- Linaro-mm-sig - lists.linaro.org

[RFC] ARM: dma-mapping: Add DMA attribute to skip iommu mapping

by Abhinav Kochhar

Adding a new dma attribute which can be used by the platform drivers to avoid creating iommu mappings. In some cases the buffers are allocated by display controller driver using dma alloc apis but are not used for scanout. Though the buffers are allocated by display controller but are only used for sharing among different devices. With this attribute the platform drivers can choose not to create iommu mapping at the time of buffer allocation and only create the mapping when they access this buffer. Change-Id: I2178b3756170982d814e085ca62474d07b616a21 Signed-off-by: Abhinav Kochhar <abhinav.k(a)samsung.com> --- arch/arm/mm/dma-mapping.c | 8 +++++--- include/linux/dma-attrs.h | 1 + 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index c0f0f43..e73003c 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1279,9 +1279,11 @@ static void *arm_iommu_alloc_attrs(struct device *dev, size_t size, if (!pages) return NULL; - *handle = __iommu_create_mapping(dev, pages, size); - if (*handle == DMA_ERROR_CODE) - goto err_buffer; + if (!dma_get_attr(DMA_ATTR_NO_IOMMU_MAPPING, attrs)) { + *handle = __iommu_create_mapping(dev, pages, size); + if (*handle == DMA_ERROR_CODE) + goto err_buffer; + } if (dma_get_attr(DMA_ATTR_NO_KERNEL_MAPPING, attrs)) return pages; diff --git a/include/linux/dma-attrs.h b/include/linux/dma-attrs.h index c8e1831..1f04419 100644 --- a/include/linux/dma-attrs.h +++ b/include/linux/dma-attrs.h @@ -15,6 +15,7 @@ enum dma_attr { DMA_ATTR_WEAK_ORDERING, DMA_ATTR_WRITE_COMBINE, DMA_ATTR_NON_CONSISTENT, + DMA_ATTR_NO_IOMMU_MAPPING, DMA_ATTR_NO_KERNEL_MAPPING, DMA_ATTR_SKIP_CPU_SYNC, DMA_ATTR_FORCE_CONTIGUOUS, -- 1.7.8.6

12 years, 4 months

3
2
0 0

Attendance at ELC San Francisco

by Jesse Barker

Hi all, With all of the attention that the Common Display Framework is getting, I was wondering if it was worth having a BoF discussion at ELC next month in San Francisco. This will be only a couple of weeks after FOSDEM, but given the pace that things seem to be moving, that could be a great opportunity either to have a follow-on discussion, or simply to involve a slightly different cross-section of the community in a face-to-face discussion. If folks could let me know that they'll be at ELC and are interested in a BoF there, I'll look into getting it set up. cheers, Jesse

12 years, 4 months

3
3
0 0

[PATCH] gpu: ion: Add ion_share_dma_buf_kernel

by Johan Mossberg

ion_share_dma_buf_kernel enables you to share ion buffers via dma buf for kernel only use cases. Useful for example when a GPU driver using ion wants to share its output buffers with a 3d party display controller driver supporting dma buf. Signed-off-by: Johan Mossberg <johan.mossberg(a)stericsson.com> --- drivers/gpu/ion/ion.c | 18 ++++++++++++++++-- include/linux/ion.h | 8 ++++++++ 2 files changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c index 748ff7d..31bb5e1 100644 --- a/drivers/gpu/ion/ion.c +++ b/drivers/gpu/ion/ion.c @@ -991,12 +991,12 @@ struct dma_buf_ops dma_buf_ops = { .kunmap = ion_dma_buf_kunmap, }; -int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) +struct dma_buf *ion_share_dma_buf_kernel(struct ion_client *client, + struct ion_handle *handle) { struct ion_buffer *buffer; struct dma_buf *dmabuf; bool valid_handle; - int fd; mutex_lock(&client->lock); valid_handle = ion_handle_validate(client, handle); @@ -1013,6 +1013,20 @@ int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) ion_buffer_put(buffer); return PTR_ERR(dmabuf); } + + return dmabuf; +} +EXPORT_SYMBOL(ion_share_dma_buf_kernel); + +int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) +{ + struct dma_buf *dmabuf; + int fd; + + dmabuf = ion_share_dma_buf_kernel(client, handle); + if (IS_ERR(dmabuf)) + return PTR_ERR(dmabuf); + fd = dma_buf_fd(dmabuf, O_CLOEXEC); if (fd < 0) dma_buf_put(dmabuf); diff --git a/include/linux/ion.h b/include/linux/ion.h index 3ccc75a..cfb831d 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -210,6 +210,14 @@ void *ion_map_kernel(struct ion_client *client, struct ion_handle *handle); void ion_unmap_kernel(struct ion_client *client, struct ion_handle *handle); /** + * ion_share_dma_buf_kernel() - share buffer as dma-buf + * @client: the client + * @handle: the handle + */ +struct dma_buf *ion_share_dma_buf_kernel(struct ion_client *client, + struct ion_handle *buf); + +/** * ion_share_dma_buf() - given an ion client, create a dma-buf fd * @client: the client * @handle: the handle -- 1.8.0

12 years, 4 months

2
3
0 0

Re: [Linaro-mm-sig] [PATCHv2] gpu: ion: Add ion_share_dma_buf_kernel

by Johan Mossberg

Hi Rebecca, Is this patch accepted for ion mainline or do you have more comments/questions? /Johan Mossberg On 12/13/2012 10:24 AM, Johan Mossberg wrote: > ion_share_dma_buf_kernel enables you to share ion buffers via dma buf > for kernel only use cases. Useful for example when a GPU driver using > ion wants to share its output buffers with a 3d party display > controller driver supporting dma buf. > > Signed-off-by: Johan Mossberg <johan.mossberg(a)stericsson.com> > --- > drivers/gpu/ion/ion.c | 22 ++++++++++++++++++---- > include/linux/ion.h | 8 ++++++++ > 2 files changed, 26 insertions(+), 4 deletions(-) > > diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c > index 3872095..e7b0d0b 100644 > --- a/drivers/gpu/ion/ion.c > +++ b/drivers/gpu/ion/ion.c > @@ -955,19 +955,19 @@ struct dma_buf_ops dma_buf_ops = { > .kunmap = ion_dma_buf_kunmap, > }; > > -int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) > +struct dma_buf *ion_share_dma_buf_kernel(struct ion_client *client, > + struct ion_handle *handle) > { > struct ion_buffer *buffer; > struct dma_buf *dmabuf; > bool valid_handle; > - int fd; > > mutex_lock(&client->lock); > valid_handle = ion_handle_validate(client, handle); > mutex_unlock(&client->lock); > if (!valid_handle) { > WARN(1, "%s: invalid handle passed to share.\n", __func__); > - return -EINVAL; > + return ERR_PTR(-EINVAL); > } > > buffer = handle->buffer; > @@ -975,8 +975,22 @@ int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) > dmabuf = dma_buf_export(buffer, &dma_buf_ops, buffer->size, O_RDWR); > if (IS_ERR(dmabuf)) { > ion_buffer_put(buffer); > - return PTR_ERR(dmabuf); > + return dmabuf; > } > + > + return dmabuf; > +} > +EXPORT_SYMBOL(ion_share_dma_buf_kernel); > + > +int ion_share_dma_buf(struct ion_client *client, struct ion_handle *handle) > +{ > + struct dma_buf *dmabuf; > + int fd; > + > + dmabuf = ion_share_dma_buf_kernel(client, handle); > + if (IS_ERR(dmabuf)) > + return PTR_ERR(dmabuf); > + > fd = dma_buf_fd(dmabuf, O_CLOEXEC); > if (fd < 0) > dma_buf_put(dmabuf); > diff --git a/include/linux/ion.h b/include/linux/ion.h > index a7d399c..8720e9b 100644 > --- a/include/linux/ion.h > +++ b/include/linux/ion.h > @@ -205,6 +205,14 @@ void *ion_map_kernel(struct ion_client *client, struct ion_handle *handle); > void ion_unmap_kernel(struct ion_client *client, struct ion_handle *handle); > > /** > + * ion_share_dma_buf_kernel() - share buffer as dma-buf > + * @client: the client > + * @handle: the handle > + */ > +struct dma_buf *ion_share_dma_buf_kernel(struct ion_client *client, > + struct ion_handle *buf); > + > +/** > * ion_share_dma_buf() - given an ion client, create a dma-buf fd > * @client: the client > * @handle: the handle > -- > 1.8.0 > > > _______________________________________________ > Linaro-mm-sig mailing list > Linaro-mm-sig(a)lists.linaro.org > http://lists.linaro.org/mailman/listinfo/linaro-mm-sig > >

12 years, 5 months

2
1
0 0

[PATCH v2] arm: dma mapping: export arm iommu functions

by Prathyush K

This patch adds EXPORT_SYMBOL_GPL calls to the three arm iommu functions - arm_iommu_create_mapping, arm_iommu_free_mapping and arm_iommu_attach_device. These three functions are arm specific wrapper functions for creating/freeing/using an iommu mapping and they are called by various drivers. If any of these drivers need to be built as dynamic modules, these functions need to be exported. Changelog v2: using EXPORT_SYMBOL_GPL as suggested by Marek. Signed-off-by: Prathyush K <prathyush.k(a)samsung.com> --- arch/arm/mm/dma-mapping.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 6b2fb87..226ebcf 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1797,6 +1797,7 @@ err2: err: return ERR_PTR(err); } +EXPORT_SYMBOL_GPL(arm_iommu_create_mapping); static void release_iommu_mapping(struct kref *kref) { @@ -1813,6 +1814,7 @@ void arm_iommu_release_mapping(struct dma_iommu_mapping *mapping) if (mapping) kref_put(&mapping->kref, release_iommu_mapping); } +EXPORT_SYMBOL_GPL(arm_iommu_release_mapping); /** * arm_iommu_attach_device @@ -1841,5 +1843,6 @@ int arm_iommu_attach_device(struct device *dev, pr_debug("Attached IOMMU controller to %s device.\n", dev_name(dev)); return 0; } +EXPORT_SYMBOL_GPL(arm_iommu_attach_device); #endif -- 1.8.0

12 years, 5 months

1
0
0 0

[RFC] ARM: DMA-Mapping: add a new attribute to clear buffer

by daeinki＠gmail.com

From: Inki Dae <inki.dae(a)samsung.com> This patch adds a new attribute, DMA_ATTR_SKIP_BUFFER_CLEAR to skip buffer clearing. The buffer clearing also flushes CPU cache so this operation has performance deterioration a little bit. With this patch, allocated buffer region is cleared as default. So if you want to skip the buffer clearing, just set this attribute. But this flag should be used carefully because this use might get access to some vulnerable content such as security data. So with this patch, we make sure that all pages will be somehow cleared before exposing to userspace. For example, let's say that the security data had been stored in some memory and freed without clearing it. And then malicious process allocated the region though some buffer allocator such as gem and ion without clearing it, and requested blit operation with cleared another buffer though gpu or other drivers. At this time, the malicious process could access the security data. Signed-off-by: Inki Dae <inki.dae(a)samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park(a)samsung.com> --- arch/arm/mm/dma-mapping.c | 6 ++++-- include/linux/dma-attrs.h | 1 + 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 6b2fb87..fbe9dff 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1058,7 +1058,8 @@ static struct page **__iommu_alloc_buffer(struct device *dev, size_t size, if (!page) goto error; - __dma_clear_buffer(page, size); + if (!dma_get_attr(DMA_ATTR_SKIP_BUFFER_CLEAR, attrs)) + __dma_clear_buffer(page, size); for (i = 0; i < count; i++) pages[i] = page + i; @@ -1082,7 +1083,8 @@ static struct page **__iommu_alloc_buffer(struct device *dev, size_t size, pages[i + j] = pages[i] + j; } - __dma_clear_buffer(pages[i], PAGE_SIZE << order); + if (!dma_get_attr(DMA_ATTR_SKIP_BUFFER_CLEAR, attrs)) + __dma_clear_buffer(pages[i], PAGE_SIZE << order); i += 1 << order; count -= 1 << order; } diff --git a/include/linux/dma-attrs.h b/include/linux/dma-attrs.h index c8e1831..2592c05 100644 --- a/include/linux/dma-attrs.h +++ b/include/linux/dma-attrs.h @@ -18,6 +18,7 @@ enum dma_attr { DMA_ATTR_NO_KERNEL_MAPPING, DMA_ATTR_SKIP_CPU_SYNC, DMA_ATTR_FORCE_CONTIGUOUS, + DMA_ATTR_SKIP_BUFFER_CLEAR, DMA_ATTR_MAX, }; -- 1.7.4.1

12 years, 5 months

3
2
0 0

[PATCH] arm: dma mapping: export arm iommu functions

by Prathyush K

This patch adds EXPORT_SYMBOL calls to the three arm iommu functions - arm_iommu_create_mapping, arm_iommu_free_mapping and arm_iommu_attach_device. These functions can now be called from dynamic modules. Signed-off-by: Prathyush K <prathyush.k(a)samsung.com> --- arch/arm/mm/dma-mapping.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 6b2fb87..c0f0f43 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -1797,6 +1797,7 @@ err2: err: return ERR_PTR(err); } +EXPORT_SYMBOL(arm_iommu_create_mapping); static void release_iommu_mapping(struct kref *kref) { @@ -1813,6 +1814,7 @@ void arm_iommu_release_mapping(struct dma_iommu_mapping *mapping) if (mapping) kref_put(&mapping->kref, release_iommu_mapping); } +EXPORT_SYMBOL(arm_iommu_release_mapping); /** * arm_iommu_attach_device @@ -1841,5 +1843,6 @@ int arm_iommu_attach_device(struct device *dev, pr_debug("Attached IOMMU controller to %s device.\n", dev_name(dev)); return 0; } +EXPORT_SYMBOL(arm_iommu_attach_device); #endif -- 1.8.0

12 years, 5 months

5
4
0 0

[PATCH] [RFC] dma-buf: implement vmap refcounting in the interface logic

by Daniel Vetter

All drivers which implement this need to have some sort of refcount to allow concurrent vmap usage. Hence implement this in the dma-buf core. To protect against concurrent calls we need a lock, which potentially causes new funny locking inversions. But this shouldn't be a problem for exporters with statically allocated backing storage, and more dynamic drivers have decent issues already anyway. Inspired by some refactoring patches from Aaron Plattner, who implemented the same idea, but only for drm/prime drivers. v2: Check in dma_buf_release that no dangling vmaps are left. Suggested by Aaron Plattner. We might want to do similar checks for attachments, but that's for another patch. Also fix up ERR_PTR return for vmap. Cc: Aaron Plattner <aplattner(a)nvidia.com> Signed-off-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> --- Compile-tested only - Aaron has been bugging me too a bit too often about this on irc. Cheers, Daniel --- Documentation/dma-buf-sharing.txt | 6 +++++- drivers/base/dma-buf.c | 42 ++++++++++++++++++++++++++++++++++----- include/linux/dma-buf.h | 4 +++- 3 files changed, 45 insertions(+), 7 deletions(-) diff --git a/Documentation/dma-buf-sharing.txt b/Documentation/dma-buf-sharing.txt index 0188903..4966b1b 100644 --- a/Documentation/dma-buf-sharing.txt +++ b/Documentation/dma-buf-sharing.txt @@ -302,7 +302,11 @@ Access to a dma_buf from the kernel context involves three steps: void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) The vmap call can fail if there is no vmap support in the exporter, or if it - runs out of vmalloc space. Fallback to kmap should be implemented. + runs out of vmalloc space. Fallback to kmap should be implemented. Note that + the dma-buf layer keeps a reference count for all vmap access and calls down + into the exporter's vmap function only when no vmapping exists, and only + unmaps it once. Protection against concurrent vmap/vunmap calls is provided + by taking the dma_buf->lock mutex. 3. Finish access diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index a3f79c4..36af5de 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -39,6 +39,8 @@ static int dma_buf_release(struct inode *inode, struct file *file) dmabuf = file->private_data; + BUG_ON(dmabuf->vmapping_counter); + dmabuf->ops->release(dmabuf); kfree(dmabuf); return 0; @@ -482,12 +484,34 @@ EXPORT_SYMBOL_GPL(dma_buf_mmap); */ void *dma_buf_vmap(struct dma_buf *dmabuf) { + void *ptr; + if (WARN_ON(!dmabuf)) return NULL; - if (dmabuf->ops->vmap) - return dmabuf->ops->vmap(dmabuf); - return NULL; + if (!dmabuf->ops->vmap) + return NULL; + + mutex_lock(&dmabuf->lock); + if (dmabuf->vmapping_counter) { + dmabuf->vmapping_counter++; + BUG_ON(!dmabuf->vmap_ptr); + ptr = dmabuf->vmap_ptr; + goto out_unlock; + } + + BUG_ON(dmabuf->vmap_ptr); + + ptr = dmabuf->ops->vmap(dmabuf); + if (IS_ERR_OR_NULL(ptr)) + goto out_unlock; + + dmabuf->vmap_ptr = ptr; + dmabuf->vmapping_counter = 1; + +out_unlock: + mutex_unlock(&dmabuf->lock); + return ptr; } EXPORT_SYMBOL_GPL(dma_buf_vmap); @@ -501,7 +525,15 @@ void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) if (WARN_ON(!dmabuf)) return; - if (dmabuf->ops->vunmap) - dmabuf->ops->vunmap(dmabuf, vaddr); + BUG_ON(!dmabuf->vmap_ptr); + BUG_ON(dmabuf->vmapping_counter > 0); + + mutex_lock(&dmabuf->lock); + if (--dmabuf->vmapping_counter == 0) { + if (dmabuf->ops->vunmap) + dmabuf->ops->vunmap(dmabuf, vaddr); + dmabuf->vmap_ptr = NULL; + } + mutex_unlock(&dmabuf->lock); } EXPORT_SYMBOL_GPL(dma_buf_vunmap); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index bd2e52c..e3bf2f6 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -119,8 +119,10 @@ struct dma_buf { struct file *file; struct list_head attachments; const struct dma_buf_ops *ops; - /* mutex to serialize list manipulation and attach/detach */ + /* mutex to serialize list manipulation, attach/detach and vmap/unmap */ struct mutex lock; + unsigned vmapping_counter; + void *vmap_ptr; void *priv; }; -- 1.7.11.7

12 years, 5 months

3
6
0 0

[GIT PULL]: dma-buf updates for 3.8

by Sumit Semwal

Hi Linus, A fairly small dma-buf pull request for 3.8 - only 2 patches. Could you please pull? Thanks! ~Sumit. The following changes since commit f01af9f85855e38fbd601e033a8eac204cc4cc1c: Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc (2012-12-19 20:31:02 -0800) are available in the git repository at: git://git.linaro.org/people/sumitsemwal/linux-dma-buf.git tags/tag-for-linus-3.8 for you to fetch changes up to ada65c74059f8c104f1b467c126205471634c435: dma-buf: remove fallback for !CONFIG_DMA_SHARED_BUFFER (2012-12-20 12:05:06 +0530) ---------------------------------------------------------------- 3.8: dma-buf minor updates ---------------------------------------------------------------- Maarten Lankhorst (1): dma-buf: remove fallback for !CONFIG_DMA_SHARED_BUFFER Rob Clark (1): dma-buf: might_sleep() in dma_buf_unmap_attachment() drivers/base/dma-buf.c | 2 + include/linux/dma-buf.h | 99 ----------------------------------------------- 2 files changed, 2 insertions(+), 99 deletions(-)

12 years, 5 months

1
0
0 0

[PATCH] dma-buf: Add debugfs support

by sumit.semwal＠ti.com

From: Sumit Semwal <sumit.semwal(a)linaro.org> Add debugfs support to make it easier to print debug information about the dma-buf buffers. Signed-off-by: Sumit Semwal <sumit.semwal(a)ti.com> --- drivers/base/dma-buf.c | 149 +++++++++++++++++++++++++++++++++++++++++++++++ include/linux/dma-buf.h | 6 +- 2 files changed, 154 insertions(+), 1 deletion(-) diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index a3f79c4..49fcf0f 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -27,9 +27,18 @@ #include <linux/dma-buf.h> #include <linux/anon_inodes.h> #include <linux/export.h> +#include <linux/debugfs.h> +#include <linux/seq_file.h> static inline int is_dma_buf_file(struct file *); +struct dma_buf_list { + struct list_head head; + struct mutex lock; +}; + +static struct dma_buf_list db_list; + static int dma_buf_release(struct inode *inode, struct file *file) { struct dma_buf *dmabuf; @@ -40,6 +49,11 @@ static int dma_buf_release(struct inode *inode, struct file *file) dmabuf = file->private_data; dmabuf->ops->release(dmabuf); + + mutex_lock(&db_list.lock); + list_del(&dmabuf->list_node); + mutex_unlock(&db_list.lock); + kfree(dmabuf); return 0; } @@ -120,6 +134,10 @@ struct dma_buf *dma_buf_export(void *priv, const struct dma_buf_ops *ops, mutex_init(&dmabuf->lock); INIT_LIST_HEAD(&dmabuf->attachments); + mutex_lock(&db_list.lock); + list_add(&dmabuf->list_node, &db_list.head); + mutex_unlock(&db_list.lock); + return dmabuf; } EXPORT_SYMBOL_GPL(dma_buf_export); @@ -505,3 +523,134 @@ void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) dmabuf->ops->vunmap(dmabuf, vaddr); } EXPORT_SYMBOL_GPL(dma_buf_vunmap); + +static int dma_buf_init_debugfs(void); +static void dma_buf_uninit_debugfs(void); + +static void __init dma_buf_init(void) +{ + mutex_init(&db_list.lock); + INIT_LIST_HEAD(&db_list.head); + dma_buf_init_debugfs(); +} + +static void __exit dma_buf_deinit(void) +{ + dma_buf_uninit_debugfs(); +} + +#ifdef CONFIG_DEBUG_FS +static int dma_buf_describe(struct seq_file *s) +{ + int ret; + struct dma_buf *buf_obj; + struct dma_buf_attachment *attach_obj; + int count = 0, attach_count; + size_t size = 0; + + ret = mutex_lock_interruptible(&db_list.lock); + + if (ret) + return ret; + + seq_printf(s, "\nDma-buf Objects:\n"); + seq_printf(s, "\tsize\tflags\tmode\tcount\n"); + + list_for_each_entry(buf_obj, &db_list.head, list_node) { + seq_printf(s, "\t"); + + seq_printf(s, "%08zu\t%08x\t%08x\t%08d\n", + buf_obj->size, buf_obj->file->f_flags, + buf_obj->file->f_mode, + buf_obj->file->f_count.counter); + + seq_printf(s, "\t\tAttached Devices:\n"); + attach_count = 0; + + list_for_each_entry(attach_obj, &buf_obj->attachments, node) { + seq_printf(s, "\t\t"); + + seq_printf(s, "%s\n", attach_obj->dev->init_name); + attach_count++; + } + + seq_printf(s, "\n\t\tTotal %d devices attached\n", + attach_count); + + count++; + size += buf_obj->size; + } + + seq_printf(s, "\nTotal %d objects, %zu bytes\n", count, size); + + mutex_unlock(&db_list.lock); + return 0; +} + +static int dma_buf_show(struct seq_file *s, void *unused) +{ + void (*func)(struct seq_file *) = s->private; + func(s); + return 0; +} + +static int dma_buf_debug_open(struct inode *inode, struct file *file) +{ + return single_open(file, dma_buf_show, inode->i_private); +} + +static const struct file_operations dma_buf_debug_fops = { + .open = dma_buf_debug_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static struct dentry *dma_buf_debugfs_dir; + +static int dma_buf_init_debugfs(void) +{ + int err = 0; + dma_buf_debugfs_dir = debugfs_create_dir("dma_buf", NULL); + if (IS_ERR(dma_buf_debugfs_dir)) { + err = PTR_ERR(dma_buf_debugfs_dir); + dma_buf_debugfs_dir = NULL; + return err; + } + + err = dma_buf_debugfs_create_file("bufinfo", dma_buf_describe); + + if (err) + pr_debug("dma_buf: debugfs: failed to create node bufinfo\n"); + + return err; +} + +static void dma_buf_uninit_debugfs(void) +{ + if (dma_buf_debugfs_dir) + debugfs_remove_recursive(dma_buf_debugfs_dir); +} + +int dma_buf_debugfs_create_file(const char *name, + int (*write)(struct seq_file *)) +{ + struct dentry *d; + + d = debugfs_create_file(name, S_IRUGO, dma_buf_debugfs_dir, + write, &dma_buf_debug_fops); + + if (IS_ERR(d)) + return PTR_ERR(d); + + return 0; +} +#else +static inline int dma_buf_init_debugfs(void) +{ + return 0; +} +static inline void dma_buf_uninit_debugfs(void) +{ +} +#endif diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index bd2e52c..160453f 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -112,6 +112,7 @@ struct dma_buf_ops { * @file: file pointer used for sharing buffers across, and for refcounting. * @attachments: list of dma_buf_attachment that denotes all devices attached. * @ops: dma_buf_ops associated with this buffer object. + * @list_node: node for dma_buf accounting and debugging. * @priv: exporter specific private data for this buffer object. */ struct dma_buf { @@ -121,6 +122,8 @@ struct dma_buf { const struct dma_buf_ops *ops; /* mutex to serialize list manipulation and attach/detach */ struct mutex lock; + + struct list_head list_node; void *priv; }; @@ -183,5 +186,6 @@ int dma_buf_mmap(struct dma_buf *, struct vm_area_struct *, unsigned long); void *dma_buf_vmap(struct dma_buf *); void dma_buf_vunmap(struct dma_buf *, void *vaddr); - +int dma_buf_debugfs_create_file(const char *name, + int (*write)(struct seq_file *)); #endif /* __DMA_BUF_H__ */ -- 1.7.10.4

12 years, 5 months

8
11
0 0

drm prime/dma-buf import lifetime issue

by Dave Airlie

So I've gotten back to playing with prime for a day, and found some old intel/radeon tests I had failing, Tracked it down to a lifetime issue with the current code and can think of two fixes, The problem scenario is 1. i915: create gem object 2. i915: export gem object to prime 3. radeon: import gem object 4. close prime fd 5. radeon: unref object 6. i915: unref object So we end up at this point, with a imported buffer record for the dma_buf on the i915 file private. Now if a subsequent test (without closing the drm fd) reallocates the dma_buf with the same address, we can end up seeing that. So why doesn't that reference get cleaned up? So the reference gets added above at step 2, and when radeon unrefs the object, i915 gets the dma-buf release callback, however at that stage we don't actually have the file priv to remove the pointer from, so it dangles there. Possible fixes: a) take a reference on the dma_buf attached to the gem handle when we export it, keep it until the gem handle goes away. I'm unsure if this could create zombie objects, since the dma buf has a reference on the gem object, but since the gem handle is separate to the object it might work. b) don't keep track of dma_buf, keep track of gem objects, when we get a lookup, check inside the gem object, since currently we NULL out the export_dma_buf when we hit the release path, apart from the fact I'm sure the locking is foobar, c) scan all the file privs for all the devices, no. Anyone else any better plans? Dave.

12 years, 5 months

2
2
0 0

[PATCH, resend] dma-buf: remove fallback for !CONFIG_DMA_SHARED_BUFFER

by Maarten Lankhorst

Documentation says that code requiring dma-buf should add it to select, so inline fallbacks are not going to be used. A link error will make it obvious what went wrong, instead of silently doing nothing at runtime. Signed-off-by: Maarten Lankhorst <maarten.lankhorst(a)canonical.com> Reviewed-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> Reviewed-by: Rob Clark <rob.clark(a)linaro.org> --- Resubmit since it seemed to have gotten lost last cycle. include/linux/dma-buf.h | 99 ------------------------------------------------- 1 file changed, 99 deletions(-) diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index eb48f38..bd2e52c 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -156,7 +156,6 @@ static inline void get_dma_buf(struct dma_buf *dmabuf) get_file(dmabuf->file); } -#ifdef CONFIG_DMA_SHARED_BUFFER struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, struct device *dev); void dma_buf_detach(struct dma_buf *dmabuf, @@ -184,103 +183,5 @@ int dma_buf_mmap(struct dma_buf *, struct vm_area_struct *, unsigned long); void *dma_buf_vmap(struct dma_buf *); void dma_buf_vunmap(struct dma_buf *, void *vaddr); -#else - -static inline struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf, - struct device *dev) -{ - return ERR_PTR(-ENODEV); -} - -static inline void dma_buf_detach(struct dma_buf *dmabuf, - struct dma_buf_attachment *dmabuf_attach) -{ - return; -} - -static inline struct dma_buf *dma_buf_export(void *priv, - const struct dma_buf_ops *ops, - size_t size, int flags) -{ - return ERR_PTR(-ENODEV); -} - -static inline int dma_buf_fd(struct dma_buf *dmabuf, int flags) -{ - return -ENODEV; -} - -static inline struct dma_buf *dma_buf_get(int fd) -{ - return ERR_PTR(-ENODEV); -} - -static inline void dma_buf_put(struct dma_buf *dmabuf) -{ - return; -} - -static inline struct sg_table *dma_buf_map_attachment( - struct dma_buf_attachment *attach, enum dma_data_direction write) -{ - return ERR_PTR(-ENODEV); -} - -static inline void dma_buf_unmap_attachment(struct dma_buf_attachment *attach, - struct sg_table *sg, enum dma_data_direction dir) -{ - return; -} - -static inline int dma_buf_begin_cpu_access(struct dma_buf *dmabuf, - size_t start, size_t len, - enum dma_data_direction dir) -{ - return -ENODEV; -} - -static inline void dma_buf_end_cpu_access(struct dma_buf *dmabuf, - size_t start, size_t len, - enum dma_data_direction dir) -{ -} - -static inline void *dma_buf_kmap_atomic(struct dma_buf *dmabuf, - unsigned long pnum) -{ - return NULL; -} - -static inline void dma_buf_kunmap_atomic(struct dma_buf *dmabuf, - unsigned long pnum, void *vaddr) -{ -} - -static inline void *dma_buf_kmap(struct dma_buf *dmabuf, unsigned long pnum) -{ - return NULL; -} - -static inline void dma_buf_kunmap(struct dma_buf *dmabuf, - unsigned long pnum, void *vaddr) -{ -} - -static inline int dma_buf_mmap(struct dma_buf *dmabuf, - struct vm_area_struct *vma, - unsigned long pgoff) -{ - return -ENODEV; -} - -static inline void *dma_buf_vmap(struct dma_buf *dmabuf) -{ - return NULL; -} - -static inline void dma_buf_vunmap(struct dma_buf *dmabuf, void *vaddr) -{ -} -#endif /* CONFIG_DMA_SHARED_BUFFER */ #endif /* __DMA_BUF_H__ */ -- 1.8.0

12 years, 5 months

1
0
0 0

[PATCH] gpu: ion: Update debugfs show functions to show heap specific info

by Nishanth Peethambaran

The debugfs show functions for client and heap were showing info for the heap type instead of showing info of the individual heap. Change-Id: Id5afe7963c8ddfafae1f959ce48dd5c2a5fcca07 Signed-off-by: Nishanth Peethambaran <nishanth(a)broadcom.com> --- drivers/gpu/ion/ion.c | 18 +++++++++--------- include/linux/ion.h | 4 ++-- 2 files changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c index 6aa817a..48cda5d 100644 --- a/drivers/gpu/ion/ion.c +++ b/drivers/gpu/ion/ion.c @@ -413,7 +413,7 @@ struct ion_handle *ion_alloc(struct ion_client *client, size_t len, /* if the client doesn't support this heap type */ if (!((1 << heap->type) & client->heap_mask)) continue; - /* if the caller didn't specify this heap type */ + /* if the caller didn't specify this heap id */ if (!((1 << heap->id) & heap_mask)) continue; buffer = ion_buffer_create(heap, dev, len, align, flags); @@ -597,11 +597,11 @@ static int ion_debug_client_show(struct seq_file *s, void *unused) for (n = rb_first(&client->handles); n; n = rb_next(n)) { struct ion_handle *handle = rb_entry(n, struct ion_handle, node); - enum ion_heap_type type = handle->buffer->heap->type; + int id = handle->buffer->heap->id; - if (!names[type]) - names[type] = handle->buffer->heap->name; - sizes[type] += handle->buffer->size; + if (!names[id]) + names[id] = handle->buffer->heap->name; + sizes[id] += handle->buffer->size; } mutex_unlock(&client->lock); @@ -1176,7 +1176,7 @@ static const struct file_operations ion_fops = { }; static size_t ion_debug_heap_total(struct ion_client *client, - enum ion_heap_type type) + int id) { size_t size = 0; struct rb_node *n; @@ -1186,7 +1186,7 @@ static size_t ion_debug_heap_total(struct ion_client *client, struct ion_handle *handle = rb_entry(n, struct ion_handle, node); - if (handle->buffer->heap->type == type) + if (handle->buffer->heap->id == id) size += handle->buffer->size; } mutex_unlock(&client->lock); @@ -1207,7 +1207,7 @@ static int ion_debug_heap_show(struct seq_file *s, void *unused) for (n = rb_first(&dev->clients); n; n = rb_next(n)) { struct ion_client *client = rb_entry(n, struct ion_client, node); - size_t size = ion_debug_heap_total(client, heap->type); + size_t size = ion_debug_heap_total(client, heap->id); if (!size) continue; if (client->task) { @@ -1228,7 +1228,7 @@ static int ion_debug_heap_show(struct seq_file *s, void *unused) for (n = rb_first(&dev->buffers); n; n = rb_next(n)) { struct ion_buffer *buffer = rb_entry(n, struct ion_buffer, node); - if (buffer->heap->type != heap->type) + if (buffer->heap->id != heap->id) continue; total_size += buffer->size; if (!buffer->handle_count) { diff --git a/include/linux/ion.h b/include/linux/ion.h index a7d399c..d8168fb 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -135,7 +135,7 @@ void ion_client_destroy(struct ion_client *client); * @len: size of the allocation * @align: requested allocation alignment, lots of hardware blocks have * alignment requirements of some kind - * @heap_mask: mask of heaps to allocate from, if multiple bits are set + * @heap_mask: mask of heap ids to allocate from, if multiple bits are set * heaps will be tried in order from lowest to highest order bit * @flags: heap flags, the low 16 bits are consumed by ion, the high 16 * bits are passed on to the respective heap and can be heap @@ -236,7 +236,7 @@ struct ion_handle *ion_import_dma_buf(struct ion_client *client, int fd); * struct ion_allocation_data - metadata passed from userspace for allocations * @len: size of the allocation * @align: required alignment of the allocation - * @heap_mask: mask of heaps to allocate from + * @heap_mask: mask of heap ids to allocate from * @flags: flags passed to heap * @handle: pointer that will be populated with a cookie to use to refer * to this allocation -- 1.7.0.4

12 years, 5 months

1
0
0 0

[PATCH] gpu: ion: Fix incorrect usage of heap id in ion_alloc

by Johan Mossberg

The heap id is compared against the heap mask in ion_alloc which is incorrect, the heap type should be used. Signed-off-by: Johan Mossberg <johan.mossberg(a)stericsson.com> --- drivers/gpu/ion/ion.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c index fc152b9..c811380f 100644 --- a/drivers/gpu/ion/ion.c +++ b/drivers/gpu/ion/ion.c @@ -414,7 +414,7 @@ struct ion_handle *ion_alloc(struct ion_client *client, size_t len, if (!((1 << heap->type) & client->heap_mask)) continue; /* if the caller didn't specify this heap type */ - if (!((1 << heap->id) & heap_mask)) + if (!((1 << heap->type) & heap_mask)) continue; buffer = ion_buffer_create(heap, dev, len, align, flags); if (!IS_ERR_OR_NULL(buffer)) -- 1.8.0

12 years, 5 months

3
9
0 0

[PATCH] gpu: ion: Fix memory leak of sg_table in carveout heap

by Nishanth Peethambaran

Change-Id: I03ce713510a942d5a290ec56c1c7490602632fd9 Signed-off-by: Nishanth Peethambaran <nishanth(a)broadcom.com> --- drivers/gpu/ion/ion_carveout_heap.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/drivers/gpu/ion/ion_carveout_heap.c b/drivers/gpu/ion/ion_carveout_heap.c index ce8d311..e8164a6 100644 --- a/drivers/gpu/ion/ion_carveout_heap.c +++ b/drivers/gpu/ion/ion_carveout_heap.c @@ -107,6 +107,7 @@ void ion_carveout_heap_unmap_dma(struct ion_heap *heap, struct ion_buffer *buffer) { sg_free_table(buffer->sg_table); + kfree(buffer->sg_table); } void *ion_carveout_heap_map_kernel(struct ion_heap *heap, -- 1.7.0.4

12 years, 5 months

1
0
0 0

[PATCH] gpu: ion: Use heap->id as the identifier instead of heap->type

by Nishanth Peethambaran

The heap mask used during client create was a bitmask of heap types and the heap mask used during buffer allocation was a bitmask of heap ids. Now, both of them assume a bitmask of heap ids. The debugfs show functions for client and heap were showing info for the heap type instead of showing info of the individual heap. Change-Id: I9c2c142a63865f3d4250cd681dc5cde9638ca09f Signed-off-by: Nishanth Peethambaran <nishanth(a)broadcom.com> --- drivers/gpu/ion/ion.c | 26 +++++++++++++------------- include/linux/ion.h | 8 ++++---- 2 files changed, 17 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c index 6aa817a..b808972 100644 --- a/drivers/gpu/ion/ion.c +++ b/drivers/gpu/ion/ion.c @@ -62,7 +62,7 @@ struct ion_device { * @dev: backpointer to ion device * @handles: an rb tree of all the handles in this client * @lock: lock protecting the tree of handles - * @heap_mask: mask of all supported heaps + * @heap_mask: mask of all supported heap ids * @name: used for debugging * @task: used for debugging * @@ -398,7 +398,7 @@ struct ion_handle *ion_alloc(struct ion_client *client, size_t len, align, heap_mask, flags); /* * traverse the list of heaps available in this system in priority - * order. If the heap type is supported by the client, and matches the + * order. If the heap id is supported by the client, and matches the * request of the caller allocate from it. Repeat until allocate has * succeeded or all heaps have been tried */ @@ -410,10 +410,10 @@ struct ion_handle *ion_alloc(struct ion_client *client, size_t len, down_read(&dev->lock); for (n = rb_first(&dev->heaps); n != NULL; n = rb_next(n)) { struct ion_heap *heap = rb_entry(n, struct ion_heap, node); - /* if the client doesn't support this heap type */ - if (!((1 << heap->type) & client->heap_mask)) + /* if the client doesn't support this heap id */ + if (!((1 << heap->id) & client->heap_mask)) continue; - /* if the caller didn't specify this heap type */ + /* if the caller didn't specify this heap id */ if (!((1 << heap->id) & heap_mask)) continue; buffer = ion_buffer_create(heap, dev, len, align, flags); @@ -597,11 +597,11 @@ static int ion_debug_client_show(struct seq_file *s, void *unused) for (n = rb_first(&client->handles); n; n = rb_next(n)) { struct ion_handle *handle = rb_entry(n, struct ion_handle, node); - enum ion_heap_type type = handle->buffer->heap->type; + int id = handle->buffer->heap->id; - if (!names[type]) - names[type] = handle->buffer->heap->name; - sizes[type] += handle->buffer->size; + if (!names[id]) + names[id] = handle->buffer->heap->name; + sizes[id] += handle->buffer->size; } mutex_unlock(&client->lock); @@ -1176,7 +1176,7 @@ static const struct file_operations ion_fops = { }; static size_t ion_debug_heap_total(struct ion_client *client, - enum ion_heap_type type) + int id) { size_t size = 0; struct rb_node *n; @@ -1186,7 +1186,7 @@ static size_t ion_debug_heap_total(struct ion_client *client, struct ion_handle *handle = rb_entry(n, struct ion_handle, node); - if (handle->buffer->heap->type == type) + if (handle->buffer->heap->id == id) size += handle->buffer->size; } mutex_unlock(&client->lock); @@ -1207,7 +1207,7 @@ static int ion_debug_heap_show(struct seq_file *s, void *unused) for (n = rb_first(&dev->clients); n; n = rb_next(n)) { struct ion_client *client = rb_entry(n, struct ion_client, node); - size_t size = ion_debug_heap_total(client, heap->type); + size_t size = ion_debug_heap_total(client, heap->id); if (!size) continue; if (client->task) { @@ -1228,7 +1228,7 @@ static int ion_debug_heap_show(struct seq_file *s, void *unused) for (n = rb_first(&dev->buffers); n; n = rb_next(n)) { struct ion_buffer *buffer = rb_entry(n, struct ion_buffer, node); - if (buffer->heap->type != heap->type) + if (buffer->heap->id != heap->id) continue; total_size += buffer->size; if (!buffer->handle_count) { diff --git a/include/linux/ion.h b/include/linux/ion.h index a7d399c..f0399ae 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -72,7 +72,7 @@ struct ion_buffer; /** * struct ion_platform_heap - defines a heap in the given platform * @type: type of the heap from ion_heap_type enum - * @id: unique identifier for heap. When allocating (lower numbers + * @id: unique identifier for heap. When allocating (lower numbers * will be allocated from first) * @name: used for debug purposes * @base: base address of heap in physical memory if applicable @@ -114,7 +114,7 @@ void ion_reserve(struct ion_platform_data *data); /** * ion_client_create() - allocate a client and returns it * @dev: the global ion device - * @heap_mask: mask of heaps this client can allocate from + * @heap_mask: mask of heap ids this client can allocate from * @name: used for debugging */ struct ion_client *ion_client_create(struct ion_device *dev, @@ -135,7 +135,7 @@ void ion_client_destroy(struct ion_client *client); * @len: size of the allocation * @align: requested allocation alignment, lots of hardware blocks have * alignment requirements of some kind - * @heap_mask: mask of heaps to allocate from, if multiple bits are set + * @heap_mask: mask of heap ids to allocate from, if multiple bits are set * heaps will be tried in order from lowest to highest order bit * @flags: heap flags, the low 16 bits are consumed by ion, the high 16 * bits are passed on to the respective heap and can be heap @@ -236,7 +236,7 @@ struct ion_handle *ion_import_dma_buf(struct ion_client *client, int fd); * struct ion_allocation_data - metadata passed from userspace for allocations * @len: size of the allocation * @align: required alignment of the allocation - * @heap_mask: mask of heaps to allocate from + * @heap_mask: mask of heap ids to allocate from * @flags: flags passed to heap * @handle: pointer that will be populated with a cookie to use to refer * to this allocation -- 1.7.0.4

12 years, 5 months

1
0
0 0

[PATCH] dma_buf: Cleanup dma_buf_fd

by Borislav Petkov

Remove redundant 'error' variable. Signed-off-by: Borislav Petkov <bp(a)alien8.de> --- drivers/base/dma-buf.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/base/dma-buf.c b/drivers/base/dma-buf.c index 460e22dee36d..a384b63be757 100644 --- a/drivers/base/dma-buf.c +++ b/drivers/base/dma-buf.c @@ -134,15 +134,14 @@ EXPORT_SYMBOL_GPL(dma_buf_export); */ int dma_buf_fd(struct dma_buf *dmabuf, int flags) { - int error, fd; + int fd; if (!dmabuf || !dmabuf->file) return -EINVAL; - error = get_unused_fd_flags(flags); - if (error < 0) - return error; - fd = error; + fd = get_unused_fd_flags(flags); + if (fd < 0) + return fd; fd_install(fd, dmabuf->file); -- 1.8.0

12 years, 5 months

1
0
0 0

[GIT PULL] CMA and DMA-mapping updates for v3.8

by Marek Szyprowski

Hi Linus, I would like to ask for pulling another set of Contiguous Memory Allocator and DMA-mapping framework updates for v3.8. The following changes since commit 29594404d7fe73cd80eaa4ee8c43dcc53970c60e: Linux 3.7 (2012-12-10 19:30:57 -0800) are available in the git repository at: git://git.linaro.org/people/mszyprowski/linux-dma-mapping.git for-v3.8 for you to fetch changes up to 4009793e15d44469da1547a46ab129cc08ffa503: drivers: cma: represent physical addresses as phys_addr_t (2012-12-11 09:28:09 +0100) ---------------------------------------------------------------- This pull request consists only of two patches. First fixes long standing issue with dmapools (the code predates current GIT history), which forced all allocations to use GFP_ATOMIC flag, ignoring the flags passed by the caller. The second patch changes CMA code to correctly use phys_addr_t type what enables support for LPAE systems. Thanks! Best regards Marek Szyprowski Samsung Poland R&D Center Patch summary: Marek Szyprowski (1): mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls Vitaly Andrianov (1): drivers: cma: represent physical addresses as phys_addr_t drivers/base/dma-contiguous.c | 24 ++++++++++-------------- include/linux/dma-contiguous.h | 4 ++-- mm/dmapool.c | 31 +++++++------------------------ 3 files changed, 19 insertions(+), 40 deletions(-)

12 years, 5 months

1
0
0 0

[PATCH] gpu: ion: Export ion_client_create

by Johan Mossberg

Will enable modules to allocate memory with ion. Signed-off-by: Johan Mossberg <johan.mossberg(a)stericsson.com> --- drivers/gpu/ion/ion.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/ion/ion.c b/drivers/gpu/ion/ion.c index 100e5fe..fc152b9 100644 --- a/drivers/gpu/ion/ion.c +++ b/drivers/gpu/ion/ion.c @@ -688,6 +688,7 @@ struct ion_client *ion_client_create(struct ion_device *dev, return client; } +EXPORT_SYMBOL(ion_client_create); void ion_client_destroy(struct ion_client *client) { -- 1.8.0

12 years, 5 months

1
0
0 0

[PATCH v8 0/3] Add CMA heap for ION memory manager

by Benjamin Gaignard

The goal of those patches is to allow ION clients (drivers or userland applications) to use Contiguous Memory Allocator (CMA). To get more info about CMA: http://lists.linaro.org/pipermail/linaro-mm-sig/2012-February/001328.html patches version 8: - fix memory leak when release sg_table - remove virt_to_phys from ion_cma_phys patches version 7: - rebased on Android kernel - fix ion Makefile - add ion_cma_map_kernel function - remove CONFIG_CMA compilation flags from ion_heap.c patches version 6: - add private field in ion_platform_heap to pass the device linked with CMA. - rework CMA heap to use private field. - prepare CMA heap for incoming dma_common_get_sgtable function http://lists.linaro.org/pipermail/linaro-mm-sig/2012-June/002109.html - simplify ion-ux500 driver. patches version 5: - port patches on android kernel 3.4 where ION use dmabuf - add ion_cma_heap_map_dma and ion_cma_heap_unmap_dma functions patches version 4: - add ION_HEAP_TYPE_DMA heap type in ion_heap_type enum. - CMA heap is now a "native" ION heap. - add ion_heap_create_full function to keep backward compatibilty. - clean up included files in CMA heap - ux500-ion is using ion_heap_create_full instead of ion_heap_create patches version 3: - add a private field in ion_heap structure instead of expose ion_device structure to all heaps - ion_cma_heap is no more a platform driver - ion_cma_heap use ion_heap private field to store the device pointer and make the link with reserved CMA regions - provide ux500-ion driver and configuration file for snowball board to give an example of how use CMA heaps patches version 2: - fix comments done by Andy Green Benjamin Gaignard (3): gpu: ion: fix ion_platform_data definition gpu: ion: add private field in ion_heap and ion_platform_heap structure gpu: ion: add CMA heap drivers/gpu/ion/Kconfig | 5 ++ drivers/gpu/ion/Makefile | 1 + drivers/gpu/ion/ion_cma_heap.c | 192 ++++++++++++++++++++++++++++++++++++++++ drivers/gpu/ion/ion_heap.c | 7 ++ drivers/gpu/ion/ion_priv.h | 16 ++++ include/linux/ion.h | 7 +- 6 files changed, 227 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/ion/ion_cma_heap.c -- 1.7.10

12 years, 6 months

1
3
0 0

[PATCH v7 3/3] gpu: ion: add CMA heap

by benjamin.gaignard＠linaro.org

From: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> New heap type ION_HEAP_TYPE_DMA where allocation is done with dma_alloc_coherent API. device coherent_dma_mask must be set to DMA_BIT_MASK(32). ion_platform_heap private field is used to retrieve the device linked to CMA, if NULL the default CMA area is used. ion_cma_get_sgtable is a copy of dma_common_get_sgtable function which should be in kernel 3.5 Signed-off-by: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> --- drivers/gpu/ion/Kconfig | 5 ++ drivers/gpu/ion/Makefile | 1 + drivers/gpu/ion/ion_cma_heap.c | 187 ++++++++++++++++++++++++++++++++++++++++ drivers/gpu/ion/ion_heap.c | 6 ++ drivers/gpu/ion/ion_priv.h | 14 +++ include/linux/ion.h | 3 + 6 files changed, 216 insertions(+) create mode 100644 drivers/gpu/ion/ion_cma_heap.c diff --git a/drivers/gpu/ion/Kconfig b/drivers/gpu/ion/Kconfig index b5bfdb4..ea0a896 100644 --- a/drivers/gpu/ion/Kconfig +++ b/drivers/gpu/ion/Kconfig @@ -5,6 +5,11 @@ menuconfig ION help Chose this option to enable the ION Memory Manager. +config ION_CMA + bool + default y + depends on CMA && ION + config ION_TEGRA tristate "Ion for Tegra" depends on ARCH_TEGRA && ION diff --git a/drivers/gpu/ion/Makefile b/drivers/gpu/ion/Makefile index d1ddebb..fddb85c 100644 --- a/drivers/gpu/ion/Makefile +++ b/drivers/gpu/ion/Makefile @@ -1,3 +1,4 @@ obj-$(CONFIG_ION) += ion.o ion_heap.o ion_page_pool.o ion_system_heap.o \ ion_carveout_heap.o +obj-$(CONFIG_ION_CMA) += ion_cma_heap.o obj-$(CONFIG_ION_TEGRA) += tegra/ diff --git a/drivers/gpu/ion/ion_cma_heap.c b/drivers/gpu/ion/ion_cma_heap.c new file mode 100644 index 0000000..543f891 --- /dev/null +++ b/drivers/gpu/ion/ion_cma_heap.c @@ -0,0 +1,187 @@ +/* + * drivers/gpu/ion/ion_cma_heap.c + * + * Copyright (C) Linaro 2012 + * Author: <benjamin.gaignard(a)linaro.org> for ST-Ericsson. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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. + * + */ + +#include <linux/device.h> +#include <linux/ion.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/err.h> +#include <linux/dma-mapping.h> + +/* for ion_heap_ops structure */ +#include "ion_priv.h" + +#define ION_CMA_ALLOCATE_FAILED -1 + +struct ion_cma_buffer_info { + void *cpu_addr; + dma_addr_t handle; + struct sg_table *table; +}; + +/* + * Create scatter-list for the already allocated DMA buffer. + * This function could be replaced by dma_common_get_sgtable + * as soon as it will avalaible. + */ +int ion_cma_get_sgtable(struct device *dev, struct sg_table *sgt, + void *cpu_addr, dma_addr_t handle, size_t size) +{ + struct page *page = virt_to_page(cpu_addr); + int ret; + + ret = sg_alloc_table(sgt, 1, GFP_KERNEL); + if (unlikely(ret)) + return ret; + + sg_set_page(sgt->sgl, page, PAGE_ALIGN(size), 0); + return 0; +} + +/* ION CMA heap operations functions */ +static int ion_cma_allocate(struct ion_heap *heap, struct ion_buffer *buffer, + unsigned long len, unsigned long align, + unsigned long flags) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info; + + dev_dbg(dev, "Request buffer allocation len %ld\n", len); + + info = kzalloc(sizeof(struct ion_cma_buffer_info), GFP_KERNEL); + if (!info) { + dev_err(dev, "Can't allocate buffer info\n"); + return ION_CMA_ALLOCATE_FAILED; + } + + info->cpu_addr = dma_alloc_coherent(dev, len, &(info->handle), 0); + + if (!info->cpu_addr) { + dev_err(dev, "Fail to allocate buffer\n"); + goto err; + } + + info->table = kmalloc(sizeof(struct sg_table), GFP_KERNEL); + if (!info->table) { + dev_err(dev, "Fail to allocate sg table\n"); + goto err; + } + + ion_cma_get_sgtable(dev, + info->table, info->cpu_addr, info->handle, len); + + /* keep this for memory release */ + buffer->priv_virt = info; + dev_dbg(dev, "Allocate buffer %p\n", buffer); + return 0; + +err: + kfree(info); + return ION_CMA_ALLOCATE_FAILED; +} + +static void ion_cma_free(struct ion_buffer *buffer) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Release buffer %p\n", buffer); + /* release memory */ + dma_free_coherent(dev, buffer->size, info->cpu_addr, info->handle); + /* release sg table */ + kfree(info->table); + kfree(info); +} + +/* return physical address in addr */ +static int ion_cma_phys(struct ion_heap *heap, struct ion_buffer *buffer, + ion_phys_addr_t *addr, size_t *len) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Return buffer %p physical address 0x%x\n", buffer, + virt_to_phys(info->cpu_addr)); + + *addr = virt_to_phys(info->cpu_addr); + *len = buffer->size; + + return 0; +} + +struct sg_table *ion_cma_heap_map_dma(struct ion_heap *heap, + struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return info->table; +} + +void ion_cma_heap_unmap_dma(struct ion_heap *heap, + struct ion_buffer *buffer) +{ + return; +} + +static int ion_cma_mmap(struct ion_heap *mapper, struct ion_buffer *buffer, + struct vm_area_struct *vma) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return dma_mmap_coherent(dev, vma, info->cpu_addr, info->handle, + buffer->size); +} + +void *ion_cma_map_kernel(struct ion_heap *heap, struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + /* kernel memory mapping has been done at allocation time */ + return info->cpu_addr; +} + +static struct ion_heap_ops ion_cma_ops = { + .allocate = ion_cma_allocate, + .free = ion_cma_free, + .map_dma = ion_cma_heap_map_dma, + .unmap_dma = ion_cma_heap_unmap_dma, + .phys = ion_cma_phys, + .map_user = ion_cma_mmap, + .map_kernel = ion_cma_map_kernel, +}; + +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *data) +{ + struct ion_heap *heap; + + heap = kzalloc(sizeof(struct ion_heap), GFP_KERNEL); + + if (!heap) + return ERR_PTR(-ENOMEM); + + heap->ops = &ion_cma_ops; + /* set device as private heaps data, later it will be + * used to make the link with reserved CMA memory */ + heap->priv = data->priv; + heap->type = ION_HEAP_TYPE_DMA; + return heap; +} + +void ion_cma_heap_destroy(struct ion_heap *heap) +{ + kfree(heap); +} diff --git a/drivers/gpu/ion/ion_heap.c b/drivers/gpu/ion/ion_heap.c index f9ab1df..c960230 100644 --- a/drivers/gpu/ion/ion_heap.c +++ b/drivers/gpu/ion/ion_heap.c @@ -32,6 +32,9 @@ struct ion_heap *ion_heap_create(struct ion_platform_heap *heap_data) case ION_HEAP_TYPE_CARVEOUT: heap = ion_carveout_heap_create(heap_data); break; + case ION_HEAP_TYPE_DMA: + heap = ion_cma_heap_create(heap_data); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap_data->type); @@ -66,6 +69,9 @@ void ion_heap_destroy(struct ion_heap *heap) case ION_HEAP_TYPE_CARVEOUT: ion_carveout_heap_destroy(heap); break; + case ION_HEAP_TYPE_DMA: + ion_cma_heap_destroy(heap); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap->type); diff --git a/drivers/gpu/ion/ion_priv.h b/drivers/gpu/ion/ion_priv.h index a31203b..046f1a2 100644 --- a/drivers/gpu/ion/ion_priv.h +++ b/drivers/gpu/ion/ion_priv.h @@ -203,6 +203,20 @@ ion_phys_addr_t ion_carveout_allocate(struct ion_heap *heap, unsigned long size, unsigned long align); void ion_carveout_free(struct ion_heap *heap, ion_phys_addr_t addr, unsigned long size); + +#ifdef CONFIG_CMA +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *heap); +void ion_cma_heap_destroy(struct ion_heap *heap); +#else +static inline struct ion_heap *ion_cma_heap_create(struct ion_platform_heap + *heap) +{ + return NULL; +} + +static inline void ion_cma_heap_destroy(struct ion_heap *heap) {}; +#endif + /** * The carveout heap returns physical addresses, since 0 may be a valid * physical address, this is used to indicate allocation failed diff --git a/include/linux/ion.h b/include/linux/ion.h index b85f4b1..3ccc75a 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -27,6 +27,7 @@ struct ion_handle; * @ION_HEAP_TYPE_CARVEOUT: memory allocated from a prereserved * carveout heap, allocations are physically * contiguous + * @ION_HEAP_TYPE_DMA: memory allocated via DMA API * @ION_NUM_HEAPS: helper for iterating over heaps, a bit mask * is used to identify the heaps, so only 32 * total heap types are supported @@ -35,6 +36,7 @@ enum ion_heap_type { ION_HEAP_TYPE_SYSTEM, ION_HEAP_TYPE_SYSTEM_CONTIG, ION_HEAP_TYPE_CARVEOUT, + ION_HEAP_TYPE_DMA, ION_HEAP_TYPE_CUSTOM, /* must be last so device specific heaps always are at the end of this enum */ ION_NUM_HEAPS = 16, @@ -43,6 +45,7 @@ enum ion_heap_type { #define ION_HEAP_SYSTEM_MASK (1 << ION_HEAP_TYPE_SYSTEM) #define ION_HEAP_SYSTEM_CONTIG_MASK (1 << ION_HEAP_TYPE_SYSTEM_CONTIG) #define ION_HEAP_CARVEOUT_MASK (1 << ION_HEAP_TYPE_CARVEOUT) +#define ION_HEAP_TYPE_DMA_MASK (1 << ION_HEAP_TYPE_DMA) /** * heap flags - the lower 16 bits are used by core ion, the upper 16 -- 1.7.10

12 years, 6 months

4
3
0 0

[PATCH v7 3/3] gpu: ion: add CMA heap

by Benjamin Gaignard

New heap type ION_HEAP_TYPE_DMA where allocation is done with dma_alloc_coherent API. device coherent_dma_mask must be set to DMA_BIT_MASK(32). ion_platform_heap private field is used to retrieve the device linked to CMA, if NULL the default CMA area is used. ion_cma_get_sgtable is a copy of dma_common_get_sgtable function which should be in kernel 3.5 Signed-off-by: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> --- drivers/gpu/ion/Kconfig | 5 ++ drivers/gpu/ion/Makefile | 1 + drivers/gpu/ion/ion_cma_heap.c | 192 ++++++++++++++++++++++++++++++++++++++++ drivers/gpu/ion/ion_heap.c | 6 ++ drivers/gpu/ion/ion_priv.h | 14 +++ include/linux/ion.h | 3 + 6 files changed, 221 insertions(+) create mode 100644 drivers/gpu/ion/ion_cma_heap.c diff --git a/drivers/gpu/ion/Kconfig b/drivers/gpu/ion/Kconfig index b5bfdb4..ea0a896 100644 --- a/drivers/gpu/ion/Kconfig +++ b/drivers/gpu/ion/Kconfig @@ -5,6 +5,11 @@ menuconfig ION help Chose this option to enable the ION Memory Manager. +config ION_CMA + bool + default y + depends on CMA && ION + config ION_TEGRA tristate "Ion for Tegra" depends on ARCH_TEGRA && ION diff --git a/drivers/gpu/ion/Makefile b/drivers/gpu/ion/Makefile index d1ddebb..fddb85c 100644 --- a/drivers/gpu/ion/Makefile +++ b/drivers/gpu/ion/Makefile @@ -1,3 +1,4 @@ obj-$(CONFIG_ION) += ion.o ion_heap.o ion_page_pool.o ion_system_heap.o \ ion_carveout_heap.o +obj-$(CONFIG_ION_CMA) += ion_cma_heap.o obj-$(CONFIG_ION_TEGRA) += tegra/ diff --git a/drivers/gpu/ion/ion_cma_heap.c b/drivers/gpu/ion/ion_cma_heap.c new file mode 100644 index 0000000..b44e0e0 --- /dev/null +++ b/drivers/gpu/ion/ion_cma_heap.c @@ -0,0 +1,192 @@ +/* + * drivers/gpu/ion/ion_cma_heap.c + * + * Copyright (C) Linaro 2012 + * Author: <benjamin.gaignard(a)linaro.org> for ST-Ericsson. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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. + * + */ + +#include <linux/device.h> +#include <linux/ion.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/err.h> +#include <linux/dma-mapping.h> + +/* for ion_heap_ops structure */ +#include "ion_priv.h" + +#define ION_CMA_ALLOCATE_FAILED -1 + +struct ion_cma_buffer_info { + void *cpu_addr; + dma_addr_t handle; + struct sg_table *table; +}; + +/* + * Create scatter-list for the already allocated DMA buffer. + * This function could be replaced by dma_common_get_sgtable + * as soon as it will avalaible. + */ +int ion_cma_get_sgtable(struct device *dev, struct sg_table *sgt, + void *cpu_addr, dma_addr_t handle, size_t size) +{ + struct page *page = virt_to_page(cpu_addr); + int ret; + + ret = sg_alloc_table(sgt, 1, GFP_KERNEL); + if (unlikely(ret)) + return ret; + + sg_set_page(sgt->sgl, page, PAGE_ALIGN(size), 0); + return 0; +} + +/* ION CMA heap operations functions */ +static int ion_cma_allocate(struct ion_heap *heap, struct ion_buffer *buffer, + unsigned long len, unsigned long align, + unsigned long flags) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info; + + dev_dbg(dev, "Request buffer allocation len %ld\n", len); + + info = kzalloc(sizeof(struct ion_cma_buffer_info), GFP_KERNEL); + if (!info) { + dev_err(dev, "Can't allocate buffer info\n"); + return ION_CMA_ALLOCATE_FAILED; + } + + info->cpu_addr = dma_alloc_coherent(dev, len, &(info->handle), 0); + + if (!info->cpu_addr) { + dev_err(dev, "Fail to allocate buffer\n"); + goto err; + } + + info->table = kmalloc(sizeof(struct sg_table), GFP_KERNEL); + if (!info->table) { + dev_err(dev, "Fail to allocate sg table\n"); + goto free_mem; + } + + if (ion_cma_get_sgtable + (dev, info->table, info->cpu_addr, info->handle, len)) + goto free_table; + + /* keep this for memory release */ + buffer->priv_virt = info; + dev_dbg(dev, "Allocate buffer %p\n", buffer); + return 0; + +free_table: + kfree(info->table); +free_mem: + dma_free_coherent(dev, len, info->cpu_addr, info->handle); +err: + kfree(info); + return ION_CMA_ALLOCATE_FAILED; +} + +static void ion_cma_free(struct ion_buffer *buffer) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Release buffer %p\n", buffer); + /* release memory */ + dma_free_coherent(dev, buffer->size, info->cpu_addr, info->handle); + /* release sg table */ + sg_free_table(info->table); + kfree(info->table); + kfree(info); +} + +/* return physical address in addr */ +static int ion_cma_phys(struct ion_heap *heap, struct ion_buffer *buffer, + ion_phys_addr_t *addr, size_t *len) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Return buffer %p physical address 0x%x\n", buffer, + virt_to_phys(info->cpu_addr)); + + *addr = virt_to_phys(info->cpu_addr); + *len = buffer->size; + + return 0; +} + +struct sg_table *ion_cma_heap_map_dma(struct ion_heap *heap, + struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return info->table; +} + +void ion_cma_heap_unmap_dma(struct ion_heap *heap, struct ion_buffer *buffer) +{ + return; +} + +static int ion_cma_mmap(struct ion_heap *mapper, struct ion_buffer *buffer, + struct vm_area_struct *vma) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return dma_mmap_coherent(dev, vma, info->cpu_addr, info->handle, + buffer->size); +} + +void *ion_cma_map_kernel(struct ion_heap *heap, struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + /* kernel memory mapping has been done at allocation time */ + return info->cpu_addr; +} + +static struct ion_heap_ops ion_cma_ops = { + .allocate = ion_cma_allocate, + .free = ion_cma_free, + .map_dma = ion_cma_heap_map_dma, + .unmap_dma = ion_cma_heap_unmap_dma, + .phys = ion_cma_phys, + .map_user = ion_cma_mmap, + .map_kernel = ion_cma_map_kernel, +}; + +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *data) +{ + struct ion_heap *heap; + + heap = kzalloc(sizeof(struct ion_heap), GFP_KERNEL); + + if (!heap) + return ERR_PTR(-ENOMEM); + + heap->ops = &ion_cma_ops; + /* set device as private heaps data, later it will be + * used to make the link with reserved CMA memory */ + heap->priv = data->priv; + heap->type = ION_HEAP_TYPE_DMA; + return heap; +} + +void ion_cma_heap_destroy(struct ion_heap *heap) +{ + kfree(heap); +} diff --git a/drivers/gpu/ion/ion_heap.c b/drivers/gpu/ion/ion_heap.c index f9ab1df..c960230 100644 --- a/drivers/gpu/ion/ion_heap.c +++ b/drivers/gpu/ion/ion_heap.c @@ -32,6 +32,9 @@ struct ion_heap *ion_heap_create(struct ion_platform_heap *heap_data) case ION_HEAP_TYPE_CARVEOUT: heap = ion_carveout_heap_create(heap_data); break; + case ION_HEAP_TYPE_DMA: + heap = ion_cma_heap_create(heap_data); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap_data->type); @@ -66,6 +69,9 @@ void ion_heap_destroy(struct ion_heap *heap) case ION_HEAP_TYPE_CARVEOUT: ion_carveout_heap_destroy(heap); break; + case ION_HEAP_TYPE_DMA: + ion_cma_heap_destroy(heap); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap->type); diff --git a/drivers/gpu/ion/ion_priv.h b/drivers/gpu/ion/ion_priv.h index a31203b..046f1a2 100644 --- a/drivers/gpu/ion/ion_priv.h +++ b/drivers/gpu/ion/ion_priv.h @@ -203,6 +203,20 @@ ion_phys_addr_t ion_carveout_allocate(struct ion_heap *heap, unsigned long size, unsigned long align); void ion_carveout_free(struct ion_heap *heap, ion_phys_addr_t addr, unsigned long size); + +#ifdef CONFIG_CMA +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *heap); +void ion_cma_heap_destroy(struct ion_heap *heap); +#else +static inline struct ion_heap *ion_cma_heap_create(struct ion_platform_heap + *heap) +{ + return NULL; +} + +static inline void ion_cma_heap_destroy(struct ion_heap *heap) {}; +#endif + /** * The carveout heap returns physical addresses, since 0 may be a valid * physical address, this is used to indicate allocation failed diff --git a/include/linux/ion.h b/include/linux/ion.h index b85f4b1..3ccc75a 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -27,6 +27,7 @@ struct ion_handle; * @ION_HEAP_TYPE_CARVEOUT: memory allocated from a prereserved * carveout heap, allocations are physically * contiguous + * @ION_HEAP_TYPE_DMA: memory allocated via DMA API * @ION_NUM_HEAPS: helper for iterating over heaps, a bit mask * is used to identify the heaps, so only 32 * total heap types are supported @@ -35,6 +36,7 @@ enum ion_heap_type { ION_HEAP_TYPE_SYSTEM, ION_HEAP_TYPE_SYSTEM_CONTIG, ION_HEAP_TYPE_CARVEOUT, + ION_HEAP_TYPE_DMA, ION_HEAP_TYPE_CUSTOM, /* must be last so device specific heaps always are at the end of this enum */ ION_NUM_HEAPS = 16, @@ -43,6 +45,7 @@ enum ion_heap_type { #define ION_HEAP_SYSTEM_MASK (1 << ION_HEAP_TYPE_SYSTEM) #define ION_HEAP_SYSTEM_CONTIG_MASK (1 << ION_HEAP_TYPE_SYSTEM_CONTIG) #define ION_HEAP_CARVEOUT_MASK (1 << ION_HEAP_TYPE_CARVEOUT) +#define ION_HEAP_TYPE_DMA_MASK (1 << ION_HEAP_TYPE_DMA) /** * heap flags - the lower 16 bits are used by core ion, the upper 16 -- 1.7.10

12 years, 6 months

1
0
0 0

[PATCH v7 3/3] gpu: ion: add CMA heap

by benjamin.gaignard＠linaro.org

From: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> New heap type ION_HEAP_TYPE_DMA where allocation is done with dma_alloc_coherent API. device coherent_dma_mask must be set to DMA_BIT_MASK(32). ion_platform_heap private field is used to retrieve the device linked to CMA, if NULL the default CMA area is used. ion_cma_get_sgtable is a copy of dma_common_get_sgtable function which should be in kernel 3.5 Change-Id: I31bb25360f69f3744c1984f6d83f9d939090a81f Signed-off-by: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> --- drivers/gpu/ion/Kconfig | 5 ++ drivers/gpu/ion/Makefile | 1 + drivers/gpu/ion/ion_cma_heap.c | 189 ++++++++++++++++++++++++++++++++++++++++ drivers/gpu/ion/ion_heap.c | 6 ++ drivers/gpu/ion/ion_priv.h | 14 +++ include/linux/ion.h | 3 + 6 files changed, 218 insertions(+) create mode 100644 drivers/gpu/ion/ion_cma_heap.c diff --git a/drivers/gpu/ion/Kconfig b/drivers/gpu/ion/Kconfig index b5bfdb4..ea0a896 100644 --- a/drivers/gpu/ion/Kconfig +++ b/drivers/gpu/ion/Kconfig @@ -5,6 +5,11 @@ menuconfig ION help Chose this option to enable the ION Memory Manager. +config ION_CMA + bool + default y + depends on CMA && ION + config ION_TEGRA tristate "Ion for Tegra" depends on ARCH_TEGRA && ION diff --git a/drivers/gpu/ion/Makefile b/drivers/gpu/ion/Makefile index d1ddebb..fddb85c 100644 --- a/drivers/gpu/ion/Makefile +++ b/drivers/gpu/ion/Makefile @@ -1,3 +1,4 @@ obj-$(CONFIG_ION) += ion.o ion_heap.o ion_page_pool.o ion_system_heap.o \ ion_carveout_heap.o +obj-$(CONFIG_ION_CMA) += ion_cma_heap.o obj-$(CONFIG_ION_TEGRA) += tegra/ diff --git a/drivers/gpu/ion/ion_cma_heap.c b/drivers/gpu/ion/ion_cma_heap.c new file mode 100644 index 0000000..22bbf2d --- /dev/null +++ b/drivers/gpu/ion/ion_cma_heap.c @@ -0,0 +1,189 @@ +/* + * drivers/gpu/ion/ion_cma_heap.c + * + * Copyright (C) Linaro 2012 + * Author: <benjamin.gaignard(a)linaro.org> for ST-Ericsson. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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. + * + */ + +#include <linux/device.h> +#include <linux/ion.h> +#include <linux/slab.h> +#include <linux/errno.h> +#include <linux/err.h> +#include <linux/dma-mapping.h> + +/* for ion_heap_ops structure */ +#include "ion_priv.h" + +#define ION_CMA_ALLOCATE_FAILED -1 + +struct ion_cma_buffer_info { + void *cpu_addr; + dma_addr_t handle; + struct sg_table *table; +}; + +/* + * Create scatter-list for the already allocated DMA buffer. + * This function could be replaced by dma_common_get_sgtable + * as soon as it will avalaible. + */ +int ion_cma_get_sgtable(struct device *dev, struct sg_table *sgt, + void *cpu_addr, dma_addr_t handle, size_t size) +{ + struct page *page = virt_to_page(cpu_addr); + int ret; + + ret = sg_alloc_table(sgt, 1, GFP_KERNEL); + if (unlikely(ret)) + return ret; + + sg_set_page(sgt->sgl, page, PAGE_ALIGN(size), 0); + return 0; +} + +/* ION CMA heap operations functions */ +static int ion_cma_allocate(struct ion_heap *heap, struct ion_buffer *buffer, + unsigned long len, unsigned long align, + unsigned long flags) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info; + + dev_dbg(dev, "Request buffer allocation len %ld\n", len); + + info = kzalloc(sizeof(struct ion_cma_buffer_info), GFP_KERNEL); + if (!info) { + dev_err(dev, "Can't allocate buffer info\n"); + return ION_CMA_ALLOCATE_FAILED; + } + + info->cpu_addr = dma_alloc_coherent(dev, len, &(info->handle), 0); + + if (!info->cpu_addr) { + dev_err(dev, "Fail to allocate buffer\n"); + goto err; + } + + info->table = kmalloc(sizeof(struct sg_table), GFP_KERNEL); + if (!info->table) { + dev_err(dev, "Fail to allocate sg table\n"); + goto err; + } + + if (ion_cma_get_sgtable + (dev, info->table, info->cpu_addr, info->handle, len)) + goto free_table; + + /* keep this for memory release */ + buffer->priv_virt = info; + dev_dbg(dev, "Allocate buffer %p\n", buffer); + return 0; + +free_table: + sg_free_table(info->table); +err: + kfree(info); + return ION_CMA_ALLOCATE_FAILED; +} + +static void ion_cma_free(struct ion_buffer *buffer) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Release buffer %p\n", buffer); + /* release memory */ + dma_free_coherent(dev, buffer->size, info->cpu_addr, info->handle); + /* release sg table */ + sg_free_table(info->table); + kfree(info); +} + +/* return physical address in addr */ +static int ion_cma_phys(struct ion_heap *heap, struct ion_buffer *buffer, + ion_phys_addr_t *addr, size_t *len) +{ + struct device *dev = heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + dev_dbg(dev, "Return buffer %p physical address 0x%x\n", buffer, + virt_to_phys(info->cpu_addr)); + + *addr = virt_to_phys(info->cpu_addr); + *len = buffer->size; + + return 0; +} + +struct sg_table *ion_cma_heap_map_dma(struct ion_heap *heap, + struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return info->table; +} + +void ion_cma_heap_unmap_dma(struct ion_heap *heap, struct ion_buffer *buffer) +{ + return; +} + +static int ion_cma_mmap(struct ion_heap *mapper, struct ion_buffer *buffer, + struct vm_area_struct *vma) +{ + struct device *dev = buffer->heap->priv; + struct ion_cma_buffer_info *info = buffer->priv_virt; + + return dma_mmap_coherent(dev, vma, info->cpu_addr, info->handle, + buffer->size); +} + +void *ion_cma_map_kernel(struct ion_heap *heap, struct ion_buffer *buffer) +{ + struct ion_cma_buffer_info *info = buffer->priv_virt; + /* kernel memory mapping has been done at allocation time */ + return info->cpu_addr; +} + +static struct ion_heap_ops ion_cma_ops = { + .allocate = ion_cma_allocate, + .free = ion_cma_free, + .map_dma = ion_cma_heap_map_dma, + .unmap_dma = ion_cma_heap_unmap_dma, + .phys = ion_cma_phys, + .map_user = ion_cma_mmap, + .map_kernel = ion_cma_map_kernel, +}; + +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *data) +{ + struct ion_heap *heap; + + heap = kzalloc(sizeof(struct ion_heap), GFP_KERNEL); + + if (!heap) + return ERR_PTR(-ENOMEM); + + heap->ops = &ion_cma_ops; + /* set device as private heaps data, later it will be + * used to make the link with reserved CMA memory */ + heap->priv = data->priv; + heap->type = ION_HEAP_TYPE_DMA; + return heap; +} + +void ion_cma_heap_destroy(struct ion_heap *heap) +{ + kfree(heap); +} diff --git a/drivers/gpu/ion/ion_heap.c b/drivers/gpu/ion/ion_heap.c index f9ab1df..c960230 100644 --- a/drivers/gpu/ion/ion_heap.c +++ b/drivers/gpu/ion/ion_heap.c @@ -32,6 +32,9 @@ struct ion_heap *ion_heap_create(struct ion_platform_heap *heap_data) case ION_HEAP_TYPE_CARVEOUT: heap = ion_carveout_heap_create(heap_data); break; + case ION_HEAP_TYPE_DMA: + heap = ion_cma_heap_create(heap_data); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap_data->type); @@ -66,6 +69,9 @@ void ion_heap_destroy(struct ion_heap *heap) case ION_HEAP_TYPE_CARVEOUT: ion_carveout_heap_destroy(heap); break; + case ION_HEAP_TYPE_DMA: + ion_cma_heap_destroy(heap); + break; default: pr_err("%s: Invalid heap type %d\n", __func__, heap->type); diff --git a/drivers/gpu/ion/ion_priv.h b/drivers/gpu/ion/ion_priv.h index a31203b..046f1a2 100644 --- a/drivers/gpu/ion/ion_priv.h +++ b/drivers/gpu/ion/ion_priv.h @@ -203,6 +203,20 @@ ion_phys_addr_t ion_carveout_allocate(struct ion_heap *heap, unsigned long size, unsigned long align); void ion_carveout_free(struct ion_heap *heap, ion_phys_addr_t addr, unsigned long size); + +#ifdef CONFIG_CMA +struct ion_heap *ion_cma_heap_create(struct ion_platform_heap *heap); +void ion_cma_heap_destroy(struct ion_heap *heap); +#else +static inline struct ion_heap *ion_cma_heap_create(struct ion_platform_heap + *heap) +{ + return NULL; +} + +static inline void ion_cma_heap_destroy(struct ion_heap *heap) {}; +#endif + /** * The carveout heap returns physical addresses, since 0 may be a valid * physical address, this is used to indicate allocation failed diff --git a/include/linux/ion.h b/include/linux/ion.h index b85f4b1..3ccc75a 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -27,6 +27,7 @@ struct ion_handle; * @ION_HEAP_TYPE_CARVEOUT: memory allocated from a prereserved * carveout heap, allocations are physically * contiguous + * @ION_HEAP_TYPE_DMA: memory allocated via DMA API * @ION_NUM_HEAPS: helper for iterating over heaps, a bit mask * is used to identify the heaps, so only 32 * total heap types are supported @@ -35,6 +36,7 @@ enum ion_heap_type { ION_HEAP_TYPE_SYSTEM, ION_HEAP_TYPE_SYSTEM_CONTIG, ION_HEAP_TYPE_CARVEOUT, + ION_HEAP_TYPE_DMA, ION_HEAP_TYPE_CUSTOM, /* must be last so device specific heaps always are at the end of this enum */ ION_NUM_HEAPS = 16, @@ -43,6 +45,7 @@ enum ion_heap_type { #define ION_HEAP_SYSTEM_MASK (1 << ION_HEAP_TYPE_SYSTEM) #define ION_HEAP_SYSTEM_CONTIG_MASK (1 << ION_HEAP_TYPE_SYSTEM_CONTIG) #define ION_HEAP_CARVEOUT_MASK (1 << ION_HEAP_TYPE_CARVEOUT) +#define ION_HEAP_TYPE_DMA_MASK (1 << ION_HEAP_TYPE_DMA) /** * heap flags - the lower 16 bits are used by core ion, the upper 16 -- 1.7.10

12 years, 6 months

2
1
0 0

[PATCH v7 2/3] gpu: ion: add private field in ion_heap and ion_platform_heap structure

by benjamin.gaignard＠linaro.org

From: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> copy private field from platform configuration to internal heap structure. Signed-off-by: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> --- drivers/gpu/ion/ion_heap.c | 1 + drivers/gpu/ion/ion_priv.h | 2 ++ include/linux/ion.h | 2 ++ 3 files changed, 5 insertions(+) diff --git a/drivers/gpu/ion/ion_heap.c b/drivers/gpu/ion/ion_heap.c index 8ce3c19..f9ab1df 100644 --- a/drivers/gpu/ion/ion_heap.c +++ b/drivers/gpu/ion/ion_heap.c @@ -47,6 +47,7 @@ struct ion_heap *ion_heap_create(struct ion_platform_heap *heap_data) heap->name = heap_data->name; heap->id = heap_data->id; + heap->priv = heap_data->priv; return heap; } diff --git a/drivers/gpu/ion/ion_priv.h b/drivers/gpu/ion/ion_priv.h index 21c1963..a31203b 100644 --- a/drivers/gpu/ion/ion_priv.h +++ b/drivers/gpu/ion/ion_priv.h @@ -117,6 +117,7 @@ struct ion_heap_ops { * allocating. These are specified by platform data and * MUST be unique * @name: used for debugging + * @priv: heap private data * @debug_show: called when heap debug file is read to add any * heap specific debug info to output * @@ -132,6 +133,7 @@ struct ion_heap { struct ion_heap_ops *ops; int id; const char *name; + void *priv; int (*debug_show)(struct ion_heap *heap, struct seq_file *, void *); }; diff --git a/include/linux/ion.h b/include/linux/ion.h index 6541a11..b85f4b1 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -77,6 +77,7 @@ struct ion_buffer; * @name: used for debug purposes * @base: base address of heap in physical memory if applicable * @size: size of the heap in bytes if applicable + * @priv: heap private data * * Provided by the board file. */ @@ -86,6 +87,7 @@ struct ion_platform_heap { const char *name; ion_phys_addr_t base; size_t size; + void *priv; }; /** -- 1.7.10

12 years, 6 months

1
0
0 0

[PATCH v7 1/3] gpu: ion: fix ion_platform_data definition

by benjamin.gaignard＠linaro.org

From: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> fix ion_platform_heap to make is use an usual way in board configuration file. Signed-off-by: Benjamin Gaignard <benjamin.gaignard(a)linaro.org> --- include/linux/ion.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/ion.h b/include/linux/ion.h index a7d399c..6541a11 100644 --- a/include/linux/ion.h +++ b/include/linux/ion.h @@ -97,7 +97,7 @@ struct ion_platform_heap { */ struct ion_platform_data { int nr; - struct ion_platform_heap heaps[]; + struct ion_platform_heap *heaps; }; /** -- 1.7.10

12 years, 6 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig