)]}'
{
  "commit": "58bca4a8fa90fcf9069379653b396b2cec642f7f",
  "tree": "483c535136b5b168f36326956453e80ce5aa6543",
  "parents": [
    "64ebe987311853ea857a244439de5b947a4b1b07",
    "64d70fe5d3640e1a89790ed21120921278f8cb86"
  ],
  "author": {
    "name": "Linus Torvalds",
    "email": "torvalds@linux-foundation.org",
    "time": "Wed Apr 04 17:13:43 2012 -0700"
  },
  "committer": {
    "name": "Linus Torvalds",
    "email": "torvalds@linux-foundation.org",
    "time": "Wed Apr 04 17:13:43 2012 -0700"
  },
  "message": "Merge branch \u0027for-linus\u0027 of git://git.linaro.org/people/mszyprowski/linux-dma-mapping\n\nPull DMA mapping branch from Marek Szyprowski:\n \"Short summary for the whole series:\n\n  A few limitations have been identified in the current dma-mapping\n  design and its implementations for various architectures.  There exist\n  more than one function for allocating and freeing the buffers:\n  currently these 3 are used dma_{alloc, free}_coherent,\n  dma_{alloc,free}_writecombine, dma_{alloc,free}_noncoherent.\n\n  For most of the systems these calls are almost equivalent and can be\n  interchanged.  For others, especially the truly non-coherent ones\n  (like ARM), the difference can be easily noticed in overall driver\n  performance.  Sadly not all architectures provide implementations for\n  all of them, so the drivers might need to be adapted and cannot be\n  easily shared between different architectures.  The provided patches\n  unify all these functions and hide the differences under the already\n  existing dma attributes concept.  The thread with more references is\n  available here:\n\n    http://www.spinics.net/lists/linux-sh/msg09777.html\n\n  These patches are also a prerequisite for unifying DMA-mapping\n  implementation on ARM architecture with the common one provided by\n  dma_map_ops structure and extending it with IOMMU support.  More\n  information is available in the following thread:\n\n    http://thread.gmane.org/gmane.linux.kernel.cross-arch/12819\n\n  More works on dma-mapping framework are planned, especially in the\n  area of buffer sharing and managing the shared mappings (together with\n  the recently introduced dma_buf interface: commit d15bd7ee445d\n  \"dma-buf: Introduce dma buffer sharing mechanism\").\n\n  The patches in the current set introduce a new alloc/free methods\n  (with support for memory attributes) in dma_map_ops structure, which\n  will later replace dma_alloc_coherent and dma_alloc_writecombine\n  functions.\"\n\nPeople finally started piping up with support for merging this, so I\u0027m\nmerging it as the last of the pending stuff from the merge window.\nLooks like pohmelfs is going to wait for 3.5 and more external support\nfor merging.\n\n* \u0027for-linus\u0027 of git://git.linaro.org/people/mszyprowski/linux-dma-mapping:\n  common: DMA-mapping: add NON-CONSISTENT attribute\n  common: DMA-mapping: add WRITE_COMBINE attribute\n  common: dma-mapping: introduce mmap method\n  common: dma-mapping: remove old alloc_coherent and free_coherent methods\n  Hexagon: adapt for dma_map_ops changes\n  Unicore32: adapt for dma_map_ops changes\n  Microblaze: adapt for dma_map_ops changes\n  SH: adapt for dma_map_ops changes\n  Alpha: adapt for dma_map_ops changes\n  SPARC: adapt for dma_map_ops changes\n  PowerPC: adapt for dma_map_ops changes\n  MIPS: adapt for dma_map_ops changes\n  X86 \u0026 IA64: adapt for dma_map_ops changes\n  common: dma-mapping: introduce generic alloc() and free() methods\n",
  "tree_diff": []
}