kernel/linux.git/drivers/dma/dmaengine.c, branch v7.0-rc7

Convert 'alloc_obj' family to use the new default GFP_KERNEL argument

2026-02-22T01:09:51+00:00

This was done entirely with mindless brute force, using git grep -l '\

treewide: Replace kmalloc with kmalloc_obj for non-scalar types

2026-02-21T09:02:28+00:00

This is the result of running the Coccinelle script from scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to avoid scalar types (which need careful case-by-case checking), and instead replace kmalloc-family calls that allocate struct or union object instances: Single allocations: kmalloc(sizeof(TYPE), ...) are replaced with: kmalloc_obj(TYPE, ...) Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...) are replaced with: kmalloc_objs(TYPE, COUNT, ...) Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...) are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...) (where TYPE may also be *VAR) The resulting allocations no longer return "void *", instead returning "TYPE *". Signed-off-by: Kees Cook

dmaengine: Add devm_dma_request_chan()

2025-06-26T22:18:04+00:00

Expand the arsenal of devm functions for DMA devices, this time for requesting channels. Signed-off-by: Bence Csókás Link: https://lore.kernel.org/r/20250610082256.400492-2-csokas.bence@prolan.hu Signed-off-by: Vinod Koul

dmaengine: Unify checks in dma_request_chan()

2025-02-10T10:58:53+00:00

Use dev_fwnode() to simplify the check logic for Device Tree and ACPI in dma_request_chan(). Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20250205145757.889247-5-andriy.shevchenko@linux.intel.com Signed-off-by: Vinod Koul

dmaengine: Add a comment on why it's okay when kasprintf() fails

2025-02-10T10:58:53+00:00

Add a comment in dma_request_chan() to clarify kasprintf() missing return value check and it is correct functionality. Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20250205145757.889247-4-andriy.shevchenko@linux.intel.com Signed-off-by: Vinod Koul

dmaengine: avoid non-constant format string

2024-08-29T17:19:53+00:00

Using an arbitrary string as a printf-style format can be a security problem if that string contains % characters, as the optionalal -Wformat-security flag points out: drivers/dma/dmaengine.c: In function '__dma_async_device_channel_register': drivers/dma/dmaengine.c:1073:17: error: format not a string literal and no format arguments [-Werror=format-security] 1073 | dev_set_name(&chan->dev->device, name); | ^~~~~~~~~~~~ Change this newly added instance to use "%s" as the format instead to pass the actual name. Fixes: 10b8e0fd3f72 ("dmaengine: add channel device name to channel registration") Signed-off-by: Arnd Bergmann Reviewed-by: Dave Jiang Link: https://lore.kernel.org/r/20240719102319.546622-1-arnd@kernel.org Signed-off-by: Vinod Koul

dmaengine: add channel device name to channel registration

2024-06-11T18:25:34+00:00

Channel device name is used for sysfs, but also by dmatest filter function. With dynamic channel registration, channels can be registered after dma controller registration. Users may want to have specific channel names. If name is NULL, the channel name relies on previous implementation, dmachan. Signed-off-by: Amelie Delaunay Link: https://lore.kernel.org/r/20240531150712.2503554-11-amelie.delaunay@foss.st.com Signed-off-by: Vinod Koul

dmaengine: fix NULL pointer in channel unregistration function

2023-12-21T16:14:29+00:00

__dma_async_device_channel_register() can fail. In case of failure, chan->local is freed (with free_percpu()), and chan->local is nullified. When dma_async_device_unregister() is called (because of managed API or intentionally by DMA controller driver), channels are unconditionally unregistered, leading to this NULL pointer: [ 1.318693] Unable to handle kernel NULL pointer dereference at virtual address 00000000000000d0 [...] [ 1.484499] Call trace: [ 1.486930] device_del+0x40/0x394 [ 1.490314] device_unregister+0x20/0x7c [ 1.494220] __dma_async_device_channel_unregister+0x68/0xc0 Look at dma_async_device_register() function error path, channel device unregistration is done only if chan->local is not NULL. Then add the same condition at the beginning of __dma_async_device_channel_unregister() function, to avoid NULL pointer issue whatever the API used to reach this function. Fixes: d2fb0a043838 ("dmaengine: break out channel registration") Signed-off-by: Amelie Delaunay Reviewed-by: Dave Jiang Link: https://lore.kernel.org/r/20231213160452.2598073-1-amelie.delaunay@foss.st.com Signed-off-by: Vinod Koul

dmaengine: Simplify dma_async_device_register()

2023-08-21T13:06:46+00:00

There are a lot of duplicate codes for checking if the dma has some capability. Define a temporary macro that is used to check if the dma claims some capability and if the corresponding function is implemented. Signed-off-by: Yajun Deng Reviewed-by: Dave Jiang Link: https://lore.kernel.org/r/20230815072346.2798927-1-yajun.deng@linux.dev Signed-off-by: Vinod Koul

dmaengine: Actually use devm_add_action_or_reset()

2023-03-17T17:43:42+00:00

It appears that the commit a1beaa50b583 ("dmaengine: Simplify dmaenginem_async_device_register() function") mentions devm_add_action_or_reset() the actual change utilised devm_add_action() call by mistake. Fix the issue by switching to devm_add_action_or_reset(). Fixes: a1beaa50b583 ("dmaengine: Simplify dmaenginem_async_device_register() function") Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20230213112138.32118-1-andriy.shevchenko@linux.intel.com Signed-off-by: Vinod Koul