)]}'
{
  "log": [
    {
      "commit": "c7c57b763c1b16922bc589bff93d48101e1c452c",
      "tree": "6f0876c341b58cad330b4befd97997845aa64b4a",
      "parents": [
        "e353aa8e16e11fa41face83eadb7f75f5e4ddd30"
      ],
      "author": {
        "name": "yuanhecai",
        "email": "yuanhecai@loongson.cn",
        "time": "Wed Apr 01 01:32:27 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Apr 14 21:53:53 2026"
      },
      "message": "LoongArch: Defer filter pointer retrieval to fix out-of-bounds\n\nChange-Id: I9790c89630d5a1d2c0ecbce484a6e56003be73c1\n"
    },
    {
      "commit": "e353aa8e16e11fa41face83eadb7f75f5e4ddd30",
      "tree": "4d27d98ba9e53aa7bd68168a327870c2a2408722",
      "parents": [
        "b1f431c1ebdd386d227599c712d0a7b68fcfc8d0"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Apr 14 18:36:30 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Apr 14 18:36:30 2026"
      },
      "message": "vpx_temporal_svc_encoder: remove unused global\n\nFixes a warning with recent versions of clang:\n\n```\nexamples/vpx_temporal_svc_encoder.c|35 col 20| warning: variable\n\u0027exec_name\u0027 set but not used [-Wunused-but-set-global]\n```\n\nChange-Id: I25e2963f666a8b36b88028dbf49e6e486741e5e7\n"
    },
    {
      "commit": "b1f431c1ebdd386d227599c712d0a7b68fcfc8d0",
      "tree": "42b32247017d4a274af1a8f6aa84583271d2d6f2",
      "parents": [
        "a5e2e6528374c8ae7cfe8ef6b7aad4fc327c8688"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Apr 08 22:10:19 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Apr 09 18:29:59 2026"
      },
      "message": "decode_api_test.cc: fix visual studio warnings\n\nAfter:\ndb2e24852 vp9_decodeframe.c: move ref frame index check earlier\n\n```\ntest\\decode_api_test.cc(216,3): warning\n  C4267: \u0027argument\u0027: conversion from \u0027size_t\u0027 to \u0027unsigned int\u0027,\n  possible loss of data\ntest\\decode_api_test.cc(233,3): warning\n  C4267: \u0027argument\u0027: conversion from \u0027size_t\u0027 to \u0027unsigned int\u0027,\n  possible loss of data\n```\n\nChange-Id: I19de723cb5747ee1b4235c4596c6fd987062f930\n"
    },
    {
      "commit": "a5e2e6528374c8ae7cfe8ef6b7aad4fc327c8688",
      "tree": "ae0aff9e9f729a2960d5c43f639bb51ea73e56ae",
      "parents": [
        "0226b9516c98f26d7b68970bbccda3abdacbc7bb"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Thu Apr 09 00:32:51 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Thu Apr 09 02:31:06 2026"
      },
      "message": "vp8_de_alloc_frame_buffers: more cleanup\n\nBug: 500638476\nChange-Id: I61d7c47f76119409160d99972f8ba432863e7b54\n"
    },
    {
      "commit": "0226b9516c98f26d7b68970bbccda3abdacbc7bb",
      "tree": "ec1d5f18baf84c6a5773129fbe6c8a8bba96a286",
      "parents": [
        "e914ae9e8d253d2230009778846475334eb451e5"
      ],
      "author": {
        "name": "Jan-Ivar Bruaroey",
        "email": "jbruaroey@mozilla.com",
        "time": "Wed Apr 08 19:58:54 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Apr 08 22:19:42 2026"
      },
      "message": "vp8_de_alloc_frame_buffers: clear mi/postproc_state\n\nThis avoids leaving the decoder in an inconsistent state, which can lead\nto a use-after-free (UAF) through `VP8_COMMON::mi`,\n`VP8_COMMON::frame_to_show`, and `VP8_COMMON::postproc_state` with\n`VP8_MFQE` enabled when decoding proceeds after an earlier allocation\nfailure.\n\nBug: 500638476\nChange-Id: I0adb05c63db897093df2aa283b51888543a107ea\n"
    },
    {
      "commit": "e914ae9e8d253d2230009778846475334eb451e5",
      "tree": "90dd8110c138c874ec8cf6ebd81af1575a5a350e",
      "parents": [
        "db2e2485279f61bbcfe688507c099f1cd39ed0f9"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Tue Apr 07 20:47:52 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Tue Apr 07 22:15:33 2026"
      },
      "message": "Change asm volatile to __asm__ __volatile__\n\nThe keywords with double underscores are not considered as GNU\nextensions and can be compiled with -std\u003dc11 rather than -std\u003dgnu11.\n\nChange-Id: I2f84ba87265a398a3676b4fd20f9197964f0edf1\n"
    },
    {
      "commit": "db2e2485279f61bbcfe688507c099f1cd39ed0f9",
      "tree": "459f917357a273b593fdda26468a3c3437d216de",
      "parents": [
        "606b96d664dfb65b5d50df9a8b7774ac8d0065ab"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Apr 06 22:31:47 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Apr 06 22:39:26 2026"
      },
      "message": "vp9_decodeframe.c: move ref frame index check earlier\n\nIn 75b6cfe1c Prevent read to invalid RefBuffer, reference buffer index\nchecking was added to `setup_frame_size_with_refs()`. Prior to this\ncall, however, `cm-\u003eframe_refs[]` are populated using the index from\n`cm-\u003eref_frame_map[]`, which may be `-1` (`INVALID_IDX`). In that case\n`pbi-\u003ecommon.buffer_pool-\u003eframe_bufs[]` would be accessed with a\nnegative index, potentially resulting in a crash.\n\nThe checks from `setup_frame_size_with_refs()` are moved to\n`read_uncompressed_header()` before populating `frame_refs[]`.\n\nTest and fix based on a patch from kaixuanli0131 at gmail.\n\nBug: 499206650\nChange-Id: I25b0810ed5c619facf9226edd069439ca2ce1166\n"
    },
    {
      "commit": "606b96d664dfb65b5d50df9a8b7774ac8d0065ab",
      "tree": "baf7c6f1f6c9a38ea2961bc228884cd28bd69b1a",
      "parents": [
        "7074f15477e61d9818442edaf4ea6ba79aa9baac"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sat Apr 04 01:08:11 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sat Apr 04 17:36:55 2026"
      },
      "message": "Enable the ISO C11 standard\n\nC11 allows us to use static_assert().\n\nThe public headers should remain C99.\n\nChange-Id: I6e0b80c4c26e1dee41dc2ef69215fe7f0ea611bb\n"
    },
    {
      "commit": "7074f15477e61d9818442edaf4ea6ba79aa9baac",
      "tree": "2e8f752e2a636f92542cac0f04aca98d1d4ef5ad",
      "parents": [
        "47ac1ec7f3de7d7cb3d070844c427c8f1fa9d6fc"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Apr 02 21:10:19 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Apr 02 21:10:19 2026"
      },
      "message": "{vp9_cx_iface,vp9_decoder}.c: add some missing includes\n\nChange-Id: I1780d97849a30095911a727d944f01974ba9220a\n"
    },
    {
      "commit": "47ac1ec7f3de7d7cb3d070844c427c8f1fa9d6fc",
      "tree": "60f4b0a588167f91f471ba1fdb6cec72c9d5eaef",
      "parents": [
        "c18b190ca003858b321e170e027ef09779ff746e"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Apr 01 21:04:36 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Apr 02 01:44:53 2026"
      },
      "message": "vp9_decodeframe: promote row-mt calcs to size_t\n\nThis avoids integer overflow when decoding very large frames. `size_t`\nis acceptable for 32-bit as a check in `vp9_dec_alloc_row_mt_mem()`,\nwhich was added in:\n  d9a548c01 vp9: fix for out-of-bounds write in decode coeffs\nwould have already failed. The calculations fixed in this change are\nsimilar to the ones in the earlier change.\n\nBug: 496830326\nChange-Id: I8ee5164b15a6a00060cbab69ffe729f22e465f17\n"
    },
    {
      "commit": "c18b190ca003858b321e170e027ef09779ff746e",
      "tree": "e2be2d99e4beb52fe0a7f0eeb54822961431e15c",
      "parents": [
        "d9a548c01fb6d3eab69459ea8faa89e1e5f1e8af"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Apr 01 18:38:03 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Apr 01 18:38:03 2026"
      },
      "message": "test/android/Android.mk: build tests w/-std\u003dc++17\n\nC++17 is a requirement for building Googletest since:\n  f14de9d29 Update to GoogleTest 1.17.0\n\nChange-Id: I3b624ee3261557037074be51246d89d45b4e8ada\n"
    },
    {
      "commit": "d9a548c01fb6d3eab69459ea8faa89e1e5f1e8af",
      "tree": "df65a6a26843bd2e1b9dc5ca7831755a06307139",
      "parents": [
        "59d0c766707371bd8c35a189c556ce530cc2fc8b"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Mar 30 22:23:06 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Wed Apr 01 01:05:20 2026"
      },
      "message": "vp9: fix for out-of-bounds write in decode coeffs\n\nAdd a check and cast on num_sbs.\n\nBug: 496830326\n\nChange-Id: Ib18a07eed289da7253a322f6504e5034d7f17351\n"
    },
    {
      "commit": "59d0c766707371bd8c35a189c556ce530cc2fc8b",
      "tree": "c40f76b7b0af62a7a8842c2ba7409aeacd586787",
      "parents": [
        "f14de9d2973c80b4131265dcc146ed37ad5727cb"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Mar 31 19:30:26 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Mar 31 19:38:09 2026"
      },
      "message": "gen_msvs_vcxproj: set C++ language standard to C++17\n\nThis is required since:\n  f14de9d29 Update to GoogleTest 1.17.0\n\nVisual Studio defaults to `/std:c++14`.\n\nFixes:\n googletest\\src\\include\\gtest\\internal\\gtest-port.h(273,1):\n error C1189: #error:  C++ versions less than C++17 are not supported.\n\nChange-Id: I1becf88a9063c9889c3097928658a1cfe41e58dd\n"
    },
    {
      "commit": "f14de9d2973c80b4131265dcc146ed37ad5727cb",
      "tree": "6a12fd646a21a37420829ef8afee50f747281c15",
      "parents": [
        "d20e271c144ee80e5c6025b26bc01a8b8c59a187"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sun Mar 29 01:31:48 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Tue Mar 31 02:35:24 2026"
      },
      "message": "Update to GoogleTest 1.17.0\n\nNew Windows code in gtest-port.cc makes full use of the Notification\nclass, so we can no longer use a stub Notification class. Apply a\npatch to implement the Notification class on Windows by using a Windows\nmanual-reset event object.\n\nChange-Id: Ifd7d994b9303e5c96b3e1877648fb35830c91e17\n"
    },
    {
      "commit": "d20e271c144ee80e5c6025b26bc01a8b8c59a187",
      "tree": "62ef756eefd85ff96bb2edb1165a315c033ac84d",
      "parents": [
        "97a59b0460cab900841cb177c2b0e03fbbd25b0e"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Mar 30 21:17:30 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Mar 30 21:25:35 2026"
      },
      "message": "vp9; move source input check to validate_img\n\nAnd remove the check:\nctx-\u003eoxcf.input_bit_depth \u003e 8\n\nBug: 488585490\nChange-Id: Ibf73572d9db20e8a731fffe363675a7f8173d4f7\n"
    },
    {
      "commit": "97a59b0460cab900841cb177c2b0e03fbbd25b0e",
      "tree": "1e23bc814aae6cdb894f11f22f2357683ea685fd",
      "parents": [
        "4d868e2856b2903eb851ad0133d95ad27d32eb49"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sun Mar 29 18:07:28 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Mon Mar 30 20:36:50 2026"
      },
      "message": "Allow patch files to have trailing spaces\n\nA patch file generated by the \"git diff\" command may have lines\nconsisting of a single space character.\n\nChange-Id: If0086318107886005273940466bb190b023c74e8\n"
    },
    {
      "commit": "4d868e2856b2903eb851ad0133d95ad27d32eb49",
      "tree": "7135a1ea91246f7f86c41f474f37784f62e62b8e",
      "parents": [
        "8cf7442ad1260a98b9b22ad043f4eb9203b6f04e"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sat Mar 28 06:54:35 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Mon Mar 30 20:33:51 2026"
      },
      "message": "encoder_encode: Omit two res \u003d\u003d VPX_CODEC_OK tests\n\nAt those two points, `res` is equal to VPX_CODEC_OK, so\nres \u003d\u003d VPX_CODEC_OK is true.\n\nChange-Id: Ibfdc40378f4db8cdd6f0121430048c109438d5ad\n"
    },
    {
      "commit": "8cf7442ad1260a98b9b22ad043f4eb9203b6f04e",
      "tree": "f0fcf02ada744b941ca0172a3c8070144b65b4a2",
      "parents": [
        "15dca2fe657ea80f929bc99cf6edba333a110322"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sun Mar 29 06:00:15 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Sun Mar 29 06:00:15 2026"
      },
      "message": "Remove bit_depth member of struct TestSSEFuncs\n\nThe bit_depth member of struct TestSSEFuncs is not used.\n\nChange-Id: Ie179e07145c0a7ef46ea77ac63c2f4c3c3ba46dd\n"
    },
    {
      "commit": "15dca2fe657ea80f929bc99cf6edba333a110322",
      "tree": "7c29749a09cb7f8dcaf227c585458fda55ee9058",
      "parents": [
        "090dd8b8cb3a11dec5250bef62aba9b99eb46636"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Mar 27 20:45:27 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Mar 27 20:45:27 2026"
      },
      "message": "README: update bug tracker link\n\nUse a direct link to the creation page to avoid libvpx bugs being filed\nunder the top-level WebM component when the \u0027Create Issue\u0027 button is\nused.\n\nChange-Id: I70c788172d18580dd750a77f040189817c9a5076\n"
    },
    {
      "commit": "090dd8b8cb3a11dec5250bef62aba9b99eb46636",
      "tree": "4fda5d2ebecd0bf1fc52b26b81617c5c485831c0",
      "parents": [
        "3fce57ecc905d95a4619f33d09851d68c5a88663"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Mar 23 22:09:28 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Wed Mar 25 23:17:34 2026"
      },
      "message": "vp9: Add check to validate source input\n\nReturn invalid_params.\n\nThe check is wrapped around a new control:\nVP9E_SET_VALIDATE_INPUT_HBD,\nand is enabled by default.\n\nCheck is done only CONFIG_VP9_HIGHBITDEPTH build\nand for bitdepth \u003e 8.\n\nBug: 488585490\nChange-Id: Ic79dff1bd314d4b197087d3c46d315ece8fd355c\n"
    },
    {
      "commit": "3fce57ecc905d95a4619f33d09851d68c5a88663",
      "tree": "09c49864bfa93aaf233629a7bb9eb5efb8b30c49",
      "parents": [
        "d51fafc086f90e858a4f8de47b81df54d6b69da5"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Mar 23 23:20:46 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Mar 23 23:20:46 2026"
      },
      "message": "encode_api_test: replace kHeight w/kWidth for stride\n\nin InvalidUVStrides. Stride is associated with width, height was used\nmistakenly; they have the same value in this test.\n\nChange-Id: I88634a94af40ea3c0ff846a76f267003011653c9\n"
    },
    {
      "commit": "d51fafc086f90e858a4f8de47b81df54d6b69da5",
      "tree": "1b7b078cd2e8c8cdab046b0b471f8fbed54dc1b6",
      "parents": [
        "915643b9486b0d73aba95213bf08905c30c7159c"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Mar 16 23:27:39 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Mar 18 22:24:38 2026"
      },
      "message": "vpx_image,img_alloc_helper: verify vpx_img_fmt_t\n\nExplicitly check for known values rather than just failing when the\nformat is equal to `VPX_IMG_FMT_NONE`.\n\nChange-Id: Iae9715dc0ca7de8484b14f538ae567aace96bfe3\n"
    },
    {
      "commit": "915643b9486b0d73aba95213bf08905c30c7159c",
      "tree": "4f72a0f0220c84281945b104f8dfec474ab5aa71",
      "parents": [
        "3894fccb5e0c7ebfd01a7478cc468b9b0baaf4e8"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Sun Mar 15 01:21:45 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Mar 18 22:24:38 2026"
      },
      "message": "vp[89] encoders: reject unequal UV strides\n\nThe `vpx_image_t` from the external API is converted to an internal\n`YV12_BUFFER_CONFIG`, which assumes U and V strides are the same. This\navoids a potential crash or corrupted chroma if they differ.\n\nBug: 492213293\nChange-Id: I3c878bdd330a8f7f838ac3bce8b66201c49f537d\n"
    },
    {
      "commit": "3894fccb5e0c7ebfd01a7478cc468b9b0baaf4e8",
      "tree": "68288476745031b68730c4672ebfda458bbd8754",
      "parents": [
        "42d580255a7b842c66f655f035ff3038c0c71d14"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Sun Mar 15 01:19:41 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Mar 18 22:24:38 2026"
      },
      "message": "vp9,encoder_encode: fix validate_img() check\n\nFail early on error return. Previously only a positive return was\nchecked and `res` was later reused to capture the return from\n`pick_quickcompress_mode()`, losing any failure from `validate_img()`.\n\nBug: 492213293\nChange-Id: I595db1c1b4e5eea595c1acff6580ef00a2375c06\n"
    },
    {
      "commit": "42d580255a7b842c66f655f035ff3038c0c71d14",
      "tree": "8d056fdbf486595b1a37d1f50aa48ba56bbb96d9",
      "parents": [
        "e1dd14963a005523d0eb3a72954921934b3a8be7"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Mar 12 02:15:52 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Mar 17 02:15:28 2026"
      },
      "message": "encode_api_test: add repro for b/488585490\n\nThis causes a read overflow in vp9_cat6_high10_high_cost[].\n\nNote this differs from the settings used by WebCodecs and the sequence\nin the original POC. Using a single frame (6) from that sequence and\nlossless encoding produces a more extreme overflow.\n\nAn additional test case may be added, but converting the POC to a unit\ntest wasn\u0027t causing a failure as encoding diverged on frame 6 from that\nobserved in Chrome (likely incomplete settings transfer).\n\nBug: 488585490\nChange-Id: Id105d73c54556b17590af44cfa710a8f5cd3f9b2\n"
    },
    {
      "commit": "e1dd14963a005523d0eb3a72954921934b3a8be7",
      "tree": "5a8b56cfa320a2d66605e2830d983128926b35c2",
      "parents": [
        "e7d204077442c53ecd3dc2d496a0e6df4358c357"
      ],
      "author": {
        "name": "CodeByMoriarty",
        "email": "alacambradev@gmail.com",
        "time": "Sun Mar 08 10:53:45 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Mar 12 17:57:16 2026"
      },
      "message": "y4minput: fix integer overflow in buffer size calculations\n\nReject Y4M input files with dimensions large enough to cause integer\noverflow in the buffer size computations in y4m_input_open(). The\nbuffer sizes (dst_buf_sz, dst_buf_read_sz, aux_buf_sz) are computed\nusing int arithmetic that silently wraps for large dimension values,\ncausing malloc() to allocate a drastically undersized buffer.\nSubsequent frame processing then constructs plane pointers far past\nthe allocated buffer, leading to heap out-of-bounds access.\n\nAdd a check in parse_tags() that (int64_t)pic_w * pic_h \u003c\u003d INT_MAX/6,\nwhich provides enough headroom for the most demanding buffer formula\n(2 * 3 * pic_w * pic_h for 444 format at high bit-depth). Also reject\nnon-positive dimensions early in y4m_parse_tags().\n\nBug: 491509033, 491509034\nChange-Id: I0d36273b07eecee1bcda24b93e05db2e60647c7f\n"
    },
    {
      "commit": "e7d204077442c53ecd3dc2d496a0e6df4358c357",
      "tree": "be08b0ed3d84ec28311f86a753302e70ccadface",
      "parents": [
        "3068fa0ab44330e4e0397dd865fb5f9664aa5e24"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Wed Mar 11 19:48:52 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Wed Mar 11 19:48:52 2026"
      },
      "message": "y4minput.c: Set alpha plane to NULL, stride to 0\n\nWe do not allocate any space for the alpha plane in _y4m-\u003edst_buf. Set\n_img-\u003eplanes[VPX_PLANE_ALPHA] to NULL and _img-\u003estride[VPX_PLANE_ALPHA]\nto 0 to detect accidental use of the alpha plane.\n\nChange-Id: I58c11f52892281b2a8dba10726e6e95b3f8cd778\n"
    },
    {
      "commit": "3068fa0ab44330e4e0397dd865fb5f9664aa5e24",
      "tree": "8157a2b0d270ab4440404f6d3361aac2b384bd4c",
      "parents": [
        "9a2d3d1f46afbdfa9b9820a9fd3aacb084e65e2f"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Tue Mar 10 17:57:28 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Tue Mar 10 23:49:13 2026"
      },
      "message": "vp9-svc: validate spatial_layer_id\n\nValidate spatial_layer_id in ctrl_set_svc_layer_id()\n\nBug: 490810420\nChange-Id: I60d7a8524a4bacc693f6a6a607ad5f2deeec357a\n"
    },
    {
      "commit": "9a2d3d1f46afbdfa9b9820a9fd3aacb084e65e2f",
      "tree": "83006fa912b1051adb57d5bbb664d11d9a8a51da",
      "parents": [
        "ab5ec7a1852f634b81d29ade3c5fa74056498973"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Mar 02 23:04:57 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Mar 03 03:16:22 2026"
      },
      "message": "vp9_scale_references: fail if no free buffer is available\n\nIf `get_free_fb()` fails to return a buffer for use with scaling, fail\nimmediately. This avoids heap overflows (read) in code that assumes the\nframe has been scaled to the current resolution\n(`vp9_int_pro_motion_estimation()` and `combined_motion_search()`).\n\nA test case will be added in a follow up.\n\nThis matches the behavior in libaom since:\n  4a8c004b80 Move the assertion back to get_free_fb().\n  ...\n  Also change the scale_references() function to not ignore\n  get_free_fb() failure.\n\nBug: 487259772\nChange-Id: If8028e40173b06b67cc161735af730eeb99ac57e\n"
    },
    {
      "commit": "ab5ec7a1852f634b81d29ade3c5fa74056498973",
      "tree": "ab5b5d4da2dc1db84fbf85d59e70284d1d50a350",
      "parents": [
        "4fcebeabe58e79255d291acb4cead4ed7953149e"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Mar 02 22:51:36 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Mar 03 00:49:50 2026"
      },
      "message": "vp9_pick_inter_mode: fix buf offsets w/scaled refs\n\nWhen calculating SAD for `NEWMV` with the `LAST_FRAME`, check whether a\nscaled reference frame is available and update `xd-\u003eplane[].pre[0]`\noffsets accordingly. This matches the behavior in\n`combined_motion_search()`.\n\nThis fixes a heap overflow (read) when calculating the SAD when using\nscaled references.\n\nBug: 487259772\nChange-Id: I13cb02faa13d95b4f9dc4c8a49363c55d5e90efa\n"
    },
    {
      "commit": "4fcebeabe58e79255d291acb4cead4ed7953149e",
      "tree": "28748e7855a42c5ef2cb8da1154bc79d646de8ba",
      "parents": [
        "bb9e8fa1988d6176cd6eed0682537042a066988c"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 26 18:13:10 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 26 18:28:12 2026"
      },
      "message": "yuv_video_source: fix stride/width check w/16bpp\n\nIn the previous commit:\n  bb9e8fa198 yuv_video_source: fix reading files w/odd widths\nbitdepth was not taken into account when checking whether stride and\nwidth differed. This fixes test cases using yuv files containing\nVPX_IMG_FMT_I4??16 (notably VP9/EndToEndTestLarge.EndtoEndPSNR*).\n\nBug: 486910181\nChange-Id: I6a50d8879c1044b1d01b4e34e4b01792f776e808\n"
    },
    {
      "commit": "bb9e8fa1988d6176cd6eed0682537042a066988c",
      "tree": "ab87d94dd9e77123a7e39dce2c64b33810b3a7a3",
      "parents": [
        "6845d7229e1b1d9315f506a7c559ea1003d832ab"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 24 19:33:49 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Feb 25 18:16:47 2026"
      },
      "message": "yuv_video_source: fix reading files w/odd widths\n\nPreviously, the code assumed the allocation made by `vpx_img_alloc` was\nunpadded and read all of the data for a frame as a single block. Odd\ndimensions will be padded to a power of 2 with YUV 4:2:0, so the reads\nneed to be done row-wise in this case.\n\nThis fixes the valgrind issues in OssFuzz471723682; the test is\nre-enabled in this change. The original issue still reproduces if the\nfix from db312f06fb (vp8,calc_pframe_target_size: fix integer overflow)\nis reverted.\n\nBug: 486910181\nChange-Id: Iba6ef07be3ab92c8293ed92a0f0af82d80c90b76\n"
    },
    {
      "commit": "6845d7229e1b1d9315f506a7c559ea1003d832ab",
      "tree": "94be182c5d0ce95fcea4a24948bf77ea512bd78b",
      "parents": [
        "30f3852521b11b5e361ec1eaeef5a12730bfe90f"
      ],
      "author": {
        "name": "Andrey Filipenkov",
        "email": "kambalaDecapitator@gmail.com",
        "time": "Tue Feb 24 11:58:59 2026"
      },
      "committer": {
        "name": "Andrey Filipenkov",
        "email": "kambalaDecapitator@gmail.com",
        "time": "Tue Feb 24 19:03:19 2026"
      },
      "message": "add support for msvc v145 / VS 2026\n\nChange-Id: I24769c0794a9d82bc01c6c3998c155dc95b8377b\n"
    },
    {
      "commit": "30f3852521b11b5e361ec1eaeef5a12730bfe90f",
      "tree": "ab973ed884f1ed064cd971799f4d9b337fd24888",
      "parents": [
        "3ac99950b20b6693b03a48ebc300aba1d0f0dcfc"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Feb 23 19:26:11 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Feb 23 22:56:12 2026"
      },
      "message": "encode_api_test: disable OssFuzz471723682\n\nThis test (added in\ndb312f06fb vp8,calc_pframe_target_size: fix integer overflow) causes\nissues under valgrind related to the use of uninitialized values in\n`vpx_skin_pixel()`.\n\nBug: 486910181\nChange-Id: Iac1d5436fde18ac17e5dabd57a6a7518a73fd40f\n"
    },
    {
      "commit": "3ac99950b20b6693b03a48ebc300aba1d0f0dcfc",
      "tree": "62fc1bddab2b5161d69d177e6c3764b79f877094",
      "parents": [
        "73b417df365deba04acacf80830f2acf6d8f60a9"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 19 02:50:07 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Mon Feb 23 20:37:02 2026"
      },
      "message": "vpx_enc_fuzzer: condition max frames on deadline\n\nUse 150 frames for `VPX_DL_BEST_QUALITY` and 300 for\n`VPX_DL_GOOD_QUALITY` and `VPX_DL_REALTIME`. This may help reduce the\nlikelihood of timeouts with `VPX_DL_BEST_QUALITY`.\n\nBug: 484258317, 484277503\nChange-Id: I06aff5ec0e48e9d0cc03047045525cebb5991e0f\n"
    },
    {
      "commit": "73b417df365deba04acacf80830f2acf6d8f60a9",
      "tree": "c110a3e4ad1dea332476d15c8c871a453749b586",
      "parents": [
        "3c8cd520908217eb35691fdc3da280ce9c2aac2d"
      ],
      "author": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Thu Feb 19 13:57:49 2026"
      },
      "committer": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Mon Feb 23 09:20:51 2026"
      },
      "message": "Make direct calls to Neon variance paths in Neon subpel variance\n\nNow that we have Neon DotProd paths for subpel variance as well as\nvariance, we no longer need to have an indirect call to the variance\nfunction in Armv8.0 Neon subpel variance paths. Make a direct call to\nthe Neon variance path in these Neon subpel functions.\n\nThis is a port from SVT-AV1:\nhttps://gitlab.com/AOMediaCodec/SVT-AV1/-/merge_requests/2608\n\nOriginally authored by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e\n\nChange-Id: If94c9af098356ec857255496a435b7a4cbb3c522\n"
    },
    {
      "commit": "3c8cd520908217eb35691fdc3da280ce9c2aac2d",
      "tree": "fd7cd1c61592a7641b9d25364aa65522d6d0480f",
      "parents": [
        "6837859d04b3e6ad8b4b37ca339a34a7920c6f82"
      ],
      "author": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Thu Feb 19 13:57:49 2026"
      },
      "committer": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Mon Feb 23 09:20:44 2026"
      },
      "message": "Add Armv8.4 Neon DotProd subpel variance paths\n\nFor block widths \u003e\u003d 8, merge the source pixel bilinear interpolation\ninto the variance kernel. For 4x8 and 4x16 blocks - where merging\nbilinear interpolation into the variance kernel is not beneficial - keep\nthe original approach but with a direct call to the Neon DotProd\nvariance kernel. (The Armv8.0 Neon implementation remains fastest for\n4x4 blocks.)\n\nAlso add the relevant unit tests.\n\nThis is a port from SVT-AV1:\nhttps://gitlab.com/AOMediaCodec/SVT-AV1/-/merge_requests/2608\n\nOriginally authored by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e and\nGerda Zsejke More \u003cgerdazsejke.more@arm.com\u003e\n\nChange-Id: I9e3143c51c2898b570de027d908b25023fff0f9e\n"
    },
    {
      "commit": "6837859d04b3e6ad8b4b37ca339a34a7920c6f82",
      "tree": "852cca5d6be885f6f827dea66d9a1724808e92f7",
      "parents": [
        "db312f06fb50089e699f13a9c516daf44ce0f7d8"
      ],
      "author": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Thu Feb 19 13:57:48 2026"
      },
      "committer": {
        "name": "Li Zhang",
        "email": "li.zhang2@arm.com",
        "time": "Thu Feb 19 13:57:48 2026"
      },
      "message": "Remove Armv8.4 Neon DotProd 4x4 variance kernel\n\nThe Armv8.0 Neon implementation is faster for 4x4 blocks.\n\nThis is a port from SVT-AV1:\nhttps://gitlab.com/AOMediaCodec/SVT-AV1/-/merge_requests/2608\n\nOriginally authored by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e\n\nChange-Id: I9fd19b430ee6af7ddc9b1bec39a911dc3e813dbd\n"
    },
    {
      "commit": "db312f06fb50089e699f13a9c516daf44ce0f7d8",
      "tree": "0843c722a4633caab5a0fa4c51352ce7627c0293",
      "parents": [
        "3115db72821d14cac82a37c186fd092962b7f2f4"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Feb 18 23:52:30 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 19 03:30:13 2026"
      },
      "message": "vp8,calc_pframe_target_size: fix integer overflow\n\nWhen calculating `bits_in_section` with a high `cpi-\u003einter_frame_target`\nbitrate, the result may exceed a signed integer. This change first using\n64-bit math to calculate the product and then clamps the value to\n`INT_MAX`.\n\nBug: 471723682\nChange-Id: I216f46b1eb9ea48d7dc2b255b2ef0aaf8de1be8e\n"
    },
    {
      "commit": "3115db72821d14cac82a37c186fd092962b7f2f4",
      "tree": "d3fdd8befec8085ff0e8fad6bbb9fa1b7e31a333",
      "parents": [
        "1bb63fff1df29d1bdd13796b4309e176b60234d0"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Feb 18 23:14:33 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 19 03:29:19 2026"
      },
      "message": "examples/*.c: fix incorrect die_codec() usage\n\n`die_codec()` should only be used with an initialized codec context.\nWhen reporting errors with `vpx_codec_enc_config_default()`, `die()`\nshould be used instead. This fixes a potential crash on failure.\n\nChange-Id: Id7b5b0733728510b4c659c6fa6980da56d6c68a3\n"
    },
    {
      "commit": "1bb63fff1df29d1bdd13796b4309e176b60234d0",
      "tree": "54517d1a93da3c7c090653f60b02e11e0532f6ee",
      "parents": [
        "66cbb0134741180798f0f2d30e02baed039c9812"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 17 21:48:27 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 17 21:48:27 2026"
      },
      "message": "**README.libvpx: add missing fields / sync w/libaom\n\nThis fixes some warnings in a Chromium tool\n(depot_tools/metadata/scan.py).\n\nChange-Id: Ie809c530f358e51e7b30a12953c0d57ebe634977\n"
    },
    {
      "commit": "66cbb0134741180798f0f2d30e02baed039c9812",
      "tree": "3981e9e0a27120500c4caab40b10c6ccea3c7f19",
      "parents": [
        "eccfce582fad249e2bdb2ee36f9e0807a39df5ca"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 12 22:54:40 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 17 21:29:40 2026"
      },
      "message": "vp8_decode_frame: improve header / partition 0 check\n\n`first_partition_length_in_bytes` is\nconstructed from 24-bits, it can only be positive or 0. The only check\nthat is needed is that `data_end - data` is greater than or equal to the\nvalue.\n\nThis change moves the `first_partition_length_in_bytes` check after the\nparsing of the keyframe data to simplify the calculation. The size\nchecks done during keyframe parsing are now merged into a single check.\n\nChange-Id: I15f70ff8cdfed3ad7064ee9cf09e71cd0675aec1\n"
    },
    {
      "commit": "eccfce582fad249e2bdb2ee36f9e0807a39df5ca",
      "tree": "a5af8716d82669031e074ad038f1a1994177abb6",
      "parents": [
        "ad17f6195cc6df6e4ac52bbb62b6b63e17c948a0"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Feb 13 23:10:24 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Feb 13 23:10:24 2026"
      },
      "message": "rtcd.pl: use a fixed year for *rtcd.h Copyright\n\nThis allows builds to be stable across the years, which can be important\nfor package maintainers. The date chosen, 2017, matches the date in\nrtcd.pl.\n\nChange-Id: I8acfe29c6260cf85240d2e428ce2448c2b250624\n"
    },
    {
      "commit": "ad17f6195cc6df6e4ac52bbb62b6b63e17c948a0",
      "tree": "359df1fc6cfc1b03a14031862ba3357d50613812",
      "parents": [
        "dcf4253fca380bcb51f146d0447dc425797809e9"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Feb 12 22:51:19 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Feb 13 00:12:19 2026"
      },
      "message": "vp8_decode_frame: validate partition 0 size\n\nThe first partition size can not be 0. This avoids some slower than\nnecessary decodes with corrupt / fuzzed data.\n\nBug: 473051902\nChange-Id: I41662e873baf8d278d805ed1f7e85079c28a4bd8\n"
    },
    {
      "commit": "dcf4253fca380bcb51f146d0447dc425797809e9",
      "tree": "3a18a2772ccd902cba9170fc57c3a6dc74a79974",
      "parents": [
        "a0ac06f5d3c83f890cb502c80470ef0502343fec"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 21:56:19 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 21:56:19 2026"
      },
      "message": "vpx_enc_fuzzer: limit number of encoded frames\n\nPreviously the encode loop was only bounded by the data size. With the\nsmall resolutions being fuzzed this might result in hundreds of\nthousands of frames to encode, resulting in a timeout. The frames are\nnow limited to 300.\n\nBug: 482163265, 482328006, 482494306, 482594458\nChange-Id: I026de66933a267e1ce7e979005eea181bfd54a1e\n"
    },
    {
      "commit": "a0ac06f5d3c83f890cb502c80470ef0502343fec",
      "tree": "c8a6a5ad49e1a0ad8087ae8f6d3f3daa41401c99",
      "parents": [
        "11eb7346db677aabb67f8186a1504bd68127bbd8"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 00:46:32 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 00:48:06 2026"
      },
      "message": "Revert \"Increase alignment for wider SIMD optimization\"\n\nThis reverts commit 27b08ed69c760e405cbba3d90d81c47ef42ba59d.\n\nThis causes failures in Chrome:\n\n```\nthird_party/blink/tools/run_web_tests.py --debug \\\n  --no-show-results --nocheck-sys-deps \\\n  --no-retry-failures media/video-colorspace-yuv422.html\n...\n[1/1] media/video-colorspace-yuv422.html failed unexpectedly (image diff) 1.1559s\nFinally stop servers and clean up\n\n0 tests ran as expected, 1 didn\u0027t:\n    media/video-colorspace-yuv422.html\n```\n\nChange-Id: I9795e12525bb2c8d834073c62386d88bc2d37aaf\n"
    },
    {
      "commit": "11eb7346db677aabb67f8186a1504bd68127bbd8",
      "tree": "489adcde0d131a900aaa6a073166fb07e19fde5a",
      "parents": [
        "99e542aa7d3a72227176817c697571a9efb336e8"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 00:45:47 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Feb 10 00:48:05 2026"
      },
      "message": "Revert \"vpxdsp: [x86] AVX2 and AVX512 for tm intra predictor\"\n\nThis reverts commit 473c80779480192732d2baf802070627f16e09ab.\n\nThis is necessary to allow reverting the parent:\n  27b08ed69c Increase alignment for wider SIMD optimization\nwhich causes failures in Chrome tests.\n\nChange-Id: Ieddf8070aacd68da510996acd1e7f7aabd36a7df\n"
    },
    {
      "commit": "99e542aa7d3a72227176817c697571a9efb336e8",
      "tree": "48dd3bc27b2e2659cd2956d299aed190c7c5e423",
      "parents": [
        "473c80779480192732d2baf802070627f16e09ab"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Tue Feb 03 03:24:36 2026"
      },
      "committer": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Fri Feb 06 04:57:06 2026"
      },
      "message": "vpxdsp: [x86] small tweak to vpx_highbd_tm_predictor_4x4_sse2\n\n1. Fix its speed test that was broken since Day 1\n2. movhpd to movhps, identical behavior but one byte shorter\n\nChange-Id: I204d8819907708dbbd63c1aef6db94f1d950b495\n"
    },
    {
      "commit": "473c80779480192732d2baf802070627f16e09ab",
      "tree": "e90b275121533b0a4358bb502e6fa726a9d0681c",
      "parents": [
        "27b08ed69c760e405cbba3d90d81c47ef42ba59d"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Fri Jan 23 04:40:53 2026"
      },
      "committer": {
        "name": "Jie Meng",
        "email": "zuxy.meng@gmail.com",
        "time": "Fri Feb 06 00:47:44 2026"
      },
      "message": "vpxdsp: [x86] AVX2 and AVX512 for tm intra predictor\n\nAVX2 version of 8x8, 16x16 and 32x32, and AVX512 version of highbd 32x32\n\nTest results with ./test_intra_pred_speed on a 7840hs (best of 5 runs):\n\n                 Baseline/SSE2   AVX2    AVX512\n8x8              86ms            76ms\n16x16            53ms            36ms\n32x32            32ms            18ms\n\nhighbd-8x8       98ms            79ms\nhighbd-16x16     57ms            33ms\nhighbd-32x32     47ms            24ms    22ms\n\nChange-Id: I2648e2c3c1463e79045385bcd9aef16b992d1bea\n"
    },
    {
      "commit": "27b08ed69c760e405cbba3d90d81c47ef42ba59d",
      "tree": "e656b3faec653f8f8ff3ce1dbfd450cfd5a2a8be",
      "parents": [
        "a56f868e78c8050b0c5eecded49eff7fd5cb3860"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Sat Jan 24 04:24:03 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Feb 06 00:47:03 2026"
      },
      "message": "Increase alignment for wider SIMD optimization\n\nIncrease pred plane alignment to 64 byte for high bd and 32 byte\nfor 8-bit to facilitate AVX512/AVX2 optimizations for intra pred,\navoiding much of unaligned memory accesses.\n\nMore consistent performance numbers from test_intra_pred_speed\nas a result.\n\nChange-Id: I048ea46b546b02c37138068720bd05109bf74a35\n"
    },
    {
      "commit": "a56f868e78c8050b0c5eecded49eff7fd5cb3860",
      "tree": "5ce4c6eb3ea40b40bf148f1743eedc90692f872e",
      "parents": [
        "f84e792eb99c80049af9f34e0ab309b14b08ff9b"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Feb 04 19:02:40 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Feb 04 20:11:01 2026"
      },
      "message": "vpx_enc_fuzzer: add missing break statements\n\nFixes unintended fall through, resulting in limited resolution coverage,\nand the related `-Wimplicit-fallthrough`:\n\n```\nexamples/vpx_enc_fuzzer.cc:187:5: error: unannotated fall-through\nbetween switch labels [-Werror,-Wimplicit-fallthrough]\n  187 |     case 1: cfg.g_w \u003d 1; cfg.g_h \u003d 48;\n      |     ^\nexamples/vpx_enc_fuzzer.cc:187:5: note: insert \u0027[[fallthrough]];\u0027 to\nsilence this warning\n  187 |     case 1: cfg.g_w \u003d 1; cfg.g_h \u003d 48;\n      |     ^\n      |     [[fallthrough]];\nexamples/vpx_enc_fuzzer.cc:187:5: note: insert \u0027break;\u0027 to avoid\nfall-through\n  187 |     case 1: cfg.g_w \u003d 1; cfg.g_h \u003d 48;\n      |     ^\n      |     break;\n```\n\nChange-Id: Id9fd5e06de8dfd99a34a57aeedfcf7be103096f4\n"
    },
    {
      "commit": "f84e792eb99c80049af9f34e0ab309b14b08ff9b",
      "tree": "e4474d8cc9ca678e34443b08531560b17e4c2a1f",
      "parents": [
        "80b0b05bfb86931622e8df6469bf19677a98f75b"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Feb 03 19:42:14 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Feb 04 14:06:07 2026"
      },
      "message": "Test for ioc: use more general rounding\n\nBug: 475394382\nChange-Id: Ibf36e6ef2236a5ccc689339e9c70f368b36f76f0\n"
    },
    {
      "commit": "80b0b05bfb86931622e8df6469bf19677a98f75b",
      "tree": "c0ce220db37be8342cfe37eb8281503185696192",
      "parents": [
        "0f5732b4c6e856ead3244d64bb09862d7183b3b2"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Thu Jan 29 01:52:44 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 20:20:03 2026"
      },
      "message": "builtin_assume range of token values\n\nI ran this with and without this change:\n\n```\nperf record -g -- taskset 0x1 ./vpxenc -o /dev/null --ivf --codec\u003dvp9 niklas_1280_720_30.y4m\n```\n\nI observed a ~0.06 percentage point reduction in UBSan samples in\nvp9_optimize_b from 0.35% to 0.29%.\n\nBug: 478049916\nChange-Id: I0f810354315ecbcc8b3b07b13f783c60f0ffc441\n"
    },
    {
      "commit": "0f5732b4c6e856ead3244d64bb09862d7183b3b2",
      "tree": "1512faddf5204eb1629ee786e8d01fdd4ce57ecf",
      "parents": [
        "d4c26badd0352dd155937e6a8eabb9d2f32e3312"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Mon Feb 02 23:32:07 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 20:20:03 2026"
      },
      "message": "Test that tokens are within range\n\nBug: 478049916\nChange-Id: Ia52aad96a7ab5ff495764b2697a4dddfebda030b\n"
    },
    {
      "commit": "d4c26badd0352dd155937e6a8eabb9d2f32e3312",
      "tree": "b214a5e67dfa4749225fbd1178bda0fdfb8c51f7",
      "parents": [
        "1801b9684ea674c98a97eef0d522f997c54e75ae"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 00:22:18 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 20:20:03 2026"
      },
      "message": "Also builtin_assume value of ctx_cur\n\nIf I run this with and without this change:\n\n```\nperf record -g -- taskset 0x1 ./vpxenc -o /dev/null --ivf --codec\u003dvp9 niklas_1280_720_30.y4m\n```\n\nI get a 0.165 percentage point reduction in UBSan samples in\nvp9_optimize_b, from 0.374% to 0.210%.\n\nBug: 478049916\nChange-Id: I5f4f931c0631507cef4068a55291e94e22bf2dd1\n"
    },
    {
      "commit": "1801b9684ea674c98a97eef0d522f997c54e75ae",
      "tree": "e8fd71bbe78a11744aea20fb0943f02d499a16e5",
      "parents": [
        "53b5de7d75742b0b5dff237c7ea3d96577050e4f"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 02:16:15 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Feb 03 17:07:00 2026"
      },
      "message": "cleanup: remove unnecessary copy of ctx\n\nChange-Id: I159b15a5904a2d7c52407c96e3d7957b4e2605dc\n"
    },
    {
      "commit": "53b5de7d75742b0b5dff237c7ea3d96577050e4f",
      "tree": "275d038fc5fc30cff7f69b6314c1d621add3b643",
      "parents": [
        "75d62e849507e78e254d0c2ba47f50174c18d293"
      ],
      "author": {
        "name": "Jingning Han",
        "email": "jingning@google.com",
        "time": "Thu Jan 22 23:25:24 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Fri Jan 30 15:58:28 2026"
      },
      "message": "Avoid ioc in intermediate step\n\nBUG\u003db/475394382\n\nChange-Id: Ic3f9daf5ebbbe83867eb61bdaa240932902431e3\n"
    },
    {
      "commit": "75d62e849507e78e254d0c2ba47f50174c18d293",
      "tree": "c59ded1e3dd5d1321555b0b935e1324a4cbd6793",
      "parents": [
        "0fe8cfa2b896499cb9f0e66a27f4ef22225fd45f"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Thu Jan 29 00:52:08 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Thu Jan 29 01:12:02 2026"
      },
      "message": "builtin_assume energy values are within bounds\n\nThis helps optimize codegen for fsanitize\u003darray-bounds.\n\nA relatively unscientific run of `perf record -g -- ./vpxenc -o\n/dev/null --ivf --codec\u003dvp9 niklas_1280_720_30.y4m` show a reduction\nof ~25 % for samples in __ubsan_check_array_bounds in vp9_optimize_b\nafter this change. This is a much bigger  change than any other\nfunction.\n\nBug: 478049916\nChange-Id: I8bf7f4cd781093e5678789b21954d22641663a95\n"
    },
    {
      "commit": "0fe8cfa2b896499cb9f0e66a27f4ef22225fd45f",
      "tree": "6c76ae176983672953812299c04345ebbdbceca6",
      "parents": [
        "e83e25f791932202256479052f18bdd03a091147"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Thu Jan 29 00:41:29 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Thu Jan 29 01:08:07 2026"
      },
      "message": "Add test for maximum value of vp9_pt_energy_class\n\nBug: 478049916\nChange-Id: I26e6ec5ba192f0a152b6f67a3861359097cd2ef6\n"
    },
    {
      "commit": "e83e25f791932202256479052f18bdd03a091147",
      "tree": "2de81c478b2e16a7ae7d68428a861d2dd6d59999",
      "parents": [
        "c6144e1284bc38397eb5f1449a9e6012c11686bb"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Fri Jan 23 01:20:44 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Tue Jan 27 20:14:23 2026"
      },
      "message": "__builtin_assume that scan values are in range\n\nWe verify this using a unit test. This can improve codegen with\nfsanitize\u003darray-bounds.\n\nThis function shows up high in CPU overhead for\nfsanitize\u003darray-bounds.\n\nThis helps us get rid of some cmp and branch against 0x400:\nhttps://gist.github.com/fmayer/4c1753b93109f3cb35cdc33a8cab5c0c\n\nA relatively unscientific run of `perf record -g -- ./vpxenc -o\n/dev/null --ivf --codec\u003dvp9 niklas_1280_720_30.y4m` show a halfing of\nsamples in __ubsan_check_array_bounds in vp9_optimize_b\nafter this change. This is a much bigger  change than any other\nfunction.\n\nBug: 478049916\nChange-Id: Icfccc94c3b442cf6a6535246b6084dd2b9bc0555\n"
    },
    {
      "commit": "c6144e1284bc38397eb5f1449a9e6012c11686bb",
      "tree": "05a8e4431a28e039f291a83e8a5dc3675aea50f3",
      "parents": [
        "754659a7f01f2b89f88b1109956eb75f7f588be5"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Jan 08 21:03:04 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 27 20:12:17 2026"
      },
      "message": "configure: add --disable-x86-asm\n\nSimilar to --disable-neon-asm, this disables most handwritten assembly,\nleaving the intrinsics optimizations. This is meant for use with MSan,\nwhich will report false positives with the default configuration. Note\nthis doesn\u0027t remove all assembly, emms_mmx.asm remains, but doesn\u0027t\naffect any memory buffers.\n\nBug: 42330861\nChange-Id: Ib4324b74e6bad214af172f5ebafa6063de6e2fd1\n"
    },
    {
      "commit": "754659a7f01f2b89f88b1109956eb75f7f588be5",
      "tree": "98b20618da709e0a95eb0eeefe0f296f5dd0a058",
      "parents": [
        "952c4d8a665d1ae797cf768c987c7df4df50c4d4"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Sun Jan 18 05:50:12 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 27 18:58:11 2026"
      },
      "message": "vpxdsp [x86]: AVX2 for 16x16 high bd d207 and d63 intra pred\n\nTested with ./test_intra_pred_speed\n--gtest_filter\u003d*.TestHighbdIntraPred16 on a Zen 4 (7840hs),\nvpx_highbd_d207_predictor_16x16_avx2 is 70% faster while\nvpx_highbd_d63_predictor_16x16_avx2 is 33% faster than their SSSE3\ncounterparts.\n\nChange-Id: Ia0034b96d9578bd89829f34da787e097828e6ec7\n"
    },
    {
      "commit": "952c4d8a665d1ae797cf768c987c7df4df50c4d4",
      "tree": "947411d76057b3a488caa7d1c52a37af0146be4e",
      "parents": [
        "5de804c9b4d8d24882a31f5be287f1f688729db5"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Jan 23 19:48:32 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Jan 23 19:56:23 2026"
      },
      "message": "vp8: fix use_dc_pred condition in mb_activity_measure\n\nPossible fix for issue below.\n\nTo use dc_pred make sure at least one block exist on\ntop and on left. This fixes the asan buffer overflow\nin issue below.\n\nOnly happens in best quality mode where this is\nused in vp8_optimize_mby,\n\nBug: 475394383\nChange-Id: I3c059028587851c4ccee1ba32975008cc41cccb0\n"
    },
    {
      "commit": "5de804c9b4d8d24882a31f5be287f1f688729db5",
      "tree": "57d23cca19156c46f9864d72ff6c7e3997a0388a",
      "parents": [
        "d5f35ac8d93cba7f7a3f7ddb8f9dc8bd28f785e1"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Fri Jan 23 00:42:29 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Fri Jan 23 05:24:28 2026"
      },
      "message": "Add test that all scan values are \u003c\u003d MAX_SCAN_VALUE\n\nBug: 478049916\nChange-Id: I8b0c578cd647a8025ad05669ea9238fe2d94ca08\n"
    },
    {
      "commit": "d5f35ac8d93cba7f7a3f7ddb8f9dc8bd28f785e1",
      "tree": "3d3d4a68eb5f4b184c04338021e6e8a6437a17be",
      "parents": [
        "f4ed22f7460bdda1424d400df463e863444face0"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Thu Jan 22 02:03:55 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Fri Jan 23 02:41:04 2026"
      },
      "message": "write_superframe_index: return 0 if buffer is full\n\nwrite_superframe_index() should return the number of bytes written to\nctx-\u003epending_cx_data. If ctx-\u003epending_cx_data is full,\nwrite_superframe_index() doesn\u0027t write the optional superframe index, so\nit should return 0 in this case. Add an assertion that would have\ndetected this bug. Add and clarify comments for code related to this\nbug.\n\nAlso fix the buffer full check. The check should not assume that\nctx-\u003epending_cx_data is equal to ctx-\u003ecx_data, and the check had an\noff-by-one error.\n\nThe bug was introduced when write_superframe_index() was added in the\nfollowing CLs:\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/44659\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/45268\n\nBug: oss-fuzz:476466137\nChange-Id: Ie113568cf25acc73f8af640a3c51cfdb5b900613\n"
    },
    {
      "commit": "f4ed22f7460bdda1424d400df463e863444face0",
      "tree": "4f00a5babd97447c343bf98a402110d73a9efe43",
      "parents": [
        "09863354ead7ebc9e4ec233da4b274f71daf392c"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Jan 16 23:13:57 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Fri Jan 23 01:32:01 2026"
      },
      "message": "highbd_variance64: use unsigned math w/sse calculation\n\nThis makes the operation defined should out of range input be used in\ne.g., fuzzing tests. In practice, the library expects 8/10/12-bit values\nto be in range ([0, 255], [0, 1023], [0, 4095]). This allows data types\nto be chosen that facilitate SIMD optimizations.\n\nBug: 475394379\nChange-Id: I2468a4f70427bdfc9f572170b309c5cbc694d4ca\n"
    },
    {
      "commit": "09863354ead7ebc9e4ec233da4b274f71daf392c",
      "tree": "505b28ac933cf171c15ac82ce12ffe611294c5f5",
      "parents": [
        "92be056a95cdbe0cd4148f3555a9050bff8652ae"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Sat Jan 17 04:14:27 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Jan 22 17:55:01 2026"
      },
      "message": "vpxdsp: [x86] vpx_highbd_d207_predictor_32x32 for AVX2 and AVX512\n\nWhen tested with \u0027./test_intra_pred_speed\n--gtest_filter\u003d*.TestHighbdIntraPred32\u0027 on a Zen 4, both are about 90%\nfaster than the SSSE3 implementation\n\nChange-Id: I115861bad93821e0681ec95c8d98db847e904b19\n"
    },
    {
      "commit": "92be056a95cdbe0cd4148f3555a9050bff8652ae",
      "tree": "63615702c60611362257b7031d5af9868a29e210",
      "parents": [
        "8786a5adc3ad412f55e99cb1a4c6da7518ae26e5"
      ],
      "author": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Thu Jan 22 03:11:59 2026"
      },
      "committer": {
        "name": "Wan-Teh Chang",
        "email": "wtc@google.com",
        "time": "Thu Jan 22 17:25:56 2026"
      },
      "message": "update_level_info: Declare `size` param as size_t\n\nThe `size` parameter of update_level_info() is always dereferenced to\nread its current value, so it doesn\u0027t need to be a pointer.\n\nChange-Id: Ibebbe8c3725769682c87ebb6ee445d3603939929\n"
    },
    {
      "commit": "8786a5adc3ad412f55e99cb1a4c6da7518ae26e5",
      "tree": "142cae49d7811c7f603eba7f46b65466fc1a9fd6",
      "parents": [
        "807c698468ee6264d6b206d74b9465b5acd97b5c",
        "1024874c5919305883187e2953de8fcb4c3d7fa6"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Jan 21 19:54:45 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Jan 21 19:55:10 2026"
      },
      "message": "Merge tag \u0027v1.16.0\u0027\n\nRelease v1.16.0 Xenonetta Duck\n\n2026-01-21 v1.16.0 \"Xenonetta Duck\"\n\n  This release includes Arm SVE2 and Neon optimizations for 12-tap filters,\n  AVX512 implementations for SAD, support for per-frame and per-spatial-layer\n  PSNR calculation, and numerous bug fixes.\n\n  - Upgrading:\n    This release is ABI incompatible with the previous release.\n    Unit tests require C++17 to build.\n    Support for 32-bit iOS targets (armv7, armv7s, and i386) has been removed.\n\n  - Enhancement:\n    Optimized Arm SVE2 and Neon implementations for 12-tap convolution filters.\n    Optimized Neon High Bitdepth (HBD) SAD and sad_avg functions.\n    Added Arm Neon DotProd and I8MM implementations for vpx_convolve12.\n    Added AVX512 implementations for SAD64 and sad_skip functions.\n    Added SSSE3 and AVX2 implementations for 12-tap temporal filter prediction.\n    Added support for per-frame and per-spatial-layer PSNR calculation.\n\n    Adjusted temporal filter strength to improve visual quality and reduce block\n    artifacts.\n\n    Added support for darwin24 (macOS 15) and darwin25 (macOS 26).\n    libwebm is upgraded to commit b4f01ea.\n\n  - Bug fixes:\n    Fix to heap buffer overflow in vp9_deblock, vp9_post_proc_frame, and\n    vp9_pack_bitstream.\n\n    Fix to integer overflow in vp9_highbd_post_proc, vp9_rc_regulate_q,\n    tiny_ssim, and vp9_calc_pframe_target_size_one_pass_cbr.\n\n    Fix to use-of-uninitialized-value in vp9_highbd_post_proc, mfqe, and\n    vp8_datarate_test.\n\n    Fix to out-of-bounds in log_tile_cols_from_picsize_level.\n    Fix to double free on initialization failure in vpx_codec_enc_init_multi.\n    Fix to division-by-zero crash in vpxenc with 0 FPS numerator input.\n    Fix to various build failures for Arm/SVE2, macOS cross-compilation, and\n    Xcode 16.\n\nChange-Id: I2f8be65249d1ced1c84b58e1f451273cb031d58c\n"
    },
    {
      "commit": "807c698468ee6264d6b206d74b9465b5acd97b5c",
      "tree": "ab46c59172870a0de6c35d83f54373bfffac2afa",
      "parents": [
        "dea485bea6333a59bf4c8f3a786134114a0cd3bb"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 20 21:12:00 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 20 21:12:00 2026"
      },
      "message": "vp9_postproc: fix limits allocation size\n\nAfter:\n929869588a vp9: Fix to heap overflow for postproc with resize\n\nUse `unscaled_width` in the allocation call. In 929869588a the\nassignment to `cm-\u003epostproc_state.limit_size` was moved after the\nallocation and will be 0 on the first call.\n\nBug: 475854355\nChange-Id: I20dacd38f6911f423585d20c7e1e8066b18eda24\n"
    },
    {
      "commit": "dea485bea6333a59bf4c8f3a786134114a0cd3bb",
      "tree": "935043af151286549631b67d2a7e73ca653ce408",
      "parents": [
        "5af00838c0e83eea72003f398d8a1a328143f576"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Sat Jan 17 00:51:59 2026"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Sat Jan 17 00:59:12 2026"
      },
      "message": "Fix size of token cache\n\nWe use values from scan[] to index into it. The largest value\nin any scan is 1024 from the last value of vp9_default_iscan_32x32.\n\n```\nvp9_default_iscan_32x32[1024]) \u003d {\n[...]\n  968,  974,  989,  997,  1003, 1007, 1015, 1019, 1022, 1024,\n};\n```\n\nI have a change that can verify that all scans are \u003c\u003d the\nMAX_SCAN_VALUE, but that involves some macro hackery that might\nbe controversial, so merging this fix first.\n\nChange-Id: I7ae3b5640ca7d0d6525173777442dc267e5b5457\n"
    },
    {
      "commit": "5af00838c0e83eea72003f398d8a1a328143f576",
      "tree": "07ab0fa9be61cce31305ad53176a61cfd96cd006",
      "parents": [
        "4b77b98417cf32d6f6074a16e0994581d82880cc"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Jan 14 23:05:19 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Jan 15 20:01:55 2026"
      },
      "message": "vp9_ext_ratectrl_test: fix test w/CONFIG_VP9_DECODER\u003d0\n\n`received_show_frame_count_` is incremented at decode time; skip the\ncheck when `--disable-vp9-decoder` is used in the configuration.\n\nChange-Id: Ibb21cd53763467e2439245c073f55df194dd621f\n"
    },
    {
      "commit": "4b77b98417cf32d6f6074a16e0994581d82880cc",
      "tree": "d75f5e198d575e0f8e080b5912a747d1718615c0",
      "parents": [
        "7900df4ee02cb885d16b2543dfa8a54e386a32a6"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Sun Dec 21 06:10:29 2025"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Jan 15 20:01:37 2026"
      },
      "message": "Add \"edx\" to x86_readtsc()\u0027s clobber list\n\nrdtsc always writes to eax and edx at the same time. Use\nx86_readtsc() w/o the correct clobber list could lead to segfault\nif edx is later read.\n\nChange-Id: I5359ed60d92479873df729bb3f0ec9e7765396d6\n"
    },
    {
      "commit": "7900df4ee02cb885d16b2543dfa8a54e386a32a6",
      "tree": "4105b46396a49cc0372bb94bfc86268266a389e5",
      "parents": [
        "929869588adaccb3b44ec043937e8b3158adb796"
      ],
      "author": {
        "name": "zuxy",
        "email": "zuxy.meng@gmail.com",
        "time": "Sun Dec 21 05:27:38 2025"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Thu Jan 15 20:01:31 2026"
      },
      "message": "vpxdsp: [x86] vpx_highbd_d63_predictor_32x32 for AVX2 and AVX512\n\nWhen tested with test_intra_pred_speed on a Zen 4, the AVX2\nversion is ~45% faster and the AVX512 is ~90% faster than the\ncurrent SSSE3 implementation\n\nChange-Id: I2a8537d004f508b4e966a9f9af8735e87c0a7035\n"
    },
    {
      "commit": "929869588adaccb3b44ec043937e8b3158adb796",
      "tree": "c17b14a326d5844a99c4e71f32d9d0365a42d6a8",
      "parents": [
        "02ff212b9b9f8e5a67961a26e60297fc51f7a9cc"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Wed Jan 14 23:04:23 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Thu Jan 15 19:55:57 2026"
      },
      "message": "vp9: Fix to heap overflow for postproc with resize\n\nThe allocation size for postproc_stat-\u003elimits is\nun_scaled_source-\u003ey_width and is stored in limits_size\nin vp9_post_proc_frame(). To handle resize we need to\ncheck and re-allocate postproc_state-\u003elimits if the\nallocated size is below the current\nun_scaled_source-\u003ey_width.\n\nUpdate the allocation in vp9_postproc.c and set\nthe limits_size at the time of allocation.\n\nBug: 475854355\nChange-Id: Ia924a3019bd267c0018f0b2903dd45932320c307\n"
    },
    {
      "commit": "02ff212b9b9f8e5a67961a26e60297fc51f7a9cc",
      "tree": "0fa5fccb535769022697447de6499a420763662c",
      "parents": [
        "6aa37dfddba550bcf3dd02ada23854d5990b5ed4"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Jan 14 21:23:07 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Jan 14 21:23:07 2026"
      },
      "message": "fastssim.c: remove \u0027;\u0027 from #define\n\nFixes:\n../vpx_dsp/fastssim.c|450 col 52| warning: empty expression statement\nhas no effect; remove unnecessary \u0027;\u0027 to silence this warning\n[-Wextra-semi-stmt]\n\nChange-Id: I924cdc89f461f5d314e6340ffd0c407ce90dc854\n"
    },
    {
      "commit": "6aa37dfddba550bcf3dd02ada23854d5990b5ed4",
      "tree": "c83c06f04c328666e5423636ac322bb757554b9e",
      "parents": [
        "f72b9f6a85fdc4fbbec6402146fb653130330e8f"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 13 21:51:41 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 13 21:51:41 2026"
      },
      "message": "vpx_{encod,decod}er.h: clarify vpx_codec_*_init_ver comment\n\nExplicitly note that `vpx_codec_destroy()` must be used to free\nresources.\n\nBug: 475394380\nChange-Id: I19768776a91ff3d3986045f463cd53a13cb1ab2e\n"
    },
    {
      "commit": "f72b9f6a85fdc4fbbec6402146fb653130330e8f",
      "tree": "253f58cd20ef315412e64ae1e73ec4fbcd8e2510",
      "parents": [
        "1a400b808407a49d6ee626e8e6015c5cd9df6298"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Jan 12 19:27:11 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Jan 12 21:59:51 2026"
      },
      "message": "vp9: Fixes for CONFIG_REALTIME_ONLY\n\nRe-application of change below (which was reverted),\nwith fix to skip tests with speed \u003c 5 for realtime mode,\nsince the change clamps speed to 5 for realtime encoding:\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/7417233\n\nFor CONFIG_REALTIME_ONLY build the speed setting\nis clamped to 5 (done in both the default config\nand in the control for setting the speed/cpu_used),\nbut for !CONFIG_REALTIME_ONLY and realtime encoding\n(VPX_DL_REALTIME) the speed is not clamped and can\nbe set down to speeds \u003c 5.\n\nFix this by adding the clamp in pick_quickcompress_mode()\nwhere the REALTIME mode is set.\n\nAnd fix the threshold for the ResizeCspWorks test to make it pass\nat speed 5.\n\nBug: 454987857\nChange-Id: Ie6c953926d3ae22afe1ca525a8738b3632cdf7ee\n"
    },
    {
      "commit": "1a400b808407a49d6ee626e8e6015c5cd9df6298",
      "tree": "8a4e3dd2282473fb4b330034ad08e137e0ba7ee4",
      "parents": [
        "b21ac87cee116662ebf9ca875f39148075b62c52"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Jan 09 21:25:33 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Jan 09 22:17:02 2026"
      },
      "message": "Revert \"vp9: Fixes for CONFIG_REALTIME_ONLY\"\n\nThis reverts commit 3383144ae67b9539279dd72ea236d29270e56da4.\n\nReason for revert:\nBreaks the nightly tests. Will follow-up with the proper fix.\n\nOriginal change\u0027s description:\n\u003e vp9: Fixes for CONFIG_REALTIME_ONLY\n\u003e\n\u003e For CONFIG_REALTIME_ONLY build the speed setting\n\u003e is clamped to 5 (done in both the default config\n\u003e and in the control for setting the speed/cpu_used),\n\u003e but for !CONFIG_REALTIME_ONLY and realtime encoding\n\u003e (VPX_DL_REALTIME) the speed is not clamped and can\n\u003e be set down to speeds \u003c 5.\n\u003e\n\u003e Fix this by adding the clamp in pick_quickcompress_mode()\n\u003e where the REALTIME mode is set.\n\u003e\n\u003e Fix the kCpuUsedVectors[] in vp9_end_to_end_test.cc,\n\u003e as the psnr thresholds (kPsnrThreshold[cpu_used_][encoding_mode_])\n\u003e for the tests are set based on these cpu_used settings, and this\n\u003e was causing failures for CONFIG_REALTIME_ONLY (which was encoding\n\u003e with speed 5 but using psnr-thresholds set for speed \u003c 5).\n\u003e\n\u003e And fix the threshold for the ResizeCspWorks test to make it pass\n\u003e at speed 5.\n\u003e\n\u003e Bug: 454987857\n\u003e Change-Id: I1669da872ac42b65b8a2fc3a6e2d795e54e33d8a\n\nBug: 454987857\nNo-Presubmit: true\nNo-Tree-Checks: true\nNo-Try: true\nChange-Id: I0f8ec9c8d42baea655054d38b466a76e12d97588\n"
    },
    {
      "commit": "b21ac87cee116662ebf9ca875f39148075b62c52",
      "tree": "ac667247de62eea4df4a7505f199a146b2a64ac9",
      "parents": [
        "3383144ae67b9539279dd72ea236d29270e56da4"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Jan 09 18:30:40 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Fri Jan 09 18:30:40 2026"
      },
      "message": "vp9_end_to_end_test: fix -Wunused-const-variable warning\n\nAfter:\nb0be221b68 fix tests w/CONFIG_REALTIME_ONLY\n\nFixes the following warning with clang:\n../test/vp9_end_to_end_test.cc:67:22: warning: unused variable\n  \u0027k4x2VideoTestVectors\u0027 [-Wunused-const-variable]\n   67 | const TestVideoParam k4x2VideoTestVectors[] \u003d {\n\nChange-Id: Ib8df06de295197a72f6fd0169fb4c2969fe4cefe\n"
    },
    {
      "commit": "3383144ae67b9539279dd72ea236d29270e56da4",
      "tree": "4d949e70d23f2a0d4d17e9b50bfbed05c5717f1e",
      "parents": [
        "b0be221b6811038e1579a4281241a55549d7611d"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Thu Jan 08 22:56:08 2026"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Jan 09 00:53:01 2026"
      },
      "message": "vp9: Fixes for CONFIG_REALTIME_ONLY\n\nFor CONFIG_REALTIME_ONLY build the speed setting\nis clamped to 5 (done in both the default config\nand in the control for setting the speed/cpu_used),\nbut for !CONFIG_REALTIME_ONLY and realtime encoding\n(VPX_DL_REALTIME) the speed is not clamped and can\nbe set down to speeds \u003c 5.\n\nFix this by adding the clamp in pick_quickcompress_mode()\nwhere the REALTIME mode is set.\n\nFix the kCpuUsedVectors[] in vp9_end_to_end_test.cc,\nas the psnr thresholds (kPsnrThreshold[cpu_used_][encoding_mode_])\nfor the tests are set based on these cpu_used settings, and this\nwas causing failures for CONFIG_REALTIME_ONLY (which was encoding\nwith speed 5 but using psnr-thresholds set for speed \u003c 5).\n\nAnd fix the threshold for the ResizeCspWorks test to make it pass\nat speed 5.\n\nBug: 454987857\nChange-Id: I1669da872ac42b65b8a2fc3a6e2d795e54e33d8a\n"
    },
    {
      "commit": "1024874c5919305883187e2953de8fcb4c3d7fa6",
      "tree": "5e48c010122b13fe684e4419d45feee3157d7387",
      "parents": [
        "e1bd87f99dd0541dbaa85aa14ac35fe6e9be2a21"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Thu Jan 08 16:00:11 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Thu Jan 08 16:01:40 2026"
      },
      "message": "Update CHANGELOG for xenonetta\n\nChange-Id: I1b71c43ae46232b405dfdca9569ff825d57eb069\n"
    },
    {
      "commit": "b0be221b6811038e1579a4281241a55549d7611d",
      "tree": "b261cf227934c9b0bdb25ff957a82666ef49349f",
      "parents": [
        "ce95ea77668648ee37d462cda18df9d9d461a00e"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Jan 07 19:47:12 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Wed Jan 07 21:44:48 2026"
      },
      "message": "fix tests w/CONFIG_REALTIME_ONLY\n\nThe fixes include:\n- removing use of modes other than libvpx_test::kRealTime\n- removing superframe_test.cc and tile_independence_test.cc from the\n  build as the tests were tuned for two pass mode\n- correcting the CONFIG_REALTIME_ONLY define in cq_test.cc (-Wundef is\n  not used with C++ due to warnings in third_party/ code)\n\nBug: 454987857\nChange-Id: I9d03819162090442c997d15c730ac39d19836432\n"
    },
    {
      "commit": "e1bd87f99dd0541dbaa85aa14ac35fe6e9be2a21",
      "tree": "daa7ab8412d51b7638675063cec1e0c39343793a",
      "parents": [
        "4078082ebae32d912ce441c67d96bfb6a23693a3"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Jan 06 16:24:25 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Jan 07 18:58:20 2026"
      },
      "message": "Update AUTHORS and version for release\n\nVersion update:\n\nv1.15.2\nc\u003d11, a\u003d0, r\u003d1\n\nRule 3: code changed \u003d\u003e r\u003d2\nRule 4: interface changed \u003d\u003e c\u003d12, a\u003d0, r\u003d0\nRule 5: NA\nRule 6: interface changed \u003d\u003e a\u003d0\n\nv1.16.0\nc\u003d12, a\u003d0, r\u003d0\n\nBug: 459857079\nChange-Id: I6d8ce4f8ff5cf13425b03438d34991dbcd4f5cf3\n"
    },
    {
      "commit": "ce95ea77668648ee37d462cda18df9d9d461a00e",
      "tree": "301cf4cca07bdafa2129a854c9835781ca207ef6",
      "parents": [
        "98f6c832ed467ea53de90c3b376a31ee536a2d85"
      ],
      "author": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 06 20:37:17 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 06 23:08:53 2026"
      },
      "message": "init_vpx_test: sync append_negative_gtest_filter w/libaom\n\nThis is primarily a port of:\nc4ec0329c6 expand gtest cpu extension filter\n\nwhich extended the patterns to include `OPT_` as in `SSE2_TO_C_8` from\nvp9_intrapred_test.cc. This will become important if AVX512\noptimizations are added as the test machines currently do not support\nit.\n\nChange-Id: I9c8d5be8422bd4f985589e2a1312f4c89b4d4c27\n"
    },
    {
      "commit": "98f6c832ed467ea53de90c3b376a31ee536a2d85",
      "tree": "228919c068e54257870be706fe119df6ae5cadb5",
      "parents": [
        "1a3634a26f85a850712d197cc462361e05be8f9b"
      ],
      "author": {
        "name": "Darryl Pogue",
        "email": "darryl@dpogue.ca",
        "time": "Thu Jan 01 06:12:37 2026"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Jan 06 21:14:30 2026"
      },
      "message": "Check for pthread_setname_np during configure\n\nThis is not available on older versions of Mac OS X (\u003c10.6).\n\nChange-Id: Ia9403650721c5ff60e57ce8ad79902d24de54dec\n"
    },
    {
      "commit": "4078082ebae32d912ce441c67d96bfb6a23693a3",
      "tree": "064d38e83b58f3fa0cb820ccba7ff475d99bc563",
      "parents": [
        "61902a30f643f2439958192fe2a974428c09fff3"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Jan 06 18:06:13 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Jan 06 19:22:11 2026"
      },
      "message": "Ext RC: Bump ABI versions\n\nThis is for the following two CLs:\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/5902473\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/5997450\n\nChange-Id: I13b94505db09713076d2d75a11770eb088dbd537\n(cherry picked from commit 1a3634a26f85a850712d197cc462361e05be8f9b)\n"
    },
    {
      "commit": "1a3634a26f85a850712d197cc462361e05be8f9b",
      "tree": "c26254db83097ce8af830f49c013800a1559bde8",
      "parents": [
        "d5399cdd6abea1169a246f28c5ca9a50975d7ba9"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Jan 06 18:06:13 2026"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Tue Jan 06 19:19:57 2026"
      },
      "message": "Ext RC: Bump ABI versions\n\nThis is for the following two CLs:\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/5902473\nhttps://chromium-review.googlesource.com/c/webm/libvpx/+/5997450\n\nChange-Id: I13b94505db09713076d2d75a11770eb088dbd537\n"
    },
    {
      "commit": "d5399cdd6abea1169a246f28c5ca9a50975d7ba9",
      "tree": "fcac06dd94857d1002a7ef7ce0d36f4449c25253",
      "parents": [
        "61902a30f643f2439958192fe2a974428c09fff3"
      ],
      "author": {
        "name": "Alex Davicenko",
        "email": "alex.davicenko@arm.com",
        "time": "Mon Dec 01 13:52:26 2025"
      },
      "committer": {
        "name": "Jonathan Wright",
        "email": "jonathan.wright@arm.com",
        "time": "Fri Dec 19 15:19:16 2025"
      },
      "message": "Optimize Neon HBD sadx4d and sad_skipx4d functions\n\nOptimize vpx_highbd_sadx4d_neon vpx_highbd_sad_skip_x4d_neon functions\nby accumulating into 16-bit vectors and widening only at the point of\noverflow.\n\nCo-authored-by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e\nChange-Id: Ic371b099efe0c2993887ea47bc9c915e0cdd13d9\n"
    },
    {
      "commit": "61902a30f643f2439958192fe2a974428c09fff3",
      "tree": "5a8aacff11ec1228f924ecbf20ad3cdec37222e9",
      "parents": [
        "dca6330a24c7a4cb0306331582b25b58e2f903e8"
      ],
      "author": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Dec 17 15:22:39 2025"
      },
      "committer": {
        "name": "Jerome Jiang",
        "email": "jianj@google.com",
        "time": "Wed Dec 17 15:22:39 2025"
      },
      "message": "Add unit test for a 4x2 test video\n\nBug: b/468489421\nChange-Id: I2657a2b4012325bda6d3e469ab4f5d7766f30d63\n"
    },
    {
      "commit": "dca6330a24c7a4cb0306331582b25b58e2f903e8",
      "tree": "1f564b6cf06f5aae4254969dacd34b4ae95b46ee",
      "parents": [
        "53cb0b40341695e07f697d2c9d29ad2dc83681fb"
      ],
      "author": {
        "name": "Jingning Han",
        "email": "jingning@google.com",
        "time": "Tue Dec 16 22:26:21 2025"
      },
      "committer": {
        "name": "Jingning Han",
        "email": "jingning@google.com",
        "time": "Tue Dec 16 22:26:21 2025"
      },
      "message": "Clear a first pass motion search corner case\n\nWhen setting the motion search parameters, align the frame size\nto be same or above a coding block size for consistency with other\nrelated use cases in the encoder.\n\nBUG\u003db/468489421\n\nChange-Id: I0af2651b21671f3b8c42984cbb1cff1f62072b60\n"
    },
    {
      "commit": "53cb0b40341695e07f697d2c9d29ad2dc83681fb",
      "tree": "35c6f78fbea020f0f4c91556f66e20d889f9b19d",
      "parents": [
        "cdd9b1ea25aca22200996a726eafd25d4fcd56d1"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Tue Dec 16 06:33:53 2025"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Tue Dec 16 19:15:14 2025"
      },
      "message": "vp9-postproc: memset the post_proc_buffer\n\nThis prevents use of uninitialized value in\nthe vp9_highbd_post_proc.\n\nBug: b:468724242\nChange-Id: Ib55c0b28650d8798a5d07699dd7a9ce78bd850ca\n"
    },
    {
      "commit": "cdd9b1ea25aca22200996a726eafd25d4fcd56d1",
      "tree": "84f83a41f7f43ca34ef99b9de355f2a6b4fcb58e",
      "parents": [
        "ddf807ad4c77d65e2cf13317cd22b9bd7baefaf3"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Dec 15 22:31:56 2025"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Tue Dec 16 01:02:52 2025"
      },
      "message": "vp9: fix integer overflow in vp9_highbd_post_proc\n\nMake the same fixes to:\nvp9_highbd_mbpost_proc_across_ip_c()\nvp9_highbd_mbpost_proc_down_c()\n\nBug: b:468736065\nChange-Id: If90015b43163bf3b9f62e3d5b49d3f07e78a0b2e\n"
    },
    {
      "commit": "ddf807ad4c77d65e2cf13317cd22b9bd7baefaf3",
      "tree": "0bb1d3926b520615f709c20ae3263e877e79dfad",
      "parents": [
        "363aa07c9a4ac8fd8f0ec6f45b3735ca0c0cb753"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Dec 15 19:59:55 2025"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Dec 15 22:04:07 2025"
      },
      "message": "vp9: fix to heap buffer overflow in vp9_deblock\n\nFor vp9_post_proc: set the limits_size and use\nit in the memset of limits in vp9_deblock.\n\nBug: b:468697910\nChange-Id: I0f04750cf2f1701e7fe18262b030411ccc6a4b4b\n"
    },
    {
      "commit": "363aa07c9a4ac8fd8f0ec6f45b3735ca0c0cb753",
      "tree": "cff15b1354aa4b47e1519b6a25f4e295766497a1",
      "parents": [
        "70a4f20f9931d7128dfa5ebe0bef5c3021a435c3"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Dec 15 19:41:08 2025"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Mon Dec 15 22:03:11 2025"
      },
      "message": "vp9: fix to heap buffer overflow in vp9_post_proc_frame\n\npost_proc_buffer_int.frame_size should be\nused in memset of post_proc_buffer_int.\n\nBug: b:468843217\nChange-Id: Id84e19a0032782b400ee87b56ea0464e0455ed31\n"
    },
    {
      "commit": "70a4f20f9931d7128dfa5ebe0bef5c3021a435c3",
      "tree": "099a2ead04f611b88ccf4d972d3257ef85b90847",
      "parents": [
        "d98e70839f01aeeee875d343475574386869d47a"
      ],
      "author": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Fri Dec 12 23:17:48 2025"
      },
      "committer": {
        "name": "Marco Paniconi",
        "email": "marpan@google.com",
        "time": "Sat Dec 13 18:37:31 2025"
      },
      "message": "vp8-postproc: Fix use-of-uninitialized-value in mfqe\n\nBug: b:467517001\nChange-Id: Iffe22253107c9814f1e8dfcaff8be899a7b36185\n"
    },
    {
      "commit": "d98e70839f01aeeee875d343475574386869d47a",
      "tree": "7e26bef1f660289993c8b866b4e80ff91270a205",
      "parents": [
        "c4c48fd32bc092333b9f8e53882cf12d9b7fde33"
      ],
      "author": {
        "name": "Alex Davicenko",
        "email": "alex.davicenko@arm.com",
        "time": "Fri Dec 05 14:12:18 2025"
      },
      "committer": {
        "name": "Jonathan Wright",
        "email": "jonathan.wright@arm.com",
        "time": "Thu Dec 11 16:48:10 2025"
      },
      "message": "Optimize Neon HBD sad_avg functions\n\nOptimize vpx_highbd_sad_avg_neon functions by accumulating into 16-bit\nvectors and widening only at the point of overflow.\n\nCo-authored-by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e\nChange-Id: I21b3e2c4e35d29a96700d7005820f5e45bee3030\n"
    },
    {
      "commit": "c4c48fd32bc092333b9f8e53882cf12d9b7fde33",
      "tree": "a151a2f2a9889e53ca455e722c4840b32703d18d",
      "parents": [
        "d6e5cd7132d59d3cea86d06c9aae26f95fe00adb"
      ],
      "author": {
        "name": "Alex Davicenko",
        "email": "alex.davicenko@arm.com",
        "time": "Fri Dec 05 14:12:09 2025"
      },
      "committer": {
        "name": "Jonathan Wright",
        "email": "jonathan.wright@arm.com",
        "time": "Thu Dec 11 16:47:56 2025"
      },
      "message": "Optimize Neon HBD sad and sad_skip functions\n\nOptimize vpx_highbd_sad_neon and vpx_highbd_sad_skip_neon functions\nby accumulating into 16-bit vectors and widening only at the point of\noverflow.\n\nCo-authored-by: Jonathan Wright \u003cJonathan.Wright@arm.com\u003e\nChange-Id: I3f4289899cf4cd309b254d2933a6ea8682ebd8a0\n"
    },
    {
      "commit": "d6e5cd7132d59d3cea86d06c9aae26f95fe00adb",
      "tree": "0c6f9b57ec9c97cd8eacc90ef1d5e9771a217d69",
      "parents": [
        "b98c79655e5405ba36110297573f9c1fdbf2bc06"
      ],
      "author": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Mon Dec 08 21:42:09 2025"
      },
      "committer": {
        "name": "Florian Mayer",
        "email": "fmayer@google.com",
        "time": "Wed Dec 10 01:08:58 2025"
      },
      "message": "Fix out of bounds in log_tile_cols_from_picsize_level\n\nThis accidentally used enum values as array indices, and as such would\nrun from 10 to 255 on a 14 element array.\n\nBug: b/465388500\nChange-Id: Ibaa1f3444934fdf574ad908d57aee9590b21b089\n"
    },
    {
      "commit": "b98c79655e5405ba36110297573f9c1fdbf2bc06",
      "tree": "3eb8411869f2fa819025b376527aac3bea0bb309",
      "parents": [
        "df44a8bdabfdef2acbf5df0aa68acbc4ec19a236"
      ],
      "author": {
        "name": "Jonathan Wright",
        "email": "jonathan.wright@arm.com",
        "time": "Tue Dec 09 13:54:49 2025"
      },
      "committer": {
        "name": "James Zern",
        "email": "jzern@google.com",
        "time": "Tue Dec 09 22:38:58 2025"
      },
      "message": "configure: add support for darwin25 (macOS 26.x)\n\nAdd support for darwin25 platforms in configure scripts and update\nREADME.\n\nChange-Id: Idfc34f404296c1a257dff1d690a28d7b9a722432\n"
    }
  ],
  "next": "df44a8bdabfdef2acbf5df0aa68acbc4ec19a236"
}