)]}' { "log": [ { "commit": "a018d6e2dc9b3f0b1a48d1deade8160e44589189", "tree": "e4c1bc45207cc50b3dfeb97fb5f9f277c41b09c2", "parents": [ "5dbdde43f63058c0f0579441aa4d94d34ac5c5c0" ], "author": { "name": "Eric Holk", "email": "eholk@chromium.org", "time": "Fri Feb 02 20:58:25 2018" }, "committer": { "name": "Eric Holk", "email": "eholk@chromium.org", "time": "Fri Feb 02 22:00:45 2018" }, "message": "Add runtime back to check-lit dependencies\n\nChange-Id: I43717bac3c07fab6502d0d28b93c518125a49852\nReviewed-on: https://chromium-review.googlesource.com/900123\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "5dbdde43f63058c0f0579441aa4d94d34ac5c5c0", "tree": "d1370eb2734d5181cf09ed3094451f45c676af9e", "parents": [ "cb642c8aed91b0f3a00dd96d4b932dc02b41e93c" ], "author": { "name": "Eric Holk", "email": "eholk@chromium.org", "time": "Fri Feb 02 00:20:01 2018" }, "committer": { "name": "Eric Holk", "email": "eholk@chromium.org", "time": "Fri Feb 02 20:29:17 2018" }, "message": "Handle mul instructions with two immediates on x86-32 and x86-64\n\nWe were hitting an assert in the assembler due to an instruction that\nhad not been properly legalized. Now we ensure that mul instructions\ntake at most one immediate.\n\nIt is not clear why these instructions were not removed by upstream\noptimizations. This is possibly a result of values only being known at\nlink time but somehow missed by LTO.\n\nThis CL includes some minor cleanup as well:\n* Running lit tests does not actually require the Subzero runtime, so\n that is removed from the dependencies list.\n* Removes unused capture clauses that were causing warnings treated as\n errors on newer compilers.\n\nChange-Id: I8dc561c82e309fedb46475dcacf14dcb7b535b99\nReviewed-on: https://chromium-review.googlesource.com/896962\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nReviewed-by: Karl Schimpf \u003ckschimpf@chromium.org\u003e\n" }, { "commit": "cb642c8aed91b0f3a00dd96d4b932dc02b41e93c", "tree": "355699174b1df200e54feaa07d44fed8bfbb8a77", "parents": [ "10626a8a29834052366a825e1fcbe9f87ae3fef6" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Oct 03 14:34:46 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Oct 04 02:06:46 2017" }, "message": "Refactor vector register moves.\n\nEliminates code duplication, and moves between identical S registers\nare elided.\n\nChange-Id: Iec5d626c72f3c521a199a755e8804e26a5b4a782\nReviewed-on: https://chromium-review.googlesource.com/697946\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "10626a8a29834052366a825e1fcbe9f87ae3fef6", "tree": "4abfc99ba7d9ce4eff9b02f1ac4459e803066007", "parents": [ "e4707fe3d3e0b0185d57708449517e40b5cb97df" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 02 14:44:03 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Oct 03 16:00:23 2017" }, "message": "Optimize common vector shuffle patterns for ARM32.\n\nUse VDUP for replicating a single element.\nUse VZIP for interleaving vectors.\nUse VMOV Dd, Dm for rearranging quadword vectors.\n\nBug b/67106219\n\nChange-Id: I0de1457454c1db6d467bf870288b7af7cb59ac09\nReviewed-on: https://chromium-review.googlesource.com/695004\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "e4707fe3d3e0b0185d57708449517e40b5cb97df", "tree": "b0a557d0e3852a93817a454029ca74dad5715d2c", "parents": [ "507495ac472a5bcbdf408447150f73e7d79cf375" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 02 21:41:13 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Oct 03 13:41:30 2017" }, "message": "Fix vector packing pseudo-instruction.\n\nWhen the first source register is different from the destination\nregister, but equal to the second source, we need to first narrow\nthe second source to prevent overwriting data we still need. Similar\nwhen the destination is equal to the first source.\n\nBug b/37496082\n\nChange-Id: I908a8e125a77ec4bf1eb5eab9e48c6112ee4ca13\nReviewed-on: https://chromium-review.googlesource.com/696032\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "507495ac472a5bcbdf408447150f73e7d79cf375", "tree": "7328b5ed7073c153500f5ac9b594400a009fbe93", "parents": [ "f45aa6e560f26c350bc7e9f22a8c752aadf72d74" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Sep 27 19:06:35 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Oct 02 12:47:40 2017" }, "message": "Implement vector intrinsics for ARM32.\n\nMultiplyAddPairs is implemented using VMULL+VPADD.\nMultiplyHighSigned/Unsigned is implemented using VMULL+VSHRN.\nSubVectorLoad/Store is implemented using VLDR/VLD1/VSTR/VST1.\nVectorPackSigned/Unsigned is implemented using two VQMOVN.\n\nBug b/37496078\nBug b/37496856\nBug b/37496321\nBug b/37496082\n\nChange-Id: I141fd901d53da24ce780f503dc7ad17b94fc6ba8\nReviewed-on: https://chromium-review.googlesource.com/693049\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "f45aa6e560f26c350bc7e9f22a8c752aadf72d74", "tree": "6a3fe07d5e947db658022f080435e756c96777ae", "parents": [ "86e21263a330d6ce8a6a5b62cc238f77284c4ddc" ], "author": { "name": "Casey Dahlin", "email": "sadmac@google.com", "time": "Wed Jun 28 20:58:58 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Sep 29 00:16:09 2017" }, "message": "Support saturating vector add and subtract on ARM32.\n\nBug b/37495545\n\nChange-Id: I1d025f894bb7cf08dcaafd950605781633596ab3\nReviewed-on: https://chromium-review.googlesource.com/689098\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "86e21263a330d6ce8a6a5b62cc238f77284c4ddc", "tree": "7f46c47be2999b527e74d0d7c87ecd4935244694", "parents": [ "b4ab3a2782f441823d5d64c39a82ab4de0cecc88" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Sep 27 18:33:11 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Thu Sep 28 17:25:43 2017" }, "message": "Support vector shift by constant for ARM32.\n\nThis also fixes the encoding of right shifts with unsigned elements.\n\nBug b/37496338\n\nChange-Id: I9a1dc91359daea5f4391a137b7f9e03bd941146b\nReviewed-on: https://chromium-review.googlesource.com/688057\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "b4ab3a2782f441823d5d64c39a82ab4de0cecc88", "tree": "93153bb8545863821bb8cd11fb6f1e1124158a5a", "parents": [ "83670e622ac9f4285356063e076e0638e3e3ba17" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Apr 18 19:14:16 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Sep 27 19:11:33 2017" }, "message": "Add intrinsic lowering stubs for ARM.\n\nChange-Id: I24a2ac536c24732538cc5c7f2aad4e7763989f9e\nReviewed-on: https://chromium-review.googlesource.com/568673\nReviewed-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "83670e622ac9f4285356063e076e0638e3e3ba17", "tree": "4a80e6fa1d1f05cab764443f64e4db412525273f", "parents": [ "ca3f7bfe49aec3dcddd26ede7c547f8662b3b99a" ], "author": { "name": "Aaron Gable", "email": "agable@chromium.org", "time": "Wed Jul 26 22:11:04 2017" }, "committer": { "name": "Aaron Gable", "email": "agable@chromium.org", "time": "Tue Aug 01 20:04:08 2017" }, "message": "Upload pnacl-subzero changes to Gerrit by default\n\nR\u003dbradnelson@chromium.org\n\nBug: 665593\nChange-Id: I1b9a5fd31c5e426bd193a32a2feb291eb17af5a6\nReviewed-on: https://chromium-review.googlesource.com/587993\nReviewed-by: Brad Nelson \u003cbradnelson@chromium.org\u003e\n" }, { "commit": "ca3f7bfe49aec3dcddd26ede7c547f8662b3b99a", "tree": "5fa1b33e9f994f17e38e793f25b3998e14b0e977", "parents": [ "fb705a6d55003b2c32772ae49e25b0babcff5acc" ], "author": { "name": "Lei Zhang", "email": "thestig@chromium.org", "time": "Tue May 23 22:08:59 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue May 23 22:08:59 2017" }, "message": "Do not inline GlobalContext TLS methods.\n\nThe gold linker thinks that causes an ODR violation.\n\nBUG\u003dchromium:449754\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2896133003 .\n\nPatch from Lei Zhang \u003cthestig@chromium.org\u003e.\n" }, { "commit": "fb705a6d55003b2c32772ae49e25b0babcff5acc", "tree": "0003a7fb4cc1d78774492565490c63334244a1ea", "parents": [ "6c629dc303ddbb76edc70832073bc22657783079" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon May 01 19:31:29 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue May 02 14:22:38 2017" }, "message": "Fix lowering ARM shift by zero as a move.\n\nARM does not support shifting by 0. An immediate value of 0 is interpreted as\nshifting by 32. See section A8.4.1 Constant shifts of the ARMv7-A/R reference\nmanual.\n\nChange-Id: I289a7c6091c04387700dc2e9b3f959639bd919ce\nReviewed-on: https://chromium-review.googlesource.com/491949\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "6c629dc303ddbb76edc70832073bc22657783079", "tree": "8568ddd5d634fa91ea9f9078b14b1b07acccce02", "parents": [ "47b6ba6db7c8b8dd36ac54dbd19ae7b4a5b77424" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Apr 25 20:41:00 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Apr 25 21:01:16 2017" }, "message": "Fix potential null pointer dereference.\n\nWhen the compiler does not perform return value optimization, the\nLockPtr\u003c\u003e destructor of the temporary object is called after the move\nconstructor has set the Lock to null, thus causing a null pointer\ndereference in the destructor. This can be replicated using the\n-fno-elide-constructors build flag.\n\nChange-Id: Ie00c3f93364fdf78ea1993469b9a606b3c87ebdc\nReviewed-on: https://chromium-review.googlesource.com/486985\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "47b6ba6db7c8b8dd36ac54dbd19ae7b4a5b77424", "tree": "7330b1065d6a8f252450b2902bc9d2e8b73ccc96", "parents": [ "eb568f5569b3a25346eb601898baf6fe774f6d4a" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Feb 13 16:27:34 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Feb 13 16:57:59 2017" }, "message": "Optimize floating-point minimum/maximum operations.\n\nCombine a floating-point compare + select into min/max operations on\nx86 using SSE2 instructions, where equivalent. Only handles the most\ncommon cases for now (but for both scalars and vectors):\n\nx \u003e y ? x : y -\u003e max\nx \u003c y ? x : y -\u003e min\n\nBUG\u003dswiftshader:19\n\nChange-Id: Ic87bfa90cefd8014af5624d85a2ecef1c891e328\nReviewed-on: https://chromium-review.googlesource.com/439814\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "eb568f5569b3a25346eb601898baf6fe774f6d4a", "tree": "be6f1e4a22458cb8c6357e0ad537ad63a6658c48", "parents": [ "e965530494ef466f9f039df950b37a64a1a6472b" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Feb 13 14:49:43 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Feb 13 16:55:06 2017" }, "message": "Implement minps/maxps x86 instructions.\n\nBUG\u003dswiftshader:19\n\nChange-Id: I6954db6f1ab3456a199d164676d7f01f91ff7514\nReviewed-on: https://chromium-review.googlesource.com/400090\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "e965530494ef466f9f039df950b37a64a1a6472b", "tree": "c7df8976bbffb07abb405e90a5bc0777c169f481", "parents": [ "71f6c8deb228ebf7fa586c6252d917b14bf54f98" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Feb 13 15:21:20 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Feb 13 16:45:59 2017" }, "message": "Add boolean type property function.\n\nThis will be used by a subsequent min/max optimization patch.\n\nBUG\u003dswiftshader:19\n\nChange-Id: I524806d478ecc13de539d166940e16764ce4a7d9\nReviewed-on: https://chromium-review.googlesource.com/441215\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "71f6c8deb228ebf7fa586c6252d917b14bf54f98", "tree": "cbc7a1972b919c021cc817bd3b395ba11615682b", "parents": [ "c48bb8b02c98ae49438e43aa1143a958784822a5" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Fri Feb 10 14:45:40 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Feb 10 14:45:40 2017" }, "message": "[SubZero] Generate spin-lock for atomic load store\n\nThe patch generates LL-SC sequence for atomic load store operations.\n64-bit atomic operations are lowered to __sync_*_8 functions.\n\nR\u003dstichnot@chromium.org\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n\nReview-Url: https://codereview.chromium.org/2682673002 .\n" }, { "commit": "c48bb8b02c98ae49438e43aa1143a958784822a5", "tree": "208054c90426d8304c19ffcb5093f5f97bf0394b", "parents": [ "59b0dad77008c3f053c2675bde5d98fc9f825ea8" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Feb 02 13:50:06 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Feb 03 16:09:28 2017" }, "message": "Fix swizzle of second vector.\n\nChange-Id: I6ad399f250a49ae5e1c49f3afafbab7ca58fb81b\nReviewed-on: https://chromium-review.googlesource.com/435715\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "59b0dad77008c3f053c2675bde5d98fc9f825ea8", "tree": "43c1e526bb01c1e7fe7d22e3877fc1dffbe9b747", "parents": [ "36bcf2d917adf63b67fd26dd24b2d50e92c53f4e" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Feb 01 17:03:02 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Feb 01 17:03:02 2017" }, "message": "Subzero: Add a new document describing the register allocator.\n\nThere\u0027s some good work in Subzero\u0027s register allocation, which deserves to be captured in a single place.\n\nBUG\u003d none\nR\u003dkschimpf@google.com\n\nReview URL: https://codereview.chromium.org/2277493003 .\n" }, { "commit": "36bcf2d917adf63b67fd26dd24b2d50e92c53f4e", "tree": "82230621641b16f64fd2c0d3586e4d06b930c631", "parents": [ "e986b318b8295be4e361fe55c21bef6939151e49" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Jan 27 09:48:46 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 31 15:08:11 2017" }, "message": "Optimize single-vector shuffling.\n\nChange-Id: Id3d40a72cb74c75ef4431e6af8855e08bde2bb5c\nReviewed-on: https://chromium-review.googlesource.com/433329\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "e986b318b8295be4e361fe55c21bef6939151e49", "tree": "466fa969cfc30618b686988b7d39d738dd786dc1", "parents": [ "aaf04f586f4e94a04e20f47c3d7268a607b8a1e0" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Jan 27 08:56:42 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 31 15:07:55 2017" }, "message": "Perform address optimization of sub-vector load/store.\n\nChange-Id: I3459f9a5472aba37e1b7016b27403094e17bb9f7\nReviewed-on: https://chromium-review.googlesource.com/433372\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "aaf04f586f4e94a04e20f47c3d7268a607b8a1e0", "tree": "e250fd581876c782485622020a578b12967de539", "parents": [ "fc8f6bfae75430b00d8d6fbf78e62da4c3abed9d" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Jan 26 22:10:27 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 31 15:07:18 2017" }, "message": "Optimize x86-64 addressing with a 32-bit unsigned index.\n\nChange-Id: I65aff3da87dfb9c3e5db58621a1a02944a6065e8\nReviewed-on: https://chromium-review.googlesource.com/433365\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "fc8f6bfae75430b00d8d6fbf78e62da4c3abed9d", "tree": "af8dbe6c31511006c713941fabaa14c87084e1bf", "parents": [ "8bd18e1be3eb25d60a4696bb948ab41f6ce6afd6" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Thu Jan 26 06:53:08 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Jan 26 06:53:08 2017" }, "message": "[SubZero] Fix code generation of AtomicCmpxchg\n\nThe patch fixes a code generation issue occurred in the 2nd iteration of LL-SC loop.\n\nR\u003dstichnot@chromium.org\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n\nReview-Url: https://codereview.chromium.org/2656723003 .\n" }, { "commit": "8bd18e1be3eb25d60a4696bb948ab41f6ce6afd6", "tree": "e608131110b73ef8796cbbec85ec47f33ca2cd52", "parents": [ "5c4d677c58b0bb1fc22c8515c4141c8b77cf6bf1" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Jan 20 20:21:45 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 24 22:16:10 2017" }, "message": "Work around empty set default parameter compilation issue.\n\nOlder versions of Clang don\u0027t allow using an empty set as the default\nparameter of a class with an explicit constructor.\n\"error: chosen constructor is explicit in copy-initialization\"\n\nBUG\u003dchromium:630728\n\nChange-Id: I580073788ce3346d1ecffab336a0fcee210b2e0f\nReviewed-on: https://chromium-review.googlesource.com/431080\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "5c4d677c58b0bb1fc22c8515c4141c8b77cf6bf1", "tree": "e528c8ace885df7676662536e63ca167f0611eba", "parents": [ "dbf81e0c0e364173176159e0e2548e9948b197a4" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Jan 18 21:57:52 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Jan 19 00:13:47 2017" }, "message": "Fix signed/unsigned comparison warning.\n\nChange-Id: Idf81fb96dd32df8f96b5bc688bdce290265ff372\nReviewed-on: https://chromium-review.googlesource.com/430230\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "dbf81e0c0e364173176159e0e2548e9948b197a4", "tree": "6b7ed1d2810dec370076348e345062b22d4ceec2", "parents": [ "4e679e51f645b99940f43c80fec7f146d59eb00f" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Sat Jan 14 17:53:55 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 17 21:17:31 2017" }, "message": "Implement an intrinsic for nearbyint.\n\nThe round intrinsic gets translated to roundps on x86, which is SSE4.1\nonly. cvtps2pd + cvtdq2ps can be used as an SSE2 fallback. cvtps2pd\nalso corresponds to LLVM\u0027s nearbyint intrinsic.\n\nBUG\u003dswiftshader:20\n\nChange-Id: I8b5896c443f202a5b25125b4e5049b0b3d3a11b0\nReviewed-on: https://chromium-review.googlesource.com/428491\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "4e679e51f645b99940f43c80fec7f146d59eb00f", "tree": "2811ec154a3e4a11093b51639aac2ee13b617579", "parents": [ "a551dfce9bd44034b4fd588cd4b06b3b0435df8d" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Jan 12 22:01:06 2017" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 17 18:40:29 2017" }, "message": "Support 16-byte aligned stack on 32-bit Windows.\n\nOn Windows x86-32, the ABI only guarantees the stack to be 4-byte\naligned. We therefore need the stack pointer to be explicitly\naligned when using vectors. This demands using a frame pointer (to\naccess function arguments). Also, we had to change accessing spilled\nvariables from the stack pointer instead of the frame pointer so they\nare also aligned. This change does not affect PNaCl. Projects using\nthe Microsoft ABI should define SUBZERO_USE_MICROSOFT_ABI.\n\nBUG\u003dswiftshader:29\n\nChange-Id: I186ce9435244d6fa9494ec514a91122b6be130b3\nReviewed-on: https://chromium-review.googlesource.com/427348\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "a551dfce9bd44034b4fd588cd4b06b3b0435df8d", "tree": "14734aeb86dd08826b81b47e9ee8b784ada1522c", "parents": [ "298d14e243e0a83abf89bc9d1d29cc95905b0236" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Dec 14 22:37:10 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Jan 17 18:39:55 2017" }, "message": "Always align the stack to the fixed alloca requirements.\n\nLocal variables that use a fixed alloca stack slot are assigned offsets\nstarting at 0, before the prolog is written. Therefore the stack pointer\nneeds to be aligned to the alloca\u0027s maximum alignment requirement. This\nrequired the following changes:\n- Add FixedAllocaSizeBytes to SpillAreaSizeBytes before aligning it.\n- Compute the maximum alignment requirement from FixedAllocaAlignBytes\n and SpillAreaAlignmentBytes, and prior NeedsStackAlignment uses.\n- Always align the stack pointer to this maximum.\n- Affected lit tests have been rebased. Note that in some cases the\n frame size is now bigger than necessary. This is due to\n FixedAllocaSizeBytes being padding to be a multiple of the alignment.\n This isn\u0027t strictly necessary since the spill areas take care of their\n own alignment.\n\nBUG\u003dswiftshader:29\n\nChange-Id: Ief30acda91c958d072528b8b59c2e933f68adbb1\nReviewed-on: https://chromium-review.googlesource.com/419816\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "298d14e243e0a83abf89bc9d1d29cc95905b0236", "tree": "6268779b83a1e4f6ecffc352cb6d42eb530ddb27", "parents": [ "2bbda7f48c30683eea26e4901054c58b79221d53" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Wed Jan 11 13:58:27 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Jan 11 13:58:27 2017" }, "message": "Subzero, MIPS32: Atomic intrinsics fixes\n\nThis patch introduces changes to the MIPS32 intrinsic functions to\ncomply with PNaCl smoke tests.\n\nAlso made a change regarding addressing relative to frame pointer,\nsince it differs in MIPS compared to ARM and x86.\n\nR\u003dstichnot@chromium.org\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n\nReview-Url: https://codereview.chromium.org/2619363003 .\n" }, { "commit": "2bbda7f48c30683eea26e4901054c58b79221d53", "tree": "02e58ef7dc5449bac3c626fa6416893cabbf16c8", "parents": [ "4b1bdae3955769f62e563e826065d10ddfe040ea" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Wed Jan 11 05:48:58 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Jan 11 05:48:58 2017" }, "message": "[SubZero] Fix code generation issues occurred in Cross-test and PNaCL smoke-tests\n\nThe patch fixes various code generation issues found during testing of Cross-test and PNaCL smoke-test framework.\n\n 1)\tTo keep track of branches to same label, relative position of the branch from previous branch is used.\n 2)\tFixed encoding of conditional mov instructions\n 3)\tAdded MovFP64ToI64 instruction for f64 to i64 move\n 4)\tHandled vector-types in Phi nodes\n 5)\tFixed alignment of spilled vector arguments on stack\n 6)\tSave-restore FP registers\n 7)\tFixed code generation for Zext and Sext operations\n 8)\tFixed InsertElement for vi16x8 type\n\nR\u003dstichnot@chromium.org\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n\nReview-Url: https://codereview.chromium.org/2619943003 .\n" }, { "commit": "4b1bdae3955769f62e563e826065d10ddfe040ea", "tree": "fb46e6bc84a8b6a474595b4ab44fdb6031977beb", "parents": [ "f2c91aa563c9ec867e88e633ecda66c01380c9ae" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Jan 06 23:19:17 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Sat Jan 07 06:40:10 2017" }, "message": "Fix Mac OS compilation.\n\nChange-Id: I140c17b1b48156ae5dd5ca6cf4ef41f3bc03f16a\nReviewed-on: https://chromium-review.googlesource.com/425780\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "f2c91aa563c9ec867e88e633ecda66c01380c9ae", "tree": "50c5bf1957e2e4538336ca6d7f8e27d19276bbfe", "parents": [ "c78be4ec9c4a1d34d0abed50de0de37a87b4accc" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Jan 06 21:23:38 2017" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Jan 06 21:46:43 2017" }, "message": "Fix inadvertent use of the Microsoft x86-64 calling convention.\n\nChange-Id: I3d60dd5d2020b0fe234587810bdddbf855fa9e4a\nReviewed-on: https://chromium-review.googlesource.com/425829\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "c78be4ec9c4a1d34d0abed50de0de37a87b4accc", "tree": "0933fdf4f30e71f56e208ef3ff7f96934557c4fd", "parents": [ "d722ff2022b623710d9f89f090f162d7f5d521f5" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Dec 27 04:10:25 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Dec 27 04:10:25 2016" }, "message": "Subzero: Fix a build error against LLVM trunk.\n\nLLVM trunk commit a5479197371e169bc676e0e570f2a994329f8773 split include/llvm/Bitcode/ReaderWriter.h into two separate header files. This CL hacks it back, in a sense.\n\nWith this change (along with the rest of the recent fixes), Subzero builds against the latest LLVM trunk - 759dd39154f0bbf1adc87bf613c95f9564f64aa8 at this writing.\n\nBUG\u003d none\nR\u003dkschimpf@google.com\n\nReview-Url: https://codereview.chromium.org/2604753002 .\n" }, { "commit": "d722ff2022b623710d9f89f090f162d7f5d521f5", "tree": "be12fc4b1e698388a1aab6343d58f7693c41f46f", "parents": [ "79810753afd2c8e7b0d04359481fa2ae4dac4bf2" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Dec 27 04:10:07 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Dec 27 04:10:07 2016" }, "message": "Subzero: Fix build problem against LLVM trunk.\n\nLLVM commit 4ac35e82723d4bdbc094f73b2d33aaaa86a00d6a removed some code that Subzero still needs, specifically StreamingMemoryObject and its parents.\n\nDeal with this by adding those files back to the Subzero repo.\n\nThis allows Subzero to build through LLVM c75e3c2fd2bb9ca010f4e1c32acbd142adc32c7f.\n\nBUG\u003d none\nR\u003dkschimpf@google.com\n\nReview-Url: https://codereview.chromium.org/2605653002 .\n" }, { "commit": "79810753afd2c8e7b0d04359481fa2ae4dac4bf2", "tree": "4c1377779206cde6815f291023a83938456c1277", "parents": [ "f8c695b31745752d359a19d26d7a7250ab78cc9c" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Dec 26 22:16:00 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Dec 26 22:16:00 2016" }, "message": "Subzero: Fix some build problems against LLVM trunk.\n\nThis allows Subzero to build through LLVM e8516587a2604386a8faaab28c410663c2ec884c.\n\n1. Mirror an internal change to the llvm::cl implementation.\n\n2. Something changed in llvm::format() that now requires an explicit conversion from RegNumT.\n\nBUG\u003d none\nR\u003dkschimpf@google.com\n\nReview-Url: https://codereview.chromium.org/2602713002 .\n" }, { "commit": "f8c695b31745752d359a19d26d7a7250ab78cc9c", "tree": "f987f701975b91658192ec5ab9bf130f9532339f", "parents": [ "35bbca3f007c4272447d0088b1da40a5e6393c3c" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Dec 26 17:56:08 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Dec 26 17:56:08 2016" }, "message": "Subzero: Fix a build issue against LLVM trunk.\n\nThe existing code gives a build error against LLVM trunk.\n\nActually, the latest LLVM trunk causes a lot more Subzero build problems, but if we roll LLVM back to 1d79fff9e65e77f84bf80c2cf4f0155bd167c90d, everything builds except for this one issue.\n\nBUG\u003d none\nR\u003dkschimpf@google.com\n\nReview-Url: https://codereview.chromium.org/2598263002 .\n" }, { "commit": "35bbca3f007c4272447d0088b1da40a5e6393c3c", "tree": "13b6efaa28a6f5b6a2d01f6bd21b065c44ce5e6a", "parents": [ "373913fa0a4c7c4dd23ada9e6b4e1780fdea0096" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Sat Dec 24 04:00:19 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Sat Dec 24 04:00:19 2016" }, "message": "Subzero: Fix multiply defined symbols in Windows/g++ build.\n\nBased on build failure messages in https://build.chromium.org/p/tryserver.nacl/builders/nacl-toolchain-win7-pnacl-x86_64/builds/3609/steps/llvm_i686_w64_mingw32%20%28build%29/logs/stdio .\n\nBUG\u003d none\nR\u003djpp@chromium.org\n\nReview URL: https://codereview.chromium.org/2601653002 .\n" }, { "commit": "373913fa0a4c7c4dd23ada9e6b4e1780fdea0096", "tree": "fd8f820e273ee67320923f2852c6e8a7d64a201e", "parents": [ "579b1b3a84da15d233c9ab4e3d3dc35cff4edf86" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Dec 23 14:07:07 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Dec 23 14:07:07 2016" }, "message": "Subzero: Legalize the movzx argument.\n\nThe movzx operand must be a register or memory operand. An immediate operand is not allowed.\n\nBUG\u003d https://bugs.chromium.org/p/nativeclient/issues/detail?id\u003d4384\nR\u003djpp@chromium.org\n\nReview URL: https://codereview.chromium.org/2598153002 .\n" }, { "commit": "579b1b3a84da15d233c9ab4e3d3dc35cff4edf86", "tree": "e0cab5c8b81bfdb8781532e6abc478c1da05aaea", "parents": [ "a29da906bf910a32de3a81ef43d6ec43f138f2ab" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Dec 09 19:56:03 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Dec 09 22:29:22 2016" }, "message": "Generalize vector shuffling to accept any operand.\n\nThe arguments get legalized to Reg or Mem, so we can allow constants\nas well (including undef values). This change makes all instruction\u0027s\nsource arguments Ice::Operands.\n\nBUG\u003dswiftshader:24\n\nChange-Id: I1659cdfdb1b8a12c4acc7c473211d8a67bfd5868\nReviewed-on: https://chromium-review.googlesource.com/418504\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "a29da906bf910a32de3a81ef43d6ec43f138f2ab", "tree": "80e8111be1920b7933ac74853dd8180f37edff05", "parents": [ "e1e17832e9c85b8b2be2a9efe98697c9d851345a" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Nov 30 15:39:01 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Nov 30 17:03:42 2016" }, "message": "Ensure that the sub-vector load destination is a register.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I7e10342fa1ef9bce22bc8c445240fc34a68e8f47\nReviewed-on: https://chromium-review.googlesource.com/414992\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "e1e17832e9c85b8b2be2a9efe98697c9d851345a", "tree": "edd08b1f0d4067aeb94a39e5cb8d1df11eaa0b96", "parents": [ "8be69751c0300f5a35a064a71a4c2e385755065f" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Nov 25 04:38:33 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Nov 28 15:36:16 2016" }, "message": "Fix skipping deleted instructions before replacing operands.\n\nFixes hitting a (benign) assert in replaceSource().\n\nChange-Id: I7f984d484133e619717d004f20cd671a54473185\nReviewed-on: https://chromium-review.googlesource.com/414490\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "8be69751c0300f5a35a064a71a4c2e385755065f", "tree": "3a4e90bd1e74fa241d4e962a693013ad70f3e4d6", "parents": [ "ef18fc51dd5c1b2640c291a363cf1e4712142eaa" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Mon Nov 28 14:57:27 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Nov 28 14:57:27 2016" }, "message": "[SubZero] Fix size of arguments on stack\n\nThis patch fixes size of arguments on stack\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2533563002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "ef18fc51dd5c1b2640c291a363cf1e4712142eaa", "tree": "ceb31688d5ddf4c6af202a0bd36abb690842a063", "parents": [ "2220990ebb0419cca8938bd023a13da4eff47aa1" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Nov 24 14:54:12 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Thu Nov 24 20:17:55 2016" }, "message": "Match sub-vector load/store operand order to regular load/store.\n\nBUG\u003dswiftshader:15\n\nChange-Id: If608ab4903d97daa0ad342d02f496ac3fa6471d9\nReviewed-on: https://chromium-review.googlesource.com/414389\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "2220990ebb0419cca8938bd023a13da4eff47aa1", "tree": "7146f09cffd24c09ee3eb93cc2f6557f0b09cda1", "parents": [ "8208e756c4b6b488d2fc053072b579c63d5d73ae" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Nov 23 20:37:06 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Nov 23 22:22:03 2016" }, "message": "Fix offset adjustment in x86 address optimization.\n\nChange-Id: I469a7ddaa658d79fc491112b63972bd9b056689d\nReviewed-on: https://chromium-review.googlesource.com/414186\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nReviewed-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "8208e756c4b6b488d2fc053072b579c63d5d73ae", "tree": "edd447111ca6b8384a40dc5d10bfa3aea0fabdac", "parents": [ "becb85f1fa938e3046576ad2c6e9abe60b7288e7" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Tue Nov 22 18:25:37 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Nov 22 18:25:37 2016" }, "message": "Subzero, MIPS32: Changes for improving sandbox crosstest results\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2519863002 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "becb85f1fa938e3046576ad2c6e9abe60b7288e7", "tree": "156083af07157dc899c42a18570927d52d348bb6", "parents": [ "6e03343bc3076bc76e994a5027c11b014c0af839" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Fri Nov 18 20:15:46 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Nov 18 20:15:46 2016" }, "message": "[Subzero][MIPS] Implements atomic intrinsics for MIPS32\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2504253002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "6e03343bc3076bc76e994a5027c11b014c0af839", "tree": "d2fd187f498d9ee38e4f01ab44d1a6eddda5a7d2", "parents": [ "83425dec5ecae21e092a9a440845ce99a13ded69" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Thu Nov 10 01:31:51 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Nov 10 01:31:51 2016" }, "message": "Subzero, MIPS32: Sandbox initial patch\n\nInitial patch regarding the Subzero MIPS32 sandboxing stage.\nAt the moment, the results of the crosstests with vector tests\ndisabled are as follows:\n\nASM mode:\n19 passing / 5 failing\ntest_bitmanip: O2\ntest_calling_conv: Om1, O2\ntest_sync_atomic: Om1, O2\n\nELF mode:\n15 passing / 9 failing\ntest_bitmanip: O2\ntest_calling_conv: Om1, O2\ntest_global: Om1, O2\ntest_stacksave: Om1, O2\ntest_sync_atomic: Om1, O2\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2482123002 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "83425dec5ecae21e092a9a440845ce99a13ded69", "tree": "4fe4f0547a16b3dbef9703e54d36da06db8bd867", "parents": [ "21f78bb1b7cd9040ce5baea3be51f7be49a1bb1f" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Nov 01 21:18:51 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Nov 09 14:07:43 2016" }, "message": "Support 64-bit jump tables with LP64 data model.\n\nBUG\u003dswiftshader:9\n\nChange-Id: I779abfe7775632e1108e9d608bf21a63c8cefe9e\nReviewed-on: https://chromium-review.googlesource.com/407882\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "21f78bb1b7cd9040ce5baea3be51f7be49a1bb1f", "tree": "d35847364b680a3d48d024715a37a9e928bf45dd", "parents": [ "3da9f6571608b337af7c9d2ad9e0ca938bb49ae7" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Mon Nov 07 04:53:48 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Nov 07 04:53:48 2016" }, "message": "[SubZero] Utilize instructions with immediate operands\n\nThis patch optimizes code generation of instructions with 16-bit immediate operands\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2478113003 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "3da9f6571608b337af7c9d2ad9e0ca938bb49ae7", "tree": "324340a8d88deb73f881b94090fcccc87a568ea2", "parents": [ "f0d12c30600aad72ee5184cb47ddf2b67e7ccbee" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Fri Nov 04 05:54:06 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Nov 04 05:54:06 2016" }, "message": "[SubZero] Generate MIPS.abiflags section\n\nThe patch generates MIPS.abiflags section. This section contains a\nversioned data structure with essential information required for\nloader to determine the requirements of the application.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2471883005 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "f0d12c30600aad72ee5184cb47ddf2b67e7ccbee", "tree": "53d9e01174a95195c496241de1a2f9e05a135e14", "parents": [ "f8c9977b34f0c053204f0401b056f4529e23c2a6" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Oct 27 19:17:41 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Thu Nov 03 03:19:02 2016" }, "message": "Implement floating-point rounding intrinsic.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I8e53f2fdb8208f8be0f4cdff3241b4a5efe9bc8a\nReviewed-on: https://chromium-review.googlesource.com/404352\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "f8c9977b34f0c053204f0401b056f4529e23c2a6", "tree": "44a7bee2a250d0c2c084ee2903e7de2e2957419b", "parents": [ "0c4c07d0d83aedf38f0529ecc2337ccfa5a18ba6" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Wed Nov 02 12:59:06 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Nov 02 12:59:06 2016" }, "message": "Subzero, MIPS32: Stacksave/Stackrestore implementation\n\nImplemets Stacksave/Stackrestore; test_stacksave runs successfully\nwhen jal implementation is present, both in forceasm as well as in\nelf mode\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2455933002 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "0c4c07d0d83aedf38f0529ecc2337ccfa5a18ba6", "tree": "d7a4e6aabe97ddf305860b0486eb580039c75105", "parents": [ "67a49b5b3606815c5636d9a061a372c367f28685" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Wed Nov 02 06:53:52 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Nov 02 06:53:52 2016" }, "message": "[SubZero] Fix code generation for vector type\n\nThe patch fixes legalizeToReg issues in vector code generation.\nThe patch also generates JALR for pointer to function and corrects encoding of FP conditional move instruction.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2468133002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "67a49b5b3606815c5636d9a061a372c367f28685", "tree": "119fae32d068d1234211c04a9a5dc3afca600685", "parents": [ "13cde0f5d3af36944075ef510c3b05aeacf959b3" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 26 17:18:35 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Nov 01 16:51:53 2016" }, "message": "Implement saturated vector add/subtract.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Ic120eddd1761e33b7d76bf3ed8ec5ca74634f958\nReviewed-on: https://chromium-review.googlesource.com/403477\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "13cde0f5d3af36944075ef510c3b05aeacf959b3", "tree": "27affd6bd1e29672d11876d51f520d70eb3eea04", "parents": [ "956cfd60e8ee9efcc374cd19513bbf35a20e52ee" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 26 14:36:11 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Nov 01 12:53:33 2016" }, "message": "Implement integer vector multiply intrinsics.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Ib822b50c0a14e5ebc114db9759cbeecbb9f7a3c1\nReviewed-on: https://chromium-review.googlesource.com/403472\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "956cfd60e8ee9efcc374cd19513bbf35a20e52ee", "tree": "98f5b853316d575c3af3d9769f7d61bfc045c174", "parents": [ "7ad028ea364d4795fd3fa441e36341db6849941e" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 31 18:28:09 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Nov 01 02:17:04 2016" }, "message": "Generalize the Sqrt intrinsic to process vectors.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Ib89d628c85696c20a249b8810cd357a292d10402\nReviewed-on: https://chromium-review.googlesource.com/405293\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "7ad028ea364d4795fd3fa441e36341db6849941e", "tree": "9d40c2d67acacaaf5bea42fab3e12f42272c929a", "parents": [ "73ae4fdc32345ad1dc44cc5268204174555b2b55" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Mon Oct 31 13:41:46 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 31 13:41:46 2016" }, "message": "This patch enables running a couple more of lit tests for MIPS32\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2448193008 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "73ae4fdc32345ad1dc44cc5268204174555b2b55", "tree": "38738817f9cc8764dae05bf7058b0ff8a59ad70e", "parents": [ "130aca718ed16be20e616e591a53de1d69f00d12" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Oct 28 15:24:50 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 28 19:05:28 2016" }, "message": "Preserve rsi and rdi when using Microsoft x86-64 calling convention.\n\nAlso, their priority is lowered so that registers which are scratch on both\nUnix and Windows are preferred by the register allocator.\n\nBUG\u003dswiftshader:22\n\nChange-Id: Id55d8c8b8c106947e3041a082099069d7c6c6ed0\nReviewed-on: https://chromium-review.googlesource.com/404503\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "130aca718ed16be20e616e591a53de1d69f00d12", "tree": "c0d908cd0921357a3eca9f2b36d9aba901b87154", "parents": [ "0dabe18f9251b797cbc6aa9c014a7a79e58fae07" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Fri Oct 28 12:30:54 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Oct 28 12:30:54 2016" }, "message": "[SubZero] Generate relocations for MIPS\n\nThe patch generate Hi, Lo, Jal and data relocations. Instruction encoding of instructions like ldc1, sdc1 etc. has been added.\n\nFollowing tests from cross-test framework were tested successfully: (non-vector, OBJ mode, Om1, O2)\n\nmem_intrin\nTotalTests\u003d114300 Passes\u003d114300 Failures\u003d0\n\nsimple_loop\nTotalTests\u003d102 Passes\u003d102 Failures\u003d0\n\ntest_arith\nTotalTests\u003d49489704 Passes\u003d49489704 Failures\u003d0\n\ntest_bitmanip\nTotalTests\u003d1200 Passes\u003d1200 Failures\u003d0\n\ntest_cast\nTotalTests\u003d7444 Passes\u003d7444 Failures\u003d0\n\ntest_fcmp\nTotalTests\u003d123904 Passes\u003d123904 Failures\u003d0\n\ntest_global\nTotalTests\u003d270 Passes\u003d270 Failures\u003d0\n\ntest_icmp\nTotalTests\u003d3341520 Passes\u003d3341520 Failures\u003d0\n\ntest_strengthreduce\nTotalTests\u003d240 Passes\u003d240 Failures\u003d0\n\nFollowing tests are disabled as they are either all-vectors or contain unimplemented intrinsic lowering:\n\ntest_calling_conv\ntest_select\ntest_stacksave\ntest_sync_atomic\ntest_vector_ops\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2446273003 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "0dabe18f9251b797cbc6aa9c014a7a79e58fae07", "tree": "0b02f38c1a3c3eac97e39607e5baf2ce818119a9", "parents": [ "a7979bfd84bc01c0891169e8d058063fe699cc0e" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Fri Oct 28 12:23:57 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Fri Oct 28 12:23:57 2016" }, "message": "Subzero, MIPS32: Remove --skip-unimplemented from lit tests\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2433243003 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "a7979bfd84bc01c0891169e8d058063fe699cc0e", "tree": "405424780343a6c407a831f91dacfaace260ff9b", "parents": [ "a3688eaffa0b11e21b7dc48bbba514fc3f999198" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Wed Oct 26 13:16:19 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Oct 26 13:16:19 2016" }, "message": "[SubZero] Fix f64 to/from i64 moves\n\nThe allocation of Hi/Lo part of i64 on stack has been corrected as per MIPS32 ABI. The patch also fixes ZEXT issues occurred while lowering unsigned operations.\nFollowing tests from cross-test framework were testing successfully: (non-vector, ASM mode, Om1, O2)\n\nmem_intrin\nTotalTests\u003d114300 Passes\u003d114300 Failures\u003d0\n\nsimple_loop\nTotalTests\u003d102 Passes\u003d102 Failures\u003d0\n\ntest_arith\nTotalTests\u003d49489704 Passes\u003d49489704 Failures\u003d0\n\ntest_bitmanip\nTotalTests\u003d1200 Passes\u003d1200 Failures\u003d0\n\ntest_cast\nTotalTests\u003d3722 Passes\u003d3722 Failures\u003d0\n\ntest_fcmp\nTotalTests\u003d123904 Passes\u003d123904 Failures\u003d0\n\ntest_global\nTotalTests\u003d270 Passes\u003d270 Failures\u003d0\n\ntest_icmp\nTotalTests\u003d3341520 Passes\u003d3341520 Failures\u003d0\n\ntest_strengthreduce\nTotalTests\u003d240 Passes\u003d240 Failures\u003d0\n\nFollowing tests are disabled as they are either all-vectors or contain unimplemented intrinsic lowering:\n\ntest_calling_conv\ntest_select\ntest_stacksave\ntest_sync_atomic\ntest_vector_ops\n\nThere are couple of fixes to ARM32 and X86 specific files occurred due to compile-time errors.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2432373002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "a3688eaffa0b11e21b7dc48bbba514fc3f999198", "tree": "d7835c9aa4b7510fd08ec052330fe258a3b7d7f7", "parents": [ "61593fb9d46f1b3d98878d4b90dcb07009b93fbc" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 26 02:23:07 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Wed Oct 26 04:53:16 2016" }, "message": "Fix two-vector unpack case.\n\nBug swiftshader:15\n\nChange-Id: I351268b44491091c271d6c7c5b644cd21ffb623b\nReviewed-on: https://chromium-review.googlesource.com/403409\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "61593fb9d46f1b3d98878d4b90dcb07009b93fbc", "tree": "ac9a098d1fdeda707f572fdb67ab1c941711e145", "parents": [ "c9e91afc85fe6d682edd75ce9e7a1d924721651d" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 24 15:04:22 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Oct 24 16:57:52 2016" }, "message": "Fix unit tests.\n\nChange-Id: I70899be0455958aaad6af8d8218f1db50591beae\nReviewed-on: https://chromium-review.googlesource.com/401385\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "c9e91afc85fe6d682edd75ce9e7a1d924721651d", "tree": "f2de08fe5e0f7189699e9ea22aa22df4ce79dd4a", "parents": [ "d0e30307fdd0c0c4bdaeab8d18dc53fb6aa845ea" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 19 18:19:41 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 21 17:48:18 2016" }, "message": "Remove verified asserts.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I3c3314f3787d42835a9483c7b797dc1dbdc0b76a\nReviewed-on: https://chromium-review.googlesource.com/400663\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "d0e30307fdd0c0c4bdaeab8d18dc53fb6aa845ea", "tree": "23e622c3985d24f47f87ba4627a47069ebdcf0ac", "parents": [ "e3cabdafc42cb8120ebb268378e9facd9a8107bd" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Oct 21 16:47:59 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 21 17:46:38 2016" }, "message": "Assert that PNaCl bitcode only uses 128-bit vector casts.\n\nChange-Id: I5aee2c998842f95ccc44d5c0fed90aa289bdf67b\nReviewed-on: https://chromium-review.googlesource.com/401639\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "e3cabdafc42cb8120ebb268378e9facd9a8107bd", "tree": "1b8ecb9c460dcc618e75297085c83984e410480e", "parents": [ "8b8af8248849994d22b1c13c5e0d12b251338075" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Oct 18 20:27:27 2016" }, "committer": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Oct 20 20:26:36 2016" }, "message": "Implement vector sign mask intrinsic.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I7fec56518a5b4e07d2189ab01a0a223b038564c1\n" }, { "commit": "8b8af8248849994d22b1c13c5e0d12b251338075", "tree": "407b094c412045837cac4a88a9aa3a7908171e05", "parents": [ "ef8210d9cf35b11d6ce8210a43e127466098826b" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 17 21:42:29 2016" }, "committer": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 19 20:28:41 2016" }, "message": "Implement bitcast between i32 and (emulated) v4i8.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Ic795def8a914508ab0d850c846b73b343ace45de\n" }, { "commit": "ef8210d9cf35b11d6ce8210a43e127466098826b", "tree": "552432e7049d422bdb1786f1eac554b823471dfa", "parents": [ "32f9ccef8dee33110cc9ebdca2853d6a078752ed" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 17 21:42:29 2016" }, "committer": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 19 20:28:31 2016" }, "message": "Implement vector packing intrinsics.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Id95a08f82c47ec20bb958358c01f389b6fb5565b\n" }, { "commit": "32f9ccef8dee33110cc9ebdca2853d6a078752ed", "tree": "f2b865e3d66f0400979e68c57010d0d295097270", "parents": [ "7145e69306813c75e1d9260ed468c266f9e038f8" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 19 05:24:27 2016" }, "committer": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Oct 19 20:26:59 2016" }, "message": "Fix 64-bit pointer type for non-x32 ABIs.\n\nBUG\u003dswiftshader:9\n\nChange-Id: Ife06416736d47acba4f2cff1ea8b17be61134752\n" }, { "commit": "7145e69306813c75e1d9260ed468c266f9e038f8", "tree": "9a965053189f015efa2bd5b539380b38fa164082", "parents": [ "1448d95c891c926f480e4b898c6cb5716983e582" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Oct 19 12:49:47 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Oct 19 12:49:47 2016" }, "message": "Subzero: Fix compiler warnings.\n\nsrc/IceTargetLoweringX86BaseImpl.h:6093:13: error: unused variable \u0027Src1RM\u0027 [-Werror,-Wunused-variable]\n auto *Src1RM \u003d legalize(Src1, Legal_Reg | Legal_Mem);\n ^\n\nsrc/IceTargetLoweringX86BaseImpl.h:4007:3: error: default label in switch which covers all enumeration values [-Werror,-Wcovered-switch-default]\n default:\n ^\n\nsrc/IceTargetLoweringMIPS32.cpp:4065:3: error: default label in switch which covers all enumeration values [-Werror,-Wcovered-switch-default]\n default:\n ^\n\nsrc/IceTargetLoweringARM32.cpp:4975:3: error: default label in switch which covers all enumeration values [-Werror,-Wcovered-switch-default]\n default:\n ^\n\nBUG\u003d none\nR\u003dcapn@chromium.org\n\nReview URL: https://codereview.chromium.org/2434643002 .\n" }, { "commit": "1448d95c891c926f480e4b898c6cb5716983e582", "tree": "76cc9d3c9810401e8861977a5ff0e0bf2e92ee24", "parents": [ "89be8873c07f0a4a67697288804d8b9b28ef5520" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Oct 14 20:37:09 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Tue Oct 18 16:55:40 2016" }, "message": "Optimize shuffles corresponding to x86 punpckh instructions.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I04a7c4206f3936c604ec623e43834c2a153fd3cb\nReviewed-on: https://chromium-review.googlesource.com/399379\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "89be8873c07f0a4a67697288804d8b9b28ef5520", "tree": "99e1f5e7f52598c760b24aac57243815370e4d5f", "parents": [ "0e90622d85cf110ceabefb03bb0019571accc522" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Tue Oct 18 14:32:21 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Oct 18 14:32:21 2016" }, "message": "[Subzero][MIPS32] Account for variable alloca alignment bytes in addProlog\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2425673002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "0e90622d85cf110ceabefb03bb0019571accc522", "tree": "d25b7d0d0c589a283b54c981ab00b2e890b0b188", "parents": [ "7638e275845ae4e6e2087198dde9b969d5a7f0d2" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 17 17:02:47 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Oct 17 19:08:39 2016" }, "message": "Generate error on unexpected intrisics.\n\nChange-Id: I5a02aee156a64f48baca356f0a5263123f570741\nReviewed-on: https://chromium-review.googlesource.com/399590\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "7638e275845ae4e6e2087198dde9b969d5a7f0d2", "tree": "4c9ec6d9bbff6f30f6ab3ac81906ea49db18d71a", "parents": [ "f52cea4bb987e8c43a881f243dbb8f41f31797d5" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Oct 06 15:33:55 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Oct 17 19:05:16 2016" }, "message": "Add x86 vector packing instructions.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I0d40fab6287130143693e8e4752859b7142a503d\nReviewed-on: https://chromium-review.googlesource.com/394007\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "f52cea4bb987e8c43a881f243dbb8f41f31797d5", "tree": "adf6d9728918168812fb7c6ac2ed6a2605f46614", "parents": [ "3a01f337a848f4200542393d60d32c7e79ff42ea" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Fri Oct 14 20:29:11 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Mon Oct 17 17:04:32 2016" }, "message": "Fix unpacking from a single vector.\n\nBoth vector arguments were being used in a punpckl instruction, while\nthe shuffle mask repeats elements from just the first vector.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I8e29c252ee4957692c4949e724ae67253b423e89\nReviewed-on: https://chromium-review.googlesource.com/399419\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "3a01f337a848f4200542393d60d32c7e79ff42ea", "tree": "67a89741c3971bfdc3e2832d92e0fb755f7755a2", "parents": [ "45e4d5edca13d208bb65cb0423264ab9d9c2f08a" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Mon Oct 17 13:33:50 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 17 13:33:50 2016" }, "message": "[SubZero] Implement Fcmp, ICmp, Cast and Select for vector type\n\nThe patch scalarizes Fcmp, ICmp, Cast and Select for operands of vector type.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2412053002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "45e4d5edca13d208bb65cb0423264ab9d9c2f08a", "tree": "b6a4f2ea43558139a3f2274febb216863cd17210", "parents": [ "468323715da610429d8d4f2525387fbd2f2165eb" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Mon Oct 17 04:50:43 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 17 04:50:43 2016" }, "message": "[SubZero] Handle relocatable constants for MIPS\n\nThe patch generates HI/LO modifiers for relocatable constants.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2420033002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "468323715da610429d8d4f2525387fbd2f2165eb", "tree": "208bd3ebbd6d5f48258ef86912bb6ac6abae6448", "parents": [ "ec92917338c13f4791c441deb4e447fd9ef60228" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Sun Oct 16 14:20:44 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Sun Oct 16 14:20:44 2016" }, "message": "[Subzero][MIPS32] Fix alloca alignment and offset for Om1 and O2 optimization\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2417233002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "ec92917338c13f4791c441deb4e447fd9ef60228", "tree": "5975ed63735ea51afad9c4a0105ceb533db53088", "parents": [ "b001cc4c0bea8e13be1d6733ef0f96fe5706fc21" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Sun Oct 16 14:13:38 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Sun Oct 16 14:13:38 2016" }, "message": "[SubZero] Legalize load, store for MIPS post lower\n\nThis patch legalizes load, store instructions post lowering.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2411193003 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "b001cc4c0bea8e13be1d6733ef0f96fe5706fc21", "tree": "94e97d006048c95ce042bcf081921c05607e42b8", "parents": [ "46f4feaafea68c18e4384ac4bad2e012eec52e5f" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Wed Oct 12 06:36:01 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Oct 12 06:36:01 2016" }, "message": "[Subzero][MIPS32] Implement bitcast operation for both 32-bit and 64-bit operands\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2404803002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "46f4feaafea68c18e4384ac4bad2e012eec52e5f", "tree": "3a71e531908545258dc9ec2edbfc2e5234e2068f", "parents": [ "b09353904b93e8914d459a286b6b9badb8b1983b" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Thu Oct 06 21:25:39 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 07 13:20:56 2016" }, "message": "Support running unit tests on Windows.\n\nBUG\u003dswiftshader:7\n\nChange-Id: I83e51a3256365700dbaf550ed4b50c2352612f7d\nReviewed-on: https://chromium-review.googlesource.com/394887\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "b09353904b93e8914d459a286b6b9badb8b1983b", "tree": "1c703146446c127ebf385575b235be16a41b4e07", "parents": [ "71c6937378c27d04edd59399817873dc090025a7" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Oct 04 14:13:03 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 07 03:58:37 2016" }, "message": "Optimize x86 vector shift by constant.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I4b7b97f3de18c201a502d0bc38a2c845a1caf278\nReviewed-on: https://chromium-review.googlesource.com/392627\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\n" }, { "commit": "71c6937378c27d04edd59399817873dc090025a7", "tree": "a298fd6bc7cd9f206dd54c490e14014c46e371e9", "parents": [ "acfb3df0bdd163dc2c5f94f450be2e3c7ec0fff1" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Tue Oct 04 04:07:21 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 07 03:56:22 2016" }, "message": "Optimize lowering of x86 byte and word vector unpack.\n\nBUG\u003dswiftshader:15\n\nChange-Id: Id0d3bed46d00336fc31501c41a26ebe2d4ddd697\nReviewed-on: https://chromium-review.googlesource.com/392626\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "acfb3df0bdd163dc2c5f94f450be2e3c7ec0fff1", "tree": "6bf0e222ce2338e9b5c397f50130ec2603814e25", "parents": [ "533a514f4d045a6cd9d2dd9fccb54f86094e6ef3" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Mon Oct 03 14:46:30 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Fri Oct 07 03:55:58 2016" }, "message": "Implement intrinsics for loading/storing subvectors.\n\nThis enables emulating 64-bit and 32-bit vectors using 128-bit\nvectors internally (x86 only for now). Note that these Intrinsics\nare not part of the PNaCL specification.\n\nBUG\u003dswiftshader:15\n\nChange-Id: I61a666243832c2856e60eb477d42a72dec07d01d\nReviewed-on: https://chromium-review.googlesource.com/392246\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "533a514f4d045a6cd9d2dd9fccb54f86094e6ef3", "tree": "c4999cafd43c2b9dcd159f0eb309e49ebdbe357c", "parents": [ "afe5fe227a89523c708b9543bbba1db90be36c82" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Oct 06 21:24:38 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Oct 06 21:24:38 2016" }, "message": "Subzero: Fix \"make -f Makefile.standalone check-lit FORCEASM\u003d1\".\n\nhttps://codereview.chromium.org/2384983002/ apparently didn\u0027t test the full \"make presubmit\", otherwise this would have been caught.\n\nBUG\u003d none\nTBR\u003djpp@chromium.org\n\nReview URL: https://codereview.chromium.org/2399873003 .\n" }, { "commit": "afe5fe227a89523c708b9543bbba1db90be36c82", "tree": "3ecb5fd8ddc7a43b76c97d52a778ef49f653bb4c", "parents": [ "033dda7ec0c31f6de17890ccb05859766269c548" ], "author": { "name": "Stefan Maksimovic", "email": "makdstefan@gmail.com", "time": "Wed Oct 05 17:45:19 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Oct 05 17:45:19 2016" }, "message": "Subzero, MIPS32: Fix conditional mov instructions\n\nThis patch implements changes needed for conditional mov instructions\nto fix problem with failing crosstest and invalid register allocation.\nProblem is visible from icmp test examples, causing cross test for icmp\nto fail. Eg:\n\nIncorrect, before this change:\n674:\t00653026\txor\ta2,v1,a1\n678:\t00a3182b\tsltu\tv1,a1,v1\n67c:\t0082102b\tsltu\tv0,a0,v0\n680:\t0043180a\tmovz\tv1,v0,v0\n\nCorrect, aftrer this change:\n674:\t00653026\txor\ta2,v1,a1\n678:\t00a3182b\tsltu\tv1,a1,v1\n67c:\t0082102b\tsltu\tv0,a0,v0\n680:\t0046180a\tmovz\tv1,v0,a2\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2394773004 .\n\nPatch from Stefan Maksimovic \u003cmakdstefan@gmail.com\u003e.\n" }, { "commit": "033dda7ec0c31f6de17890ccb05859766269c548", "tree": "7073e33e1ba489bbf47a32a7026d2faf8acd34af", "parents": [ "958ddb75696187bd45682d464b5b2eb8ca378e06" ], "author": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 03 14:56:25 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 03 14:56:25 2016" }, "message": "Subzero: Remove --skip-unimplemented from ARM lit tests.\n\nARM support is complete, so clean up some of the lit tests:\n\n1. Remove --skip-unimplemented\n2. Use --filetype\u003dobj instead of \u003dasm, and remove --assemble\n3. Remove --need\u003dallow_dump requirement\n4. Remove related TODOs.\n5. Fix some CHECK lines because objdump output is slightly different from filetype\u003dasm output.\n\nBUG\u003d none\nR\u003djpp@chromium.org\n\nReview URL: https://codereview.chromium.org/2384983002 .\n" }, { "commit": "958ddb75696187bd45682d464b5b2eb8ca378e06", "tree": "a199438bd00a3bb3cc36b41879d5b37066e288c6", "parents": [ "9309756d32e760188e111dec5672050637cd0eca" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Mon Oct 03 14:52:48 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Mon Oct 03 14:52:48 2016" }, "message": "[SubZero] Vector types support for MIPS\n\nThis patch implements vector operations on MIPS32 using VariableVecOn32 method (on the lines of Variable64On32).\nVector operations are scalarized prior to lowering. Each vector variable is split into 4 containers to hold a variable of vector type.\nFor MIPS32, four GP/FP registers are used to hold a vector variable. Arguments are passed in GP registers irrespective of the type of the vector variable.\n\nLit test vector-mips.ll has been added to test this implementation.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2380023002 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n" }, { "commit": "9309756d32e760188e111dec5672050637cd0eca", "tree": "3126bc4b09dcacb8cb4e030903bef710088028a0", "parents": [ "269eed47c46a019745960292d3b440584cd50efd" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Thu Sep 29 13:52:29 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Sep 29 13:52:29 2016" }, "message": "[Subzero][MIPS] Implement conditional branches with 64-bit integer compares\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2384433002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "269eed47c46a019745960292d3b440584cd50efd", "tree": "ec8bf1b66ab2637a5e96bf26d9c6a6e2a22d9847", "parents": [ "f53580b063d9a206e72a2b4166aff2c62e71555d" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Thu Sep 29 13:22:25 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Thu Sep 29 13:22:25 2016" }, "message": "[Subzero][MIPS] Add RUN command line with -Om1 in test 64bit.pnacl.ll\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2376233004 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "f53580b063d9a206e72a2b4166aff2c62e71555d", "tree": "7c430001a7dca4177e8706edcfe6d0b6a6a249f1", "parents": [ "fe93fdd5e29d7470eee6d41eb70075ad9415d227" ], "author": { "name": "Nicolas Capens", "email": "capn@google.com", "time": "Wed Sep 28 11:20:30 2016" }, "committer": { "name": "Nicolas Capens", "email": "nicolascapens@google.com", "time": "Thu Sep 29 01:32:05 2016" }, "message": "Don\u0027t emit address size prefixes for native x86-64 ABI.\n\nAddress size prefixes are used in 64-bit x86 for PNaCl\u0027s use of the\nx32 ABI with ILP32 data model. Don\u0027t emit them for any other ABI.\n\nBUG\u003dswiftshader:9\n\nChange-Id: I1351db086d44ce4b144b3428866a54e84637b9a4\nReviewed-on: https://chromium-review.googlesource.com/390409\nReviewed-by: Jim Stichnoth \u003cstichnot@chromium.org\u003e\nTested-by: Nicolas Capens \u003cnicolascapens@google.com\u003e\n" }, { "commit": "fe93fdd5e29d7470eee6d41eb70075ad9415d227", "tree": "67b52a46b40e2f424dd1952cfed32441c17695c5", "parents": [ "70b6ed47d1a4692408a0adbe025329c6adcb5c10" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Wed Sep 28 13:38:44 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Sep 28 13:38:44 2016" }, "message": "Subzero, MIPS32: SRAV instruction encoding\n\nImplements SRAV instruction encoding\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2375923002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "70b6ed47d1a4692408a0adbe025329c6adcb5c10", "tree": "70533c27c8fc184248a92323389fa2bd8b52fdde", "parents": [ "b0f09fc8c53e20fe65c4a1119a1a4ac56f4fbfa0" ], "author": { "name": "Sagar Thakur", "email": "sagar.thakur@imgtec.com", "time": "Wed Sep 28 13:36:47 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Sep 28 13:36:47 2016" }, "message": "[Subzero][MIPS] Implement 64-bit integer compare operations\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2369323002 .\n\nPatch from Sagar Thakur \u003csagar.thakur@imgtec.com\u003e.\n" }, { "commit": "b0f09fc8c53e20fe65c4a1119a1a4ac56f4fbfa0", "tree": "e0fc248c96fe30aadf4541e95969a9627d7e8454", "parents": [ "6163c62fdaca2c1f192afb0ce1da96daf9e53306" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Wed Sep 28 03:43:11 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Sep 28 03:43:11 2016" }, "message": "Subzero, MIPS32: MOVZ instruction encoding\n\nImplements MOVZ instruction encoding\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2377783002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "6163c62fdaca2c1f192afb0ce1da96daf9e53306", "tree": "5b9ff406e1d25221ba412ab7813f3f9e5e61c683", "parents": [ "175cb1381a25d5fdd961bfeab4dea0dfb92a527a" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Wed Sep 28 03:38:30 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Wed Sep 28 03:38:30 2016" }, "message": "Subzero, MIPS32: DIVU instruction encoding\n\nImplements DIVU instruction encoding\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2377733002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "175cb1381a25d5fdd961bfeab4dea0dfb92a527a", "tree": "9e9f7e04a4cf327c2ab15985edb368a0f65472e9", "parents": [ "6fd9c0e3f5b4c7e9e22dc982bbebbb5a891b35e9" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Tue Sep 27 14:00:17 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Sep 27 14:00:17 2016" }, "message": "Subzero, MIPS32: Intrinsic call Bswap for i16, i32 and i64\n\nImplements intrinsic call llvm.bswap for i16, i32 and i64\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2368343003 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "6fd9c0e3f5b4c7e9e22dc982bbebbb5a891b35e9", "tree": "389b0d78f0ee0354398fbd748305a6f505a1f05d", "parents": [ "58eeedf7a97346c147c6206fe6364d79be3f7a04" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Tue Sep 27 03:58:17 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Sep 27 03:58:17 2016" }, "message": "Subzero, MIPS32: Intrinsic calls Ctlz and Cttz for i64\n\nImplements intrinsic calls llvm.ctlz and llvm.cttz for i64.\nAlso adds test cases for constant operands.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2364093002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "58eeedf7a97346c147c6206fe6364d79be3f7a04", "tree": "463b3346a666bc70acc2ede500cba2ba0ab7e9f9", "parents": [ "4c49b108f7d568c549646f2ede010ab57d41275f" ], "author": { "name": "Jaydeep Patil", "email": "jaydeep.patil@imgtec.com", "time": "Tue Sep 27 03:48:18 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Sep 27 03:48:18 2016" }, "message": "Subzero, MIPS32: Binding intrablock labels, unconditional branch\n\nThis patch was supposed to be a part of patch with instruction encodings.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2367743004 .\n\nPatch from Jaydeep Patil \u003cjaydeep.patil@imgtec.com\u003e.\n\nAuthor: Jaydeep Patil\n" }, { "commit": "4c49b108f7d568c549646f2ede010ab57d41275f", "tree": "ca10baeeb568cd0a0293f5f311fa704d5055da38", "parents": [ "623f8ce32c8dd33f9c755dde229967dfddd3353b" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Tue Sep 27 03:12:17 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Sep 27 03:12:17 2016" }, "message": "Subzero, MIPS32: Filling missing bits from genTargetHelperCallFor\n\nImplements missing calls to runtime libraries, covering mostly\ndata casting.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2363333002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" }, { "commit": "623f8ce32c8dd33f9c755dde229967dfddd3353b", "tree": "cd0130e74c072bbc940d60d37e170fb743a8d5b9", "parents": [ "98405d3938a1a0aeda6033eb13ee77e44b45ed33" ], "author": { "name": "Srdjan Obucina", "email": "Srdjan.Obucina@imgtec.com", "time": "Tue Sep 27 03:03:20 2016" }, "committer": { "name": "Jim Stichnoth", "email": "stichnot@chromium.org", "time": "Tue Sep 27 03:03:20 2016" }, "message": "Subzero, MIPS32: Cross-testing enabled for MIPS32\n\nEnables running crosstests for MIPS32 target.\n\nR\u003dstichnot@chromium.org\n\nReview URL: https://codereview.chromium.org/2085303002 .\n\nPatch from Srdjan Obucina \u003cSrdjan.Obucina@imgtec.com\u003e.\n" } ], "next": "98405d3938a1a0aeda6033eb13ee77e44b45ed33" }