_attrs *attrs) +static void sn_dma_unmap_sg(struct device *dev, struct scatterlist *sgl, + int nhwentries, enum dma_data_direction dir, + struct dma_attrs *attrs) { int i; struct pci_dev *pdev = to_pci_dev(dev); @@ -250,15 +247,14 @@ void sn_dma_unmap_sg_attrs(struct device *dev, struct scatterlist *sgl, BUG_ON(dev->bus != &pci_bus_type); for_each_sg(sgl, sg, nhwentries, i) { - provider->dma_unmap(pdev, sg->dma_address, direction); + provider->dma_unmap(pdev, sg->dma_address, dir); sg->dma_address = (dma_addr_t) NULL; sg->dma_length = 0; } } -EXPORT_SYMBOL(sn_dma_unmap_sg_attrs); /** - * sn_dma_map_sg_attrs - map a scatterlist for DMA + * sn_dma_map_sg - map a scatterlist for DMA * @dev: device to map for * @sg: scatterlist to map * @nhwentries: number of entries @@ -272,8 +268,9 @@ EXPORT_SYMBOL(sn_dma_unmap_sg_attrs); * * Maps each entry of @sg for DMA. */ -int sn_dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl, - int nhwentries, int direction, struct dma_attrs *attrs) +static int sn_dma_map_sg(struct device *dev, struct scatterlist *sgl, + int nhwentries, enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long phys_addr; struct scatterlist *saved_sg = sgl, *sg; @@ -310,8 +307,7 @@ int sn_dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl, * Free any successfully allocated entries. */ if (i > 0) - sn_dma_unmap_sg_attrs(dev, saved_sg, i, - direction, attrs); + sn_dma_unmap_sg(dev, saved_sg, i, dir, attrs); return 0; } @@ -320,41 +316,36 @@ int sn_dma_map_sg_attrs(struct device *dev, struct scatterlist *sgl, return nhwentries; } -EXPORT_SYMBOL(sn_dma_map_sg_attrs); -void sn_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, - size_t size, int direction) +static void sn_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, + size_t size, enum dma_data_direction dir) { BUG_ON(dev->bus != &pci_bus_type); } -EXPORT_SYMBOL(sn_dma_sync_single_for_cpu); -void sn_dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, - size_t size, int direction) +static void sn_dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, + size_t size, + enum dma_data_direction dir) { BUG_ON(dev->bus != &pci_bus_type); } -EXPORT_SYMBOL(sn_dma_sync_single_for_device); -void sn_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, - int nelems, int direction) +static void sn_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, + int nelems, enum dma_data_direction dir) { BUG_ON(dev->bus != &pci_bus_type); } -EXPORT_SYMBOL(sn_dma_sync_sg_for_cpu); -void sn_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, - int nelems, int direction) +static void sn_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, + int nelems, enum dma_data_direction dir) { BUG_ON(dev->bus != &pci_bus_type); } -EXPORT_SYMBOL(sn_dma_sync_sg_for_device); -int sn_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) +static int sn_dma_mapping_error(struct device *dev, dma_addr_t dma_addr) { return 0; } -EXPORT_SYMBOL(sn_dma_mapping_error); u64 sn_dma_get_required_mask(struct device *dev) { @@ -471,3 +462,23 @@ int sn_pci_legacy_write(struct pci_bus *bus, u16 port, u32 val, u8 size) out: return ret; } + +static struct dma_map_ops sn_dma_ops = { + .alloc_coherent = sn_dma_alloc_coherent, + .free_coherent = sn_dma_free_coherent, + .map_page = sn_dma_map_page, + .unmap_page = sn_dma_unmap_page, + .map_sg = sn_dma_map_sg, + .unmap_sg = sn_dma_unmap_sg, + .sync_single_for_cpu = sn_dma_sync_single_for_cpu, + .sync_sg_for_cpu = sn_dma_sync_sg_for_cpu, + .sync_single_for_device = sn_dma_sync_single_for_device, + .sync_sg_for_device = sn_dma_sync_sg_for_device, + .mapping_error = sn_dma_mapping_error, + .dma_supported = sn_dma_supported, +}; + +void sn_dma_init(void) +{ + dma_ops = &sn_dma_ops; +} diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 34bc3a89228..45161b81631 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -40,6 +40,7 @@ config X86 select HAVE_GENERIC_DMA_COHERENT if X86_32 select HAVE_EFFICIENT_UNALIGNED_ACCESS select USER_STACKTRACE_SUPPORT + select HAVE_DMA_API_DEBUG select HAVE_KERNEL_GZIP select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA diff --git a/arch/x86/include/asm/device.h b/arch/x86/include/asm/device.h index 3c034f48fdb..4994a20acbc 100644 --- a/arch/x86/include/asm/device.h +++ b/arch/x86/include/asm/device.h @@ -6,7 +6,7 @@ struct dev_archdata { void *acpi_handle; #endif #ifdef CONFIG_X86_64 -struct dma_mapping_ops *dma_ops; +struct dma_map_ops *dma_ops; #endif #ifdef CONFIG_DMAR void *iommu; /* hook for IOMMU specific extension */ diff --git a/arch/x86/include/asm/dma-mapping.h b/arch/x86/include/asm/dma-mapping.h index 132a134d12f..cea7b74963e 100644 --- a/arch/x86/include/asm/dma-mapping.h +++ b/arch/x86/include/asm/dma-mapping.h @@ -7,6 +7,8 @@ */ #include <linux/scatterlist.h> +#include <linux/dma-debug.h> +#include <linux/dma-attrs.h> #include <asm/io.h> #include <asm/swiotlb.h> #include <asm-generic/dma-coherent.h> @@ -16,47 +18,9 @@ extern int iommu_merge; extern struct device x86_dma_fallback_dev; extern int panic_on_overflow; -struct dma_mapping_ops { - int (*mapping_error)(struct device *dev, - dma_addr_t dma_addr); - void* (*alloc_coherent)(struct device *dev, size_t size, - dma_addr_t *dma_handle, gfp_t gfp); - void (*free_coherent)(struct device *dev, size_t size, - void *vaddr, dma_addr_t dma_handle); - dma_addr_t (*map_single)(struct device *hwdev, phys_addr_t ptr, - size_t size, int direction); - void (*unmap_single)(struct device *dev, dma_addr_t addr, - size_t size, int direction); - void (*sync_single_for_cpu)(struct device *hwdev, - dma_addr_t dma_handle, size_t size, - int direction); - void (*sync_single_for_device)(struct device *hwdev, - dma_addr_t dma_handle, size_t size, - int direction); - void (*sync_single_range_for_cpu)(struct device *hwdev, - dma_addr_t dma_handle, unsigned long offset, - size_t size, int direction); - void (*sync_single_range_for_device)(struct device *hwdev, - dma_addr_t dma_handle, unsigned long offset, - size_t size, int direction); - void (*sync_sg_for_cpu)(struct device *hwdev, - struct scatterlist *sg, int nelems, - int direction); - void (*sync_sg_for_device)(struct device *hwdev, - struct scatterlist *sg, int nelems, - int direction); - int (*map_sg)(struct device *hwdev, struct scatterlist *sg, - int nents, int direction); - void (*unmap_sg)(struct device *hwdev, - struct scatterlist *sg, int nents, - int direction); - int (*dma_supported)(struct device *hwdev, u64 mask); - int is_phys; -}; - -extern struct dma_mapping_ops *dma_ops; - -static inline struct dma_mapping_ops *get_dma_ops(struct device *dev) +extern struct dma_map_ops *dma_ops; + +static inline struct dma_map_ops *get_dma_ops(struct device *dev) { #ifdef CONFIG_X86_32 return dma_ops; @@ -71,7 +35,7 @@ static inline struct dma_mapping_ops *get_dma_ops(struct device *dev) /* Make sure we keep the same behaviour */ static inline int dma_mapping_error(struct device *dev, dma_addr_t dma_addr) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); if (ops->mapping_error) return ops->mapping_error(dev, dma_addr); @@ -90,137 +54,167 @@ extern void *dma_generic_alloc_coherent(struct device *dev, size_t size, static inline dma_addr_t dma_map_single(struct device *hwdev, void *ptr, size_t size, - int direction) + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); - - BUG_ON(!valid_dma_direction(direction)); - return ops->map_single(hwdev, virt_to_phys(ptr), size, direction); + struct dma_map_ops *ops = get_dma_ops(hwdev); + dma_addr_t addr; + + BUG_ON(!valid_dma_direction(dir)); + addr = ops->map_page(hwdev, virt_to_page(ptr), + (unsigned long)ptr & ~PAGE_MASK, size, + dir, NULL); + debug_dma_map_page(hwdev, virt_to_page(ptr), + (unsigned long)ptr & ~PAGE_MASK, size, + dir, addr, true); + return addr; } static inline void dma_unmap_single(struct device *dev, dma_addr_t addr, size_t size, - int direction) + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); - BUG_ON(!valid_dma_direction(direction)); - if (ops->unmap_single) - ops->unmap_single(dev, addr, size, direction); + BUG_ON(!valid_dma_direction(dir)); + if (ops->unmap_page) + ops->unmap_page(dev, addr, size, dir, NULL); + debug_dma_unmap_page(dev, addr, size, dir, true); } static inline int dma_map_sg(struct device *hwdev, struct scatterlist *sg, - int nents, int direction) + int nents, enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); + int ents; + + BUG_ON(!valid_dma_direction(dir)); + ents = ops->map_sg(hwdev, sg, nents, dir, NULL); + debug_dma_map_sg(hwdev, sg, nents, ents, dir); - BUG_ON(!valid_dma_direction(direction)); - return ops->map_sg(hwdev, sg, nents, direction); + return ents; } static inline void dma_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents, - int direction) + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); + debug_dma_unmap_sg(hwdev, sg, nents, dir); if (ops->unmap_sg) - ops->unmap_sg(hwdev, sg, nents, direction); + ops->unmap_sg(hwdev, sg, nents, dir, NULL); } static inline void dma_sync_single_for_cpu(struct device *hwdev, dma_addr_t dma_handle, - size_t size, int direction) + size_t size, enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_single_for_cpu) - ops->sync_single_for_cpu(hwdev, dma_handle, size, direction); + ops->sync_single_for_cpu(hwdev, dma_handle, size, dir); + debug_dma_sync_single_for_cpu(hwdev, dma_handle, size, dir); flush_write_buffers(); } static inline void dma_sync_single_for_device(struct device *hwdev, dma_addr_t dma_handle, - size_t size, int direction) + size_t size, enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_single_for_device) - ops->sync_single_for_device(hwdev, dma_handle, size, direction); + ops->sync_single_for_device(hwdev, dma_handle, size, dir); + debug_dma_sync_single_for_device(hwdev, dma_handle, size, dir); flush_write_buffers(); } static inline void dma_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dma_handle, - unsigned long offset, size_t size, int direction) + unsigned long offset, size_t size, + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_single_range_for_cpu) ops->sync_single_range_for_cpu(hwdev, dma_handle, offset, - size, direction); + size, dir); + debug_dma_sync_single_range_for_cpu(hwdev, dma_handle, + offset, size, dir); flush_write_buffers(); } static inline void dma_sync_single_range_for_device(struct device *hwdev, dma_addr_t dma_handle, unsigned long offset, size_t size, - int direction) + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_single_range_for_device) ops->sync_single_range_for_device(hwdev, dma_handle, - offset, size, direction); + offset, size, dir); + debug_dma_sync_single_range_for_device(hwdev, dma_handle, + offset, size, dir); flush_write_buffers(); } static inline void dma_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg, - int nelems, int direction) + int nelems, enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_sg_for_cpu) - ops->sync_sg_for_cpu(hwdev, sg, nelems, direction); + ops->sync_sg_for_cpu(hwdev, sg, nelems, dir); + debug_dma_sync_sg_for_cpu(hwdev, sg, nelems, dir); flush_write_buffers(); } static inline void dma_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg, - int nelems, int direction) + int nelems, enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(hwdev); + struct dma_map_ops *ops = get_dma_ops(hwdev); - BUG_ON(!valid_dma_direction(direction)); + BUG_ON(!valid_dma_direction(dir)); if (ops->sync_sg_for_device) - ops->sync_sg_for_device(hwdev, sg, nelems, direction); + ops->sync_sg_for_device(hwdev, sg, nelems, dir); + debug_dma_sync_sg_for_device(hwdev, sg, nelems, dir); flush_write_buffers(); } static inline dma_addr_t dma_map_page(struct device *dev, struct page *page, size_t offset, size_t size, - int direction) + enum dma_data_direction dir) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); + dma_addr_t addr; - BUG_ON(!valid_dma_direction(direction)); - return ops->map_single(dev, page_to_phys(page) + offset, - size, direction); + BUG_ON(!valid_dma_direction(dir)); + addr = ops->map_page(dev, page, offset, size, dir, NULL); + debug_dma_map_page(dev, page, offset, size, dir, addr, false); + + return addr; } static inline void dma_unmap_page(struct device *dev, dma_addr_t addr, - size_t size, int direction) + size_t size, enum dma_data_direction dir) { - dma_unmap_single(dev, addr, size, direction); + struct dma_map_ops *ops = get_dma_ops(dev); + + BUG_ON(!valid_dma_direction(dir)); + if (ops->unmap_page) + ops->unmap_page(dev, addr, size, dir, NULL); + debug_dma_unmap_page(dev, addr, size, dir, false); } static inline void @@ -266,7 +260,7 @@ static inline void * dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, gfp_t gfp) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); void *memory; gfp &= ~(__GFP_DMA | __GFP_HIGHMEM | __GFP_DMA32); @@ -285,20 +279,24 @@ dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_handle, if (!ops->alloc_coherent) return NULL; - return ops->alloc_coherent(dev, size, dma_handle, - dma_alloc_coherent_gfp_flags(dev, gfp)); + memory = ops->alloc_coherent(dev, size, dma_handle, + dma_alloc_coherent_gfp_flags(dev, gfp)); + debug_dma_alloc_coherent(dev, size, *dma_handle, memory); + + return memory; } static inline void dma_free_coherent(struct device *dev, size_t size, void *vaddr, dma_addr_t bus) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); WARN_ON(irqs_disabled()); /* for portability */ if (dma_release_from_coherent(dev, get_order(size), vaddr)) return; + debug_dma_free_coherent(dev, size, vaddr, bus); if (ops->free_coherent) ops->free_coherent(dev, size, vaddr, bus); } diff --git a/arch/x86/include/asm/iommu.h b/arch/x86/include/asm/iommu.h index a6ee9e6f530..af326a2975b 100644 --- a/arch/x86/include/asm/iommu.h +++ b/arch/x86/include/asm/iommu.h @@ -3,7 +3,7 @@ extern void pci_iommu_shutdown(void); extern void no_iommu_init(void); -extern struct dma_mapping_ops nommu_dma_ops; +extern struct dma_map_ops nommu_dma_ops; extern int force_iommu, no_iommu; extern int iommu_detected; diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 6e9c1f320ac..c611ad64137 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -105,7 +105,7 @@ obj-$(CONFIG_MICROCODE) += microcode.o obj-$(CONFIG_X86_CHECK_BIOS_CORRUPTION) += check.o -obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o # NB rename without _64 +obj-$(CONFIG_SWIOTLB) += pci-swiotlb.o ### # 64 bit specific files diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 5113c080f0c..c5962fe3796 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -22,10 +22,9 @@ #include <linux/bitops.h> #include <linux/debugfs.h> #include <linux/scatterlist.h> +#include <linux/dma-mapping.h> #include <linux/iommu-helper.h> -#ifdef CONFIG_IOMMU_API #include <linux/iommu.h> -#endif #include <asm/proto.h> #include <asm/iommu.h> #include <asm/gart.h> @@ -1297,8 +1296,10 @@ static void __unmap_single(struct amd_iommu *iommu, /* * The exported map_single function for dma_ops. */ -static dma_addr_t map_single(struct device *dev, phys_addr_t paddr, - size_t size, int dir) +static dma_addr_t map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long flags; struct amd_iommu *iommu; @@ -1306,6 +1307,7 @@ static dma_addr_t map_single(struct device *dev, phys_addr_t paddr, u16 devid; dma_addr_t addr; u64 dma_mask; + phys_addr_t paddr = page_to_phys(page) + offset; INC_STATS_COUNTER(cnt_map_single); @@ -1340,8 +1342,8 @@ out: /* * The exported unmap_single function for dma_ops. */ -static void unmap_single(struct device *dev, dma_addr_t dma_addr, - size_t size, int dir) +static void unmap_page(struct device *dev, dma_addr_t dma_addr, size_t size, + enum dma_data_direction dir, struct dma_attrs *attrs) { unsigned long flags; struct amd_iommu *iommu; @@ -1390,7 +1392,8 @@ static int map_sg_no_iommu(struct device *dev, struct scatterlist *sglist, * lists). */ static int map_sg(struct device *dev, struct scatterlist *sglist, - int nelems, int dir) + int nelems, enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long flags; struct amd_iommu *iommu; @@ -1457,7 +1460,8 @@ unmap: * lists). */ static void unmap_sg(struct device *dev, struct scatterlist *sglist, - int nelems, int dir) + int nelems, enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long flags; struct amd_iommu *iommu; @@ -1644,11 +1648,11 @@ static void prealloc_protection_domains(void) } } -static struct dma_mapping_ops amd_iommu_dma_ops = { +static struct dma_map_ops amd_iommu_dma_ops = { .alloc_coherent = alloc_coherent, .free_coherent = free_coherent, - .map_single = map_single, - .unmap_single = unmap_single, + .map_page = map_page, + .unmap_page = unmap_page, .map_sg = map_sg, .unmap_sg = unmap_sg, .dma_supported = amd_iommu_dma_supported, diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index d28bbdc35e4..755c21e906f 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -380,8 +380,9 @@ static inline struct iommu_table *find_iommu_table(struct device *dev) return tbl; } -static void calgary_unmap_sg(struct device *dev, - struct scatterlist *sglist, int nelems, int direction) +static void calgary_unmap_sg(struct device *dev, struct scatterlist *sglist, + int nelems,enum dma_data_direction dir, + struct dma_attrs *attrs) { struct iommu_table *tbl = find_iommu_table(dev); struct scatterlist *s; @@ -404,7 +405,8 @@ static void calgary_unmap_sg(struct device *dev, } static int calgary_map_sg(struct device *dev, struct scatterlist *sg, - int nelems, int direction) + int nelems, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct iommu_table *tbl = find_iommu_table(dev); struct scatterlist *s; @@ -429,15 +431,14 @@ static int calgary_map_sg(struct device *dev, struct scatterlist *sg, s->dma_address = (entry << PAGE_SHIFT) | s->offset; /* insert into HW table */ - tce_build(tbl, entry, npages, vaddr & PAGE_MASK, - direction); + tce_build(tbl, entry, npages, vaddr & PAGE_MASK, dir); s->dma_length = s->length; } return nelems; error: - calgary_unmap_sg(dev, sg, nelems, direction); + calgary_unmap_sg(dev, sg, nelems, dir, NULL); for_each_sg(sg, s, nelems, i) { sg->dma_address = bad_dma_address; sg->dma_length = 0; @@ -445,10 +446,12 @@ error: return 0; } -static dma_addr_t calgary_map_single(struct device *dev, phys_addr_t paddr, - size_t size, int direction) +static dma_addr_t calgary_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { - void *vaddr = phys_to_virt(paddr); + void *vaddr = page_address(page) + offset; unsigned long uaddr; unsigned int npages; struct iommu_table *tbl = find_iommu_table(dev); @@ -456,17 +459,18 @@ static dma_addr_t calgary_map_single(struct device *dev, phys_addr_t paddr, uaddr = (unsigned long)vaddr; npages = iommu_num_pages(uaddr, size, PAGE_SIZE); - return iommu_alloc(dev, tbl, vaddr, npages, direction); + return iommu_alloc(dev, tbl, vaddr, npages, dir); } -static void calgary_unmap_single(struct device *dev, dma_addr_t dma_handle, - size_t size, int direction) +static void calgary_unmap_page(struct device *dev, dma_addr_t dma_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct iommu_table *tbl = find_iommu_table(dev); unsigned int npages; - npages = iommu_num_pages(dma_handle, size, PAGE_SIZE); - iommu_free(tbl, dma_handle, npages); + npages = iommu_num_pages(dma_addr, size, PAGE_SIZE); + iommu_free(tbl, dma_addr, npages); } static void* calgary_alloc_coherent(struct device *dev, size_t size, @@ -515,13 +519,13 @@ static void calgary_free_coherent(struct device *dev, size_t size, free_pages((unsigned long)vaddr, get_order(size)); } -static struct dma_mapping_ops calgary_dma_ops = { +static struct dma_map_ops calgary_dma_ops = { .alloc_coherent = calgary_alloc_coherent, .free_coherent = calgary_free_coherent, - .map_single = calgary_map_single, - .unmap_single = calgary_unmap_single, .map_sg = calgary_map_sg, .unmap_sg = calgary_unmap_sg, + .map_page = calgary_map_page, + .unmap_page = calgary_unmap_page, }; static inline void __iomem * busno_to_bbar(unsigned char num) diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index b2542853314..c7c4776ff63 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -1,4 +1,5 @@ #include <linux/dma-mapping.h> +#include <linux/dma-debug.h> #include <linux/dmar.h> #include <linux/bootmem.h> #include <linux/pci.h> @@ -12,7 +13,7 @@ static int forbid_dac __read_mostly; -struct dma_mapping_ops *dma_ops; +struct dma_map_ops *dma_ops; EXPORT_SYMBOL(dma_ops); static int iommu_sac_force __read_mostly; @@ -44,6 +45,9 @@ struct device x86_dma_fallback_dev = { }; EXPORT_SYMBOL(x86_dma_fallback_dev); +/* Number of entries preallocated for DMA-API debugging */ +#define PREALLOC_DMA_DEBUG_ENTRIES 32768 + int dma_set_mask(struct device *dev, u64 mask) { if (!dev->dma_mask || !dma_supported(dev, mask)) @@ -224,7 +228,7 @@ early_param("iommu", iommu_setup); int dma_supported(struct device *dev, u64 mask) { - struct dma_mapping_ops *ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); #ifdef CONFIG_PCI if (mask > 0xffffffff && forbid_dac > 0) { @@ -265,6 +269,12 @@ EXPORT_SYMBOL(dma_supported); static int __init pci_iommu_init(void) { + dma_debug_init(PREALLOC_DMA_DEBUG_ENTRIES); + +#ifdef CONFIG_PCI + dma_debug_add_bus(&pci_bus_type); +#endif + calgary_iommu_init(); intel_iommu_init(); diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index d5768b1af08..b284b58c035 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -255,10 +255,13 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, } /* Map a single area into the IOMMU */ -static dma_addr_t -gart_map_single(struct device *dev, phys_addr_t paddr, size_t size, int dir) +static dma_addr_t gart_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long bus; + phys_addr_t paddr = page_to_phys(page) + offset; if (!dev) dev = &x86_dma_fallback_dev; @@ -275,8 +278,9 @@ gart_map_single(struct device *dev, phys_addr_t paddr, size_t size, int dir) /* * Free a DMA mapping. */ -static void gart_unmap_single(struct device *dev, dma_addr_t dma_addr, - size_t size, int direction) +static void gart_unmap_page(struct device *dev, dma_addr_t dma_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) { unsigned long iommu_page; int npages; @@ -298,8 +302,8 @@ static void gart_unmap_single(struct device *dev, dma_addr_t dma_addr, /* * Wrapper for pci_unmap_single working with scatterlists. */ -static void -gart_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, int dir) +static void gart_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir, struct dma_attrs *attrs) { struct scatterlist *s; int i; @@ -307,7 +311,7 @@ gart_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, int dir) for_each_sg(sg, s, nents, i) { if (!s->dma_length || !s->length) break; - gart_unmap_single(dev, s->dma_address, s->dma_length, dir); + gart_unmap_page(dev, s->dma_address, s->dma_length, dir, NULL); } } @@ -329,7 +333,7 @@ static int dma_map_sg_nonforce(struct device *dev, struct scatterlist *sg, addr = dma_map_area(dev, addr, s->length, dir, 0); if (addr == bad_dma_address) { if (i > 0) - gart_unmap_sg(dev, sg, i, dir); + gart_unmap_sg(dev, sg, i, dir, NULL); nents = 0; sg[0].dma_length = 0; break; @@ -400,8 +404,8 @@ dma_map_cont(struct device *dev, struct scatterlist *start, int nelems, * DMA map all entries in a scatterlist. * Merge chunks that have page aligned sizes into a continuous mapping. */ -static int -gart_map_sg(struct device *dev, struct scatterlist *sg, int nents, int dir) +static int gart_map_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir, struct dma_attrs *attrs) { struct scatterlist *s, *ps, *start_sg, *sgmap; int need = 0, nextneed, i, out, start; @@ -468,7 +472,7 @@ gart_map_sg(struct device *dev, struct scatterlist *sg, int nents, int dir) error: flush_gart(); - gart_unmap_sg(dev, sg, out, dir); + gart_unmap_sg(dev, sg, out, dir, NULL); /* When it was forced or merged try again in a dumb way */ if (force_iommu || iommu_merge) { @@ -521,7 +525,7 @@ static void gart_free_coherent(struct device *dev, size_t size, void *vaddr, dma_addr_t dma_addr) { - gart_unmap_single(dev, dma_addr, size, DMA_BIDIRECTIONAL); + gart_unmap_page(dev, dma_addr, size, DMA_BIDIRECTIONAL, NULL); free_pages((unsigned long)vaddr, get_order(size)); } @@ -707,11 +711,11 @@ static __init int init_k8_gatt(struct agp_kern_info *info) return -1; } -static struct dma_mapping_ops gart_dma_ops = { - .map_single = gart_map_single, - .unmap_single = gart_unmap_single, +static struct dma_map_ops gart_dma_ops = { .map_sg = gart_map_sg, .unmap_sg = gart_unmap_sg, + .map_page = gart_map_page, + .unmap_page = gart_unmap_page, .alloc_coherent = gart_alloc_coherent, .free_coherent = gart_free_coherent, }; diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c index 8b02a3936d4..c6d703b3932 100644 --- a/arch/x86/kernel/pci-nommu.c +++ b/arch/x86/kernel/pci-nommu.c @@ -25,19 +25,19 @@ check_addr(char *name, struct device *hwdev, dma_addr_t bus, size_t size) return 1; } -static dma_addr_t -nommu_map_single(struct device *hwdev, phys_addr_t paddr, size_t size, - int direction) +static dma_addr_t nommu_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { - dma_addr_t bus = paddr; + dma_addr_t bus = page_to_phys(page) + offset; WARN_ON(size == 0); - if (!check_addr("map_single", hwdev, bus, size)) - return bad_dma_address; + if (!check_addr("map_single", dev, bus, size)) + return bad_dma_address; flush_write_buffers(); return bus; } - /* 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 @@ -54,7 +54,8 @@ nommu_map_single(struct device *hwdev, phys_addr_t paddr, size_t size, * the same here. */ static int nommu_map_sg(struct device *hwdev, struct scatterlist *sg, - int nents, int direction) + int nents, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct scatterlist *s; int i; @@ -78,11 +79,11 @@ static void nommu_free_coherent(struct device *dev, size_t size, void *vaddr, free_pages((unsigned long)vaddr, get_order(size)); } -struct dma_mapping_ops nommu_dma_ops = { +struct dma_map_ops nommu_dma_ops = { .alloc_coherent = dma_generic_alloc_coherent, .free_coherent = nommu_free_coherent, - .map_single = nommu_map_single, .map_sg = nommu_map_sg, + .map_page = nommu_map_page, .is_phys = 1, }; diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb.c index d59c9174766..34f12e9996e 100644 --- a/arch/x86/kernel/pci-swiotlb_64.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -33,18 +33,11 @@ phys_addr_t swiotlb_bus_to_phys(dma_addr_t baddr) return baddr; } -int __weak swiotlb_arch_range_needs_mapping(void *ptr, size_t size) +int __weak swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size) { return 0; } -static dma_addr_t -swiotlb_map_single_phys(struct device *hwdev, phys_addr_t paddr, size_t size, - int direction) -{ - return swiotlb_map_single(hwdev, phys_to_virt(paddr), size, direction); -} - static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, dma_addr_t *dma_handle, gfp_t flags) { @@ -57,20 +50,20 @@ static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, return swiotlb_alloc_coherent(hwdev, size, dma_handle, flags); } -struct dma_mapping_ops swiotlb_dma_ops = { +struct dma_map_ops swiotlb_dma_ops = { .mapping_error = swiotlb_dma_mapping_error, .alloc_coherent = x86_swiotlb_alloc_coherent, .free_coherent = swiotlb_free_coherent, - .map_single = swiotlb_map_single_phys, - .unmap_single = swiotlb_unmap_single, .sync_single_for_cpu = swiotlb_sync_single_for_cpu, .sync_single_for_device = swiotlb_sync_single_for_device, .sync_single_range_for_cpu = swiotlb_sync_single_range_for_cpu, .sync_single_range_for_device = swiotlb_sync_single_range_for_device, .sync_sg_for_cpu = swiotlb_sync_sg_for_cpu, .sync_sg_for_device = swiotlb_sync_sg_for_device, - .map_sg = swiotlb_map_sg, - .unmap_sg = swiotlb_unmap_sg, + .map_sg = swiotlb_map_sg_attrs, + .unmap_sg = swiotlb_unmap_sg_attrs, + .map_page = swiotlb_map_page, + .unmap_page = swiotlb_unmap_page, .dma_supported = NULL, }; diff --git a/drivers/base/iommu.c b/drivers/base/iommu.c index 5e039d4f877..c2d1eed9037 100644 --- a/drivers/base/iommu.c +++ b/drivers/base/iommu.c @@ -31,7 +31,7 @@ void register_iommu(struct iommu_ops *ops) iommu_ops = ops; } -bool iommu_found() +bool iommu_found(void) { return iommu_ops != NULL; } diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index ef167b8b047..49402c39923 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -2124,11 +2124,13 @@ error: return 0; } -dma_addr_t intel_map_single(struct device *hwdev, phys_addr_t paddr, - size_t size, int dir) +static dma_addr_t intel_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) { - return __intel_map_single(hwdev, paddr, size, dir, - to_pci_dev(hwdev)->dma_mask); + return __intel_map_single(dev, page_to_phys(page) + offset, size, + dir, to_pci_dev(dev)->dma_mask); } static void flush_unmaps(void) @@ -2192,8 +2194,9 @@ static void add_unmap(struct dmar_domain *dom, struct iova *iova) spin_unlock_irqrestore(&async_umap_flush_lock, flags); } -void intel_unmap_single(struct device *dev, dma_addr_t dev_addr, size_t size, - int dir) +static void intel_unmap_page(struct device *dev, dma_addr_t dev_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) { struct pci_dev *pdev = to_pci_dev(dev); struct dmar_domain *domain; @@ -2237,8 +2240,14 @@ void intel_unmap_single(struct device *dev, dma_addr_t dev_addr, size_t size, } } -void *intel_alloc_coherent(struct device *hwdev, size_t size, - dma_addr_t *dma_handle, gfp_t flags) +static void intel_unmap_single(struct device *dev, dma_addr_t dev_addr, size_t size, + int dir) +{ + intel_unmap_page(dev, dev_addr, size, dir, NULL); +} + +static void *intel_alloc_coherent(struct device *hwdev, size_t size, + dma_addr_t *dma_handle, gfp_t flags) { void *vaddr; int order; @@ -2261,8 +2270,8 @@ void *intel_alloc_coherent(struct device *hwdev, size_t size, return NULL; } -void intel_free_coherent(struct device *hwdev, size_t size, void *vaddr, - dma_addr_t dma_handle) +static void intel_free_coherent(struct device *hwdev, size_t size, void *vaddr, + dma_addr_t dma_handle) { int order; @@ -2275,8 +2284,9 @@ void intel_free_coherent(struct device *hwdev, size_t size, void *vaddr, #define SG_ENT_VIRT_ADDRESS(sg) (sg_virt((sg))) -void intel_unmap_sg(struct device *hwdev, struct scatterlist *sglist, - int nelems, int dir) +static void intel_unmap_sg(struct device *hwdev, struct scatterlist *sglist, + int nelems, enum dma_data_direction dir, + struct dma_attrs *attrs) { int i; struct pci_dev *pdev = to_pci_dev(hwdev); @@ -2333,8 +2343,8 @@ static int intel_nontranslate_map_sg(struct device *hddev, return nelems; } -int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, - int dir) +static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, + enum dma_data_direction dir, struct dma_attrs *attrs) { void *addr; int i; @@ -2414,13 +2424,19 @@ int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int nelems, return nelems; } -static struct dma_mapping_ops intel_dma_ops = { +static int intel_mapping_error(struct device *dev, dma_addr_t dma_addr) +{ + return !dma_addr; +} + +struct dma_map_ops intel_dma_ops = { .alloc_coherent = intel_alloc_coherent, .free_coherent = intel_free_coherent, - .map_single = intel_map_single, - .unmap_single = intel_unmap_single, .map_sg = intel_map_sg, .unmap_sg = intel_unmap_sg, + .map_page = intel_map_page, + .unmap_page = intel_unmap_page, + .mapping_error = intel_mapping_error, }; static inline int iommu_domain_cache_init(void) diff --git a/include/linux/dma-debug.h b/include/linux/dma-debug.h new file mode 100644 index 00000000000..28d53cb7b5a --- /dev/null +++ b/include/linux/dma-debug.h @@ -0,0 +1,174 @@ +/* + * Copyright (C) 2008 Advanced Micro Devices, Inc. + * + * Author: Joerg Roedel <joerg.roedel@amd.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __DMA_DEBUG_H +#define __DMA_DEBUG_H + +#include <linux/types.h> + +struct device; +struct scatterlist; +struct bus_type; + +#ifdef CONFIG_DMA_API_DEBUG + +extern void dma_debug_add_bus(struct bus_type *bus); + +extern void dma_debug_init(u32 num_entries); + +extern void debug_dma_map_page(struct device *dev, struct page *page, + size_t offset, size_t size, + int direction, dma_addr_t dma_addr, + bool map_single); + +extern void debug_dma_unmap_page(struct device *dev, dma_addr_t addr, + size_t size, int direction, bool map_single); + +extern void debug_dma_map_sg(struct device *dev, struct scatterlist *sg, + int nents, int mapped_ents, int direction); + +extern void debug_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, + int nelems, int dir); + +extern void debug_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t dma_addr, void *virt); + +extern void debug_dma_free_coherent(struct device *dev, size_t size, + void *virt, dma_addr_t addr); + +extern void debug_dma_sync_single_for_cpu(struct device *dev, + dma_addr_t dma_handle, size_t size, + int direction); + +extern void debug_dma_sync_single_for_device(struct device *dev, + dma_addr_t dma_handle, + size_t size, int direction); + +extern void debug_dma_sync_single_range_for_cpu(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, + int direction); + +extern void debug_dma_sync_single_range_for_device(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, int direction); + +extern void debug_dma_sync_sg_for_cpu(struct device *dev, + struct scatterlist *sg, + int nelems, int direction); + +extern void debug_dma_sync_sg_for_device(struct device *dev, + struct scatterlist *sg, + int nelems, int direction); + +extern void debug_dma_dump_mappings(struct device *dev); + +#else /* CONFIG_DMA_API_DEBUG */ + +static inline void dma_debug_add_bus(struct bus_type *bus) +{ +} + +static inline void dma_debug_init(u32 num_entries) +{ +} + +static inline void debug_dma_map_page(struct device *dev, struct page *page, + size_t offset, size_t size, + int direction, dma_addr_t dma_addr, + bool map_single) +{ +} + +static inline void debug_dma_unmap_page(struct device *dev, dma_addr_t addr, + size_t size, int direction, + bool map_single) +{ +} + +static inline void debug_dma_map_sg(struct device *dev, struct scatterlist *sg, + int nents, int mapped_ents, int direction) +{ +} + +static inline void debug_dma_unmap_sg(struct device *dev, + struct scatterlist *sglist, + int nelems, int dir) +{ +} + +static inline void debug_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t dma_addr, void *virt) +{ +} + +static inline void debug_dma_free_coherent(struct device *dev, size_t size, + void *virt, dma_addr_t addr) +{ +} + +static inline void debug_dma_sync_single_for_cpu(struct device *dev, + dma_addr_t dma_handle, + size_t size, int direction) +{ +} + +static inline void debug_dma_sync_single_for_device(struct device *dev, + dma_addr_t dma_handle, + size_t size, int direction) +{ +} + +static inline void debug_dma_sync_single_range_for_cpu(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, + int direction) +{ +} + +static inline void debug_dma_sync_single_range_for_device(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, + int direction) +{ +} + +static inline void debug_dma_sync_sg_for_cpu(struct device *dev, + struct scatterlist *sg, + int nelems, int direction) +{ +} + +static inline void debug_dma_sync_sg_for_device(struct device *dev, + struct scatterlist *sg, + int nelems, int direction) +{ +} + +static inline void debug_dma_dump_mappings(struct device *dev) +{ +} + +#endif /* CONFIG_DMA_API_DEBUG */ + +#endif /* __DMA_DEBUG_H */ diff --git a/include/linux/dma-mapping.h b/include/linux/dma-mapping.h index ba9114ec5d3..d7d090d2103 100644 --- a/include/linux/dma-mapping.h +++ b/include/linux/dma-mapping.h @@ -3,6 +3,8 @@ #include <linux/device.h> #include <linux/err.h> +#include <linux/dma-attrs.h> +#include <linux/scatterlist.h> /* These definitions mirror those in pci.h, so they can be used * interchangeably with their PCI_ counterparts */ @@ -13,6 +15,52 @@ enum dma_data_direction { DMA_NONE = 3, }; +struct dma_map_ops { + void* (*alloc_coherent)(struct device *dev, size_t size, + dma_addr_t *dma_handle, gfp_t gfp); + void (*free_coherent)(struct device *dev, size_t size, + void *vaddr, dma_addr_t dma_handle); + dma_addr_t (*map_page)(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs); + void (*unmap_page)(struct device *dev, dma_addr_t dma_handle, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs); + int (*map_sg)(struct device *dev, struct scatterlist *sg, + int nents, enum dma_data_direction dir, + struct dma_attrs *attrs); + void (*unmap_sg)(struct device *dev, + struct scatterlist *sg, int nents, + enum dma_data_direction dir, + struct dma_attrs *attrs); + void (*sync_single_for_cpu)(struct device *dev, + dma_addr_t dma_handle, size_t size, + enum dma_data_direction dir); + void (*sync_single_for_device)(struct device *dev, + dma_addr_t dma_handle, size_t size, + enum dma_data_direction dir); + void (*sync_single_range_for_cpu)(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, + enum dma_data_direction dir); + void (*sync_single_range_for_device)(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, + enum dma_data_direction dir); + void (*sync_sg_for_cpu)(struct device *dev, + struct scatterlist *sg, int nents, + enum dma_data_direction dir); + void (*sync_sg_for_device)(struct device *dev, + struct scatterlist *sg, int nents, + enum dma_data_direction dir); + int (*mapping_error)(struct device *dev, dma_addr_t dma_addr); + int (*dma_supported)(struct device *dev, u64 mask); + int is_phys; +}; + #define DMA_BIT_MASK(n) (((n) == 64) ? ~0ULL : ((1ULL<<(n))-1)) /* diff --git a/include/linux/intel-iommu.h b/include/linux/intel-iommu.h index 78c1262e870..1d6c71d96ed 100644 --- a/include/linux/intel-iommu.h +++ b/include/linux/intel-iommu.h @@ -332,11 +332,4 @@ extern int qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr, extern int qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu); -extern void *intel_alloc_coherent(struct device *, size_t, dma_addr_t *, gfp_t); -extern void intel_free_coherent(struct device *, size_t, void *, dma_addr_t); -extern dma_addr_t intel_map_single(struct device *, phys_addr_t, size_t, int); -extern void intel_unmap_single(struct device *, dma_addr_t, size_t, int); -extern int intel_map_sg(struct device *, struct scatterlist *, int, int); -extern void intel_unmap_sg(struct device *, struct scatterlist *, int, int); - #endif diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index dedd3c0cfe3..ac9ff54f7cb 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -31,7 +31,7 @@ extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t address); extern phys_addr_t swiotlb_bus_to_phys(dma_addr_t address); -extern int swiotlb_arch_range_needs_mapping(void *ptr, size_t size); +extern int swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size); extern void *swiotlb_alloc_coherent(struct device *hwdev, size_t size, @@ -41,20 +41,13 @@ extern void swiotlb_free_coherent(struct device *hwdev, size_t size, void *vaddr, dma_addr_t dma_handle); -extern dma_addr_t -swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir); - -extern void -swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir); - -extern dma_addr_t -swiotlb_map_single_attrs(struct device *hwdev, void *ptr, size_t size, - int dir, struct dma_attrs *attrs); - -extern void -swiotlb_unmap_single_attrs(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir, struct dma_attrs *attrs); +extern dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs); +extern void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dev_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs); extern int swiotlb_map_sg(struct device *hwdev, struct scatterlist *sg, int nents, @@ -66,36 +59,38 @@ swiotlb_unmap_sg(struct device *hwdev, struct scatterlist *sg, int nents, extern int swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems, - int dir, struct dma_attrs *attrs); + enum dma_data_direction dir, struct dma_attrs *attrs); extern void swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl, - int nelems, int dir, struct dma_attrs *attrs); + int nelems, enum dma_data_direction dir, + struct dma_attrs *attrs); extern void swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir); + size_t size, enum dma_data_direction dir); extern void swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg, - int nelems, int dir); + int nelems, enum dma_data_direction dir); extern void swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir); + size_t size, enum dma_data_direction dir); extern void swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg, - int nelems, int dir); + int nelems, enum dma_data_direction dir); extern void swiotlb_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dev_addr, - unsigned long offset, size_t size, int dir); + unsigned long offset, size_t size, + enum dma_data_direction dir); extern void swiotlb_sync_single_range_for_device(struct device *hwdev, dma_addr_t dev_addr, unsigned long offset, size_t size, - int dir); + enum dma_data_direction dir); extern int swiotlb_dma_mapping_error(struct device *hwdev, dma_addr_t dma_addr); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 0626fa4856e..58bfe7e8fab 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -912,6 +912,17 @@ config DYNAMIC_DEBUG See Documentation/dynamic-debug-howto.txt for additional information. +config DMA_API_DEBUG + bool "Enable debugging of DMA-API usage" + depends on HAVE_DMA_API_DEBUG + help + Enable this option to debug the use of the DMA API by device drivers. + With this option you will be able to detect common bugs in device + drivers like double-freeing of DMA mappings or freeing mappings that + were never allocated. + This option causes a performance degredation. Use only if you want + to debug device drivers. If unsure, say N. + source "samples/Kconfig" source "lib/Kconfig.kgdb" diff --git a/lib/Makefile b/lib/Makefile index 051a33a8e02..d6edd6753f4 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -90,6 +90,8 @@ obj-$(CONFIG_DYNAMIC_DEBUG) += dynamic_debug.o obj-$(CONFIG_NLATTR) += nlattr.o +obj-$(CONFIG_DMA_API_DEBUG) += dma-debug.o + hostprogs-y := gen_crc32table clean-files := crc32table.h diff --git a/lib/dma-debug.c b/lib/dma-debug.c new file mode 100644 index 00000000000..1a992089486 --- /dev/null +++ b/lib/dma-debug.c @@ -0,0 +1,955 @@ +/* + * Copyright (C) 2008 Advanced Micro Devices, Inc. + * + * Author: Joerg Roedel <joerg.roedel@amd.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include <linux/scatterlist.h> +#include <linux/dma-mapping.h> +#include <linux/stacktrace.h> +#include <linux/dma-debug.h> +#include <linux/spinlock.h> +#include <linux/debugfs.h> +#include <linux/device.h> +#include <linux/types.h> +#include <linux/sched.h> +#include <linux/list.h> +#include <linux/slab.h> + +#include <asm/sections.h> + +#define HASH_SIZE 1024ULL +#define HASH_FN_SHIFT 13 +#define HASH_FN_MASK (HASH_SIZE - 1) + +enum { + dma_debug_single, + dma_debug_page, + dma_debug_sg, + dma_debug_coherent, +}; + +#define DMA_DEBUG_STACKTRACE_ENTRIES 5 + +struct dma_debug_entry { + struct list_head list; + struct device *dev; + int type; + phys_addr_t paddr; + u64 dev_addr; + u64 size; + int direction; + int sg_call_ents; + int sg_mapped_ents; +#ifdef CONFIG_STACKTRACE + struct stack_trace stacktrace; + unsigned long st_entries[DMA_DEBUG_STACKTRACE_ENTRIES]; +#endif +}; + +struct hash_bucket { + struct list_head list; + spinlock_t lock; +} ____cacheline_aligned_in_smp; + +/* Hash list to save the allocated dma addresses */ +static struct hash_bucket dma_entry_hash[HASH_SIZE]; +/* List of pre-allocated dma_debug_entry's */ +static LIST_HEAD(free_entries); +/* Lock for the list above */ +static DEFINE_SPINLOCK(free_entries_lock); + +/* Global disable flag - will be set in case of an error */ +static bool global_disable __read_mostly; + +/* Global error count */ +static u32 error_count; + +/* Global error show enable*/ +static u32 show_all_errors __read_mostly; +/* Number of errors to show */ +static u32 show_num_errors = 1; + +static u32 num_free_entries; +static u32 min_free_entries; + +/* number of preallocated entries requested by kernel cmdline */ +static u32 req_entries; + +/* debugfs dentry's for the stuff above */ +static struct dentry *dma_debug_dent __read_mostly; +static struct dentry *global_disable_dent __read_mostly; +static struct dentry *error_count_dent __read_mostly; +static struct dentry *show_all_errors_dent __read_mostly; +static struct dentry *show_num_errors_dent __read_mostly; +static struct dentry *num_free_entries_dent __read_mostly; +static struct dentry *min_free_entries_dent __read_mostly; + +static const char *type2name[4] = { "single", "page", + "scather-gather", "coherent" }; + +static const char *dir2name[4] = { "DMA_BIDIRECTIONAL", "DMA_TO_DEVICE", + "DMA_FROM_DEVICE", "DMA_NONE" }; + +/* + * The access to some variables in this macro is racy. We can't use atomic_t + * here because all these variables are exported to debugfs. Some of them even + * writeable. This is also the reason why a lock won't help much. But anyway, + * the races are no big deal. Here is why: + * + * error_count: the addition is racy, but the worst thing that can happen is + * that we don't count some errors + * show_num_errors: the subtraction is racy. Also no big deal because in + * worst case this will result in one warning more in the + * system log than the user configured. This variable is + * writeable via debugfs. + */ +static inline void dump_entry_trace(struct dma_debug_entry *entry) +{ +#ifdef CONFIG_STACKTRACE + if (entry) { + printk(KERN_WARNING "Mapped at:\n"); + print_stack_trace(&entry->stacktrace, 0); + } +#endif +} + +#define err_printk(dev, entry, format, arg...) do { \ + error_count += 1; \ + if (show_all_errors || show_num_errors > 0) { \ + WARN(1, "%s %s: " format, \ + dev_driver_string(dev), \ + dev_name(dev) , ## arg); \ + dump_entry_trace(entry); \ + } \ + if (!show_all_errors && show_num_errors > 0) \ + show_num_errors -= 1; \ + } while (0); + +/* + * Hash related functions + * + * Every DMA-API request is saved into a struct dma_debug_entry. To + * have quick access to these structs they are stored into a hash. + */ +static int hash_fn(struct dma_debug_entry *entry) +{ + /* + * Hash function is based on the dma address. + * We use bits 20-27 here as the index into the hash + */ + return (entry->dev_addr >> HASH_FN_SHIFT) & HASH_FN_MASK; +} + +/* + * Request exclusive access to a hash bucket for a given dma_debug_entry. + */ +static struct hash_bucket *get_hash_bucket(struct dma_debug_entry *entry, + unsigned long *flags) +{ + int idx = hash_fn(entry); + unsigned long __flags; + + spin_lock_irqsave(&dma_entry_hash[idx].lock, __flags); + *flags = __flags; + return &dma_entry_hash[idx]; +} + +/* + * Give up exclusive access to the hash bucket + */ +static void put_hash_bucket(struct hash_bucket *bucket, + unsigned long *flags) +{ + unsigned long __flags = *flags; + + spin_unlock_irqrestore(&bucket->lock, __flags); +} + +/* + * Search a given entry in the hash bucket list + */ +static struct dma_debug_entry *hash_bucket_find(struct hash_bucket *bucket, + struct dma_debug_entry *ref) +{ + struct dma_debug_entry *entry; + + list_for_each_entry(entry, &bucket->list, list) { + if ((entry->dev_addr == ref->dev_addr) && + (entry->dev == ref->dev)) + return entry; + } + + return NULL; +} + +/* + * Add an entry to a hash bucket + */ +static void hash_bucket_add(struct hash_bucket *bucket, + struct dma_debug_entry *entry) +{ + list_add_tail(&entry->list, &bucket->list); +} + +/* + * Remove entry from a hash bucket list + */ +static void hash_bucket_del(struct dma_debug_entry *entry) +{ + list_del(&entry->list); +} + +/* + * Dump mapping entries for debugging purposes + */ +void debug_dma_dump_mappings(struct device *dev) +{ + int idx; + + for (idx = 0; idx < HASH_SIZE; idx++) { + struct hash_bucket *bucket = &dma_entry_hash[idx]; + struct dma_debug_entry *entry; + unsigned long flags; + + spin_lock_irqsave(&bucket->lock, flags); + + list_for_each_entry(entry, &bucket->list, list) { + if (!dev || dev == entry->dev) { + dev_info(entry->dev, + "%s idx %d P=%Lx D=%Lx L=%Lx %s\n", + type2name[entry->type], idx, + (unsigned long long)entry->paddr, + entry->dev_addr, entry->size, + dir2name[entry->direction]); + } + } + + spin_unlock_irqrestore(&bucket->lock, flags); + } +} +EXPORT_SYMBOL(debug_dma_dump_mappings); + +/* + * Wrapper function for adding an entry to the hash. + * This function takes care of locking itself. + */ +static void add_dma_entry(struct dma_debug_entry *entry) +{ + struct hash_bucket *bucket; + unsigned long flags; + + bucket = get_hash_bucket(entry, &flags); + hash_bucket_add(bucket, entry); + put_hash_bucket(bucket, &flags); +} + +/* struct dma_entry allocator + * + * The next two functions implement the allocator for + * struct dma_debug_entries. + */ +static struct dma_debug_entry *dma_entry_alloc(void) +{ + struct dma_debug_entry *entry = NULL; + unsigned long flags; + + spin_lock_irqsave(&free_entries_lock, flags); + + if (list_empty(&free_entries)) { + printk(KERN_ERR "DMA-API: debugging out of memory " + "- disabling\n"); + global_disable = true; + goto out; + } + + entry = list_entry(free_entries.next, struct dma_debug_entry, list); + list_del(&entry->list); + memset(entry, 0, sizeof(*entry)); + +#ifdef CONFIG_STACKTRACE + entry->stacktrace.max_entries = DMA_DEBUG_STACKTRACE_ENTRIES; + entry->stacktrace.entries = entry->st_entries; + entry->stacktrace.skip = 2; + save_stack_trace(&entry->stacktrace); +#endif + num_free_entries -= 1; + if (num_free_entries < min_free_entries) + min_free_entries = num_free_entries; + +out: + spin_unlock_irqrestore(&free_entries_lock, flags); + + return entry; +} + +static void dma_entry_free(struct dma_debug_entry *entry) +{ + unsigned long flags; + + /* + * add to beginning of the list - this way the entries are + * more likely cache hot when they are reallocated. + */ + spin_lock_irqsave(&free_entries_lock, flags); + list_add(&entry->list, &free_entries); + num_free_entries += 1; + spin_unlock_irqrestore(&free_entries_lock, flags); +} + +/* + * DMA-API debugging init code + * + * The init code does two things: + * 1. Initialize core data structures + * 2. Preallocate a given number of dma_debug_entry structs + */ + +static int prealloc_memory(u32 num_entries) +{ + struct dma_debug_entry *entry, *next_entry; + int i; + + for (i = 0; i < num_entries; ++i) { + entry = kzalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + goto out_err; + + list_add_tail(&entry->list, &free_entries); + } + + num_free_entries = num_entries; + min_free_entries = num_entries; + + printk(KERN_INFO "DMA-API: preallocated %d debug entries\n", + num_entries); + + return 0; + +out_err: + + list_for_each_entry_safe(entry, next_entry, &free_entries, list) { + list_del(&entry->list); + kfree(entry); + } + + return -ENOMEM; +} + +static int dma_debug_fs_init(void) +{ + dma_debug_dent = debugfs_create_dir("dma-api", NULL); + if (!dma_debug_dent) { + printk(KERN_ERR "DMA-API: can not create debugfs directory\n"); + return -ENOMEM; + } + + global_disable_dent = debugfs_create_bool("disabled", 0444, + dma_debug_dent, + (u32 *)&global_disable); + if (!global_disable_dent) + goto out_err; + + error_count_dent = debugfs_create_u32("error_count", 0444, + dma_debug_dent, &error_count); + if (!error_count_dent) + goto out_err; + + show_all_errors_dent = debugfs_create_u32("all_errors", 0644, + dma_debug_dent, + &show_all_errors); + if (!show_all_errors_dent) + goto out_err; + + show_num_errors_dent = debugfs_create_u32("num_errors", 0644, + dma_debug_dent, + &show_num_errors); + if (!show_num_errors_dent) + goto out_err; + + num_free_entries_dent = debugfs_create_u32("num_free_entries", 0444, + dma_debug_dent, + &num_free_entries); + if (!num_free_entries_dent) + goto out_err; + + min_free_entries_dent = debugfs_create_u32("min_free_entries", 0444, + dma_debug_dent, + &min_free_entries); + if (!min_free_entries_dent) + goto out_err; + + return 0; + +out_err: + debugfs_remove_recursive(dma_debug_dent); + + return -ENOMEM; +} + +static int device_dma_allocations(struct device *dev) +{ + struct dma_debug_entry *entry; + unsigned long flags; + int count = 0, i; + + for (i = 0; i < HASH_SIZE; ++i) { + spin_lock_irqsave(&dma_entry_hash[i].lock, flags); + list_for_each_entry(entry, &dma_entry_hash[i].list, list) { + if (entry->dev == dev) + count += 1; + } + spin_unlock_irqrestore(&dma_entry_hash[i].lock, flags); + } + + return count; +} + +static int dma_debug_device_change(struct notifier_block *nb, + unsigned long action, void *data) +{ + struct device *dev = data; + int count; + + + switch (action) { + case BUS_NOTIFY_UNBIND_DRIVER: + count = device_dma_allocations(dev); + if (count == 0) + break; + err_printk(dev, NULL, "DMA-API: device driver has pending " + "DMA allocations while released from device " + "[count=%d]\n", count); + break; + default: + break; + } + + return 0; +} + +void dma_debug_add_bus(struct bus_type *bus) +{ + struct notifier_block *nb; + + nb = kzalloc(sizeof(struct notifier_block), GFP_KERNEL); + if (nb == NULL) { + printk(KERN_ERR "dma_debug_add_bus: out of memory\n"); + return; + } + + nb->notifier_call = dma_debug_device_change; + + bus_register_notifier(bus, nb); +} + +/* + * Let the architectures decide how many entries should be preallocated. + */ +void dma_debug_init(u32 num_entries) +{ + int i; + + if (global_disable) + return; + + for (i = 0; i < HASH_SIZE; ++i) { + INIT_LIST_HEAD(&dma_entry_hash[i].list); + dma_entry_hash[i].lock = SPIN_LOCK_UNLOCKED; + } + + if (dma_debug_fs_init() != 0) { + printk(KERN_ERR "DMA-API: error creating debugfs entries " + "- disabling\n"); + global_disable = true; + + return; + } + + if (req_entries) + num_entries = req_entries; + + if (prealloc_memory(num_entries) != 0) { + printk(KERN_ERR "DMA-API: debugging out of memory error " + "- disabled\n"); + global_disable = true; + + return; + } + + printk(KERN_INFO "DMA-API: debugging enabled by kernel config\n"); +} + +static __init int dma_debug_cmdline(char *str) +{ + if (!str) + return -EINVAL; + + if (strncmp(str, "off", 3) == 0) { + printk(KERN_INFO "DMA-API: debugging disabled on kernel " + "command line\n"); + global_disable = true; + } + + return 0; +} + +static __init int dma_debug_entries_cmdline(char *str) +{ + int res; + + if (!str) + return -EINVAL; + + res = get_option(&str, &req_entries); + + if (!res) + req_entries = 0; + + return 0; +} + +__setup("dma_debug=", dma_debug_cmdline); +__setup("dma_debug_entries=", dma_debug_entries_cmdline); + +static void check_unmap(struct dma_debug_entry *ref) +{ + struct dma_debug_entry *entry; + struct hash_bucket *bucket; + unsigned long flags; + + if (dma_mapping_error(ref->dev, ref->dev_addr)) { + err_printk(ref->dev, NULL, "DMA-API: device driver tries " + "to free an invalid DMA memory address\n"); + return; + } + + bucket = get_hash_bucket(ref, &flags); + entry = hash_bucket_find(bucket, ref); + + if (!entry) { + err_printk(ref->dev, NULL, "DMA-API: device driver tries " + "to free DMA memory it has not allocated " + "[device address=0x%016llx] [size=%llu bytes]\n", + ref->dev_addr, ref->size); + goto out; + } + + if (ref->size != entry->size) { + err_printk(ref->dev, entry, "DMA-API: device driver frees " + "DMA memory with different size " + "[device address=0x%016llx] [map size=%llu bytes] " + "[unmap size=%llu bytes]\n", + ref->dev_addr, entry->size, ref->size); + } + + if (ref->type != entry->type) { + err_printk(ref->dev, entry, "DMA-API: device driver frees " + "DMA memory with wrong function " + "[device address=0x%016llx] [size=%llu bytes] " + "[mapped as %s] [unmapped as %s]\n", + ref->dev_addr, ref->size, + type2name[entry->type], type2name[ref->type]); + } else if ((entry->type == dma_debug_coherent) && + (ref->paddr != entry->paddr)) { + err_printk(ref->dev, entry, "DMA-API: device driver frees " + "DMA memory with different CPU address " + "[device address=0x%016llx] [size=%llu bytes] " + "[cpu alloc address=%p] [cpu free address=%p]", + ref->dev_addr, ref->size, + (void *)entry->paddr, (void *)ref->paddr); + } + + if (ref->sg_call_ents && ref->type == dma_debug_sg && + ref->sg_call_ents != entry->sg_call_ents) { + err_printk(ref->dev, entry, "DMA-API: device driver frees " + "DMA sg list with different entry count " + "[map count=%d] [unmap count=%d]\n", + entry->sg_call_ents, ref->sg_call_ents); + } + + /* + * This may be no bug in reality - but most implementations of the + * DMA API don't handle this properly, so check for it here + */ + if (ref->direction != entry->direction) { + err_printk(ref->dev, entry, "DMA-API: device driver frees " + "DMA memory with different direction " + "[device address=0x%016llx] [size=%llu bytes] " + "[mapped with %s] [unmapped with %s]\n", + ref->dev_addr, ref->size, + dir2name[entry->direction], + dir2name[ref->direction]); + } + + hash_bucket_del(entry); + dma_entry_free(entry); + +out: + put_hash_bucket(bucket, &flags); +} + +static void check_for_stack(struct device *dev, void *addr) +{ + if (object_is_on_stack(addr)) + err_printk(dev, NULL, "DMA-API: device driver maps memory from" + "stack [addr=%p]\n", addr); +} + +static inline bool overlap(void *addr, u64 size, void *start, void *end) +{ + void *addr2 = (char *)addr + size; + + return ((addr >= start && addr < end) || + (addr2 >= start && addr2 < end) || + ((addr < start) && (addr2 >= end))); +} + +static void check_for_illegal_area(struct device *dev, void *addr, u64 size) +{ + if (overlap(addr, size, _text, _etext) || + overlap(addr, size, __start_rodata, __end_rodata)) + err_printk(dev, NULL, "DMA-API: device driver maps " + "memory from kernel text or rodata " + "[addr=%p] [size=%llu]\n", addr, size); +} + +static void check_sync(struct device *dev, dma_addr_t addr, + u64 size, u64 offset, int direction, bool to_cpu) +{ + struct dma_debug_entry ref = { + .dev = dev, + .dev_addr = addr, + .size = size, + .direction = direction, + }; + struct dma_debug_entry *entry; + struct hash_bucket *bucket; + unsigned long flags; + + bucket = get_hash_bucket(&ref, &flags); + + entry = hash_bucket_find(bucket, &ref); + + if (!entry) { + err_printk(dev, NULL, "DMA-API: device driver tries " + "to sync DMA memory it has not allocated " + "[device address=0x%016llx] [size=%llu bytes]\n", + addr, size); + goto out; + } + + if ((offset + size) > entry->size) { + err_printk(dev, entry, "DMA-API: device driver syncs" + " DMA memory outside allocated range " + "[device address=0x%016llx] " + "[allocation size=%llu bytes] [sync offset=%llu] " + "[sync size=%llu]\n", entry->dev_addr, entry->size, + offset, size); + } + + if (direction != entry->direction) { + err_printk(dev, entry, "DMA-API: device driver syncs " + "DMA memory with different direction " + "[device address=0x%016llx] [size=%llu bytes] " + "[mapped with %s] [synced with %s]\n", + addr, entry->size, + dir2name[entry->direction], + dir2name[direction]); + } + + if (entry->direction == DMA_BIDIRECTIONAL) + goto out; + + if (to_cpu && !(entry->direction == DMA_FROM_DEVICE) && + !(direction == DMA_TO_DEVICE)) + err_printk(dev, entry, "DMA-API: device driver syncs " + "device read-only DMA memory for cpu " + "[device address=0x%016llx] [size=%llu bytes] " + "[mapped with %s] [synced with %s]\n", + addr, entry->size, + dir2name[entry->direction], + dir2name[direction]); + + if (!to_cpu && !(entry->direction == DMA_TO_DEVICE) && + !(direction == DMA_FROM_DEVICE)) + err_printk(dev, entry, "DMA-API: device driver syncs " + "device write-only DMA memory to device " + "[device address=0x%016llx] [size=%llu bytes] " + "[mapped with %s] [synced with %s]\n", + addr, entry->size, + dir2name[entry->direction], + dir2name[direction]); + +out: + put_hash_bucket(bucket, &flags); + +} + +void debug_dma_map_page(struct device *dev, struct page *page, size_t offset, + size_t size, int direction, dma_addr_t dma_addr, + bool map_single) +{ + struct dma_debug_entry *entry; + + if (unlikely(global_disable)) + return; + + if (unlikely(dma_mapping_error(dev, dma_addr))) + return; + + entry = dma_entry_alloc(); + if (!entry) + return; + + entry->dev = dev; + entry->type = dma_debug_page; + entry->paddr = page_to_phys(page) + offset; + entry->dev_addr = dma_addr; + entry->size = size; + entry->direction = direction; + + if (map_single) + entry->type = dma_debug_single; + + if (!PageHighMem(page)) { + void *addr = ((char *)page_address(page)) + offset; + check_for_stack(dev, addr); + check_for_illegal_area(dev, addr, size); + } + + add_dma_entry(entry); +} +EXPORT_SYMBOL(debug_dma_map_page); + +void debug_dma_unmap_page(struct device *dev, dma_addr_t addr, + size_t size, int direction, bool map_single) +{ + struct dma_debug_entry ref = { + .type = dma_debug_page, + .dev = dev, + .dev_addr = addr, + .size = size, + .direction = direction, + }; + + if (unlikely(global_disable)) + return; + + if (map_single) + ref.type = dma_debug_single; + + check_unmap(&ref); +} +EXPORT_SYMBOL(debug_dma_unmap_page); + +void debug_dma_map_sg(struct device *dev, struct scatterlist *sg, + int nents, int mapped_ents, int direction) +{ + struct dma_debug_entry *entry; + struct scatterlist *s; + int i; + + if (unlikely(global_disable)) + return; + + for_each_sg(sg, s, mapped_ents, i) { + entry = dma_entry_alloc(); + if (!entry) + return; + + entry->type = dma_debug_sg; + entry->dev = dev; + entry->paddr = sg_phys(s); + entry->size = s->length; + entry->dev_addr = s->dma_address; + entry->direction = direction; + entry->sg_call_ents = nents; + entry->sg_mapped_ents = mapped_ents; + + if (!PageHighMem(sg_page(s))) { + check_for_stack(dev, sg_virt(s)); + check_for_illegal_area(dev, sg_virt(s), s->length); + } + + add_dma_entry(entry); + } +} +EXPORT_SYMBOL(debug_dma_map_sg); + +void debug_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, + int nelems, int dir) +{ + struct dma_debug_entry *entry; + struct scatterlist *s; + int mapped_ents = 0, i; + unsigned long flags; + + if (unlikely(global_disable)) + return; + + for_each_sg(sglist, s, nelems, i) { + + struct dma_debug_entry ref = { + .type = dma_debug_sg, + .dev = dev, + .paddr = sg_phys(s), + .dev_addr = s->dma_address, + .size = s->length, + .direction = dir, + .sg_call_ents = 0, + }; + + if (mapped_ents && i >= mapped_ents) + break; + + if (mapped_ents == 0) { + struct hash_bucket *bucket; + ref.sg_call_ents = nelems; + bucket = get_hash_bucket(&ref, &flags); + entry = hash_bucket_find(bucket, &ref); + if (entry) + mapped_ents = entry->sg_mapped_ents; + put_hash_bucket(bucket, &flags); + } + + check_unmap(&ref); + } +} +EXPORT_SYMBOL(debug_dma_unmap_sg); + +void debug_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t dma_addr, void *virt) +{ + struct dma_debug_entry *entry; + + if (unlikely(global_disable)) + return; + + if (unlikely(virt == NULL)) + return; + + entry = dma_entry_alloc(); + if (!entry) + return; + + entry->type = dma_debug_coherent; + entry->dev = dev; + entry->paddr = virt_to_phys(virt); + entry->size = size; + entry->dev_addr = dma_addr; + entry->direction = DMA_BIDIRECTIONAL; + + add_dma_entry(entry); +} +EXPORT_SYMBOL(debug_dma_alloc_coherent); + +void debug_dma_free_coherent(struct device *dev, size_t size, + void *virt, dma_addr_t addr) +{ + struct dma_debug_entry ref = { + .type = dma_debug_coherent, + .dev = dev, + .paddr = virt_to_phys(virt), + .dev_addr = addr, + .size = size, + .direction = DMA_BIDIRECTIONAL, + }; + + if (unlikely(global_disable)) + return; + + check_unmap(&ref); +} +EXPORT_SYMBOL(debug_dma_free_coherent); + +void debug_dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, + size_t size, int direction) +{ + if (unlikely(global_disable)) + return; + + check_sync(dev, dma_handle, size, 0, direction, true); +} +EXPORT_SYMBOL(debug_dma_sync_single_for_cpu); + +void debug_dma_sync_single_for_device(struct device *dev, + dma_addr_t dma_handle, size_t size, + int direction) +{ + if (unlikely(global_disable)) + return; + + check_sync(dev, dma_handle, size, 0, direction, false); +} +EXPORT_SYMBOL(debug_dma_sync_single_for_device); + +void debug_dma_sync_single_range_for_cpu(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, size_t size, + int direction) +{ + if (unlikely(global_disable)) + return; + + check_sync(dev, dma_handle, size, offset, direction, true); +} +EXPORT_SYMBOL(debug_dma_sync_single_range_for_cpu); + +void debug_dma_sync_single_range_for_device(struct device *dev, + dma_addr_t dma_handle, + unsigned long offset, + size_t size, int direction) +{ + if (unlikely(global_disable)) + return; + + check_sync(dev, dma_handle, size, offset, direction, false); +} +EXPORT_SYMBOL(debug_dma_sync_single_range_for_device); + +void debug_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, + int nelems, int direction) +{ + struct scatterlist *s; + int i; + + if (unlikely(global_disable)) + return; + + for_each_sg(sg, s, nelems, i) { + check_sync(dev, s->dma_address, s->dma_length, 0, + direction, true); + } +} +EXPORT_SYMBOL(debug_dma_sync_sg_for_cpu); + +void debug_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, + int nelems, int direction) +{ + struct scatterlist *s; + int i; + + if (unlikely(global_disable)) + return; + + for_each_sg(sg, s, nelems, i) { + check_sync(dev, s->dma_address, s->dma_length, 0, + direction, false); + } +} +EXPORT_SYMBOL(debug_dma_sync_sg_for_device); + diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 1f991acc2a0..32e2bd3b114 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -145,7 +145,7 @@ static void *swiotlb_bus_to_virt(dma_addr_t address) return phys_to_virt(swiotlb_bus_to_phys(address)); } -int __weak swiotlb_arch_range_needs_mapping(void *ptr, size_t size) +int __weak swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size) { return 0; } @@ -315,9 +315,9 @@ address_needs_mapping(struct device *hwdev, dma_addr_t addr, size_t size) return !is_buffer_dma_capable(dma_get_mask(hwdev), addr, size); } -static inline int range_needs_mapping(void *ptr, size_t size) +static inline int range_needs_mapping(phys_addr_t paddr, size_t size) { - return swiotlb_force || swiotlb_arch_range_needs_mapping(ptr, size); + return swiotlb_force || swiotlb_arch_range_needs_mapping(paddr, size); } static int is_swiotlb_buffer(char *addr) @@ -636,11 +636,14 @@ swiotlb_full(struct device *dev, size_t size, int dir, int do_panic) * Once the device is given the dma address, the device owns this memory until * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed. */ -dma_addr_t -swiotlb_map_single_attrs(struct device *hwdev, void *ptr, size_t size, - int dir, struct dma_attrs *attrs) -{ - dma_addr_t dev_addr = swiotlb_virt_to_bus(hwdev, ptr); +dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, + unsigned long offset, size_t size, + enum dma_data_direction dir, + struct dma_attrs *attrs) +{ + phys_addr_t phys = page_to_phys(page) + offset; + void *ptr = page_address(page) + offset; + dma_addr_t dev_addr = swiotlb_phys_to_bus(dev, phys); void *map; BUG_ON(dir == DMA_NONE); @@ -649,37 +652,30 @@ swiotlb_map_single_attrs(struct device *hwdev, void *ptr, size_t size, * we can safely return the device addr and not worry about bounce * buffering it. */ - if (!address_needs_mapping(hwdev, dev_addr, size) && - !range_needs_mapping(ptr, size)) + if (!address_needs_mapping(dev, dev_addr, size) && + !range_needs_mapping(virt_to_phys(ptr), size)) return dev_addr; /* * Oh well, have to allocate and map a bounce buffer. */ - map = map_single(hwdev, virt_to_phys(ptr), size, dir); + map = map_single(dev, phys, size, dir); if (!map) { - swiotlb_full(hwdev, size, dir, 1); + swiotlb_full(dev, size, dir, 1); map = io_tlb_overflow_buffer; } - dev_addr = swiotlb_virt_to_bus(hwdev, map); + dev_addr = swiotlb_virt_to_bus(dev, map); /* * Ensure that the address returned is DMA'ble */ - if (address_needs_mapping(hwdev, dev_addr, size)) + if (address_needs_mapping(dev, dev_addr, size)) panic("map_single: bounce buffer is not DMA'ble"); return dev_addr; } -EXPORT_SYMBOL(swiotlb_map_single_attrs); - -dma_addr_t -swiotlb_map_single(struct device *hwdev, void *ptr, size_t size, int dir) -{ - return swiotlb_map_single_attrs(hwdev, ptr, size, dir, NULL); -} -EXPORT_SYMBOL(swiotlb_map_single); +EXPORT_SYMBOL_GPL(swiotlb_map_page); /* * Unmap a single streaming mode DMA translation. The dma_addr and size must @@ -689,9 +685,9 @@ EXPORT_SYMBOL(swiotlb_map_single); * After this call, reads by the cpu to the buffer are guaranteed to see * whatever the device wrote there. */ -void -swiotlb_unmap_single_attrs(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir, struct dma_attrs *attrs) +void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dev_addr, + size_t size, enum dma_data_direction dir, + struct dma_attrs *attrs) { char *dma_addr = swiotlb_bus_to_virt(dev_addr); @@ -701,15 +697,7 @@ swiotlb_unmap_single_attrs(struct device *hwdev, dma_addr_t dev_addr, else if (dir == DMA_FROM_DEVICE) dma_mark_clean(dma_addr, size); } -EXPORT_SYMBOL(swiotlb_unmap_single_attrs); - -void -swiotlb_unmap_single(struct device *hwdev, dma_addr_t dev_addr, size_t size, - int dir) -{ - return swiotlb_unmap_single_attrs(hwdev, dev_addr, size, dir, NULL); -} -EXPORT_SYMBOL(swiotlb_unmap_single); +EXPORT_SYMBOL_GPL(swiotlb_unmap_page); /* * Make physical memory consistent for a single streaming mode DMA translation @@ -736,7 +724,7 @@ swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr, void swiotlb_sync_single_for_cpu(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir) + size_t size, enum dma_data_direction dir) { swiotlb_sync_single(hwdev, dev_addr, size, dir, SYNC_FOR_CPU); } @@ -744,7 +732,7 @@ EXPORT_SYMBOL(swiotlb_sync_single_for_cpu); void swiotlb_sync_single_for_device(struct device *hwdev, dma_addr_t dev_addr, - size_t size, int dir) + size_t size, enum dma_data_direction dir) { swiotlb_sync_single(hwdev, dev_addr, size, dir, SYNC_FOR_DEVICE); } @@ -769,7 +757,8 @@ swiotlb_sync_single_range(struct device *hwdev, dma_addr_t dev_addr, void swiotlb_sync_single_range_for_cpu(struct device *hwdev, dma_addr_t dev_addr, - unsigned long offset, size_t size, int dir) + unsigned long offset, size_t size, + enum dma_data_direction dir) { swiotlb_sync_single_range(hwdev, dev_addr, offset, size, dir, SYNC_FOR_CPU); @@ -778,7 +767,8 @@ EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_cpu); void swiotlb_sync_single_range_for_device(struct device *hwdev, dma_addr_t dev_addr, - unsigned long offset, size_t size, int dir) + unsigned long offset, size_t size, + enum dma_data_direction dir) { swiotlb_sync_single_range(hwdev, dev_addr, offset, size, dir, SYNC_FOR_DEVICE); @@ -803,7 +793,7 @@ EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_device); */ int swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems, - int dir, struct dma_attrs *attrs) + enum dma_data_direction dir, struct dma_attrs *attrs) { struct scatterlist *sg; int i; @@ -811,10 +801,10 @@ swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl, int nelems, BUG_ON(dir == DMA_NONE); for_each_sg(sgl, sg, nelems, i) { - void *addr = sg_virt(sg); - dma_addr_t dev_addr = swiotlb_virt_to_bus(hwdev, addr); + phys_addr_t paddr = sg_phys(sg); + dma_addr_t dev_addr = swiotlb_phys_to_bus(hwdev, paddr); - if (range_needs_mapping(addr, sg->length) || + if (range_needs_mapping(paddr, sg->length) || address_needs_mapping(hwdev, dev_addr, sg->length)) { void *map = map_single(hwdev, sg_phys(sg), sg->length, dir); @@ -850,7 +840,7 @@ EXPORT_SYMBOL(swiotlb_map_sg); */ void swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl, - int nelems, int dir, struct dma_attrs *attrs) + int nelems, enum dma_data_direction dir, struct dma_attrs *attrs) { struct scatterlist *sg; int i; @@ -858,11 +848,11 @@ swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl, BUG_ON(dir == DMA_NONE); for_each_sg(sgl, sg, nelems, i) { - if (sg->dma_address != swiotlb_virt_to_bus(hwdev, sg_virt(sg))) + if (sg->dma_address != swiotlb_phys_to_bus(hwdev, sg_phys(sg))) unmap_single(hwdev, swiotlb_bus_to_virt(sg->dma_address), sg->dma_length, dir); else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(sg_virt(sg), sg->dma_length); + dma_mark_clean(swiotlb_bus_to_virt(sg->dma_address), sg->dma_length); } } EXPORT_SYMBOL(swiotlb_unmap_sg_attrs); @@ -892,17 +882,17 @@ swiotlb_sync_sg(struct device *hwdev, struct scatterlist *sgl, BUG_ON(dir == DMA_NONE); for_each_sg(sgl, sg, nelems, i) { - if (sg->dma_address != swiotlb_virt_to_bus(hwdev, sg_virt(sg))) + if (sg->dma_address != swiotlb_phys_to_bus(hwdev, sg_phys(sg))) sync_single(hwdev, swiotlb_bus_to_virt(sg->dma_address), sg->dma_length, dir, target); else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(sg_virt(sg), sg->dma_length); + dma_mark_clean(swiotlb_bus_to_virt(sg->dma_address), sg->dma_length); } } void swiotlb_sync_sg_for_cpu(struct device *hwdev, struct scatterlist *sg, - int nelems, int dir) + int nelems, enum dma_data_direction dir) { swiotlb_sync_sg(hwdev, sg, nelems, dir, SYNC_FOR_CPU); } @@ -910,7 +900,7 @@ EXPORT_SYMBOL(swiotlb_sync_sg_for_cpu); void swiotlb_sync_sg_for_device(struct device *hwdev, struct scatterlist *sg, - int nelems, int dir) + int nelems, enum dma_data_direction dir) { swiotlb_sync_sg(hwdev, sg, nelems, dir, SYNC_FOR_DEVICE); } |