)]}' { "log": [ { "commit": "3ee225094a683a0f52ad9a00299a7f2004a2c997", "tree": "709edd9eed789e36e5501d4db10add8cc4d03f86", "parents": [ "35c021c49f44cd4e06875c4b1398a2b3464132bb" ], "author": { "name": "Chih-Hung Hsieh", "email": "chh@google.com", "time": "Tue Mar 13 23:16:53 2018 -0700" }, "committer": { "name": "Yi Kong", "email": "yikong@google.com", "time": "Fri Mar 01 10:37:33 2019 -0800" }, "message": "Revert two changes that break Android builds.\n\n* Earlier versions of lld were able to replace Android\u0027s\n lld.gold and link most files.\n* With two recent lld changes, debuggerd_test64 and some other files\n failed to link because libunwindstack.a is linked like --whole-archive.\n We need to revert the following two changes:\n * \"Put undefined symbols from shared libraries in the symbol table.\"\n lld/trunk@326242\n * \"Make undefined symbol in DSO to pull out object files from archive files.\"\n lld/trunk@325849\n\nBug: 74755833\nChange-Id: I21b0df5e3c19bdfff4ea6c0ff9e9e562bd73f921\n" }, { "commit": "35c021c49f44cd4e06875c4b1398a2b3464132bb", "tree": "972f829a5ed4f0e4a14f0a713b49a8f0db93a26f", "parents": [ "958317e226bf38ef5e647d0b056e8e95eb472c70", "14aa57da0f92683f0b8bdac0acda485a6f73edc7" ], "author": { "name": "Yi Kong", "email": "yikong@google.com", "time": "Wed Feb 27 14:18:31 2019 -0800" }, "committer": { "name": "Yi Kong", "email": "yikong@google.com", "time": "Wed Feb 27 14:18:31 2019 -0800" }, "message": "Merge 14aa57da0 for LLVM update to 353983\n\nChange-Id: Iee1792355bab030d81eb4cb645257dbac2537641\n" }, { "commit": "958317e226bf38ef5e647d0b056e8e95eb472c70", "tree": "931efffa9cb05aee46a4bcb41be035962d0418ad", "parents": [ "0f50d2239d204e2f2026b9db97f30b5d54753f0c" ], "author": { "name": "Yi Kong", "email": "yikong@google.com", "time": "Wed Feb 27 14:18:31 2019 -0800" }, "committer": { "name": "Yi Kong", "email": "yikong@google.com", "time": "Wed Feb 27 14:18:31 2019 -0800" }, "message": " revert to previous base llvm-svn.349610\n\nChange-Id: Ib0010bb41abcf296cd09c0808410dce0c5ccbfd6\n" }, { "commit": "14aa57da0f92683f0b8bdac0acda485a6f73edc7", "tree": "d876c849dbae72bc7d3815b8722cca20e2ef4958", "parents": [ "1590c11381acc085db8c7e14022e66bc8670728e" ], "author": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Wed Feb 13 21:49:55 2019 +0000" }, "committer": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Wed Feb 13 21:49:55 2019 +0000" }, "message": "ELF: Allow GOT relocs pointing to non-preemptable ifunc to resolve to an IRELATIVE where possible.\n\nNon-GOT non-PLT relocations to non-preemptible ifuncs result in the\ncreation of a canonical PLT, which now takes the identity of the IFUNC\nin the symbol table. This (a) ensures address consistency inside and\noutside the module, and (b) fixes a bug where some of these relocations\nend up pointing to the resolver.\n\nFixes (at least) PR40474 and PR40501.\n\nDifferential Revision: https://reviews.llvm.org/D57371\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353981 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "1590c11381acc085db8c7e14022e66bc8670728e", "tree": "12ba30ccffc97ab5099c27585f1989c431c9366a", "parents": [ "bb3788a37f05f73bb70240c0bd9b619f3ffa2d00" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 13 18:51:15 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 13 18:51:15 2019 +0000" }, "message": "Recover elf32-bigmips and elf32-powerpc support in LLD\n\nThis fixes a 7.0 -\u003e 8.0 regression when parsing\nOUTPUT_FORMAT(\"elf32-powerpc\"); or elf32-bigmips directive in ldscripts\nas well as an unknown emulation error when lld is invoked by clang due\nto missed elf32ppclinux case.\n\nPatch by vit9696\n\nDifferential Revision: https://reviews.llvm.org/D58005\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353968 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "bb3788a37f05f73bb70240c0bd9b619f3ffa2d00", "tree": "c26a1d000b2761fe7dce3b1abc615d65a224053f", "parents": [ "983304d126315155c2cf88e872930a2e7b4d9cac" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 13 18:48:39 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 13 18:48:39 2019 +0000" }, "message": "Show \"Unknown -z option\" error message even if --version or --help are given.\n\nPreviously, we validated -z options after we process --version or --help flags.\nSo, if one of these flags is given, we wouldn\u0027t show an \"unknown -z option\"\nerror. This patch fixes that behavior.\n\nDifferential Revision: https://reviews.llvm.org/D55446\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353967 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "983304d126315155c2cf88e872930a2e7b4d9cac", "tree": "306c682d14c4f9748e0b7747ecf9abc265d52502", "parents": [ "00d845f77b82ed67ee23e0e265066fd1480527fd" ], "author": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Tue Feb 12 17:49:04 2019 +0000" }, "committer": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Tue Feb 12 17:49:04 2019 +0000" }, "message": "[PPC64] Update tests to reflect change in printing of call operand. [NFC]\n\nThe printing of branch operands for call instructions was changed to properly\nhandle negative offsets. Updating the tests to reflect that.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353866 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "00d845f77b82ed67ee23e0e265066fd1480527fd", "tree": "d9ef824bb8c8e18c4d3355e99d30354323a0f2f1", "parents": [ "2cfa2ff84ac8faa091d3ca042f2beea2f53b58ff" ], "author": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Tue Feb 12 15:35:49 2019 +0000" }, "committer": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Tue Feb 12 15:35:49 2019 +0000" }, "message": "[PPC64] Sort .toc sections accessed with small code model relocs.\n\nA follow up to the intial patch that unblocked linking against libgcc.\nFor lld we don\u0027t need to bother tracking which objects have got based small\ncode model relocations. This is due to the fact that the compilers on\npowerpc64 use the .toc section to generate indirections to symbols (rather then\nusing got relocations) which keeps the got small. This makes overflowing a\nsmall code model got relocation very unlikely.\n\nDifferential Revision: https://reviews.llvm.org/D57245\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353849 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "2cfa2ff84ac8faa091d3ca042f2beea2f53b58ff", "tree": "f8d0c4d04f222fcae2f89c3045cfffb876b76a99", "parents": [ "83a25ddffc9d0ac9abe713a9d57b978177594c70" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Feb 11 22:01:32 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Feb 11 22:01:32 2019 +0000" }, "message": "lld: unquote possibly quoted `EXTERN(\"symbol\")` entry in linker script.\n\ngold accepts quoted strings. binutils requires quoted strings for some\nkinds of symbols, e.g.:\n\n it accepts quoted symbols with @ in name:\n\n $ echo \u0027EXTERN(\"__libc_start_main@@GLIBC_2.2.5\")\u0027 \u003e a.script\n $ g++ a.script\n /usr/lib/gcc/x86_64-redhat-linux/4.8.5/../../../../lib64/crt1.o: In function `_start\u0027:\n (.text+0x20): undefined reference to `main\u0027\n collect2: error: ld returned 1 exit status\n\n but rejects them if unquoted:\n\n $ echo \u0027EXTERN(__libc_start_main@@GLIBC_2.2.5)\u0027 \u003e a.script\n $ g++ a.script\n a.script: file not recognized: File format not recognized\n collect2: error: ld returned 1 exit status\n\nTo maintain compatibility with existing linker scripts support quoted\nstrings in lld as well.\n\nPatch by Lucian Adrian Grijincu.\n\nDifferential Revision: https://reviews.llvm.org/D57987\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353756 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "83a25ddffc9d0ac9abe713a9d57b978177594c70", "tree": "edc7bc00d93645be91e11e6a3827ef75611c4dc0", "parents": [ "6b28e9547148c343533176d6ec77db0fe9bff0bc" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Feb 11 21:38:20 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Feb 11 21:38:20 2019 +0000" }, "message": "Convert CRLF. NFC.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353751 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "6b28e9547148c343533176d6ec77db0fe9bff0bc", "tree": "08975f5f73bcf2caf92e6c872f5f935fa4ce2723", "parents": [ "902f2c13e6969c99d3c42e92d8ca7b3358606209" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Sat Feb 09 11:34:59 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Sat Feb 09 11:34:59 2019 +0000" }, "message": "[LLD][ELF] - Update test case after yaml2obj change. NFCI.\n\n.dynamic section format accepted by yaml2obj was\nchanged in r353606\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353607 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "902f2c13e6969c99d3c42e92d8ca7b3358606209", "tree": "43f6209bbd57e4751c33363fbc8ac947426e70ba", "parents": [ "8526a5f8f9c4e16bf4b95cc5d7bc97c9d3a3f96b" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Sat Feb 09 10:09:40 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Sat Feb 09 10:09:40 2019 +0000" }, "message": "[ELF] Delete a comment that is no longer correct. Fix a typo. NFC\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353605 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "8526a5f8f9c4e16bf4b95cc5d7bc97c9d3a3f96b", "tree": "6e9f649049d668f7cb5d88d8242749545e7974bb", "parents": [ "bd893e3b27d75c26c2698b6fbadc4ab07d9893d0" ], "author": { "name": "David L. Jones", "email": "dlj@google.com", "time": "Sat Feb 09 05:00:50 2019 +0000" }, "committer": { "name": "David L. Jones", "email": "dlj@google.com", "time": "Sat Feb 09 05:00:50 2019 +0000" }, "message": "[LLD][ELF] Ensure tests use -o /dev/null instead of attempting to write to the bin directory. [NFC]\n\nThis error was introduced in r353508.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353602 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "bd893e3b27d75c26c2698b6fbadc4ab07d9893d0", "tree": "8e61e8f9f5fbc18d0a9ce633eb109eb77d85ee74", "parents": [ "c8a4571b18b4e8b1de4a01ec7b145aedd55da08c" ], "author": { "name": "Bob Haarman", "email": "llvm@inglorion.net", "time": "Fri Feb 08 21:59:35 2019 +0000" }, "committer": { "name": "Bob Haarman", "email": "llvm@inglorion.net", "time": "Fri Feb 08 21:59:35 2019 +0000" }, "message": "[lld-link] better error message when failing to open archive members\n\nSummary:\nThe message \"could not get the buffer for the member defining symbol\"\nnow also contains the name of the archive and the name of the archive\nmember that we tried to open.\n\nReviewers: ruiu\n\nSubscribers: llvm-commits\n\nTags: #llvm\n\nDifferential Revision: https://reviews.llvm.org/D57974\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353572 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "c8a4571b18b4e8b1de4a01ec7b145aedd55da08c", "tree": "9893f56217c6e390c4dc332481fa65b1ba3a6033", "parents": [ "b69157f4d019ae2804c2cd032504260108973ff4" ], "author": { "name": "James Henderson", "email": "jh7370@my.bristol.ac.uk", "time": "Fri Feb 08 10:31:38 2019 +0000" }, "committer": { "name": "James Henderson", "email": "jh7370@my.bristol.ac.uk", "time": "Fri Feb 08 10:31:38 2019 +0000" }, "message": "[LLD][ELF]Add test for missing thin archive member\n\nThe code producing error messages relating to missing thin archive\nmembers was missing any testing as far as I could see, so this patch\nadds a test for it.\n\nReviewed by: ruiu\n\nDifferential Revision: https://reviews.llvm.org/D57899\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353508 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "b69157f4d019ae2804c2cd032504260108973ff4", "tree": "efb5c81b5365d4fb431532886a45843500b00956", "parents": [ "b6584c3fab115aa46dad27681af7eb3d4e5d2b35" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 22:42:16 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 22:42:16 2019 +0000" }, "message": "[WebAssembly] Refactor handling of weak undefined functions. NFC.\n\nAlso add to the docs.\n\nThis is refactor in preparation for https://reviews.llvm.org/D57909\n\nDifferential Revision: https://reviews.llvm.org/D57920\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353478 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "b6584c3fab115aa46dad27681af7eb3d4e5d2b35", "tree": "dc82a5da8c8931f500756fe32326137a7572f4e8", "parents": [ "fd4d30f6eda1909081ea7234a4a412791ad69574" ], "author": { "name": "Dan Gohman", "email": "dan433584@gmail.com", "time": "Thu Feb 07 22:00:48 2019 +0000" }, "committer": { "name": "Dan Gohman", "email": "dan433584@gmail.com", "time": "Thu Feb 07 22:00:48 2019 +0000" }, "message": "[WebAssembly] Fix imported function symbol names that differ from their import names in the .o format\n\nAdd a flag to allow symbols to have a wasm import name which differs from the\nlinker symbol name, allowing the linker to link code using the import_module\nattribute.\n\nDifferential Revision: https://reviews.llvm.org/D57632\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353473 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "fd4d30f6eda1909081ea7234a4a412791ad69574", "tree": "d57a3215b70f26d5f8b82fa4914c30ecddd908ec", "parents": [ "822543c6671de85c486e755ec6cba7a98b50e67b" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 19:05:26 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 19:05:26 2019 +0000" }, "message": "[WebAssembly] Improve docs for wasm linker\n\nDifferential Revision: https://reviews.llvm.org/D57913\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353451 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "822543c6671de85c486e755ec6cba7a98b50e67b", "tree": "12956337f1cd3ba4ef9b374b24613223cae75638", "parents": [ "620832e9bdb0c07bec847ac153291d3d1b01a92b" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Feb 07 18:12:57 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Feb 07 18:12:57 2019 +0000" }, "message": "Fix a bug in R_X86_64_PC{8,16} relocation handling.\n\nR_X86_64_PC{8,16} relocations are sign-extended, so when we check\nfor relocation overflow, we had to use checkInt instead of checkUInt.\nI confirmed that GNU linkers create the same output for the test case.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353437 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "620832e9bdb0c07bec847ac153291d3d1b01a92b", "tree": "791fbd707ae86f04d6d353dbe7992c2f89f0440a", "parents": [ "369e0b3c9953d51787e13afaaf629dbefe799845" ], "author": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Thu Feb 07 12:39:35 2019 +0000" }, "committer": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Thu Feb 07 12:39:35 2019 +0000" }, "message": "docs: add missingkeyfunction to doctree, fix title\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353399 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "369e0b3c9953d51787e13afaaf629dbefe799845", "tree": "81c6f9bc37137345359c17a7833f1684c3903c5b", "parents": [ "22c19ea7f065aeb877682a70fcdd262419ac06e4" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Thu Feb 07 07:59:43 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Thu Feb 07 07:59:43 2019 +0000" }, "message": "[LLD][ELF] - Set DF_STATIC_TLS flag for X64 target\n\nThis is the same as D57749, but for x64 target.\n\n\"ELF Handling For Thread-Local Storage\" p41 says (https://www.akkadia.org/drepper/tls.pdf):\nR_X86_64_GOTTPOFF relocation is used for IE TLS models.\nHence if linker sees this relocation we should add DF_STATIC_TLS flag.\n\nDifferential revision: https://reviews.llvm.org/D57821\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353378 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "22c19ea7f065aeb877682a70fcdd262419ac06e4", "tree": "e2b447ac5a5935ea3c827472f654aedcc99b3fed", "parents": [ "b2ea35f3ae19635cf7477bc9955a49e93c962dd7" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 01:53:50 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 01:53:50 2019 +0000" }, "message": "[WebAssembly] Honor WASM_SYMBOL_EXPORT symbol flag\n\nThis flag means that symbol should be exported in the final binary.\n\nThe reason for this change is to allow source level annotations to\ntrigger a given symbol to be exported:\nhttps://github.com/emscripten-core/emscripten/issues/7702\n\nDifferential Revision: https://reviews.llvm.org/D57869\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353364 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "b2ea35f3ae19635cf7477bc9955a49e93c962dd7", "tree": "cd9281027dfb86f728a134376f01e380a0392d07", "parents": [ "8c55491c3b88b175636b957b9f940156bd714940" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 01:21:58 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Thu Feb 07 01:21:58 2019 +0000" }, "message": "[WebAssembly] Update test expectations to match llvm-readobj change. NFC.\n\nDifferential Revision: https://reviews.llvm.org/D57868\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353358 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "8c55491c3b88b175636b957b9f940156bd714940", "tree": "9334f149fd6b4adfd8ee39c459c748f04e4fbd22", "parents": [ "0eed4f87a3b79fa02bfd2db5e21e0582eccebf38" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Wed Feb 06 21:05:29 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Wed Feb 06 21:05:29 2019 +0000" }, "message": "[MinGW] Add --insert-timestamp as complement to the --no-insert-timestamp option\n\nDifferential Revision: https://reviews.llvm.org/D57808\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353342 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "0eed4f87a3b79fa02bfd2db5e21e0582eccebf38", "tree": "62a6f61273ffb4aa678f249ca8b89130312409d7", "parents": [ "a8c0107794161cb410d2a402ff8cb5f662af0af5" ], "author": { "name": "Ed Maste", "email": "emaste@freebsd.org", "time": "Wed Feb 06 20:36:02 2019 +0000" }, "committer": { "name": "Ed Maste", "email": "emaste@freebsd.org", "time": "Wed Feb 06 20:36:02 2019 +0000" }, "message": "Correct \"varaible\" typo in comment\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353340 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "a8c0107794161cb410d2a402ff8cb5f662af0af5", "tree": "58afa4108e710d6c660fcf16def50aa1d0f6aa22", "parents": [ "019f3036f9bb0d89a908ea6f324ee53ba5db899c" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 19:28:23 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 19:28:23 2019 +0000" }, "message": "Remove a small header that is used only by one file. NFC.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353331 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "019f3036f9bb0d89a908ea6f324ee53ba5db899c", "tree": "bde9563a5f00c295fa5bf26e78c656f350042147", "parents": [ "e58280ff99d48add88ca6ea6f9dc145209ba1863" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 18:53:17 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 18:53:17 2019 +0000" }, "message": "Add comment.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353323 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "e58280ff99d48add88ca6ea6f9dc145209ba1863", "tree": "d003700bf0764395385e52aba3630e97a00aa616", "parents": [ "10a9e3c7ea1c540fe856e345c88555eaecb6fa2d" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 16:50:09 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Feb 06 16:50:09 2019 +0000" }, "message": "Support R_X86_64_PC8 and R_X86_64_PC16.\n\nThey are defined by the x86-64 ELF ABI standard.\n\nDifferential Revision: https://reviews.llvm.org/D57799\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353314 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "10a9e3c7ea1c540fe856e345c88555eaecb6fa2d", "tree": "8926f8cd3cc22a7fb3e4c53f9bd2295f728525e4", "parents": [ "918a8ad52766fd75ba3449a34bacce1728d6169e" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 14:43:30 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 14:43:30 2019 +0000" }, "message": "Recommit r353293 \"[LLD][ELF] - Set DF_STATIC_TLS flag for i386 target.\"\n\nWith the following changes:\n1) Compilation fix:\nstd::atomic\u003cbool\u003e HasStaticTlsModel \u003d false; -\u003e\nstd::atomic\u003cbool\u003e HasStaticTlsModel{false};\n\n2) Adjusted the comment in code.\n\nInitial commit message:\n\nDF_STATIC_TLS flag indicates that the shared object or executable\ncontains code using a static thread-local storage scheme.\n\nPatch checks if IE/LE relocations were used to check if the code uses\na static model. If so it sets the DF_STATIC_TLS flag.\n\nDifferential revision: https://reviews.llvm.org/D57749\n----\nModified : /lld/trunk/ELF/Arch/X86.cpp\nModified : /lld/trunk/ELF/Config.h\nModified : /lld/trunk/ELF/SyntheticSections.cpp\nAdded : /lld/trunk/test/ELF/Inputs/i386-static-tls-model1.s\nAdded : /lld/trunk/test/ELF/Inputs/i386-static-tls-model2.s\nAdded : /lld/trunk/test/ELF/Inputs/i386-static-tls-model3.s\nAdded : /lld/trunk/test/ELF/Inputs/i386-static-tls-model4.s\nAdded : /lld/trunk/test/ELF/i386-static-tls-model.s\nModified : /lld/trunk/test/ELF/i386-tls-ie-shared.s\nModified : /lld/trunk/test/ELF/tls-dynamic-i686.s\nModified : /lld/trunk/test/ELF/tls-opt-iele-i686-nopic.s\n\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353299 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "918a8ad52766fd75ba3449a34bacce1728d6169e", "tree": "30f0f7765608692210f77866f6dd4307ece09632", "parents": [ "bdddf3e48de1d498081954f9ebf2c9692cc2aec3" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 13:53:32 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 13:53:32 2019 +0000" }, "message": "Revert r353293 \"[LLD][ELF] - Set DF_STATIC_TLS flag for i386 target.\"\n\nIt broke BB:\nhttp://lab.llvm.org:8011/builders/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/builds/43450\nhttp://lab.llvm.org:8011/builders/lld-x86_64-freebsd/builds/27891\n\nError is:\ntools/lld/ELF/Config.h:84:41: error: copying member subobject of type\n\u0027std::atomic\u003cbool\u003e\u0027 invokes deleted constructor std::atomic\u003cbool\u003e HasStaticTlsModel \u003d false;\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353297 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "bdddf3e48de1d498081954f9ebf2c9692cc2aec3", "tree": "96aca404fbf0a444afe7da389cdcfcb381f2033e", "parents": [ "5865fb2366bfe26c8f46cb109f06490191d6f502" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 13:38:10 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Wed Feb 06 13:38:10 2019 +0000" }, "message": "[LLD][ELF] - Set DF_STATIC_TLS flag for i386 target.\n\nDF_STATIC_TLS flag indicates that the shared object or executable\ncontains code using a static thread-local storage scheme.\n\nPatch checks if IE/LE relocations were used to check if the code uses\na static model. If so it sets the DF_STATIC_TLS flag.\n\nDifferential revision: https://reviews.llvm.org/D57749\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353293 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "5865fb2366bfe26c8f46cb109f06490191d6f502", "tree": "30f0f7765608692210f77866f6dd4307ece09632", "parents": [ "3e9573bda10bcefc1fcc09bcc38ccce0d37fc383" ], "author": { "name": "Peter Smith", "email": "peter.smith@linaro.org", "time": "Wed Feb 06 10:03:05 2019 +0000" }, "committer": { "name": "Peter Smith", "email": "peter.smith@linaro.org", "time": "Wed Feb 06 10:03:05 2019 +0000" }, "message": "[ELF][ARM] Add test case that will fail if toPlt() is removed [NFC]\n\nWhen a thunk is created to a PLT entry, the call to the thunk is converted\nto a non-plt expression with fromPlt(). If the thunk becomes unusable we\nretarget the relocation back to its original target and try again. When we\ndo this we need to make sure that we restore the PLT form of the expression\nwith toPlt().\n\nThis change adds a test case that will fail if toPlt() is removed. We need\nto have a call to a preemptible symbol defined within the link unit. If\ntoPlt() is removed then the relocation to the thunk to the PLT entry for the\npreemptible symbol will be retargeted to the preemptible symbol itself\ninstead of its PLT entry.\n\nDifferential Revision: https://reviews.llvm.org/D57743\n\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353285 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3e9573bda10bcefc1fcc09bcc38ccce0d37fc383", "tree": "ac1d8a0e971ec81afd6089b3f3227e02b39dd5db", "parents": [ "84ab074ab01091012fb1bb997faa07e16063abcb" ], "author": { "name": "David L. Jones", "email": "dlj@google.com", "time": "Wed Feb 06 04:49:46 2019 +0000" }, "committer": { "name": "David L. Jones", "email": "dlj@google.com", "time": "Wed Feb 06 04:49:46 2019 +0000" }, "message": "[wasm-ld] Fix wasm trace test to use %t for temporary files.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353272 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "84ab074ab01091012fb1bb997faa07e16063abcb", "tree": "5335e457762fc0c488c54d280f9c5aa517e46666", "parents": [ "1906f04fadb179ffbcb2ba41b703856ba4729717" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Feb 06 02:35:18 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Feb 06 02:35:18 2019 +0000" }, "message": "[WebAssembly] Implement --trace and --trace-symbol\n\nDifferential Revision: https://reviews.llvm.org/D57725\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353264 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "1906f04fadb179ffbcb2ba41b703856ba4729717", "tree": "7f57d15059b58ab522bc0c601ec531a27b041ba9", "parents": [ "2a295842a96e4d2f4ee2b4137a274564d54f1e6e" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Wed Feb 06 02:13:40 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Wed Feb 06 02:13:40 2019 +0000" }, "message": "[PPC64] Delete the unnecessary __tls_get_addr from ppc64-tls-ld-le.s\n\nThe definition is harmful here as it suppresses R_PPC64_REL24 which is\nsupposed to follow R_PPC64_TLSLD.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353263 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "2a295842a96e4d2f4ee2b4137a274564d54f1e6e", "tree": "c1abee89619c322c23a796da8d6f3e86dcd8b080", "parents": [ "c8f96fdf2f71396b5ee06cd1c1a0972808dbf305" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Wed Feb 06 02:00:24 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Wed Feb 06 02:00:24 2019 +0000" }, "message": "[PPC64] Set the number of relocations processed for R_PPC64_TLS[GL]D to 2\n\nSummary:\nR_PPC64_TLSGD and R_PPC64_TLSLD are used as markers on TLS code sequences. After GD-to-IE or GD-to-LE relaxation, the next relocation R_PPC64_REL24 should be skipped to not create a false dependency on __tls_get_addr. When linking statically, the false dependency may cause an \"undefined symbol: __tls_get_addr\" error.\n\nR_PPC64_GOT_TLSGD16_HA\nR_PPC64_GOT_TLSGD16_LO\nR_PPC64_TLSGD R_TLSDESC_CALL\nR_PPC64_REL24 __tls_get_addr\n\nReviewers: ruiu, sfertile, syzaara, espindola\n\nReviewed By: sfertile\n\nSubscribers: emaste, nemanjai, arichardson, kbarton, jsji, llvm-commits, tamur\n\nTags: #llvm\n\nDifferential Revision: https://reviews.llvm.org/D57673\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353262 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "c8f96fdf2f71396b5ee06cd1c1a0972808dbf305", "tree": "095650092541be76a1ca2ab83c704fe466d567e6", "parents": [ "81e6ac84f29cb62902fb3bb12a7b9b7e07396087" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Feb 06 01:00:49 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Feb 06 01:00:49 2019 +0000" }, "message": "[ELF] Fix typo in test/ELF/trace.s\n\nDifferential Revision: https://reviews.llvm.org/D57796\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353254 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "81e6ac84f29cb62902fb3bb12a7b9b7e07396087", "tree": "5098a3d3a68596890641c6907a6cdbe3cac24aae", "parents": [ "44ee588274e0593a9c8ea2d71b8eb741b6965a21" ], "author": { "name": "Zachary Turner", "email": "zturner@google.com", "time": "Wed Feb 06 00:50:35 2019 +0000" }, "committer": { "name": "Zachary Turner", "email": "zturner@google.com", "time": "Wed Feb 06 00:50:35 2019 +0000" }, "message": "[PDB] Remove dots and normalize slashes with /PDBSOURCEPATH.\n\nIn a previous patch, I made changes so that PDBs which were\ngenerated on non-Windows platforms contained sensical paths\nfor the host. While this is an esoteric use case, we need\nit to be supported for certain cross compilation scenarios\nespecially with LLDB, which can debug things on non-Windows\nplatforms.\n\nHowever, this regressed a case where you specify /PDBSOURCEPATH\nand use a windows-style path. Previously, we would still remove\ndots and canonicalize slashes to backslashes, but since my\nchange intentionally tried to support non-backslash paths, this\nwas broken.\n\nThis patch fixes the situation by trying to guess which path\nstyle the user is specifying when /PDBSOURCEPATH is passed.\nIt is intentionally conservative, erring on the side of a\nWindows path style unless absolutely certain. All dots are\nremoved and slashes canonicalized to whatever the deduced\npath style is after appending the file path to the /PDBSOURCEPATH\nargument.\n\nDifferential Revision: https://reviews.llvm.org/D57769\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353250 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "44ee588274e0593a9c8ea2d71b8eb741b6965a21", "tree": "32b99d13615acb6703ecece9fddedabb49acce34", "parents": [ "5d2106efa828acd3f781ad41cb0fff759351e09c" ], "author": { "name": "Matt Davis", "email": "Matthew.Davis@sony.com", "time": "Tue Feb 05 21:01:01 2019 +0000" }, "committer": { "name": "Matt Davis", "email": "Matthew.Davis@sony.com", "time": "Tue Feb 05 21:01:01 2019 +0000" }, "message": "[llvm-readobj] Display sections that do not belong to a segment in the section-mapping\n\nSummary:\nThe following patch adds the \"None\" line to the section to segment mapping dump.\nThat line lists the sections that do not belong to any segment.\nI realize that this change differs from GNU readelf which does not display the latter information.\n\nI\u0027d rather not add this \"feature\" under a command line option. I think that might introduce confusion, since users would have to\nmake an additional decision as to if they want to see all of the section-to-segment map or just a subset of it.\n\nAnother option is to only print the \"None\" line if the `--section-mapping` option is passed; however,\nthat might also introduce some confusion, because the section-to-segment map would be different between`--program-headers`\nand the `--section-mapping` output. While the difference is just the \"None\" line, it seems that if we choose to display\nthe segment-to-section mapping, then we should always display the whole map including the sections\nthat do not belong to segments.\n\n```\nSection to Segment mapping:\n Segment Sections...\n 00\n 01 .interp\n 02 .interp .note.ABI-tag .gnu.hash\n 03 .init_array .fini_array .dynamic\n 04 .dynamic\n 05 .note.ABI-tag\n 06 .eh_frame_hdr\n 07\n 08 .init_array .fini_array .dynamic .got\n None .comment .symtab .strtab .shstrtab \u003c--- THIS LINE\n```\n\nReviewers: grimar, rupprecht, jhenderson, espindola\n\nReviewed By: rupprecht\n\nSubscribers: khemant, emaste, arichardson, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57700\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353217 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "5d2106efa828acd3f781ad41cb0fff759351e09c", "tree": "7ef0e475a352ff9eb5247ff6a8ea3920e6333d28", "parents": [ "c966069499debd5fcddeb9143df0c6f23e12de8f" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Tue Feb 05 19:19:45 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Tue Feb 05 19:19:45 2019 +0000" }, "message": "Inline a trivial function and update comment. NFC.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353200 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "c966069499debd5fcddeb9143df0c6f23e12de8f", "tree": "6163e1e0f7922384f9de3351b3dbfb566309f9ac", "parents": [ "d25c150d1c1b97bb5f583dadc376904c1b80f15b" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Tue Feb 05 16:53:33 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Tue Feb 05 16:53:33 2019 +0000" }, "message": "[WebAssembly] Use wasm-ld rather than lld with -flavor in tests. NFC.\n\nDifferential Revision: https://reviews.llvm.org/D57758\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353187 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "d25c150d1c1b97bb5f583dadc376904c1b80f15b", "tree": "e13dc115ba04ee4a077fd58b8586aca6b262ab8f", "parents": [ "3b0c327a39fa1b19d01eabdef869281485fd6bef" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Feb 05 08:16:10 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Feb 05 08:16:10 2019 +0000" }, "message": "[COFF] Avoid O(n^2) accesses into PartialSections\n\nFor MinGW, unique partial sections are much more common, e.g.\ncomdat functions get sections named e.g. text$symbol.\n\nA moderate sized example of this contains over 200K Chunks\nwhich create 174K unique PartialSections. Prior to SVN r352928\n(D57574), linking this took around 1,5 seconds for me, while\nit afterwards takes around 13 minutes. After this patch, the\nlinking time is back to what it was before.\n\nThe std::find_if in findPartialSection will do a linear scan of\nthe whole container until a match is found. To use something like\nbinary_search or the std::set container\u0027s own methods, we\u0027d need\nto already have a PartialSection*.\n\nReinstate a proper map instead of having a set with a custom sorting\ncomparator.\n\nDifferential Revision: https://reviews.llvm.org/D57666\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353146 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3b0c327a39fa1b19d01eabdef869281485fd6bef", "tree": "cc54194760af452e44a11b6fa8555605f7c57e87", "parents": [ "a78c9caa9cb2e78f6dfaf72d4e534ee71956374b" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Feb 05 08:16:06 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Feb 05 08:16:06 2019 +0000" }, "message": "[MinGW] Hook up the --no-insert-timestamp option\n\nThis fixes PR40582.\n\nPatch by Georg Koppen!\n\nDifferential Revision: https://reviews.llvm.org/D57679\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353145 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "a78c9caa9cb2e78f6dfaf72d4e534ee71956374b", "tree": "f59bc56c1df5f949756ed6cbe33f44a01654d167", "parents": [ "6d0b387c14a3ea04acbafb4b11ede92dced91bb4" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Mon Feb 04 22:34:07 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Mon Feb 04 22:34:07 2019 +0000" }, "message": "[WebAssembly] Update test to match new readobj output. NFC.\n\nDifferential Revision: https://reviews.llvm.org/D57715\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353105 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "6d0b387c14a3ea04acbafb4b11ede92dced91bb4", "tree": "5470e1d716c848f985382c4a02f6d655103c5c14", "parents": [ "66fca3a6b8cc0112a4c72237ffaea5eb4b659e3f" ], "author": { "name": "Heejin Ahn", "email": "aheejin@gmail.com", "time": "Mon Feb 04 19:13:46 2019 +0000" }, "committer": { "name": "Heejin Ahn", "email": "aheejin@gmail.com", "time": "Mon Feb 04 19:13:46 2019 +0000" }, "message": "[WebAssembly] clang-tidy (NFC)\n\nSummary:\nThis patch fixes clang-tidy warnings on wasm-only files.\nThe list of checks used is:\n`-*,clang-diagnostic-*,llvm-*,misc-*,-misc-unused-parameters,readability-identifier-naming,modernize-*`\n(LLVM\u0027s default .clang-tidy list is the same except it does not have\n`modernize-*`.)\n\nThe list of fixes are:\n- Variable names start with an uppercase letter\n- Function names start with a lowercase letter\n- Use `auto` when you use casts so the type is evident\n\nReviewers: sbc100\n\nSubscribers: dschuff, jgravelle-google, sunfish, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57499\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353076 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "66fca3a6b8cc0112a4c72237ffaea5eb4b659e3f", "tree": "c33034cbf0b90686052806a4481f95509b097727", "parents": [ "ed150dd0da9a50c49d7f77d82ddbf1f37ddb6f92" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Mon Feb 04 17:49:33 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Mon Feb 04 17:49:33 2019 +0000" }, "message": "[WebAssembly] Update relocation naming to match llvm change. NFC.\n\nDifferential Revision: https://reviews.llvm.org/D57698\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@353066 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "ed150dd0da9a50c49d7f77d82ddbf1f37ddb6f92", "tree": "7f262d3c51678b589ded7a2fa587c05bbce23c31", "parents": [ "dbc4366c26c3ff28f66054be1ef081df870f5feb" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Sat Feb 02 00:34:28 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Sat Feb 02 00:34:28 2019 +0000" }, "message": "[ELF] Default to --no-allow-shlib-undefined for executables\n\nSummary:\nThis follows the ld.bfd/gold behavior.\n\nThe error check is useful as it captures a common type of ld.so undefined symbol errors as link-time errors:\n\n // a.cc \u003d\u003e a.so (not linked with -z defs)\n void f(); // f is undefined\n void g() { f(); }\n\n // b.cc \u003d\u003e executable with a DT_NEEDED entry on a.so\n void g();\n int main() { g(); }\n\n // ld.so errors when g() is executed (lazy binding) or when the program is started (-z now)\n // symbol lookup error: ... undefined symbol: f\n\nReviewers: ruiu, grimar, pcc, espindola\n\nReviewed By: ruiu\n\nSubscribers: llvm-commits, emaste, arichardson\n\nTags: #llvm\n\nDifferential Revision: https://reviews.llvm.org/D57569\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352943 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "dbc4366c26c3ff28f66054be1ef081df870f5feb", "tree": "8e3ec0d2edc44d7982e96257c07d97a4fca3cfc7", "parents": [ "c71a5001b1af9892cbf1b0aacc28153215989e31" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Fri Feb 01 22:08:09 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Fri Feb 01 22:08:09 2019 +0000" }, "message": "[COFF] Create range extension thunks for ARM64\n\nOn ARM64, this is normally necessary only after a module exceeds\n128 MB in size (while the limit for thumb is 16 MB). For conditional\nbranches, the range limit is only 1 MB though (the same as for thumb),\nand for the tbz instruction, the range is only 32 KB, which allows for\na test much smaller than the full 128 MB.\n\nThis fixes PR40467.\n\nDifferential Revision: https://reviews.llvm.org/D57575\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352929 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "c71a5001b1af9892cbf1b0aacc28153215989e31", "tree": "ac21d896f838b23032451f2eb50ab3652871f039", "parents": [ "538dec71dd88135875a2b6e3179febca2f783065" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Fri Feb 01 22:08:03 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Fri Feb 01 22:08:03 2019 +0000" }, "message": "[COFF] Fix crashes when writing a PDB after adding thunks.\n\nWhen writing a PDB, the OutputSection of all chunks need to be set.\nThe thunks are added directly to OutputSection after the normal\nmachinery that sets it for all other chunks.\n\nThis fixes part of PR40467.\n\nDifferential Revision: https://reviews.llvm.org/D57574\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352928 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "538dec71dd88135875a2b6e3179febca2f783065", "tree": "9d37930799870a0d2daec3feebdc0a461d450ab7", "parents": [ "ef1450b9950b84a6715e366a1134b3d1f6eafde7" ], "author": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Fri Feb 01 17:06:41 2019 +0000" }, "committer": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Fri Feb 01 17:06:41 2019 +0000" }, "message": "Fix some sphinx doc errors.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352887 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "ef1450b9950b84a6715e366a1134b3d1f6eafde7", "tree": "1d6db0cf273ee00de23aa9c5b934e0258a6bd7cf", "parents": [ "bfb45938687c13f7e498b52ada676d97f5858918" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Fri Feb 01 02:29:57 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Fri Feb 01 02:29:57 2019 +0000" }, "message": "[WebAssembly] Support imports from custom module names\n\nFixes: https://bugs.llvm.org/show_bug.cgi?id\u003d37168\n\nThis is only a first pass at supporting these custom import\nmodules. In the long run we most likely want to treat these\nkinds of symbols very differently. For example, it should not\nbe possible to resolve such as symbol at static link type.\n\nDifferential Revision: https://reviews.llvm.org/D45796\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352828 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "bfb45938687c13f7e498b52ada676d97f5858918", "tree": "24ff3bdbee3e46b625873f839962dc8d23665eea", "parents": [ "696f5a142a295804e02760ef9165f35b3c6a9293" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Fri Feb 01 02:25:05 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Fri Feb 01 02:25:05 2019 +0000" }, "message": "[ELF] Support --{,no-}allow-shlib-undefined\n\nSummary:\nIn ld.bfd/gold, --no-allow-shlib-undefined is the default when linking\nan executable. This patch implements a check to error on undefined\nsymbols in a shared object, if all of its DT_NEEDED entries are seen.\n\nOur approach resembles the one used in gold, achieves a good balance to\nbe useful but not too smart (ld.bfd traces all DSOs and emulates the\nbehavior of a dynamic linker to catch more cases).\n\nThe error is issued based on the symbol table, different from undefined\nreference errors issued for relocations. It is most effective when there\nare DSOs that were not linked with -z defs (e.g. when static sanitizers\nruntime is used).\n\ngold has a comment that some system libraries on GNU/Linux may have\nspurious undefined references and thus system libraries should be\nexcluded (https://sourceware.org/bugzilla/show_bug.cgi?id\u003d6811). The\nstory may have changed now but we make --allow-shlib-undefined the\ndefault for now. Its interaction with -shared can be discussed in the\nfuture.\n\nReviewers: ruiu, grimar, pcc, espindola\n\nReviewed By: ruiu\n\nSubscribers: joerg, emaste, arichardson, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57385\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352826 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "696f5a142a295804e02760ef9165f35b3c6a9293", "tree": "1096f86833ddaf664c79c05eeac7e504680beeee", "parents": [ "a8872150a9858acc8b415534a0928c66ce13e71e" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Fri Feb 01 02:24:50 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Fri Feb 01 02:24:50 2019 +0000" }, "message": "Fix names of functions in TargetOptionsCommandFlags.h. NFC.\n\nDifferential Revision: https://reviews.llvm.org/D57555\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352825 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "a8872150a9858acc8b415534a0928c66ce13e71e", "tree": "d767f3c5f5a50f911acc52d142daeb1f2a23628e", "parents": [ "a29f0d266a02379fff68a217d083c3d331503d55" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Thu Jan 31 16:14:33 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Thu Jan 31 16:14:33 2019 +0000" }, "message": "lld-link: Allow mixing \u0027discard\u0027 and \u0027largest\u0027 comdat selections\n\ncl.exe and clang-cl.exe put vftables in a \u0027discard\u0027 comdat when building with\nRTTI disabled (/GR-) but in a \u0027largest\u0027 comdat when building with RTTI enabled.\nTo be able to link /GR- code with /GR code, lld-link needs to accept comdats\nthat have this type of comdat selection conflict.\n\nFor example, static libraries in the Visual Studio standard library are built\nwith /GR, and without this it\u0027s impossible to build client code with /GR- and\nstill link to the standard library.\n\nlink.exe also accepts merging \u0027discard\u0027 with \u0027largest\u0027, and it accepts merging\n\u0027largest\u0027 with any other selection type. lld-link is still a bit stricter since\nit only allows merging \u0027largest\u0027 with \u0027discard\u0027 for symmetry.\n\nDifferential Revision: https://reviews.llvm.org/D57515\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352765 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "a29f0d266a02379fff68a217d083c3d331503d55", "tree": "87901fc17b29d2cb0c16fc020b36a4cb97217717", "parents": [ "39b4b3429909a50430e85e0a4b99f9113b9676a7" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 20:46:18 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 20:46:18 2019 +0000" }, "message": "[LTO] Set CGOptLevel in LTO config.\n\nPreviously we were never setting this which means it was always being\nset to Default (-O2/-Os).\n\nDifferential Revision: https://reviews.llvm.org/D57422\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352667 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "39b4b3429909a50430e85e0a4b99f9113b9676a7", "tree": "f819cc6a4ffbbc2898b8d50c953e3cc6b9b73472", "parents": [ "67c4aed19347c2511f84c8b39ee36146c48e7ad0" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 18:55:15 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 18:55:15 2019 +0000" }, "message": "[WebAssembly] Fix crash with LTO + relocatable + undefined symbols\n\nChange the way we create the symbol table to be closer to how its done\non ELF. Now the output symbol table matches the internal symtab order\nand includes local and undefined symbols.\n\nFixes PR40204\n\nDifferential Revision: https://reviews.llvm.org/D56947\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352645 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "0f50d2239d204e2f2026b9db97f30b5d54753f0c", "tree": "c9ec98b3dc7f78a20ceff3a7deedc7c33eb0e963", "parents": [ "17c673662a801432dc30a4f42463ecbdc7905514" ], "author": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Mon Jan 28 19:29:41 2019 +0000" }, "committer": { "name": "Pirama Arumuga Nainar", "email": "pirama@google.com", "time": "Wed Jan 30 10:44:26 2019 -0800" }, "message": "ELF: Set sh_info on RelaIplt to point to the IgotPlt output section.\n\nPreviously we were setting it to the GotPlt output section, which is\nincorrect on ARM where this section is in .got. In static binaries\nthis can lead to sh_info being set to -1 (because there is no .got.plt)\nwhich results in various tools rejecting the output file.\n\nDifferential Revision: https://reviews.llvm.org/D57274\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352413 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "67c4aed19347c2511f84c8b39ee36146c48e7ad0", "tree": "6890511379ec706f535db358adfc84e0654d514d", "parents": [ "cf47f8dd15d304b822130957d16a5dd058485339" ], "author": { "name": "Dimitry Andric", "email": "dimitry@andric.com", "time": "Wed Jan 30 06:31:52 2019 +0000" }, "committer": { "name": "Dimitry Andric", "email": "dimitry@andric.com", "time": "Wed Jan 30 06:31:52 2019 +0000" }, "message": "Recognize FreeBSD specific BFD names in OUTPUT_FORMAT\n\nSummary:\nAfter rLLD344952 (\"Add OUTPUT_FORMAT linker script directive support\"),\nusing BFD names such as `elf64-x86-64-freebsd` the `OUTPUT_FORMAT`\nlinker script command does not work anymore, resulting in errors like:\n\n```\nld: error: /home/dim/src/clang800-import/stand/efi/loader/arch/amd64/ldscript.amd64:2: unknown output format name: elf64-x86-64-freebsd\n\u003e\u003e\u003e OUTPUT_FORMAT(\"elf64-x86-64-freebsd\", \"elf64-x86-64-freebsd\", \"elf64-x86-64-freebsd\")\n\u003e\u003e\u003e ^\n```\n\nTo fix this, recognize a `-freebsd` suffix in BFD names, and also set\n`Configuration::OSABI` to `ELFOSABI_FREEBSD` for those cases.\n\nAdd and/or update several test cases to check for the correct results of\nthese new `OUTPUT_FORMAT` arguments.\n\nReviewers: ruiu, atanasyan, grimar, hokein, emaste, espindola\n\nReviewed By: ruiu\n\nSubscribers: nemanjai, javed.absar, arichardson, krytarowski, kristof.beyls, kbarton, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57283\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352606 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "cf47f8dd15d304b822130957d16a5dd058485339", "tree": "6bffc513df82e258ae684593b12a274f2e5b8241", "parents": [ "25db38eac5ffec9ff7f24757c7daf43d885e7e35" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 30 02:17:27 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 30 02:17:27 2019 +0000" }, "message": "lld/coff: Implement some support for the comdat selection field\n\nLLD used to handle comdats as if the selection field was always set to\nIMAGE_COMDAT_SELECT_ANY. This means for obj files produced by `cl /Gy`, LLD\nwould never report a duplicate symbol error.\n\nThis change:\n- adds validation for the Selection field (should make no difference in\n practice for compiler-generated obj inputs)\n- rejects comdats that have different Selection fields in different obj files\n (likewise). This is a bit more strict but also more self-consistent thank\n link.exe (see comment in code)\n- implements handling for all the selection kinds\n\nIn practice, compilers only generate comdats with\nIMAGE_COMDAT_SELECT_NODUPLICATES (LLD now produces duplicate symbol errors for\nthese), IMAGE_COMDAT_SELECT_ANY (no behavior change), and\nIMAGE_COMDAT_SELECT_LARGEST (for RTTI data; here LLD should no longer create\nbroken executables when linking some TUs with RTTI enabled and some with it\ndisabled – but see below).\n\nThe implementation of `IMAGE_COMDAT_SELECT_LARGEST` is incomplete: If one\nSELECT_LARGEST comdat replaces an earlier one, the comdat symbol is replaced\ncorrectly, but the old section stays loaded and if /opt:ref is disabled (via\n/opt:noref or /debug) it\u0027s still written to the output. That\u0027s not ideal, but\nbetter than the current treatment of just picking any one of those comdats. I\nhope to fix this better later.\n\nFixes most of PR40094.\n\nDifferential Revision: https://reviews.llvm.org/D57324\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352590 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "25db38eac5ffec9ff7f24757c7daf43d885e7e35", "tree": "7e5e666961ffab524df296f91ba9799f8673ad1d", "parents": [ "5ba7733e4d3ce6b96b5c187df35ea7bd04940774" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 30 02:15:08 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 30 02:15:08 2019 +0000" }, "message": "lld-link: Add test I forgot to add in 332012\n\nDifferential Revision: https://reviews.llvm.org/D46693\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352589 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "5ba7733e4d3ce6b96b5c187df35ea7bd04940774", "tree": "d0472e60e7a21d83a3b25f8f2c19cf6746664da6", "parents": [ "89732389dcb5634b2f746a0808e76e4fc80bbe97" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 00:25:52 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Wed Jan 30 00:25:52 2019 +0000" }, "message": "[WebAssembly] Fix undefined weak function symbols in LTO builds\n\nSummary: Fixes PR40219\n\nSubscribers: dschuff, mehdi_amini, inglorion, jgravelle-google, aheejin, sunfish, steven_wu, dexonsmith, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57420\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352575 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "89732389dcb5634b2f746a0808e76e4fc80bbe97", "tree": "54d012fdee3a029da92df2bbb250b3c80c2bbd7e", "parents": [ "be47ca5e75bb1c7d94b0eca536d7f32f539f8a1e" ], "author": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Tue Jan 29 22:26:31 2019 +0000" }, "committer": { "name": "Sam Clegg", "email": "sbc@chromium.org", "time": "Tue Jan 29 22:26:31 2019 +0000" }, "message": "[WebAssembly] Don\u0027t load weak undefined symbols from archive files\n\nSummary: Fixes PR40494\n\nSubscribers: dschuff, jgravelle-google, aheejin, sunfish, llvm-commits\n\nDifferential Revision: https://reviews.llvm.org/D57370\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352554 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "be47ca5e75bb1c7d94b0eca536d7f32f539f8a1e", "tree": "7eabe48177a66a89936692d699ee79335cdda60a", "parents": [ "f502974fad95025a284a68b315d5becc626355e0" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Jan 29 19:24:32 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Jan 29 19:24:32 2019 +0000" }, "message": "[MinGW] Don\u0027t define names for ignored options. NFC.\n\nMove them to the same section as the newly added ignored options\nwithout a defined name.\n\nAlso move options that actually weren\u0027t ignored to the right section.\n\nDifferential Revision: https://reviews.llvm.org/D57374\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352529 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "f502974fad95025a284a68b315d5becc626355e0", "tree": "b4018ebe842520d5282c41f8f2db0d8babb06ec9", "parents": [ "3c47f28f0950d522a0741d83184bfc1b60e04425" ], "author": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Tue Jan 29 16:37:27 2019 +0000" }, "committer": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Tue Jan 29 16:37:27 2019 +0000" }, "message": "Adjust documentation for git migration.\n\nThis fixes most references to the paths:\n llvm.org/svn/\n llvm.org/git/\n llvm.org/viewvc/\n github.com/llvm-mirror/\n github.com/llvm-project/\n reviews.llvm.org/diffusion/\n\nto instead point to https://github.com/llvm/llvm-project.\n\nThis is *not* a trivial substitution, because additionally, all the\ncheckout instructions had to be migrated to instruct users on how to\nuse the monorepo layout, setting LLVM_ENABLE_PROJECTS instead of\nchecking out various projects into various subdirectories.\n\nI\u0027ve attempted to not change any scripts here, only documentation. The\nscripts will have to be addressed separately.\n\nAdditionally, I\u0027ve deleted one document which appeared to be outdated\nand unneeded:\n lldb/docs/building-with-debug-llvm.txt\n\nDifferential Revision: https://reviews.llvm.org/D57330\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352514 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3c47f28f0950d522a0741d83184bfc1b60e04425", "tree": "156bed79a5c622e930a7f5f73d565c6968be27c3", "parents": [ "3e5fe1acbce010b88c91ba5ef1bd8f5c72c65b71" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Tue Jan 29 15:50:31 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Tue Jan 29 15:50:31 2019 +0000" }, "message": "lld-link: Allow backward references between associated comdats\n\nReferences between associated comdats are invalid per COFF spec, but the newest\nWindows SDK contains obj files that have these references\n(https://bugs.chromium.org/p/chromium/issues/detail?id\u003d925943#c13). So add back\nsupport for them and add tests for them. The old code handled them fine.\n\nThis makes lld-link match the behavior of newer link.exe versions as far as I\ncan tell. (The behavior before this change matched the behavior of older\nlink.exe versions.)\n\nThis mostly reverts r352254.\n\nDifferential Revision: https://reviews.llvm.org/D57387\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352508 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3e5fe1acbce010b88c91ba5ef1bd8f5c72c65b71", "tree": "8784de2baec1c7782f20303c8b945c1581b8d3a6", "parents": [ "72af896280f5bebf7ea9e83ec6f4de82c1c6f399" ], "author": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Tue Jan 29 14:24:35 2019 +0000" }, "committer": { "name": "Fangrui Song", "email": "maskray@google.com", "time": "Tue Jan 29 14:24:35 2019 +0000" }, "message": "[ELF] Simplify. NFC\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352499 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "72af896280f5bebf7ea9e83ec6f4de82c1c6f399", "tree": "2a17818723d915e1a86f085b20fab603ef400a9a", "parents": [ "879b1438776203d10c1b3cc09d4886246c06fee0" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Tue Jan 29 11:46:00 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Tue Jan 29 11:46:00 2019 +0000" }, "message": "[ELF] - Remove dead `readBfdName` declaration. NFC.\n\n`readBfdName` was removed recently.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352482 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "879b1438776203d10c1b3cc09d4886246c06fee0", "tree": "84f2a2d33f9bb61ff8701b3810a23ab80304df3b", "parents": [ "1e8e49a1bf5bc576f7092f8a4facff7915aedcfe" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Jan 29 08:38:48 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Tue Jan 29 08:38:48 2019 +0000" }, "message": "[MinGW] Ignore the --plugin and --plugin-opt option\n\nGCC can use LLD with -fuse-ld\u003dlld for MinGW these days, but by\ndefault these options are passed to the linker (unless -fno-lto\nis passed to the GCC driver).\n\nDifferential Revision: https://reviews.llvm.org/D57304\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352459 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "1e8e49a1bf5bc576f7092f8a4facff7915aedcfe", "tree": "df048dd2e2a2ce6608f6f0a5a85c69e6740547fc", "parents": [ "793abc38e5f243222a27ffa83a26ebd8bb83493d" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Jan 28 21:45:50 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Jan 28 21:45:50 2019 +0000" }, "message": "Attempt to fix build failure with GCC 5.4.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352435 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "793abc38e5f243222a27ffa83a26ebd8bb83493d", "tree": "4b95c5fced8e624f5506514b9b37f456f8dae5f1", "parents": [ "8940eaf0e55c08a54ad18cb70b3af1e024adc682" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Mon Jan 28 21:16:15 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Mon Jan 28 21:16:15 2019 +0000" }, "message": "lld/coff: Make assoc comdat diag a bit more detailed\n\nMany different sections can have the same name, so include the indices of the\nsections mentioned in the diagnostic too.\n\nI\u0027m debugging something I can\u0027t repro locally, maybe this will help.\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352428 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "8940eaf0e55c08a54ad18cb70b3af1e024adc682", "tree": "d6896aa243d6540dcc94ed8f6ebaa93341643ce3", "parents": [ "19fbeafcc0b5a88b84a1f6cd3f42fabf484c7f72" ], "author": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Mon Jan 28 19:29:41 2019 +0000" }, "committer": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Mon Jan 28 19:29:41 2019 +0000" }, "message": "ELF: Set sh_info on RelaIplt to point to the IgotPlt output section.\n\nPreviously we were setting it to the GotPlt output section, which is\nincorrect on ARM where this section is in .got. In static binaries\nthis can lead to sh_info being set to -1 (because there is no .got.plt)\nwhich results in various tools rejecting the output file.\n\nDifferential Revision: https://reviews.llvm.org/D57274\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352413 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "19fbeafcc0b5a88b84a1f6cd3f42fabf484c7f72", "tree": "8d035675ace497cfb39b0350c1e67344ccac7949", "parents": [ "87c504deffecdc32ce4386298e1c414fefd22a36" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Jan 28 19:11:52 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Mon Jan 28 19:11:52 2019 +0000" }, "message": "Refactoring. NFC.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352407 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "87c504deffecdc32ce4386298e1c414fefd22a36", "tree": "7b129703f934658c67cd163fd5422afcf37dfec0", "parents": [ "c97e0df42f0e455fcdb87387bc27b307872e82f6" ], "author": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Mon Jan 28 15:03:47 2019 +0000" }, "committer": { "name": "George Rimar", "email": "grimar@accesssoftek.com", "time": "Mon Jan 28 15:03:47 2019 +0000" }, "message": "[LLD][ELF] - Update tests after r352366.\n\nr352366 \"[llvm-objdump] - Print LMAs when dumping section headers.\" changed the format of\nllvm-objdump output. We have to update the LLD tests.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352372 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "17c673662a801432dc30a4f42463ecbdc7905514", "tree": "088596bbe8be26199238b01182a6fe282810ff2b", "parents": [ "3abb7de662853145bda3da4422ce4c6986b8747a" ], "author": { "name": "Nick Desaulniers", "email": "ndesaulniers@google.com", "time": "Sun Jan 27 02:54:23 2019 +0000" }, "committer": { "name": "Pirama Arumuga Nainar", "email": "pirama@google.com", "time": "Sun Jan 27 22:55:04 2019 -0800" }, "message": "lld: elf: discard more specific .gnu.linkonce section\n\nSummary:\nlld discards .gnu.linonce.* sections work around a bug in glibc.\nhttps://sourceware.org/bugzilla/show_bug.cgi?id\u003d20543\n\nUnfortunately, the Linux kernel uses a section named\n.gnu.linkonce.this_module to store infomation about kernel modules. The\nkernel reads data from this section when loading kernel modules, and\nerrors if it fails to find this section. The current behavior of lld\ndiscards this section when kernel modules are linked, so kernel modules\nlinked with lld are unloadable by the linux kernel.\n\nThe Linux kernel should use a comdat section instead of .gnu.linkonce.\nThe minimum version of binutils supported by the kernel supports comdat\nsections. The kernel is also not relying on the old linkonce behavior;\nit seems to have chosen a name that contains a deprecated GNU feature.\n\nChanging the section name now in the kernel would require all kernel\nmodules to be recompiled to make use of the new section name. Instead,\nrather than discarding .gnu.linkonce.*, let\u0027s discard the more specific\nsection name to continue working around the glibc issue while supporting\nlinking Linux kernel modules.\n\nLink: https://github.com/ClangBuiltLinux/linux/issues/329\n\nReviewers: pcc, espindola\n\nReviewed By: pcc\n\nSubscribers: nathanchance, emaste, arichardson, void, srhines\n\nDifferential Revision: https://reviews.llvm.org/D57294\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352302 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "c97e0df42f0e455fcdb87387bc27b307872e82f6", "tree": "00213c9e8e5c670b49a92e7694ffc526732160ba", "parents": [ "054e77545864521127b22d0f8e73bbc9f8a5d228" ], "author": { "name": "Alexandre Ganea", "email": "alexandre.ganea@ubisoft.com", "time": "Mon Jan 28 01:45:35 2019 +0000" }, "committer": { "name": "Alexandre Ganea", "email": "alexandre.ganea@ubisoft.com", "time": "Mon Jan 28 01:45:35 2019 +0000" }, "message": "[LLD][COFF] Partial sections\n\nPersist (input) sections that make up an OutputSection. This is a supporting patch for the upcoming D54802.\n\nDifferential Revision: https://reviews.llvm.org/D55293\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352336 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "054e77545864521127b22d0f8e73bbc9f8a5d228", "tree": "8b1a3f6a1e3f688524c89650f9b7c903788d4162", "parents": [ "d83cebf5e91e0aa3526d90198728b6516d255364" ], "author": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Sun Jan 27 19:57:50 2019 +0000" }, "committer": { "name": "Martin Storsjo", "email": "martin@martin.st", "time": "Sun Jan 27 19:57:50 2019 +0000" }, "message": "[COFF] Add support for the new relocation IMAGE_REL_ARM{,64}_REL32\n\nDifferential Revision: https://reviews.llvm.org/D57292\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352325 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "d83cebf5e91e0aa3526d90198728b6516d255364", "tree": "7baa463bb6559cb1110984233e7135b5f801795a", "parents": [ "2dee08980c8e3d5d1cac7d3394b207b062da0bf4" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Sun Jan 27 03:56:37 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Sun Jan 27 03:56:37 2019 +0000" }, "message": "Follow-up to r352254: Initialize Selection field.\n\nThe diagnostic there fired spuriosly due to uninitialized memory.\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352304 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "2dee08980c8e3d5d1cac7d3394b207b062da0bf4", "tree": "5b7962ccbc5ecbf93bc0b923fc1b5d4eefcadd77", "parents": [ "35ff9b29ba45c2e7ada6c5a279ab9471057aafd7" ], "author": { "name": "Nick Desaulniers", "email": "ndesaulniers@google.com", "time": "Sun Jan 27 02:54:23 2019 +0000" }, "committer": { "name": "Nick Desaulniers", "email": "ndesaulniers@google.com", "time": "Sun Jan 27 02:54:23 2019 +0000" }, "message": "lld: elf: discard more specific .gnu.linkonce section\n\nSummary:\nlld discards .gnu.linonce.* sections work around a bug in glibc.\nhttps://sourceware.org/bugzilla/show_bug.cgi?id\u003d20543\n\nUnfortunately, the Linux kernel uses a section named\n.gnu.linkonce.this_module to store infomation about kernel modules. The\nkernel reads data from this section when loading kernel modules, and\nerrors if it fails to find this section. The current behavior of lld\ndiscards this section when kernel modules are linked, so kernel modules\nlinked with lld are unloadable by the linux kernel.\n\nThe Linux kernel should use a comdat section instead of .gnu.linkonce.\nThe minimum version of binutils supported by the kernel supports comdat\nsections. The kernel is also not relying on the old linkonce behavior;\nit seems to have chosen a name that contains a deprecated GNU feature.\n\nChanging the section name now in the kernel would require all kernel\nmodules to be recompiled to make use of the new section name. Instead,\nrather than discarding .gnu.linkonce.*, let\u0027s discard the more specific\nsection name to continue working around the glibc issue while supporting\nlinking Linux kernel modules.\n\nLink: https://github.com/ClangBuiltLinux/linux/issues/329\n\nReviewers: pcc, espindola\n\nReviewed By: pcc\n\nSubscribers: nathanchance, emaste, arichardson, void, srhines\n\nDifferential Revision: https://reviews.llvm.org/D57294\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352302 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "35ff9b29ba45c2e7ada6c5a279ab9471057aafd7", "tree": "6b30c7d05fea4756f309bb5839200dc782238cee", "parents": [ "6dc5e0c9af6478fb5894c05960969e0caf9350be" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Sat Jan 26 00:31:49 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Sat Jan 26 00:31:49 2019 +0000" }, "message": "Remove dead declaration.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352257 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "6dc5e0c9af6478fb5894c05960969e0caf9350be", "tree": "9c50c0fdf1770ed7adf53f7b6df549df26bef6eb", "parents": [ "ea65c4966a1c4622210c3341d48fddb2f94f03e8" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Sat Jan 26 00:14:52 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Sat Jan 26 00:14:52 2019 +0000" }, "message": "lld-link: Store comdat selection in SectionChunk, reject more invalid associated comdats\n\nI need the comdat selection for PR40094. To keep the patch for that smaller,\nI\u0027m adding it here, and as a first application I\u0027m using it to reject\nassociative comdats referring to earlier associative comdats. Depends on\nD56929; together with that all associative comdats referring to other\nassociative comdats are now rejected.\n\nDifferential Revision: https://reviews.llvm.org/D56931\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352254 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "ea65c4966a1c4622210c3341d48fddb2f94f03e8", "tree": "bff19f028b7aed54d48de96fb10bbb53291f8255", "parents": [ "f23961fefb2d1041034172affc68b5cf5927d139" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Fri Jan 25 21:25:25 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Fri Jan 25 21:25:25 2019 +0000" }, "message": "Simplify. NFC.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352242 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "f23961fefb2d1041034172affc68b5cf5927d139", "tree": "6df4595d21989ec5266d74448de2dff342214651", "parents": [ "3daf1150a1db840a92bf7622f509e7bf600e05f9" ], "author": { "name": "Zachary Turner", "email": "zturner@google.com", "time": "Thu Jan 24 22:25:55 2019 +0000" }, "committer": { "name": "Zachary Turner", "email": "zturner@google.com", "time": "Thu Jan 24 22:25:55 2019 +0000" }, "message": "[PDB] Increase TPI hash bucket count.\n\nPDBs contain several serialized hash tables. In the microsoft-pdb\nrepo published to support LLVM implementing PDB support, the\nprovided initializes the bucket count for the TPI and IPI streams\nto the maximum size. This occurs in tpi.cpp L33 and tpi.cpp L398.\nIn the LLVM code for generating PDBs, these streams are created with\nminimum number of buckets. This difference makes LLVM generated\nPDBs slower for when used for debugging.\n\nPatch by C.J. Hebert\nDifferential Revision: https://reviews.llvm.org/D56942\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352117 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3daf1150a1db840a92bf7622f509e7bf600e05f9", "tree": "9e0a7c6b693bd72fbd2b6e704cde75e8637b6a3f", "parents": [ "d3a56372131742afe584984c61f69c46b9f53f2c" ], "author": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Thu Jan 24 21:31:43 2019 +0000" }, "committer": { "name": "James Y Knight", "email": "jyknight@google.com", "time": "Thu Jan 24 21:31:43 2019 +0000" }, "message": "Define the _fltused symbol in a couple tests post-r352076.\n\nNormally it\u0027s defined by MSCRT, but these tests are standalone, so\nthey need to define it themselves.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352110 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "d3a56372131742afe584984c61f69c46b9f53f2c", "tree": "fd5559de69eda9ae48b78f13f4f7165d0ab85b54", "parents": [ "2c263611b857bfc71ad77c8031b05995ceabd249" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Jan 24 19:02:31 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Jan 24 19:02:31 2019 +0000" }, "message": "Fix broken export table if .rdata is merged with .text.\n\nPreviously, we assumed that .rdata is zero-filled, so when writing\nan COFF import table, we didn\u0027t write anything if the data is zero.\nThat assumption was wrong because .rdata can be merged with .text.\nIf .rdata is merged with .text, they are initialized with 0xcc which\nis a trap instruction.\n\nThis patch removes that assumption from code.\n\nShould be merged to 8.0 branch as this is a regression.\n\nFixes https://bugs.llvm.org/show_bug.cgi?id\u003d39826\n\nDifferential Revision: https://reviews.llvm.org/D57168\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352082 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "2c263611b857bfc71ad77c8031b05995ceabd249", "tree": "46e8e1038e884d44dd4ea5116077654884b1e253", "parents": [ "3c906adeb08d4becb6682e5db4d90cf56c84e95f" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Thu Jan 24 18:30:33 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Thu Jan 24 18:30:33 2019 +0000" }, "message": "lld/coff: Rename two tests from .s to .test\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352074 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3c906adeb08d4becb6682e5db4d90cf56c84e95f", "tree": "3c7c5f4f3b2fd1746692c57306fb09fa7a6dba67", "parents": [ "af940acfbe1ca38dc8b41df8a63643984c4c3da2" ], "author": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Thu Jan 24 18:17:40 2019 +0000" }, "committer": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Thu Jan 24 18:17:40 2019 +0000" }, "message": "[PPC64] Reland r351978 \u0027Sort .toc sections accessed with small code model ...\u0027\n\nGuessing that the slashes used in the scripts SECTION command was causing the\nwindows related failures in the added test.\n\nOriginal commit message:\nSmall code model global variable access on PPC64 has a very limited range of\naddressing. The instructions the relocations are used on add an offset in the\nrange [-0x8000, 0x7FFC] to the toc pointer which points to .got +0x8000, giving\nan addressable range of [.got, .got + 0xFFFC]. While user code can be recompiled\nwith medium and large code models when the binary grows too large for small code\nmodel, there are small code model relocations in the crt files and libgcc.a\nwhich are typically shipped with the distros, and the ABI dictates that linkers\nmust allow linking of relocatable object files using different code models.\n\nTo minimze the chance of relocation overflow, any file that contains a small\ncode model relocation should have its .toc section placed closer to the .got\nthen any .toc from a file without small code model relocations.\n\nDifferential Revision: https://reviews.llvm.org/D56920\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352071 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "af940acfbe1ca38dc8b41df8a63643984c4c3da2", "tree": "5a51b36c16cc22b8e4cff6f9c88d1953611bcbf1", "parents": [ "67cb2cc89518a0c3e7e90648e394c05945567a25" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Jan 24 18:17:17 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Thu Jan 24 18:17:17 2019 +0000" }, "message": "Untabify.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352070 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "67cb2cc89518a0c3e7e90648e394c05945567a25", "tree": "6781cf0c53e0c53baff54141d57c4c060ab3ac4b", "parents": [ "5500a785ffcc416cee19f4460b877250cde8129a" ], "author": { "name": "Serge Guelton", "email": "sguelton@quarkslab.com", "time": "Thu Jan 24 17:56:08 2019 +0000" }, "committer": { "name": "Serge Guelton", "email": "sguelton@quarkslab.com", "time": "Thu Jan 24 17:56:08 2019 +0000" }, "message": "Partial support of SHT_GROUP without flag\n\nThis does *not* implement full SHT_GROUP semantic, yet it is a simple step forward:\nSections within a group are still considered valid, but they do not behave as\nspecified by the standard in case of garbage collection.\n\nDifferential Revision: https://reviews.llvm.org/D56437\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@352068 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "3abb7de662853145bda3da4422ce4c6986b8747a", "tree": "c7c9c1886e7a05074c39c3b0c07de7e3d3f5e583", "parents": [ "1912d1cb847c73c30433988382cd2723fbce2c3c" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Tue Jan 15 18:30:23 2019 +0000" }, "committer": { "name": "Pirama Arumuga Nainar", "email": "pirama@google.com", "time": "Thu Jan 24 08:53:35 2019 -0800" }, "message": "Do not emit a corrupt symbol table entry for .rela_iplt_{start,end}.\n\nIf .rela.iplt does not exist, we used to emit a corrupt symbol table\nthat contains two symbols, .rela_iplt_{start,end}, pointing to a\nnonexisting section.\n\nThis patch fixes the issue by setting section index 0 to the symbols\nif .rel.iplt section does not exist.\n\nDifferential Revision: https://reviews.llvm.org/D56623\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351218 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "5500a785ffcc416cee19f4460b877250cde8129a", "tree": "c2adb62f069841e71cd236a606563afe1a1eb486", "parents": [ "ddc6decbbb141038de53e8ce219b0eaf2606c976" ], "author": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Wed Jan 23 21:46:28 2019 +0000" }, "committer": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Wed Jan 23 21:46:28 2019 +0000" }, "message": "Revert \"[PPC64] Sort .toc sections accessed with small code model ...\"\n\nThis reverts commit ca87c57a3aa4770c9cf0defd4b2feccbc342ee93.\nAdded test fails on several windows buildbots.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351985 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "ddc6decbbb141038de53e8ce219b0eaf2606c976", "tree": "4dd620f6bfe8c96a8a60047aea51cad98de95a59", "parents": [ "dd25c096870860a544d7752f0fae98c24e564635" ], "author": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Wed Jan 23 21:04:03 2019 +0000" }, "committer": { "name": "Sean Fertile", "email": "sfertile@ca.ibm.com", "time": "Wed Jan 23 21:04:03 2019 +0000" }, "message": "[PPC64] Sort .toc sections accessed with small code model relocs close to .got.\n\nSmall code model global variable access on PPC64 has a very limited range of\naddressing. The instructions the relocations are used on add an offset in the\nrange [-0x8000, 0x7FFC] to the toc pointer which points to .got +0x8000, giving\nan addressable range of [.got, .got + 0xFFFC]. While user code can be recompiled\nwith medium and large code models when the binary grows too large for small code\nmodel, there are small code model relocations in the crt files and libgcc.a\nwhich are typically shipped with the distros, and the ABI dictates that linkers\nmust allow linking of relocatable object files using different code models.\n\nTo minimze the chance of relocation overflow, any file that contains a small\ncode model relocation should have its .toc section placed closer to the .got\nthen any .toc from a file without small code model relocations.\n\nDifferential Revision: https://reviews.llvm.org/D56920\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351978 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "dd25c096870860a544d7752f0fae98c24e564635", "tree": "c2adb62f069841e71cd236a606563afe1a1eb486", "parents": [ "bdcc3eb169658f5ab10b0b6aff3f0e8dff5644a3" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 13:43:42 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 13:43:42 2019 +0000" }, "message": "Disable test better.\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351952 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "bdcc3eb169658f5ab10b0b6aff3f0e8dff5644a3", "tree": "c98b345bef11dae2a518b00e3405b5931d136e8a", "parents": [ "637390b315e1dc41396d1f757c17943fecdef816" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 12:35:08 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 12:35:08 2019 +0000" }, "message": "Disable test added in r351916.\n\nIt doesn\u0027t pass on Windows:\nhttp://lab.llvm.org:8011/builders/clang-x64-windows-msvc/builds/3627\n\nFAIL: lld :: ELF/stdout.s (1521 of 1966)\n******************** TEST \u0027lld :: ELF/stdout.s\u0027 FAILED ********************\nScript:\n--\n: \u0027RUN: at line 3\u0027; C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\llvm-mc.EXE -filetype\u003dobj -triple\u003dx86_64-unknown-linux C:\\b\\slave\\clang-x64-windows-msvc\\build\\llvm.src\\tools\\lld\\test\\ELF\\stdout.s -o C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o\n: \u0027RUN: at line 4\u0027; c:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\ld.lld.EXE C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o -o - \u003e C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1\n: \u0027RUN: at line 5\u0027; C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\llvm-objdump.EXE -d C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1 | C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\FileCheck.EXE C:\\b\\slave\\clang-x64-windows-msvc\\build\\llvm.src\\tools\\lld\\test\\ELF\\stdout.s\n: \u0027RUN: at line 10\u0027; c:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\ld.lld.EXE C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o -o C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp2\n: \u0027RUN: at line 11\u0027; diff C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1 C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp2\n--\nExit Code: 1\n\nCommand Output (stdout):\n--\n$ \":\" \"RUN: at line 3\"\n$ \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\llvm-mc.EXE\" \"-filetype\u003dobj\" \"-triple\u003dx86_64-unknown-linux\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\llvm.src\\tools\\lld\\test\\ELF\\stdout.s\" \"-o\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o\"\n$ \":\" \"RUN: at line 4\"\n$ \"c:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\ld.lld.EXE\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o\" \"-o\" \"-\"\n$ \":\" \"RUN: at line 5\"\n$ \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\llvm-objdump.EXE\" \"-d\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1\"\n$ \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\FileCheck.EXE\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\llvm.src\\tools\\lld\\test\\ELF\\stdout.s\"\n$ \":\" \"RUN: at line 10\"\n$ \"c:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\bin\\ld.lld.EXE\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp.o\" \"-o\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp2\"\n$ \":\" \"RUN: at line 11\"\n$ \"diff\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1\" \"C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp2\"\n# command output:\n*** C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp1\n--- C:\\b\\slave\\clang-x64-windows-msvc\\build\\build\\stage1\\tools\\lld\\test\\ELF\\Output\\stdout.s.tmp2\n***************\n*** 1 ****\n\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351949 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "637390b315e1dc41396d1f757c17943fecdef816", "tree": "e66f9d222d7f6963cf5532755c4621fce4694be3", "parents": [ "f6abb4b500f4f30f8ced624c7ff53e192f4c38e4" ], "author": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 02:07:10 2019 +0000" }, "committer": { "name": "Nico Weber", "email": "nicolasweber@gmx.de", "time": "Wed Jan 23 02:07:10 2019 +0000" }, "message": "lld-link: Use just one code path to process associative comdats, reject some invalid associated comdats\n\nCurrently, if an associative comdat appears after the comdat it\u0027s associated\nwith it\u0027s processed immediately, else it\u0027s deferred until the end of the object\nfile. I found this confusing to think about while working on PR40094, so this\nmakes it so that associated comdats are always processed at the end of the\nobject file. This seems to be perf-neutral and simpler.\n\nNow there\u0027s a natural place to reject the associated comdats referring to later\nassociated comdats (associated comdats referring to associated comdats is\ninvalid per COFF spec) that, so reject those. (A later patch will reject\nassociated comdats referring to earlier comdats.)\n\nDifferential Revision: https://reviews.llvm.org/D56929\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351917 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "f6abb4b500f4f30f8ced624c7ff53e192f4c38e4", "tree": "79fd4e13eed90553b1d4fe3dd8ce1dbdd1bbcdbb", "parents": [ "e67de3fa5ebbf23fd0685e1c04e3e424f1d50986" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Jan 23 02:03:26 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Wed Jan 23 02:03:26 2019 +0000" }, "message": "MemoryBlock: Do not automatically extend a given size to a multiple of page size.\n\nPreviously, MemoryBlock automatically extends a requested buffer size to a\nmultiple of page size because (I believe) doing it was thought to be harmless\nand with that you could get more memory (on average 2KiB on 4KiB-page systems)\n\"for free\".\n\nThat programming interface turned out to be error-prone. If you request N\nbytes, you usually expect that a resulting object returns N for `size()`.\nThat\u0027s not the case for MemoryBlock.\n\nLooks like there is only one place where we take the advantage of\nallocating more memory than the requested size. So, with this patch, I\nsimply removed the automatic size expansion feature from MemoryBlock\nand do it on the caller side when needed. MemoryBlock now always\nreturns a buffer whose size is equal to the requested size.\n\nDifferential Revision: https://reviews.llvm.org/D56941\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351916 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "e67de3fa5ebbf23fd0685e1c04e3e424f1d50986", "tree": "beaea1f91692810f587d3efc98914b318c739131", "parents": [ "f2a1028469505507e44e9a8b67aa64b378a81572" ], "author": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Tue Jan 22 23:54:49 2019 +0000" }, "committer": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Tue Jan 22 23:54:49 2019 +0000" }, "message": "COFF, ELF: ICF: Perform 2 rounds of relocation hash propagation.\n\nLLD\u0027s performance on PGO instrumented Windows binaries was still not\ngreat even with the fix in D56955; out of the 2m41s linker runtime,\naround 2 minutes were still being spent in ICF. I looked into this more\nclosely and discovered that the vast majority of the runtime was being\nspent segregating .pdata sections with the following relocation chain:\n\n.pdata -\u003e identical .text -\u003e unique PGO counter (not eligible for ICF)\n\nThis patch causes us to perform 2 rounds of relocation hash\npropagation, which allows the hash for the .pdata sections to\nincorporate the identifier from the PGO counter. With that, the amount\nof time spent in ICF was reduced to about 2 seconds. I also found that\nthe same change led to a significant ICF performance improvement in a\nregular release build of Chromium\u0027s chrome_child.dll, where ICF time\nwas reduced from around 1s to around 700ms.\n\nWith the same change applied to the ELF linker, median of 100 runs\nfor lld-speed-test/chrome reduced from 4.53s to 4.45s on my machine.\n\nI also experimented with increasing the number of propagation rounds\nfurther, but I did not observe any further significant performance\nimprovements linking Chromium or Firefox.\n\nDifferential Revision: https://reviews.llvm.org/D56986\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351899 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "f2a1028469505507e44e9a8b67aa64b378a81572", "tree": "05d3e563d0de02ac6582e5c9ab6af0094ab7edc9", "parents": [ "476c47495259e709b3ce0380e54c7e3b2031790c" ], "author": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Tue Jan 22 23:51:35 2019 +0000" }, "committer": { "name": "Peter Collingbourne", "email": "peter@pcc.me.uk", "time": "Tue Jan 22 23:51:35 2019 +0000" }, "message": "COFF, ELF: Adjust ICF hash computation to account for self relocations.\n\nIt turns out that sections in PGO instrumented object files on Windows\ncontain a large number of relocations pointing to themselves. With\nr347429 this can cause many sections to receive the same hash (usually\nzero) as a result of a section\u0027s hash being xor\u0027ed with itself.\n\nThis patch causes the COFF and ELF linkers to avoid this problem\nby adding the hash of the relocated section instead of xor\u0027ing it.\nOn my machine this causes the regressing test case\nprovided by Mozilla to terminate in 2m41s.\n\nDifferential Revision: https://reviews.llvm.org/D56955\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351898 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "476c47495259e709b3ce0380e54c7e3b2031790c", "tree": "f144d213b38b126926bef387519628a0ab150e05", "parents": [ "e3a5e2f228b45f46331d35b2706e2efdfcedd89a" ], "author": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Tue Jan 22 18:44:04 2019 +0000" }, "committer": { "name": "Rui Ueyama", "email": "ruiu@google.com", "time": "Tue Jan 22 18:44:04 2019 +0000" }, "message": "FileOutputBuffer: Handle \"-\" as stdout.\n\nI was honestly a bit surprised that we didn\u0027t do this before. This\npatch is to handle \"-\" as the stdout so that if you pass `-o -` to\nlld, for example, it writes an output to stdout instead of file `-`.\n\nI thought that we might want to handle this at a higher level than\nFileOutputBuffer, because if we land this patch, we can no longer\ncreate a file whose name is `-` (there\u0027s a workaround though; you can\npass `./-` instead of `-`). However, because raw_fd_ostream already\nhandles `-` as a special file name, I think it\u0027s okay and actually\nconsistent to handle `-` as a special name in FileOutputBuffer.\n\nDifferential Revision: https://reviews.llvm.org/D56940\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351852 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "e3a5e2f228b45f46331d35b2706e2efdfcedd89a", "tree": "466728941c39cdc64fd25c2f80c1f7851af16c19", "parents": [ "e4443d35d929f53a03dcfb72283c53c5fe549d23" ], "author": { "name": "James Henderson", "email": "jh7370@my.bristol.ac.uk", "time": "Tue Jan 22 09:35:47 2019 +0000" }, "committer": { "name": "James Henderson", "email": "jh7370@my.bristol.ac.uk", "time": "Tue Jan 22 09:35:47 2019 +0000" }, "message": "[LLD][ELF]Fix tests for D56910\n\nr351789 changes the output of llvm-readelf --dyn-symbols. This causes 3\nLLD tests to break. This patch fixes them.\n\nReviewed by: ruiu\n\nDifferential Revision: https://reviews.llvm.org/D56911\n\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351790 91177308-0d34-0410-b5e6-96231b3b80d8\n" }, { "commit": "e4443d35d929f53a03dcfb72283c53c5fe549d23", "tree": "5c5be326ebbbd9e0d7a58cd67a902e3e58991cf4", "parents": [ "62e03b7d6e0ae2cb49fdc7b6d927bfe90ce3bc97" ], "author": { "name": "Chandler Carruth", "email": "chandlerc@gmail.com", "time": "Mon Jan 21 09:52:34 2019 +0000" }, "committer": { "name": "Chandler Carruth", "email": "chandlerc@gmail.com", "time": "Mon Jan 21 09:52:34 2019 +0000" }, "message": "Fix typos throughout the license files that somehow I and my reviewers\nall missed!\n\nThanks to Alex Bradbury for pointing this out, and the fact that I never\nadded the intended `legacy` anchor to the developer policy. Add that\nanchor too. With hope, this will cause the links to all resolve\nsuccessfully.\n\ngit-svn-id: https://llvm.org/svn/llvm-project/lld/trunk@351731 91177308-0d34-0410-b5e6-96231b3b80d8\n" } ], "next": "62e03b7d6e0ae2cb49fdc7b6d927bfe90ce3bc97" }