Linaro-mm-sig

linaro-mm-sig@lists.linaro.org

7 participants
7696 discussions

Re: [Linaro-mm-sig] [PATCH 8/8] ARM: S5PV210: example of CMA private area for FIMC device on Goni board

by Aguirre, Sergio

Hi Marek/Kyungmin, On Fri, Aug 19, 2011 at 04:27:44PM +0200, Marek Szyprowski wrote: > This patch is an example how device private CMA area can be activated. > It creates one CMA region and assigns it to the first s5p-fimc device on > Samsung Goni S5PC110 board. > > Signed-off-by: Marek Szyprowski <m.szyprowski(a)samsung.com> > Signed-off-by: Kyungmin Park <kyungmin.park(a)samsung.com> > --- > arch/arm/mach-s5pv210/mach-goni.c | 4 ++++ > 1 files changed, 4 insertions(+), 0 deletions(-) > diff --git a/arch/arm/mach-s5pv210/mach-goni.c b/arch/arm/mach-s5pv210/mach-goni.c > index 14578f5..f766c45 100644 > --- a/arch/arm/mach-s5pv210/mach-goni.c > +++ b/arch/arm/mach-s5pv210/mach-goni.c > @@ -26,6 +26,7 @@ > #include <linux/input.h> > #include <linux/gpio.h> > #include <linux/interrupt.h> > +#include <linux/dma-contiguous.h> > > #include <asm/mach/arch.h> > #include <asm/mach/map.h> > @@ -857,6 +858,9 @@ static void __init goni_map_io(void) > static void __init goni_reserve(void) > { > s5p_mfc_reserve_mem(0x43000000, 8 << 20, 0x51000000, 8 << 20); > + > + /* Create private 16MiB contiguous memory area for s5p-fimc.0 device */ > + dma_declare_contiguous(&s5p_device_fimc0.dev, 16*SZ_1M, 0); This is broken, since according to patch #0006, dma_declare_contiguous requires a 4th param (limit) which you're not providing here. Regards, Sergio > } > > static void __init goni_machine_init(void) > -- > 1.7.1.569.g6f426

14 years, 11 months

[PATCHv14 0/9] Contiguous Memory Allocator

by Marek Szyprowski

Hello again, Like I promissed on Linaro Sprint meeting, yet another round of Contiguous Memory Allocator patches are now available. I hope that this version finally solves all pending issues and can be widely tested in ARM platform as well as finally be merged to -next kernel tree for even more testing. This version provides a solution for complete integration of CMA to DMA mapping subsystem on ARM architecture. The issue caused by double dma pages mapping and possible aliasing in coherent memory mapping has been finally resolved, both for GFP_ATOMIC case (allocations comes from reserved DMA memory pool) and non-GFP_ATOMIC case (allocations comes from CMA managed areas). All patches are prepared for Linux Kernel v3.1-rc1. A few words for these who see CMA for the first time: The Contiguous Memory Allocator (CMA) makes it possible for device drivers to allocate big contiguous chunks of memory after the system has booted. The main difference from the similar frameworks is the fact that CMA allows to transparently reuse memory region reserved for the big chunk allocation as a system memory, so no memory is wasted when no big chunk is allocated. Once the alloc request is issued, the framework will migrate system pages to create a required big chunk of physically contiguous memory. For more information you can refer to nice LWN articles: http://lwn.net/Articles/447405/ and http://lwn.net/Articles/450286/ as well as links to previous versions of the CMA framework. The CMA framework has been initially developed by Michal Nazarewicz at Samsung Poland R&D Center. Since version 9, I've taken over the development, because Michal has left the company. The current version of CMA is a set of helper functions for DMA mapping framework that handles allocation of contiguous memory blocks. The difference between this patchset and Kamezawa's alloc_contig_pages() are: 1. alloc_contig_pages() requires MAX_ORDER alignment of allocations which may be unsuitable for embeded systems where a few MiBs are required. Lack of the requirement on the alignment means that several threads might try to access the same pageblock/page. To prevent this from happening CMA uses a mutex so that only one allocating/releasing function may run at one point. 2. CMA may use its own migratetype (MIGRATE_CMA) which behaves similarly to ZONE_MOVABLE but can be put in arbitrary places. This is required for us since we need to define two disjoint memory ranges inside system RAM. (ie. in two memory banks (do not confuse with nodes)). 3. alloc_contig_pages() scans memory in search for range that could be migrated. CMA on the other hand maintains its own allocator to decide where to allocate memory for device drivers and then tries to migrate pages from that part if needed. This is not strictly required but I somehow feel it might be faster. The integration with ARM DMA-mapping subsystem is done on 2 levels. During early boot memory reserved for contiguous areas are remapped with 2-level page tables. This enables us to change cache attributes of the individual pages from such area on request. Then, DMA mapping subsystem is updated to use dma_alloc_from_contiguous() call instead of alloc_pages() and perform page attributes remapping. Current version have been tested on Samsung S5PC110 based Goni machine and s5p-fimc V4L2 driver. The driver itself uses videobuf2 dma-contig memory allocator, which in turn relies on dma_alloc_coherent() from DMA-mapping subsystem. By integrating CMA with DMA-mapping we managed to get this driver working with CMA without any single change required in the driver or videobuf2-dma-contig allocator. TODO (optional): - implement support for contiguous memory areas placed in HIGHMEM zone Best regards -- Marek Szyprowski Samsung Poland R&D Center Links to previous versions of the patchset: v13: (internal, intentionally not released) v12: <http://www.spinics.net/lists/linux-media/msg35674.html> v11: <http://www.spinics.net/lists/linux-mm/msg21868.html> v10: <http://www.spinics.net/lists/linux-mm/msg20761.html> v9: <http://article.gmane.org/gmane.linux.kernel.mm/60787> v8: <http://article.gmane.org/gmane.linux.kernel.mm/56855> v7: <http://article.gmane.org/gmane.linux.kernel.mm/55626> v6: <http://article.gmane.org/gmane.linux.kernel.mm/55626> v5: (intentionally left out as CMA v5 was identical to CMA v4) v4: <http://article.gmane.org/gmane.linux.kernel.mm/52010> v3: <http://article.gmane.org/gmane.linux.kernel.mm/51573> v2: <http://article.gmane.org/gmane.linux.kernel.mm/50986> v1: <http://article.gmane.org/gmane.linux.kernel.mm/50669> Changelog: v14: 1. Merged with "ARM: DMA: steal memory for DMA coherent mappings" patch, added support for GFP_ATOMIC allocations. 2. Added checks for NULL device pointer v13: (internal, intentionally not released) v12: 1. Fixed 2 nasty bugs in dma-contiguous allocator: - alignment argument was not passed correctly - range for dma_release_from_contiguous was not checked correctly 2. Added support for architecture specfic dma_contiguous_early_fixup() function 3. CMA and DMA-mapping integration for ARM architechture has been rewritten to take care of the memory aliasing issue that might happen for newer ARM CPUs (mapping of the same pages with different cache attributes is forbidden). TODO: add support for GFP_ATOMIC allocations basing on the "ARM: DMA: steal memory for DMA coherent mappings" patch and implement support for contiguous memory areas that are placed in HIGHMEM zone v11: 1. Removed genalloc usage and replaced it with direct calls to bitmap_* functions, dropped patches that are not needed anymore (genalloc extensions) 2. Moved all contiguous area management code from mm/cma.c to drivers/base/dma-contiguous.c 3. Renamed cm_alloc/free to dma_alloc/release_from_contiguous 4. Introduced global, system wide (default) contiguous area configured with kernel config and kernel cmdline parameters 5. Simplified initialization to just one function: dma_declare_contiguous() 6. Added example of device private memory contiguous area v10: 1. Rebased onto 3.0-rc2 and resolved all conflicts 2. Simplified CMA to be just a pure memory allocator, for use with platfrom/bus specific subsystems, like dma-mapping. Removed all device specific functions are calls. 3. Integrated with ARM DMA-mapping subsystem. 4. Code cleanup here and there. 5. Removed private context support. v9: 1. Rebased onto 2.6.39-rc1 and resolved all conflicts 2. Fixed a bunch of nasty bugs that happened when the allocation failed (mainly kernel oops due to NULL ptr dereference). 3. Introduced testing code: cma-regions compatibility layer and videobuf2-cma memory allocator module. v8: 1. The alloc_contig_range() function has now been separated from CMA and put in page_allocator.c. This function tries to migrate all LRU pages in specified range and then allocate the range using alloc_contig_freed_pages(). 2. Support for MIGRATE_CMA has been separated from the CMA code. I have not tested if CMA works with ZONE_MOVABLE but I see no reasons why it shouldn't. 3. I have added a @private argument when creating CMA contexts so that one can reserve memory and not share it with the rest of the system. This way, CMA acts only as allocation algorithm. v7: 1. A lot of functionality that handled driver->allocator_context mapping has been removed from the patchset. This is not to say that this code is not needed, it's just not worth posting everything in one patchset. Currently, CMA is "just" an allocator. It uses it's own migratetype (MIGRATE_CMA) for defining ranges of pageblokcs which behave just like ZONE_MOVABLE but dispite the latter can be put in arbitrary places. 2. The migration code that was introduced in the previous version actually started working. v6: 1. Most importantly, v6 introduces support for memory migration. The implementation is not yet complete though. Migration support means that when CMA is not using memory reserved for it, page allocator can allocate pages from it. When CMA wants to use the memory, the pages have to be moved and/or evicted as to make room for CMA. To make it possible it must be guaranteed that only movable and reclaimable pages are allocated in CMA controlled regions. This is done by introducing a MIGRATE_CMA migrate type that guarantees exactly that. Some of the migration code is "borrowed" from Kamezawa Hiroyuki's alloc_contig_pages() implementation. The main difference is that thanks to MIGRATE_CMA migrate type CMA assumes that memory controlled by CMA are is always movable or reclaimable so that it makes allocation decisions regardless of the whether some pages are actually allocated and migrates them if needed. The most interesting patches from the patchset that implement the functionality are: 09/13: mm: alloc_contig_free_pages() added 10/13: mm: MIGRATE_CMA migration type added 11/13: mm: MIGRATE_CMA isolation functions added 12/13: mm: cma: Migration support added [wip] Currently, kernel panics in some situations which I am trying to investigate. 2. cma_pin() and cma_unpin() functions has been added (after a conversation with Johan Mossberg). The idea is that whenever hardware does not use the memory (no transaction is on) the chunk can be moved around. This would allow defragmentation to be implemented if desired. No defragmentation algorithm is provided at this time. 3. Sysfs support has been replaced with debugfs. I always felt unsure about the sysfs interface and when Greg KH pointed it out I finally got to rewrite it to debugfs. v5: (intentionally left out as CMA v5 was identical to CMA v4) v4: 1. The "asterisk" flag has been removed in favour of requiring that platform will provide a "*=<regions>" rule in the map attribute. 2. The terminology has been changed slightly renaming "kind" to "type" of memory. In the previous revisions, the documentation indicated that device drivers define memory kinds and now, v3: 1. The command line parameters have been removed (and moved to a separate patch, the fourth one). As a consequence, the cma_set_defaults() function has been changed -- it no longer accepts a string with list of regions but an array of regions. 2. The "asterisk" attribute has been removed. Now, each region has an "asterisk" flag which lets one specify whether this region should by considered "asterisk" region. 3. SysFS support has been moved to a separate patch (the third one in the series) and now also includes list of regions. v2: 1. The "cma_map" command line have been removed. In exchange, a SysFS entry has been created under kernel/mm/contiguous. The intended way of specifying the attributes is a cma_set_defaults() function called by platform initialisation code. "regions" attribute (the string specified by "cma" command line parameter) can be overwritten with command line parameter; the other attributes can be changed during run-time using the SysFS entries. 2. The behaviour of the "map" attribute has been modified slightly. Currently, if no rule matches given device it is assigned regions specified by the "asterisk" attribute. It is by default built from the region names given in "regions" attribute. 3. Devices can register private regions as well as regions that can be shared but are not reserved using standard CMA mechanisms. A private region has no name and can be accessed only by devices that have the pointer to it. 4. The way allocators are registered has changed. Currently, a cma_allocator_register() function is used for that purpose. Moreover, allocators are attached to regions the first time memory is registered from the region or when allocator is registered which means that allocators can be dynamic modules that are loaded after the kernel booted (of course, it won't be possible to allocate a chunk of memory from a region if allocator is not loaded). 5. Index of new functions: +static inline dma_addr_t __must_check +cma_alloc_from(const char *regions, size_t size, + dma_addr_t alignment) +static inline int +cma_info_about(struct cma_info *info, const const char *regions) +int __must_check cma_region_register(struct cma_region *reg); +dma_addr_t __must_check +cma_alloc_from_region(struct cma_region *reg, + size_t size, dma_addr_t alignment); +static inline dma_addr_t __must_check +cma_alloc_from(const char *regions, + size_t size, dma_addr_t alignment); +int cma_allocator_register(struct cma_allocator *alloc); Patches in this patchset: mm: move some functions from memory_hotplug.c to page_isolation.c mm: alloc_contig_freed_pages() added Code "stolen" from Kamezawa. The first patch just moves code around and the second provide function for "allocates" already freed memory. mm: alloc_contig_range() added This is what Kamezawa asked: a function that tries to migrate all pages from given range and then use alloc_contig_freed_pages() (defined by the previous commit) to allocate those pages. mm: MIGRATE_CMA migration type added mm: MIGRATE_CMA isolation functions added Introduction of the new migratetype and support for it in CMA. MIGRATE_CMA works similar to ZONE_MOVABLE expect almost any memory range can be marked as one. mm: cma: Contiguous Memory Allocator added The code CMA code. Manages CMA contexts and performs memory allocations. ARM: DMA: steal memory for DMA coherent mappings Ancillary patch with significant DMA-mapping subsystem redesing, resolves double mapping issue by creating DMA exclusive memory area. ARM: integrate CMA with dma-mapping subsystem Main client of CMA frame work. CMA serves as a alloc_pages() replacement. ARM: S5PV210: example of CMA private area for FIMC device on Goni board Example of platform/board specific code that creates cma context and assigns it to particular device. Patch summary: KAMEZAWA Hiroyuki (2): mm: move some functions from memory_hotplug.c to page_isolation.c mm: alloc_contig_freed_pages() added Marek Szyprowski (3): drivers: add Contiguous Memory Allocator ARM: integrate CMA with DMA-mapping subsystem ARM: S5PV210: example of CMA private area for FIMC device on Goni board Michal Nazarewicz (3): mm: alloc_contig_range() added mm: MIGRATE_CMA migration type added mm: MIGRATE_CMA isolation functions added Russell King (1): ARM: DMA: steal memory for DMA coherent mappings arch/Kconfig | 3 + arch/arm/Kconfig | 1 + arch/arm/include/asm/device.h | 3 + arch/arm/include/asm/dma-contiguous.h | 33 +++ arch/arm/include/asm/dma-mapping.h | 3 +- arch/arm/include/asm/mach/map.h | 3 + arch/arm/include/asm/memory.h | 7 + arch/arm/mach-s5pv210/Kconfig | 1 + arch/arm/mach-s5pv210/mach-goni.c | 4 + arch/arm/mm/dma-mapping.c | 462 ++++++++++++++++++++------------- arch/arm/mm/init.c | 4 + arch/arm/mm/mm.h | 5 + arch/arm/mm/mmu.c | 53 +++- drivers/base/Kconfig | 77 ++++++ drivers/base/Makefile | 1 + drivers/base/dma-contiguous.c | 396 ++++++++++++++++++++++++++++ include/linux/dma-contiguous.h | 106 ++++++++ include/linux/mmzone.h | 41 +++- include/linux/page-isolation.h | 51 +++- mm/Kconfig | 8 +- mm/compaction.c | 10 + mm/memory_hotplug.c | 111 -------- mm/page_alloc.c | 287 +++++++++++++++++++-- mm/page_isolation.c | 129 +++++++++- 24 files changed, 1449 insertions(+), 350 deletions(-) create mode 100644 arch/arm/include/asm/dma-contiguous.h create mode 100644 drivers/base/dma-contiguous.c create mode 100644 include/linux/dma-contiguous.h -- 1.7.1.569.g6f426

14 years, 11 months

[RFC] ARM: dma_map|unmap_sg plus iommu

by Ramirez Luna, Omar

Hi, I know it is very early but here it is a tryout of the dma_map_sg and dma_unmap_sg with iommu, I made it to roughly understand what is needed to remove drivers/omap-iovmm.c (which is a virtual memory manager implementation on top of omap iommu driver). This patch is placed on top of Marek Szyprowsk initial work: ARM: DMA-mapping & IOMMU integration http://thread.gmane.org/gmane.linux.kernel.mm/63727/ It was tested on an OMAP zoom3 platform and tidspbridge driver. The patch is used to map user space buffers to dsp's iommu, get_user_pages is used to form the sg list that will be passed to dma_map_sg. While at it, I bumped into some issues that I would like to get some feedback or know if they are being considered: 1. There is no way to keep track of what virtual address are being mapped in the scatterlist, which we need to propagate to the dsp, in order that it knows where does the buffers start and end on its virtual address space. I ended up adding an iov_address to scatterlist which if accepted should be toggled/affected by the selection of CONFIG_IOMMU_API. 2. tidspbridge driver sometimes needs to map a physical address into a fixed virtual address (i.e. the start of a firmware section is expected to be at dsp va 0x20000000), there is no straight forward way to do this with the dma api given that it only expects to receive a cpu_addr, a sg or a page, by adding iov_address I could pass phys and iov addresses in a sg and overcome this limitation, but, these addresses belong to: 2a. Shared memory between ARM and DSP: this memory is allocated through memblock API which takes it out of kernel control to be later ioremap'd and iommu map'd to the dsp (this because a non-cacheable requirement), so, these physical addresses doesn't have a linear virtual address translation, which is what dma api expects. 2b. Bus addresses: of dsp peripherals which are also ioremap'd and affected by the same thing. So: kmemcheck_mark_initialized(sg_virt(s), s->length); sg_virt might be returning a wrong virtual address, which is different to what ioremap returns. I leave the code below and appreciate any comments or feedback Regards, Omar --- arch/arm/mm/dma-mapping.c | 68 +++++++++++++++++++++++++++++++++++++ drivers/iommu/omap-iommu.c | 9 ++++- include/asm-generic/scatterlist.h | 3 ++ 3 files changed, 79 insertions(+), 1 deletions(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index b6397c1..2cc4853 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1318,10 +1318,78 @@ void arm_iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr, mutex_unlock(&mapping->lock); } +int arm_iommu_map_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir) +{ + struct dma_iommu_mapping *mapping = dev->archdata.mapping; + struct scatterlist *s; + dma_addr_t iova; + size_t size = 0; + int i, j; + + BUG_ON(!valid_dma_direction(dir)); + + /* XXX do not assume al ents of PAGE_SIZE */ + size = nents * PAGE_SIZE; + iova = gen_pool_alloc(mapping->pool, size); + if (iova == 0) + return 0; + + for_each_sg(sg, s, nents, i) { + int ret; + unsigned int phys = page_to_phys(sg_page(s)); + + /* XXX Add arch flags */ + ret = iommu_map(mapping->domain, iova, phys, 0, 0); + if (ret < 0) + goto bad_mapping; + + s->iov_address = iova; + iova += PAGE_SIZE; + + /* XXX do something on error to clean iommu map*/ + s->dma_address = __dma_map_page(dev, sg_page(s), s->offset, + s->length, dir); + if (dma_mapping_error(dev, s->dma_address)) + goto bad_mapping; + } + debug_dma_map_sg(dev, sg, nents, nents, dir); + return nents; + + bad_mapping: + for_each_sg(sg, s, i, j) + __dma_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir); + return 0; + +} + +void arm_iommu_unmap_sg(struct device *dev, struct scatterlist *sg, int nents, + enum dma_data_direction dir) +{ + struct dma_iommu_mapping *mapping = dev->archdata.mapping; + dma_addr_t iova = sg_iov_address(sg); + struct scatterlist *s; + size_t size = 0; + int i; + + debug_dma_unmap_sg(dev, sg, nents, dir); + + for_each_sg(sg, s, nents, i) { + __dma_unmap_page(dev, sg_dma_address(s), sg_dma_len(s), dir); + iommu_unmap(mapping->domain, sg_iov_address(s), 0); + } + + size = nents * PAGE_SIZE; + gen_pool_free(mapping->pool, iova, size); +} + + struct arm_dma_map_ops iommu_ops = { .alloc_attrs = arm_iommu_alloc_attrs, .free_attrs = arm_iommu_free_attrs, .mmap_attrs = arm_iommu_mmap_attrs, + .map_sg = arm_iommu_map_sg, + .unmap_sg = arm_iommu_unmap_sg, }; EXPORT_SYMBOL_GPL(iommu_ops); diff --git a/drivers/iommu/omap-iommu.c b/drivers/iommu/omap-iommu.c index 9b21b80..6b2a3e1 100644 --- a/drivers/iommu/omap-iommu.c +++ b/drivers/iommu/omap-iommu.c @@ -22,6 +22,7 @@ #include <linux/mutex.h> #include <asm/cacheflush.h> +#include <asm/dma-iommu.h> #include <plat/iommu.h> #include <plat/iopgtable.h> @@ -879,9 +880,15 @@ EXPORT_SYMBOL_GPL(iommu_set_da_range); */ struct device *omap_find_iommu_device(const char *name) { - return driver_find_device(&omap_iommu_driver.driver, NULL, + struct device *dev; + + dev = driver_find_device(&omap_iommu_driver.driver, NULL, (void *)name, device_match_by_alias); + + arm_iommu_assign_device(dev, 0x204f0000, 0x304f0000); + + return dev; } EXPORT_SYMBOL_GPL(omap_find_iommu_device); diff --git a/include/asm-generic/scatterlist.h b/include/asm-generic/scatterlist.h index 5de0735..831d626 100644 --- a/include/asm-generic/scatterlist.h +++ b/include/asm-generic/scatterlist.h @@ -11,6 +11,7 @@ struct scatterlist { unsigned int offset; unsigned int length; dma_addr_t dma_address; + dma_addr_t iov_address; #ifdef CONFIG_NEED_SG_DMA_LENGTH unsigned int dma_length; #endif @@ -25,6 +26,8 @@ struct scatterlist { */ #define sg_dma_address(sg) ((sg)->dma_address) +#define sg_iov_address(sg) ((sg)->iov_address) + #ifdef CONFIG_NEED_SG_DMA_LENGTH #define sg_dma_len(sg) ((sg)->dma_length) #else -- 1.7.0.4

14 years, 11 months

Buffer sharing proof-of-concept

by Marek Szyprowski

Hello Everyone, This patchset introduces the proof-of-concept infrastructure for buffer sharing between multiple devices using file descriptors. The infrastructure has been integrated with V4L2 framework, more specifically videobuf2 and two S5P drivers FIMC (capture interface) and TV drivers, but it can be easily used by other kernel subsystems, like DRI. In this patch the buffer object has been simplified to absolute minimum - it contains only the buffer physical address (only physically contiguous buffers are supported), but this can be easily extended to complete scatter list in the future. Best regards -- Marek Szyprowski Samsung Poland R&D Center

14 years, 12 months

summit dial-in update

by Jesse Barker

Hi all, For those that tried to participate remotely yesterday, I have a couple of updates: 1) We've added an extra microphone in the middle of the room in the hopes of picking up the sound better. 2) We've gotten the room signed on to the correct IRC channel (#linaro-corpus). cheers, Jesse

15 years

dial-in for cambourne sessions

by Jesse Barker

Hi all, Sorry for the late notice, but it turns out we have remote access for the memory management sessions at the Linaro Connect in Cambourne this week,, starting right now today, same time tomorrow and Wednesday. Dial-in details are available at http://connect.linaro.org/events/event/linux-on-arm/, the room is Corpus Christi, and the channel for the room is #connect-corpus. cheers, Jesse

15 years

[PATCH/RFC 0/8] ARM: DMA-mapping framework redesign

by Marek Szyprowski

Hello, This patch series is a continuation of my works on implementing generic IOMMU support in DMA mapping framework for ARM architecture. Now I focused on the DMA mapping framework itself. It turned out that adding support for common dma_map_ops structure was not that hard as I initally thought. After some modification most of the code fits really well to the generic dma_map_ops methods. The only change required to dma_map_ops is a new alloc function. During the discussion on Linaro Memory Management meeting in Budapest we got the idea that we can have only one alloc/free/mmap function with additional attributes argument. This way all different kinds of architecture specific buffer mappings can be hidden behind the attributes without the need of creating several versions of dma_alloc_ function. I also noticed that the dma_alloc_noncoherent() function can be also implemented this way with DMA_ATTRIB_NON_COHERENT attribute. Systems that just defines dma_alloc_noncoherent as dma_alloc_coherent will just ignore such attribute. Another good use case for alloc methods with attributes is the possibility to allocate buffer without a valid kernel mapping. There are a number of drivers (mainly V4L2 and ALSA) that only exports the DMA buffers to user space. Such drivers don't touch the buffer data at all. For such buffers we can avoid the creation of a mapping in kernel virtual address space, saving precious vmalloc area. Such buffers might be allocated once a new attribute DMA_ATTRIB_NO_KERNEL_MAPPING. All the changes introduced in this patch series are intended to prepare a good ground for upcoming generic IOMMU integration to DMA mapping framework on ARM architecture. For more information about proof-of-concept IOMMU implementation in DMA mapping framework, please refer to my previous set of patches: http://www.spinics.net/lists/linux-mm/msg19856.html I've tried to split the redesign into a set of single-step changes for easier review and understanding. If there is anything that needs further clarification, please don't hesitate to ask. The patches are prepared on top of Linux Kernel v3.0-rc3. The proposed changes have been tested on Samsung Exynos4 platform. I've also tested dmabounce code (by manually registering support for DMA bounce for some of the devices available on my board), although my hardware have no such strict requirements. Would be great if one could test my patches on different ARM architectures to check if I didn't break anything. Best regards -- Marek Szyprowski Samsung Poland R&D Center Patch summary: Marek Szyprowski (8): ARM: dma-mapping: remove offset parameter to prepare for generic dma_ops ARM: dma-mapping: implement dma_map_single on top of dma_map_page ARM: dma-mapping: use asm-generic/dma-mapping-common.h ARM: dma-mapping: implement dma sg methods on top of generic dma ops ARM: dma-mapping: move all dma bounce code to separate dma ops structure ARM: dma-mapping: remove redundant code and cleanup common: dma-mapping: change alloc/free_coherent method to more generic alloc/free_attrs ARM: dma-mapping: use alloc, mmap, free from dma_ops arch/arm/Kconfig | 1 + arch/arm/common/dmabounce.c | 112 +++-- arch/arm/include/asm/device.h | 1 + arch/arm/include/asm/dma-mapping.h | 835 +++++++++++++----------------------- arch/arm/mm/dma-mapping.c | 278 +++++++------ include/linux/dma-attrs.h | 1 + include/linux/dma-mapping.h | 13 +- 7 files changed, 539 insertions(+), 702 deletions(-) rewrite arch/arm/include/asm/dma-mapping.h (66%) -- 1.7.1.569.g6f426

15 years

[PATCHv12 0/8] Contiguous Memory Allocator

by Marek Szyprowski

Hello everyone, This is yet another round of Contiguous Memory Allocator patches. Now I focused mainly on the integration of CMA to DMA mapping subsystem on ARM architecture. In this version I've tried to solve the issue of the aliasing in coherent memory mapping that was present in earlier versions of DMA mapping framework. The proposed solution should be considered as a proof-of-concept. Right now it doesn't support GFP_ATOMIC allocations. Support for them is on my TODO list and will be implemented on top of the "ARM: DMA: steal memory for DMA coherent mappings" patch by Russell King. A few words for these who see CMA for the first time: The Contiguous Memory Allocator (CMA) makes it possible for device drivers to allocate big contiguous chunks of memory after the system has booted. The main difference from the similar frameworks is the fact that CMA allows to transparently reuse memory region reserved for the big chunk allocation as a system memory, so no memory is wasted when no big chunk is allocated. Once the alloc request is issued, the framework will migrate system pages to create a required big chunk of physically contiguous memory. For more information you can refer to nice LWN articles: http://lwn.net/Articles/447405/ and http://lwn.net/Articles/450286/ as well as links to previous versions of the CMA framework. The CMA framework has been initially developed by Michal Nazarewicz at Samsung Poland R&D Center. Since version 9, I've taken over the development, because Michal has left the company. The current version of CMA is a set of helper functions for DMA mapping framework that handles allocation of contiguous memory blocks. The difference between this patchset and Kamezawa's alloc_contig_pages() are: 1. alloc_contig_pages() requires MAX_ORDER alignment of allocations which may be unsuitable for embeded systems where a few MiBs are required. Lack of the requirement on the alignment means that several threads might try to access the same pageblock/page. To prevent this from happening CMA uses a mutex so that only one allocating/releasing function may run at one point. 2. CMA may use its own migratetype (MIGRATE_CMA) which behaves similarly to ZONE_MOVABLE but can be put in arbitrary places. This is required for us since we need to define two disjoint memory ranges inside system RAM. (ie. in two memory banks (do not confuse with nodes)). 3. alloc_contig_pages() scans memory in search for range that could be migrated. CMA on the other hand maintains its own allocator to decide where to allocate memory for device drivers and then tries to migrate pages from that part if needed. This is not strictly required but I somehow feel it might be faster. The integration with ARM DMA-mapping subsystem is done on 2 levels. During early boot memory reserved for contiguous areas are remapped with 2-level page tables. This enables us to change cache attributes of the individual pages from such area on request. Then, DMA mapping subsystem is updated to use dma_alloc_from_contiguous() call instead of alloc_pages() and perform page attributes remapping. Current version have been tested on Samsung S5PC110 based Goni machine and s5p-fimc V4L2 driver. The driver itself uses videobuf2 dma-contig memory allocator, which in turn relies on dma_alloc_coherent() from DMA-mapping subsystem. By integrating CMA with DMA-mapping we managed to get this driver working with CMA without any single change required in the driver or videobuf2-dma-contig allocator. TODO: - implement GPF_ATOMIC allocations - implement support for contiguous memory areas placed in HIGHMEM zone Best regards -- Marek Szyprowski Samsung Poland R&D Center Links to previous versions of the patchset: v11: <http://www.spinics.net/lists/linux-mm/msg21868.html> v10: <http://www.spinics.net/lists/linux-mm/msg20761.html> v9: <http://article.gmane.org/gmane.linux.kernel.mm/60787> v8: <http://article.gmane.org/gmane.linux.kernel.mm/56855> v7: <http://article.gmane.org/gmane.linux.kernel.mm/55626> v6: <http://article.gmane.org/gmane.linux.kernel.mm/55626> v5: (intentionally left out as CMA v5 was identical to CMA v4) v4: <http://article.gmane.org/gmane.linux.kernel.mm/52010> v3: <http://article.gmane.org/gmane.linux.kernel.mm/51573> v2: <http://article.gmane.org/gmane.linux.kernel.mm/50986> v1: <http://article.gmane.org/gmane.linux.kernel.mm/50669> Changelog: v12: 1. Fixed 2 nasty bugs in dma-contiguous allocator: - alignment argument was not passed correctly - range for dma_release_from_contiguous was not checked correctly 2. Added support for architecture specfic dma_contiguous_early_fixup() function 3. CMA and DMA-mapping integration for ARM architechture has been rewritten to take care of the memory aliasing issue that might happen for newer ARM CPUs (mapping of the same pages with different cache attributes is forbidden). TODO: add support for GFP_ATOMIC allocations basing on the "ARM: DMA: steal memory for DMA coherent mappings" patch and implement support for contiguous memory areas that are placed in HIGHMEM zone v11: 1. Removed genalloc usage and replaced it with direct calls to bitmap_* functions, dropped patches that are not needed anymore (genalloc extensions) 2. Moved all contiguous area management code from mm/cma.c to drivers/base/dma-contiguous.c 3. Renamed cm_alloc/free to dma_alloc/release_from_contiguous 4. Introduced global, system wide (default) contiguous area configured with kernel config and kernel cmdline parameters 5. Simplified initialization to just one function: dma_declare_contiguous() 6. Added example of device private memory contiguous area v10: 1. Rebased onto 3.0-rc2 and resolved all conflicts 2. Simplified CMA to be just a pure memory allocator, for use with platfrom/bus specific subsystems, like dma-mapping. Removed all device specific functions are calls. 3. Integrated with ARM DMA-mapping subsystem. 4. Code cleanup here and there. 5. Removed private context support. v9: 1. Rebased onto 2.6.39-rc1 and resolved all conflicts 2. Fixed a bunch of nasty bugs that happened when the allocation failed (mainly kernel oops due to NULL ptr dereference). 3. Introduced testing code: cma-regions compatibility layer and videobuf2-cma memory allocator module. v8: 1. The alloc_contig_range() function has now been separated from CMA and put in page_allocator.c. This function tries to migrate all LRU pages in specified range and then allocate the range using alloc_contig_freed_pages(). 2. Support for MIGRATE_CMA has been separated from the CMA code. I have not tested if CMA works with ZONE_MOVABLE but I see no reasons why it shouldn't. 3. I have added a @private argument when creating CMA contexts so that one can reserve memory and not share it with the rest of the system. This way, CMA acts only as allocation algorithm. v7: 1. A lot of functionality that handled driver->allocator_context mapping has been removed from the patchset. This is not to say that this code is not needed, it's just not worth posting everything in one patchset. Currently, CMA is "just" an allocator. It uses it's own migratetype (MIGRATE_CMA) for defining ranges of pageblokcs which behave just like ZONE_MOVABLE but dispite the latter can be put in arbitrary places. 2. The migration code that was introduced in the previous version actually started working. v6: 1. Most importantly, v6 introduces support for memory migration. The implementation is not yet complete though. Migration support means that when CMA is not using memory reserved for it, page allocator can allocate pages from it. When CMA wants to use the memory, the pages have to be moved and/or evicted as to make room for CMA. To make it possible it must be guaranteed that only movable and reclaimable pages are allocated in CMA controlled regions. This is done by introducing a MIGRATE_CMA migrate type that guarantees exactly that. Some of the migration code is "borrowed" from Kamezawa Hiroyuki's alloc_contig_pages() implementation. The main difference is that thanks to MIGRATE_CMA migrate type CMA assumes that memory controlled by CMA are is always movable or reclaimable so that it makes allocation decisions regardless of the whether some pages are actually allocated and migrates them if needed. The most interesting patches from the patchset that implement the functionality are: 09/13: mm: alloc_contig_free_pages() added 10/13: mm: MIGRATE_CMA migration type added 11/13: mm: MIGRATE_CMA isolation functions added 12/13: mm: cma: Migration support added [wip] Currently, kernel panics in some situations which I am trying to investigate. 2. cma_pin() and cma_unpin() functions has been added (after a conversation with Johan Mossberg). The idea is that whenever hardware does not use the memory (no transaction is on) the chunk can be moved around. This would allow defragmentation to be implemented if desired. No defragmentation algorithm is provided at this time. 3. Sysfs support has been replaced with debugfs. I always felt unsure about the sysfs interface and when Greg KH pointed it out I finally got to rewrite it to debugfs. v5: (intentionally left out as CMA v5 was identical to CMA v4) v4: 1. The "asterisk" flag has been removed in favour of requiring that platform will provide a "*=<regions>" rule in the map attribute. 2. The terminology has been changed slightly renaming "kind" to "type" of memory. In the previous revisions, the documentation indicated that device drivers define memory kinds and now, v3: 1. The command line parameters have been removed (and moved to a separate patch, the fourth one). As a consequence, the cma_set_defaults() function has been changed -- it no longer accepts a string with list of regions but an array of regions. 2. The "asterisk" attribute has been removed. Now, each region has an "asterisk" flag which lets one specify whether this region should by considered "asterisk" region. 3. SysFS support has been moved to a separate patch (the third one in the series) and now also includes list of regions. v2: 1. The "cma_map" command line have been removed. In exchange, a SysFS entry has been created under kernel/mm/contiguous. The intended way of specifying the attributes is a cma_set_defaults() function called by platform initialisation code. "regions" attribute (the string specified by "cma" command line parameter) can be overwritten with command line parameter; the other attributes can be changed during run-time using the SysFS entries. 2. The behaviour of the "map" attribute has been modified slightly. Currently, if no rule matches given device it is assigned regions specified by the "asterisk" attribute. It is by default built from the region names given in "regions" attribute. 3. Devices can register private regions as well as regions that can be shared but are not reserved using standard CMA mechanisms. A private region has no name and can be accessed only by devices that have the pointer to it. 4. The way allocators are registered has changed. Currently, a cma_allocator_register() function is used for that purpose. Moreover, allocators are attached to regions the first time memory is registered from the region or when allocator is registered which means that allocators can be dynamic modules that are loaded after the kernel booted (of course, it won't be possible to allocate a chunk of memory from a region if allocator is not loaded). 5. Index of new functions: +static inline dma_addr_t __must_check +cma_alloc_from(const char *regions, size_t size, + dma_addr_t alignment) +static inline int +cma_info_about(struct cma_info *info, const const char *regions) +int __must_check cma_region_register(struct cma_region *reg); +dma_addr_t __must_check +cma_alloc_from_region(struct cma_region *reg, + size_t size, dma_addr_t alignment); +static inline dma_addr_t __must_check +cma_alloc_from(const char *regions, + size_t size, dma_addr_t alignment); +int cma_allocator_register(struct cma_allocator *alloc); Patches in this patchset: mm: move some functions from memory_hotplug.c to page_isolation.c mm: alloc_contig_freed_pages() added Code "stolen" from Kamezawa. The first patch just moves code around and the second provide function for "allocates" already freed memory. mm: alloc_contig_range() added This is what Kamezawa asked: a function that tries to migrate all pages from given range and then use alloc_contig_freed_pages() (defined by the previous commit) to allocate those pages. mm: MIGRATE_CMA migration type added mm: MIGRATE_CMA isolation functions added Introduction of the new migratetype and support for it in CMA. MIGRATE_CMA works similar to ZONE_MOVABLE expect almost any memory range can be marked as one. mm: cma: Contiguous Memory Allocator added The code CMA code. Manages CMA contexts and performs memory allocations. ARM: integrate CMA with dma-mapping subsystem Main client of CMA frame work. CMA serves as a alloc_pages() replacement. ARM: S5PV210: example of CMA private area for FIMC device on Goni board Example of platform/board specific code that creates cma context and assigns it to particular device. Patch summary: KAMEZAWA Hiroyuki (2): mm: move some functions from memory_hotplug.c to page_isolation.c mm: alloc_contig_freed_pages() added Marek Szyprowski (3): drivers: add Contiguous Memory Allocator ARM: integrate CMA with dma-mapping subsystem ARM: S5PV210: example of CMA private area for FIMC device on Goni board Michal Nazarewicz (3): mm: alloc_contig_range() added mm: MIGRATE_CMA migration type added mm: MIGRATE_CMA isolation functions added arch/Kconfig | 3 + arch/arm/Kconfig | 1 + arch/arm/include/asm/device.h | 3 + arch/arm/include/asm/dma-contiguous.h | 33 +++ arch/arm/include/asm/mach/map.h | 1 + arch/arm/mach-s5pv210/Kconfig | 1 + arch/arm/mach-s5pv210/mach-goni.c | 8 + arch/arm/mm/dma-mapping.c | 244 +++----------------- arch/arm/mm/init.c | 3 + arch/arm/mm/mmu.c | 56 +++++- drivers/base/Kconfig | 77 +++++++ drivers/base/Makefile | 1 + drivers/base/dma-contiguous.c | 396 +++++++++++++++++++++++++++++++++ include/linux/dma-contiguous.h | 102 +++++++++ include/linux/mmzone.h | 41 +++- include/linux/page-isolation.h | 51 ++++- mm/Kconfig | 8 +- mm/compaction.c | 10 + mm/memory_hotplug.c | 111 --------- mm/page_alloc.c | 287 ++++++++++++++++++++++-- mm/page_isolation.c | 129 ++++++++++- 21 files changed, 1197 insertions(+), 369 deletions(-) create mode 100644 arch/arm/include/asm/dma-contiguous.h create mode 100644 drivers/base/dma-contiguous.c create mode 100644 include/linux/dma-contiguous.h -- 1.7.1.569.g6f426

15 years

[RFC 2/2] ARM: initial proof-of-concept IOMMU mapper for DMA-mapping

by Ramirez Luna, Omar

Hi Marek, > Add initial proof of concept implementation of DMA-mapping API for > devices that have IOMMU support. Right now only dma_alloc_coherent, > dma_free_coherent and dma_mmap_coherent functions are supported. > > Signed-off-by: Marek Szyprowski <m.szyprowski(a)samsung.com> > Signed-off-by: Kyungmin Park <kyungmin.park(a)samsung.com> > --- ... > diff --git a/arch/arm/include/asm/dma-iommu.h b/arch/arm/include/asm/dma-iommu.h > new file mode 100644 > index 0000000..c246ff3 > --- /dev/null > +++ b/arch/arm/include/asm/dma-iommu.h ... > +int __init arm_iommu_assign_device(struct device *dev, dma_addr_t base, dma_addr_t size); __init causes a panic if the iommu is assigned after boot. In OMAP3 the iommu driver controls isp and dsp address spaces, it is loaded until any of those 2 drivers is needed. > diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c > index f8c6972..b6397c1 100644 > --- a/arch/arm/mm/dma-mapping.c > +++ b/arch/arm/mm/dma-mapping.c ... > +static void *arm_iommu_alloc_attrs(struct device *dev, size_t size, > + dma_addr_t *handle, gfp_t gfp, struct dma_attrs *attrs) > +{ > + struct dma_iommu_mapping *mapping = dev->archdata.mapping; > + struct page **pages; > + void *addr = NULL; > + pgprot_t prot; > + > + if (dma_get_attr(DMA_ATTR_WRITE_COMBINE, attrs)) > + prot = pgprot_writecombine(pgprot_kernel); > + else > + prot = pgprot_dmacoherent(pgprot_kernel); > + > + arm_iommu_init(dev); I found useful to call arm_iommu_init inside arm_iommu_assign_device instead. So, then gen_pool is created only once without the mapping->pool check, instead of relying on the call to ...alloc_attrs, which in my case I never use because I'm implementing iommu_map|unmap_sg functions to see how it goes with the dma mapping. Regards, Omar

15 years

Re: [Linaro-mm-sig] [PATCH 00/10] mm: Linux VM Infrastructure to support Memory Power Management

by amit kachhap

Hi All, In response to the discussion about power savings with memory regions features following measurements are done. Title: On a system with 2GB memory , 1GB is static and the other 1GB in various power states. Brief environment description: Samsung smdk-exynos board is used for this work and full board level power consumption is measured that comprises of cpu and other components. It has 2 DMC's(Dynamic memory controller) with each supporting 1 GB DDR3 memory. Power characteristics of DMC0 controlled memory remain same but memory controlled by DMC1 is changed to 4 different power states. The following numbers describe the maximum power savings measured after executing the software from DMC0 controlled memory which changes the power states of DMC1 controlled memory. Here the actual numbers are not mentioned but the percentage power savings is shown in reference to the change in overall power consumption. The memory region patches are expected to facilitate transition of memory into into one of the following low power states. 1) Percentage power savings when DMC1(1GB) moved to self refresh mode from idle/unaccess mode= 2.69% 2) Percentage power savings when DMC1(1GB) moved to precharge mode from idle/unaccess mode= 3.23% 3) Percentage power savings when DMC1(1GB) clock is gated = 6.32% The above power savings is indicative of the benefits that memory regions could provide in this platform. Thanks & Regards, Amit Daniel Kachhap Samsung India s/w operations, Bangalore On Sat, May 28, 2011 at 1:26 PM, Andrew Morton <akpm(a)linux-foundation.org> wrote: > On Fri, 27 May 2011 18:01:28 +0530 Ankita Garg <ankita(a)in.ibm.com> wrote: > >> This patchset proposes a generic memory regions infrastructure that can be >> used to tag boundaries of memory blocks which belongs to a specific memory >> power management domain and further enable exploitation of platform memory >> power management capabilities. > > A couple of quick thoughts... > > I'm seeing no estimate of how much energy we might save when this work > is completed. But saving energy is the entire point of the entire > patchset! So please spend some time thinking about that and update and > maintain the [patch 0/n] description so others can get some idea of the > benefit we might get from all of this. That estimate should include an > estimate of what proportion of machines are likely to have hardware > which can use this feature and in what timeframe. > > IOW, if it saves one microwatt on 0.001% of machines, not interested ;) > > > Also, all this code appears to be enabled on all machines? So machines > which don't have the requisite hardware still carry any additional > overhead which is added here. I can see that ifdeffing a feature like > this would be ghastly but please also have a think about the > implications of this and add that discussion also. > > If possible, it would be good to think up some microbenchmarks which > probe the worst-case performance impact and describe those and present > the results. So others can gain an understanding of the runtime costs. > > > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel(a)lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel >

15 years

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig