linux.git/kernel/dma/direct.c, branch v5.8-rc3

dma-direct: add missing set_memory_decrypted() for coherent mapping

2020-06-23T12:13:54+00:00

When a coherent mapping is created in dma_direct_alloc_pages(), it needs
to be decrypted if the device requires unencrypted DMA before returning.

Fixes: 3acac065508f ("dma-mapping: merge the generic remapping helpers into dma-direct")
Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-direct: check return value when encrypting or decrypting memory

2020-06-17T07:29:38+00:00

__change_page_attr() can fail which will cause set_memory_encrypted() and
set_memory_decrypted() to return non-zero.

If the device requires unencrypted DMA memory and decryption fails, simply
free the memory and fail.

If attempting to re-encrypt in the failure path and that encryption fails,
there is no alternative other than to leak the memory.

Fixes: c10f07aa27da ("dma/direct: Handle force decryption for DMA coherent buffers in common code")
Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-direct: re-encrypt memory if dma_direct_alloc_pages() fails

2020-06-17T07:29:38+00:00

If arch_dma_set_uncached() fails after memory has been decrypted, it needs
to be re-encrypted before freeing.

Fixes: fa7e2247c572 ("dma-direct: make uncached_kernel_address more general")
Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-direct: always align allocation size in dma_direct_alloc_pages()

2020-06-17T07:29:38+00:00

dma_alloc_contiguous() does size >> PAGE_SHIFT and set_memory_decrypted()
works at page granularity.  It's necessary to page align the allocation
size in dma_direct_alloc_pages() for consistent behavior.

This also fixes an issue when arch_dma_prep_coherent() is called on an
unaligned allocation size for dma_alloc_need_uncached() when
CONFIG_DMA_DIRECT_REMAP is disabled but CONFIG_ARCH_HAS_DMA_SET_UNCACHED
is enabled.

Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-direct: mark __dma_direct_alloc_pages static

2020-06-17T07:29:37+00:00

Signed-off-by: Christoph Hellwig

dma-direct: re-enable mmap for !CONFIG_MMU

2020-06-17T07:29:31+00:00

nommu configfs can trivially map the coherent allocations to user space,
as no actual page table setup is required and the kernel and the user
space programs share the same address space.

Fixes: 62fcee9a3bd7 ("dma-mapping: remove CONFIG_ARCH_NO_COHERENT_DMA_MMAP")
Signed-off-by: Christoph Hellwig 
Reported-by: dillon min 
Reviewed-by: Vladimir Murzin 
Tested-by: dillon min

dma-direct: atomic allocations must come from atomic coherent pools

2020-04-25T11:17:05+00:00

When a device requires unencrypted memory and the context does not allow
blocking, memory must be returned from the atomic coherent pools.

This avoids the remap when CONFIG_DMA_DIRECT_REMAP is not enabled and the
config only requires CONFIG_DMA_COHERENT_POOL.  This will be used for
CONFIG_AMD_MEM_ENCRYPT in a subsequent patch.

Keep all memory in these pools unencrypted.  When set_memory_decrypted()
fails, this prohibits the memory from being added.  If adding memory to
the genpool fails, and set_memory_encrypted() subsequently fails, there
is no alternative other than leaking the memory.

Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-pool: add additional coherent pools to map to gfp mask

2020-04-20T10:09:40+00:00

The single atomic pool is allocated from the lowest zone possible since
it is guaranteed to be applicable for any DMA allocation.

Devices may allocate through the DMA API but not have a strict reliance
on GFP_DMA memory.  Since the atomic pool will be used for all
non-blockable allocations, returning all memory from ZONE_DMA may
unnecessarily deplete the zone.

Provision for multiple atomic pools that will map to the optimal gfp
mask of the device.

When allocating non-blockable memory, determine the optimal gfp mask of
the device and use the appropriate atomic pool.

The coherent DMA mask will remain the same between allocation and free
and, thus, memory will be freed to the same atomic pool it was allocated
from.

__dma_atomic_pool_init() will be changed to return struct gen_pool *
later once dynamic expansion is added.

Signed-off-by: David Rientjes 
Signed-off-by: Christoph Hellwig

dma-direct: fix data truncation in dma_direct_get_required_mask()

2020-04-08T18:52:24+00:00

The upper 32-bit physical address gets truncated inadvertently
when dma_direct_get_required_mask() invokes phys_to_dma_direct().
This results in dma_addressing_limited() return incorrect value
when used in platforms with LPAE enabled.
Fix it here by explicitly type casting 'max_pfn' to phys_addr_t
in order to prevent overflow of intermediate value while evaluating
'(max_pfn - 1) << PAGE_SHIFT'.

Signed-off-by: Kishon Vijay Abraham I 
Signed-off-by: Christoph Hellwig

Merge tag 'dma-mapping-5.7' of git://git.infradead.org/users/hch/dma-mapping

2020-04-04T17:12:47+00:00

Pull dma-mapping updates from Christoph Hellwig:

 - fix an integer overflow in the coherent pool (Kevin Grandemange)

 - provide support for in-place uncached remapping and use that for
   openrisc

 - fix the arm coherent allocator to take the bus limit into account

* tag 'dma-mapping-5.7' of git://git.infradead.org/users/hch/dma-mapping:
  ARM/dma-mapping: merge __dma_supported into arm_dma_supported
  ARM/dma-mapping: take the bus limit into account in __dma_alloc
  ARM/dma-mapping: remove get_coherent_dma_mask
  openrisc: use the generic in-place uncached DMA allocator
  dma-direct: provide a arch_dma_clear_uncached hook
  dma-direct: make uncached_kernel_address more general
  dma-direct: consolidate the error handling in dma_direct_alloc_pages
  dma-direct: remove the cached_kernel_address hook
  dma-coherent: fix integer overflow in the reserved-memory dma allocation