)]}'
{
  "commit": "e832dee88b5ea98e7e13b8249249b97ab953e846",
  "tree": "9532d46c78a07a331ce2e0476fd890706bd31ea4",
  "parents": [
    "6935f716cb1a696f04df8f8c6e8f978993db20e4"
  ],
  "author": {
    "name": "Jonas Larsson",
    "email": "ljonas@google.com",
    "time": "Fri Jan 17 13:20:04 2020 -0800"
  },
  "committer": {
    "name": "Jonas Larsson",
    "email": "ljonas@google.com",
    "time": "Fri Jan 24 00:37:48 2020 +0000"
  },
  "message": "columbia: reduce GPU CMA carveout to 32M\n\nOut of the ~640 total memory set aside for CMA, GPU reserved 128M\non boot. That leaves less for other CMA users such as VPU\nand dma-buf backed GL textures.\n\nWhen the GPU driver needs to allocate memory that doesn\u0027t fit in the\ncarveout it will try generic CMA. By reducing the GPU carveout we can\nutilize the CMA region with greater flexibility; GPU can still alloc\nas before until CMA runs out, and if it doesn\u0027t other users can get\ntheir slices of CMA cake.\n\nGPU carveout is also never cacheable, so for performance reasons we\nwant all large allocations ever touched by CPU to be in generic CMA.\n\nenterprise already has this exact change.\n\nChange-Id: I1fdfe1d5ed0218cf613deffce0ff0a1732b1ecde\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "e200219ea8bb7553ad4f11462ca81784b1ecb937",
      "old_mode": 33188,
      "old_path": "arch/arm64/boot/dts/freescale/fsl-imx8mm.dtsi",
      "new_id": "99624c9e33f8e873d6dd89a65e8f3ade149f05d1",
      "new_mode": 33188,
      "new_path": "arch/arm64/boot/dts/freescale/fsl-imx8mm.dtsi"
    }
  ]
}