)]}'
{
  "commit": "f72b89c40b7aecdd5042c8d411c11fdb48d3d9b2",
  "tree": "b0e466820c3e15e7a76f77d563ff0cc28ce56642",
  "parents": [
    "0ea9fe43fd70b36756ac636624e94ba2b6c6695f"
  ],
  "author": {
    "name": "Ben Clayton",
    "email": "bclayton@google.com",
    "time": "Wed Jun 19 16:39:11 2019 +0100"
  },
  "committer": {
    "name": "Ben Clayton",
    "email": "bclayton@google.com",
    "time": "Thu Jun 20 21:49:41 2019 +0000"
  },
  "message": "SpirvShader: Optimize pointer OOB masking\n\nReduces the amount of LLVM IR produced when the limits are compile-time known.\n\nReduces full test execution time of dEQP-VK.ubo.random.all_out_of_order_offsets.45 by about half.\n\nFor dEQP-VK.ubo.random.all_shared_buffer.48:\n• Reduces number of pre-opt LLVM IR instructions by about 40%.\n• Reduces full test execution time by about 30%\n\nBug: b/135609394\nChange-Id: I3c46d56910bb8373a355ab8d570d3093944afc84\nReviewed-on: https://swiftshader-review.googlesource.com/c/SwiftShader/+/33053\nReviewed-by: Chris Forbes \u003cchrisforbes@google.com\u003e\nReviewed-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nTested-by: Ben Clayton \u003cbclayton@google.com\u003e\nKokoro-Presubmit: kokoro \u003cnoreply+kokoro@google.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "440054278e97240c4129acfa93254265fa56088f",
      "old_mode": 33188,
      "old_path": "src/Pipeline/SpirvShader.hpp",
      "new_id": "ba8f399a2172c0e114f6121b613f57b66e448bc3",
      "new_mode": 33188,
      "new_path": "src/Pipeline/SpirvShader.hpp"
    }
  ]
}