)]}' { "log": [ { "commit": "3d54e8335fc7f73734df572ba0dada293f525014", "tree": "3a5d0c71a3363879bffd08aea7ef2dda90d60281", "parents": [ "9b7b288ae2682af9ea39df437b2916dbdbc8ddf5" ], "author": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Fri Jun 05 18:51:26 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 09:55:57 2026 -0700" }, "message": "Revert \"[clang] Reland: fix getTemplateInstantiationArgs\" (#201864)\n\nReverts llvm/llvm-project#201373\n\nThis caused compilation errors. See comment on the original PR.\n\nGitOrigin-RevId: e668f641467892c75ea724ec21c30439a463a772\n" }, { "commit": "9b7b288ae2682af9ea39df437b2916dbdbc8ddf5", "tree": "4e3dec180bf8577ea0d0647ed693507312e2db2d", "parents": [ "823c951e91a511cd465c6bd39e83077455a73acd" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri Jun 05 17:10:25 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 09:16:16 2026 -0700" }, "message": "[NFC][lldb][Windows] extract IOHandlerProcessSTDIOWindows (#201651)\n\nCo-authored-by: Nerixyz \u003cnero.9@hotmail.de\u003e\nGitOrigin-RevId: c4f2f2535438cbdf23d9f84689ea0f518ffa4a95\n" }, { "commit": "823c951e91a511cd465c6bd39e83077455a73acd", "tree": "da29ee8e5deec81c68b36720aabfbed36f129da2", "parents": [ "0edd7d7083548af2c6e626e4e84bf6e7d22599c5" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri Jun 05 16:51:15 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 08:56:21 2026 -0700" }, "message": "[lldb][Windows] enable TestGdbRemote_qThreadStopInfo (#201868)\n\nGitOrigin-RevId: 86d0c2b10427110892de794f73c0b0db476a9211\n" }, { "commit": "0edd7d7083548af2c6e626e4e84bf6e7d22599c5", "tree": "bea28bc9ecc54a1a29ac620d8f85506396aba06d", "parents": [ "227159ed5288d75e439397c17355fc59e13db3c7" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri Jun 05 16:02:50 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 08:05:37 2026 -0700" }, "message": "[lldb][Windows] Clear stale thread stop info on resume (#201595)\n\nGitOrigin-RevId: acbd01a55d0ff71ec37654878a94111624c298dc\n" }, { "commit": "227159ed5288d75e439397c17355fc59e13db3c7", "tree": "7f8708d301f936eb53d28c4ee583ac2b3be7048f", "parents": [ "a510da87437b9c2f97780bef3ddd7853d4c9cb7f" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri Jun 05 15:30:04 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 07:35:31 2026 -0700" }, "message": "[lldb][Windows] Forward debuggee STDOUT through lldb-server via ConPTY (#201124)\n\n`lldb-server.exe` currently does not forward the debuggee\u0027s STDIO to the\nclient.\n\nThis patch wires STDOUT using a ConPTY, mirroring the existing\n`ProcessWindows` path. The two implementations share the same\ninfrastructure (PseudoConsole, ConnectionConPTY, ThreadedCommunication):\n\n- The ConPTY is setup in\n`GDBRemoteCommunicationServerLLGS::LaunchProcess`.\n- `NativeProcessWindows` owns the ConPTY STDOUT read thread. The\nread-thread callback forwards each chunk into a new\n`NativeProcessProtocol::NewProcessOutput` delegate hook.\n- `GDBRemoteCommunicationServerLLGS::NewProcessOutput` copies the data\nand posts `SendONotification` to the main loop.\n- The ConPTY is closed in `OnExitProcess`.\n\nThis patch uses a default terminal size to create the ConPTY, which\ncauses issues. This will be fixed in\nhttps://github.com/llvm/llvm-project/pull/201141.\n\nThe STDIN path will be done in a follow up PR.\n\nrdar://178725650\n\nGitOrigin-RevId: 3734a928541890a8f2b5d91587ff0bd6ebc90b46\n" }, { "commit": "a510da87437b9c2f97780bef3ddd7853d4c9cb7f", "tree": "dc5a0877f21dd45eb61b6e034e743aa6f22a435b", "parents": [ "5ffd350da2baa5aeb1f2a16c1731a4d307b4bc79" ], "author": { "name": "Nikolas Klauser", "email": "nikolasklauser@berlin.de", "time": "Fri Jun 05 13:11:27 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 04:15:25 2026 -0700" }, "message": "[lldb] Add missing includes to some tests (#201804)\n\n#195509 removed a bunch of transitive includes from libc++, causing the\ntest to fail.\n\nGitOrigin-RevId: 04cedfd597561be75d46797978f8a744a2715105\n" }, { "commit": "5ffd350da2baa5aeb1f2a16c1731a4d307b4bc79", "tree": "d7b7ac94b714e191a08241ebce22ddaa9f5f6a4e", "parents": [ "bde1485646d33ae8ab38716cfb7efe11aed39ac6" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri Jun 05 11:37:46 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 03:40:47 2026 -0700" }, "message": "[lldb][Windows] skip TestDAP_attachCommands.py and TestDAP_launch_extra_launch_commands.py (#201796)\n\nGitOrigin-RevId: 8e217eda17df80d2035d6ff27e74bfa8df77339a\n" }, { "commit": "bde1485646d33ae8ab38716cfb7efe11aed39ac6", "tree": "778822b2b1dad0366d2ae06662139d6b024338fe", "parents": [ "a08042d9db68aef75d923bb24d6d5c3ba257c359" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Fri Jun 05 09:58:40 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 01:01:15 2026 -0700" }, "message": "[lldb][NFC] Don\u0027t use C++20 designated initializer (#201075)\n\nThis source triggers the `-Wc++20-designator` warning as we\u0027re still\nusing C++ 17.\n\nGitOrigin-RevId: ec8c8183af2087e9b6e543b3331af57db0f5a9e4\n" }, { "commit": "a08042d9db68aef75d923bb24d6d5c3ba257c359", "tree": "3b7ac3030b2af025443dc1ac3f45ec2578859a39", "parents": [ "7a173df0303b06316a1296ad5ed5d8ccd71710b5" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Fri Jun 05 09:58:21 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 01:01:05 2026 -0700" }, "message": "[lldb][test] Increase polling in TestInterruptThreadNames.py (#201554)\n\nThis test runs for a very long time on my machine (11s per variation),\nand nearly all of this time is spent on the 10s sleep in this function.\n\nThere are two issues here:\n\n1. It uses the (now outdated) logic that arm64 means we have a remote\nDarwin device. This is no longer true these days as Macs also run on\narm64.\n\n2. The polling duration of 1s is still very long, and the test will\nstill spend all its time just waiting for this 1s sleep. A 100ms sleep\nthat we poll in a loop should be slow enough.\n\nGitOrigin-RevId: fdfd1c1344187d64b63504ea8e3662ae4936503a\n" }, { "commit": "7a173df0303b06316a1296ad5ed5d8ccd71710b5", "tree": "2e8554fd7a7c55710ec54ba80d83932ee72a480b", "parents": [ "d187143f17e03367b8373c1b7d1fe3af4cddc776" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Fri Jun 05 09:58:05 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 01:00:57 2026 -0700" }, "message": "[lldb][test] Assume clang supports -gmodules (#201333)\n\nWe currently spend 50ms in most dotest invocations to check if clang\nsupports `-gmodules`. The expensive part of this check is creating the\nclang process to run `clang --help`.\n\n`-gmodules` was added 11 years ago and is present in any compiler that\nhas even a remote chance in supporting the rest of our test suite. This\npatch just assumes that our compiler supports -gmodules if it is clang.\n\nGitOrigin-RevId: baccad7b70143154bf978b00d3a8bed9a03db84a\n" }, { "commit": "d187143f17e03367b8373c1b7d1fe3af4cddc776", "tree": "b293dec7dfa5b1c61f915f03733e496b06edcb49", "parents": [ "a1ec9dc6f1bb7d2c756f9e3d8d44a632415017b3" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Fri Jun 05 09:57:48 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 01:00:48 2026 -0700" }, "message": "[lldb][test] Increase polling frequency in ProcessAttach (#201532)\n\nThe test_attach_to_process_by_id_correct_executable_offset subtest\nrequires us to hit a breakpoint in an attached process. For this we\nimplement a loop that hits the breakpoint location every 2 seconds.\n\nThis patch increases the rate at which we hit this breakpoint to 50ms.\nThe reason is that a 2s interval means that this test is waiting on any\nfast system for nearly 2 seconds on the first breakpoint hit. With a\n50ms interval this subtest passed immediately.\n\nGitOrigin-RevId: 7b9435b0ed1db10c81167a5f415ae14b4b3e0565\n" }, { "commit": "a1ec9dc6f1bb7d2c756f9e3d8d44a632415017b3", "tree": "d486049d79c538e5d6ca84efe6a3b2814646984b", "parents": [ "17052c58e9628398f65c9f8e3f36946167b590ea" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Fri Jun 05 09:56:54 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 01:00:37 2026 -0700" }, "message": "[lldb][test] Make TestInterruptThreadNames not depend on debug info (#201553)\n\nThis test only reads the pthread names, which don\u0027t depend on any debug\ninfo.\n\nThis halves the runtime of this very long test from 22s to 11s.\n\nGitOrigin-RevId: 59bdd5b63b1a649d5bfc7d422c2b4173c582d114\n" }, { "commit": "17052c58e9628398f65c9f8e3f36946167b590ea", "tree": "abfac5d293b59a9b29a5fea5ba644b9c7447735d", "parents": [ "717cdc21af9d801c10c60a7e948a2008805f23f7" ], "author": { "name": "Nikolas Klauser", "email": "nikolasklauser@berlin.de", "time": "Fri Jun 05 09:45:49 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri Jun 05 00:51:11 2026 -0700" }, "message": "[libc++] Drop transitive includes by default (#195509)\n\nThis patch removes the unused transitive includes by default.\n`_LIBCPP_KEEP_TRANSITIVE_INCLUDES_LLVM23` can be defined to keep the\ntransitive includes around for an easier transition. The macro will be\nremoved in LLVM 24.\n\nThis patch implements\nhttps://discourse.llvm.org/t/rfc-remove-unused-transitive-includes-from-the-libc-headers/90157\n\nGitOrigin-RevId: ce5b2e876494cb95f02d9f915081e2b8781e74d1\n" }, { "commit": "717cdc21af9d801c10c60a7e948a2008805f23f7", "tree": "d61c30e7dad9829c19410ee6a5ca098d3682190e", "parents": [ "ff9cc69c3fdadb2d4ed2a16490dc00de0e404ab1" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Thu Jun 04 13:58:03 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 14:01:19 2026 -0700" }, "message": "[lldb][docs] Convert top-level RST docs to Markdown (NFC) (#201674)\n\nConvert the two remaining top-level RST docs, index and\npython_api_enums, to MyST Markdown. This is the final batch of an\nincremental RST -\u003e Markdown migration.\n\nAfter this change, the only RST sources left under lldb/docs/ are\nman/lldb.rst and man/lldb-server.rst, which conf.py intentionally keeps\nas RST so the man-page builder can run without myst_parser installed\n(this reduces dependencies for some llvm distributions).\n\nVerified by building the docs on origin/main and on this branch with\nidentical sphinx flags and diffing both the warnings and the rendered\nHTML. After file extension and line numbers are normalized, the warning\nsets match exactly. index.html is byte-identical; python_api_enums.html\ndiffers in a single line where CommonMark collapses two spaces after a\nperiod to one.\n\nThe diff also surfaced two semantic regressions in the conversion, fixed\nhere:\n\n- index.md and python_api_enums.md lost cross-reference behavior on\nsingle-backtick refs to `lldb` and `SBHostOS.GetLLDBPath`. RST\u0027s default\nrole is `any`, so single backticks attempted xrefs; in MyST single\nbackticks are plain code spans. Converted to explicit `{any}`...``\nsyntax.\n\nContext: https://discourse.llvm.org/t/rfc-make-myst-markdown-the-llvm-docs-format-rip-rest/\n\nAssisted-by: Claude\nGitOrigin-RevId: 7e2c7050a7b59d324f26a10da73edff5aeedd3ba\n" }, { "commit": "ff9cc69c3fdadb2d4ed2a16490dc00de0e404ab1", "tree": "b831c25fed4a0ae415273dc935d674745846a3e9", "parents": [ "b8640e26a8d73f6e2d8172ddd38535183ca8b9a4" ], "author": { "name": "Matheus Izvekov", "email": "mizvekov@gmail.com", "time": "Thu Jun 04 15:29:45 2026 -0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 14:01:07 2026 -0700" }, "message": "[clang] Reland: fix getTemplateInstantiationArgs (#201373)\n\nRelands #199528\n\nThis implements a new strategy for collecting the template arguments, by\nrelying on the qualifiers and template parameter lists to navigate the\ntemplate\ncontext of out-of-line definitions.\n\nThis greatly simplifies the signature of that function, by removing a\nbunch\nof workarounds, and simpliffying a couple that weren\u0027t removed yet.\n\nSince this now relies on qualifiers and template parameter lists,\nthis patch expends most of its effort making sure these are placed,\ntransformed and propagated to template instantiations.\n\nAlso makes the explicit specialization AST nodes stop abusing the\ntemplate\nparameter lists by storing it\u0027s own template parameter list, creating a\ndedicated field for them, similar to partial specializations.\n\nGitOrigin-RevId: bcdb732df219fc5de7c16b9712d52aa56b22fed0\n" }, { "commit": "b8640e26a8d73f6e2d8172ddd38535183ca8b9a4", "tree": "98b212e7d526e2d22c6275e38372116625c3ba75", "parents": [ "48af56ad355606bd94665643fba51fc70781a1fa" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Thu Jun 04 10:26:41 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 10:30:37 2026 -0700" }, "message": "[lldb][docs] Convert use/ RST docs to Markdown (NFC) (#201467)\n\nConvert the ten user-facing RST docs under lldb/docs/use/ to MyST\nMarkdown. This is the third batch of an incremental RST -\u003e Markdown\nmigration; PR1 covered the small leaf pages and PR2 covered the\ncontributor-facing docs under resources/.\n\nFiles: formatting, intel_pt, map, remote, symbolfilejson, symbolication,\nsymbols, troubleshooting, tutorial, variable.\n\nVerified by building the docs on origin/main and on this branch with\nidentical sphinx flags and diffing both the warnings and the rendered\nHTML. After file extension and line numbers are normalized, the warning\nsets match exactly. Seven of the ten pages are byte-identical. The three\nthat differ (symbolication, tutorial, variable) differ only in\nCommonMark collapsing two-spaces-after-period to one and MyST renaming\nauto-numbered footnote IDs (`id6` -\u003e `footnote-1`) plus adding an `\u003chr\u003e`\nseparator before footnote sections.\n\nThe diff also surfaced three semantic regressions in the conversion,\nfixed here:\n\n- variable.md lost cross-reference behavior on single-backtick refs to\n`SBValue` and `SBData`. RST\u0027s default role is `any`, so single backticks\nattempted xrefs; in MyST single backticks are plain code spans.\nConverted these occurrences to explicit `{any}`...``.\n- map.md emitted bare `[Section Name]` for the page TOC, which\nCommonMark treats as an undefined reference shortcut and falls through\nto literal text. Converted to `[Section Name](#slug)`.\n- variable.md emitted `[format name][format name]` as a similar\nundefined reference shortcut. Converted to `[format name](#format-\nname)` to match the new `(format-name)\u003d` anchor.\n\nContext:\nhttps://discourse.llvm.org/t/rfc-make-myst-markdown-the-llvm-docs-format-rip-rest/\n\nAssisted-by: Claude\nGitOrigin-RevId: 52a43992cd92b98f5059851fe43a33ba94527bd4\n" }, { "commit": "48af56ad355606bd94665643fba51fc70781a1fa", "tree": "2733d28463d782653197963da9733453276c4ad9", "parents": [ "8462a6322bbe35762a18c8517cdb3de6a5d1fa14" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Thu Jun 04 10:10:43 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 10:16:00 2026 -0700" }, "message": "[lldb] xfail tests for arm64e caused by compiler bugs (#201454)\n\nThese tests are caused by bugs in clang where arm64e support is not yet\ncomplete.\n\nGitOrigin-RevId: ff25d3158175c089b39cc98c411679ec8f7be4b6\n" }, { "commit": "8462a6322bbe35762a18c8517cdb3de6a5d1fa14", "tree": "19599ec830dc8e3134b0e0f5a455329b55d92e09", "parents": [ "bf34c6ca9687337f329474e7019fd2d8abf108a2" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Thu Jun 04 09:59:18 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 10:01:03 2026 -0700" }, "message": "[lldb] Build liblldb exports from script-interpreter imports (#201392)\n\nWith LLDB_ENABLE_DYNAMIC_SCRIPTINTERPRETERS, the script interpreter\nplugins are built as standalone shared libraries that resolve their\nlldb_private/llvm references through liblldb\u0027s re-exports. liblldb was\nfalling back to liblldb-private.exports, which allowlists the\nlldb_private/llvm namespaces wholesale.\n\nGenerate the export list per build instead. For each plugin, run llvm-nm\non its compiled objects, keep undefined references in namespaces liblldb\nowns, subtract the plugin\u0027s own definitions, and concatenate the\nper-plugin lists with liblldb.exports.\n\nThe plugin\u0027s SHARED target LINK_LIBS liblldb, which makes the plugin\u0027s\ncompile order depend on liblldb. Wiring the exports file in via\nadd_llvm_symbol_exports\u0027 built-in target dependency would form a cycle\nthrough that order-only edge. Add NO_TARGET_DEPENDENCY to\nadd_llvm_symbol_exports so we can apply the file via LINK_DEPENDS alone;\nthat still triggers a relink when the symbol set changes, without\nparticipating in compile-order tracking.\n\nWindows is unchanged: msvc_extract_private_symbols.py was already doing\nper-symbol extraction at namespace granularity.\n\nAssisted-by: Claude\nGitOrigin-RevId: 119e57630281e246d858717f8e331772b01f2908\n" }, { "commit": "bf34c6ca9687337f329474e7019fd2d8abf108a2", "tree": "567dfe28083aa90f2347f9af61bcb2ce53000c32", "parents": [ "cb54b8eebc208770980539b55fa0f6a6b7f8a73e" ], "author": { "name": "Ayush Sahay", "email": "asahay@qti.qualcomm.com", "time": "Thu Jun 04 21:55:39 2026 +0530" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 09:30:24 2026 -0700" }, "message": "[lldb][RISCV] Handle subsets of CSRs in RV32 core dump images (#142932)\n\nThe standard RISC-V ISA sets aside a 12-bit encoding space for up to\n4,096 CSRs. However, many of these may remain unutilized and needn\u0027t be\nsaved in core dump images. To address this, we\u0027ve come up with a new\nnote, _NT_CSREGMAP_, that saves subsets of CSRs as key-value pairs. This\nchange provisions support for handling the subsets of CSRs saved in\n32-bit RISC-V core dump images by building the register information for\nGPRs, FPRs, and CSRs dynamically.\n\nKindly refer to the corresponding topic ([Add RISC-V CSRs to core\ndumps](https://discourse.llvm.org/t/add-risc-v-csrs-to-core-dumps/84348))\nfor additional details.\n\nGitOrigin-RevId: a7a53bfbcff401dbc9df5544b7662a7f96ce7336\n" }, { "commit": "cb54b8eebc208770980539b55fa0f6a6b7f8a73e", "tree": "7642304e2a5292d9b6b3396a8fb3ae72a34c7cc4", "parents": [ "5ec1222a0488ff10ad9047f283fe29a128132d1a" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Thu Jun 04 15:02:10 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 07:06:09 2026 -0700" }, "message": "[lldb][Windows] Suspend new threads when stopped (#201558)\n\n`OnCreateThread` runs from the `DebuggerThread` loop after a\n`CREATE_THREAD_DEBUG_EVENT`. Each iteration of that loop ends with a\n`ContinueDebugEvent`, which on Windows resumes every thread in the\ndebuggee that *isn\u0027t* individually suspended with `SuspendThread`.\n\nIf a thread is created while the debuggee is stopped, all the existing\nthreads are suspended expect the new one. After the next\nContinueDebugEvent it just runs, while lldb\u0027s StateType still reads\neStateStopped.\n\nThis patch suspends the new thread when the debuggee is stopped.\n\nThis fixes `TestTwoHitsOneActual.py` and `TestBreakOnLambdaCapture.py`\nwhen running the test suite with `LLDB_USE_LLDB_SERVER\u003d1`.\n\nrdar://178718627\n\nGitOrigin-RevId: 78d4eac701720c252ed1c835d1f635a223bf70ce\n" }, { "commit": "5ec1222a0488ff10ad9047f283fe29a128132d1a", "tree": "65e43ffabb1ac9a021b1c74b4c29f7ff7e251605", "parents": [ "3788239eacc2fe9544e34bceee8da4dd720a5abb" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Thu Jun 04 15:01:56 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 07:05:59 2026 -0700" }, "message": "[lldb-dap][Windows] skip TestDAP_restart_console (#201587)\n\n`TestDAP_restart_console` is already failing on Windows. It reliably\ncrashes (UNRESOLVED) on some Windows version, including inside Docker\ncontainers.\n\nThis is preventing us from enabling pre-merge CI testing for lldb on\nWindows in https://github.com/llvm/llvm-project/pull/198906.\n\nThis patch skips the test entirely. See\nhttps://github.com/llvm/llvm-project/issues/200840 for more details.\n\nGitOrigin-RevId: 76ddd457c4244ea1a2cd5d3f69b5698d341a0f5b\n" }, { "commit": "3788239eacc2fe9544e34bceee8da4dd720a5abb", "tree": "928d01a61e59da56e92e82022b54f06710b56901", "parents": [ "00ae81178810a11583effc5340e5a2079522558e" ], "author": { "name": "Felipe de Azevedo Piovezan", "email": "fpiovezan@apple.com", "time": "Thu Jun 04 09:23:32 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 01:28:32 2026 -0700" }, "message": "[lldb] Use batched memory reads in ClassDescriptorV2::relative_list_entry_t (#201284)\n\nThis reduces the number of memory reads performed when reading Objective\nC classes metadata.\nNote: these addresses are indeed sequential (with a small offset between\nthem), but there are so many of them that they would not fit into a\nsingle Process::ReadMemory cache line, so this is still a win, and it\nalso puts the code into the right shape for vectorizing the next read in\nthe same loop, which will see the biggest savings.\nGitOrigin-RevId: b12cd6aa2faac1af50ee332c30a3e863dc462d7d\n" }, { "commit": "00ae81178810a11583effc5340e5a2079522558e", "tree": "a291ceefabbd1dece5fe60b0c1e88af35b38a0d7", "parents": [ "47f55bd312dc06208c2a7362333d5b09f4fcaf61" ], "author": { "name": "Felipe de Azevedo Piovezan", "email": "fpiovezan@apple.com", "time": "Thu Jun 04 08:17:14 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Thu Jun 04 00:20:50 2026 -0700" }, "message": "[lldb] Use MemoryCache in Process::ReadRangesFromMemory (#201166)\n\nThere are scenarios (especially in the ObjectiveC metadata reading) in\nwhich multiple strings are read over and over again, but through\ndifferent code paths. In order to make that part of the code use\nMultiMemRead effectively, the memory cache must be integrated into\nReadRangesFromMemory before we can migrate the string reading to\nvectorized version.\n\nGitOrigin-RevId: 96c0f5ad27abba13bc941df99686fdb8c7b59fff\n" }, { "commit": "47f55bd312dc06208c2a7362333d5b09f4fcaf61", "tree": "6fbf3a8c4e2edf18e9ac5da3a0c43c1702c9de1f", "parents": [ "80a7f3d023808574a327fdfd7480ae601cd17f71" ], "author": { "name": "Med Ismail Bennani", "email": "ismail@bennani.ma", "time": "Wed Jun 03 17:00:02 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 17:02:18 2026 -0700" }, "message": "[lldb/test] Add _excluded_variant_combinations hook in lldbtest (#201459)\n\nAdd a module-scope hook that LLDBTestCaseFactory consults during variant\nexpansion so the suite can declare combinations of variant axis values\nthat should never be generated. Each entry in\n`_excluded_variant_combinations` is a dict mapping variant_name -\u003e\nvalue; the helper `_is_excluded_variant_combination(method,\nvariant_name, value_name)` returns True when assigning the given\nvariant\u003dvalue to the method would produce a combination matching every\nentry. `_expand_test_variants` checks the predicate before generating\neach copy and drops the variant entirely so excluded crosses don\u0027t\nappear in the test matrix at all.\n\nThis is a suite-wide per-axis-combination analogue of\nNO_DEBUG_INFO_TESTCASE: instead of collapsing the entire debug-info axis\nfor one test class, it lets the suite declare narrowly that any variant\ncarrying e.g. {\"swift_module_importer\": \"noclang\", \"swift_embedded\":\n\"swiftembed\"} should be dropped, regardless of which source method or\ntest class generated it. Useful when a subset of the variant matrix is\nknown broken (or simply redundant) without resorting to broad @skipIf /\n@expectedFailureAll decorators on each affected method.\n\nThe list starts empty; the first downstream consumer\n(swiftlang/llvm-project) adds the noclang × swiftembed cross at the same\ntime it registers the swift_module_importer / swift_embedded variants.\n\nSigned-off-by: Med Ismail Bennani \u003cismail@bennani.ma\u003e\nGitOrigin-RevId: ff812a578f236d3554750391041146a6781a9f55\n" }, { "commit": "80a7f3d023808574a327fdfd7480ae601cd17f71", "tree": "e7cdf2201d31b2fec75d793c8a4ffa132da84d13", "parents": [ "7cf0b8da585011dd9c6e89ecd08ff81345a45e99" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Wed Jun 03 15:05:02 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 15:10:46 2026 -0700" }, "message": "[lldb][docs] Convert resources/ RST docs to Markdown (NFC) (#201442)\n\nConvert the ten contributor-facing RST docs under lldb/docs/resources/\nto MyST Markdown. This is the second batch of an incremental RST -\u003e\nMarkdown migration. The previous PR (#201256) covered nine small leaf\npages.\n\nFiles: build, dataformatters, debugging, extensions, formatterbytecode,\noverview, projects, qemu-testing, sbapi, test.\n\nVerified by building the docs on origin/main and on this branch with\nidentical sphinx flags and diffing both the warnings and the rendered\nHTML. After file extension and line numbers are normalized, the warning\nsets match exactly. Six of the ten pages are byte-identical. The four\nthat differ (dataformatters, projects, sbapi, test) differ only in:\n\n - CommonMark collapsing two-spaces-after-period to one.\n- MyST adding a `myst` CSS class to `\u003cdl\u003e` elements and an `\u003chr\u003e`\nseparator before footnote sections.\n - Footnote IDs renamed from auto `id3` to named `footnote-1`.\n\nThe diff also surfaced two semantic regressions in the conversion, fixed\nhere:\n\n- build.md `{ref}` to `Optional Dependencies` did not match the new\n`(optional-dependencies)\u003d` anchor; updated to use the new label.\n- test.md and build.md lost cross-reference behavior on single- backtick\nrefs to SB-class names and `lldb`. RST\u0027s default role is `any`, so\nsingle backticks attempted xrefs; in MyST single backticks are plain\ncode spans. Converted these 14 occurrences to explicit `{any}`...``\nsyntax.\n\nSeveral `eval-rst` blocks remain in formatterbytecode.md and test.md to\npreserve grid tables that MyST does not render natively.\n\nContext: https://discourse.llvm.org/t/rfc-make-myst-markdown-the-llvm-docs-format-rip-rest/\n\nAssisted-by: Claude\nGitOrigin-RevId: f70a28f4d1b5d17d94fca03fb74387ad800560de\n" }, { "commit": "7cf0b8da585011dd9c6e89ecd08ff81345a45e99", "tree": "850558213a2998da05b24dbb1ffff324623fb93b", "parents": [ "f6f0475eee52a2de451b4a23bf0d23924c094ae1" ], "author": { "name": "Martin Storsjö", "email": "martin@martin.st", "time": "Wed Jun 03 23:53:07 2026 +0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 13:55:34 2026 -0700" }, "message": "[lldb] Use the right MinGW name for the Python DLLs (#201325)\n\nIn MinGW mode, the Python DLLs have different names than they have in\nMSVC mode; they are named `libpython\u003cmajor\u003e.\u003cminor\u003e.dll` (a \"lib\" prefix\nand a dot between major and minor) and `libpython3.dll`.\n\nThis avoids a warning on startup after\n142ad481b6254104a51da7d636ad9e3f30518a32 and total failures to start up\nafter 3eb13f8db39ed42827122489c830c414cb6660e3.\n\nGitOrigin-RevId: b207fd6fd43514b2852ff64b89769a2cf4b781c2\n" }, { "commit": "f6f0475eee52a2de451b4a23bf0d23924c094ae1", "tree": "73b97ac06f5541e764563e8647a3131d3cd5c411", "parents": [ "12bdc515aa4d4df95c8b1140d6366c2e4b3642a6" ], "author": { "name": "Arseniy Obolenskiy", "email": "arseniy.obolenskiy@amd.com", "time": "Wed Jun 03 22:06:47 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 13:11:04 2026 -0700" }, "message": "[lldb] Add PARTIAL_SOURCES_INTENDED to HostTests to fix build (#201361)\n\nwith LLDB_ENABLE_PYTHON\u003dOFF\n\nPythonRuntimeLoaderTest.cpp is only added to the target when Python is\nenabled, which orphaned the source and broke configuration otherwise\n\nculprit PR: https://github.com/llvm/llvm-project/pull/200524\n\nGitOrigin-RevId: 37432c85b43654d0c3e15edd503e95c74f39afb1\n" }, { "commit": "12bdc515aa4d4df95c8b1140d6366c2e4b3642a6", "tree": "b592e2c50193250fa37f5b776af167e8fcb8db25", "parents": [ "c3bf4a11e358629c414d1c297fe900d63137e28f" ], "author": { "name": "satyanarayana reddy janga", "email": "satyajanga@fb.com", "time": "Wed Jun 03 14:26:54 2026 -0500" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 12:30:45 2026 -0700" }, "message": "[lldb-server] Add breakpoint support to accelerator plugin protocol (#200584)\n\nThis is the 2nd PR of many related to\nhttps://discourse.llvm.org/t/upstreaming-basic-support-for-accelerators/89827/6\nContinuation to https://github.com/llvm/llvm-project/pull/198907\n\nExtend the accelerator plugin infrastructure with breakpoint request and\nhit handling, allowing plugins to set breakpoints in the native process\nand respond when those breakpoints are hit.\n\nThis patch adds:\n\n- Support for jAcceleratorPluginBreakpointHit packet handler in\nGDBRemoteCommunicationServerLLGS that routes hits to the correct plugin\nby name and returns the plugin\u0027s response\n- Many related struct for defining request packet and response packet.\nNew structs: AcceleratorBreakpointByName,\nAcceleratorBreakpointByAddress, AcceleratorBreakpointInfo, SymbolValue\n,AcceleratorBreakpointHitArgs and AcceleratorBreakpointHitResponse\nstructs with JSON encode/decode for the hit round-trip.\n- breakpoints field in AcceleratorActions for plugins to request\nbreakpoints during initialization\n- Tests verifying breakpoints in initialize response and breakpoint hit\nround-trip with JSON validation\n- Packet documentation in lldbgdbremote.md\n\nGitOrigin-RevId: 8eb252007ce042204d9ebf904a353864989ffe8b\n" }, { "commit": "c3bf4a11e358629c414d1c297fe900d63137e28f", "tree": "4457ada27b729b67c3d9f1eb2acdbd5de313e6c3", "parents": [ "a616a35a7869d15e550754301454a1bc04177860" ], "author": { "name": "cmtice", "email": "cmtice@google.com", "time": "Wed Jun 03 11:38:43 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 11:40:50 2026 -0700" }, "message": "[LLDB] Add assignment to DIL. (#190223)\n\nAdd the ability for DIL to recognize and process assignment, updating\nprogram variables. Recognizes \u0027\u003d\u0027, \u0027+\u003d\u0027 and \u0027-\u003d\u0027 operators. Increment\nand decrement (\u0027++\u0027 and \u0027--\u0027) will be added in a separate (future) PR.\n\"*\u003d\" and \"/\u003d\" need to wait until DIL handles multiply and divide\noperators.\n\nGitOrigin-RevId: 0b9dc1bdd3a0186582ae7a56c3c27b6e449715fa\n" }, { "commit": "a616a35a7869d15e550754301454a1bc04177860", "tree": "8d1b97b97805225de014ece675547ac8390ae8eb", "parents": [ "9785e694cb145742ab6be78c2e6228e9d48b988b" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Wed Jun 03 11:08:56 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 11:10:48 2026 -0700" }, "message": "[lldb][docs] Convert simple RST pages to Markdown (NFC) (#201256)\n\nConvert nine short, low-risk RST docs to MyST Markdown as the first\nbatch of an incremental RST -\u003e Markdown migration. Subsequent batches\nwill cover the rest.\n\nVerified by building the docs on origin/main and on this branch with\nidentical sphinx flags and diffing the rendered HTML. Seven of nine\npages are byte-identical.\n\ncontributing.html differs in 42 lines, all attributable to `{doc}` xrefs\nreplacing RST hyperlinks to sibling pages (`reference external` -\u003e\n`reference internal`) and CommonMark collapsing two-spaces-after- period\nto one.\n\nondemand.html differs in 26 lines because two bulleted lists that\nfollowed a paragraph with no blank-line separator originally rendered as\nliteral `- ...` text in both RST and Markdown. The conversion tool\npreserved that by emitting `\\-`; this patch instead inserts a blank line\nso they render as proper `\u003cul\u003e` lists, fixing a latent doc bug.\n\nContext:\nhttps://discourse.llvm.org/t/rfc-make-myst-markdown-the-llvm-docs-format-rip-rest/\n\nAssisted-by: Claude\nGitOrigin-RevId: e43adaea1fb557e860fbb34568ac024fd649cf5c\n" }, { "commit": "9785e694cb145742ab6be78c2e6228e9d48b988b", "tree": "231927a502bde0238d31fae08ca211c58eb1416e", "parents": [ "356658c428fa5f6fad7c5625c1bf3b68d2300279" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Wed Jun 03 18:48:46 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 10:50:36 2026 -0700" }, "message": "[lldb] Preserve FIFO order for equal time MainLoop callbacks (#199056)\n\nCo-authored-by: Adrian Prantl \u003cadrian.prantl@gmail.com\u003e\nGitOrigin-RevId: 1c88bd716167a0650513319176c30dd715101a70\n" }, { "commit": "356658c428fa5f6fad7c5625c1bf3b68d2300279", "tree": "4d74e1ab59622e2771f215faf18034255d95eb95", "parents": [ "e87bdc67e7b5d09284c01d199f913e4f0b54cea2" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Wed Jun 03 17:55:43 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 10:00:32 2026 -0700" }, "message": "[lldb][windows] explicitly load python3.dll in PythonRuntimeLoader (#201399)\n\nWhen building with the Python stable API, `liblldb.dll`\u0027s delay-load\ncrashes because it needs `python3.dll` and it can\u0027t find it\nLoading `python310.dll` via `LoadLibrary` with a full path doesn\u0027t add\nPython\u0027s directory to the DLL search path for subsequent loads.\nThis patch also explicitly loads `python3.dll` from the same directory\nin `PythonRuntimeLoader` to fix the issue.\n\nThis is needed for https://github.com/llvm/llvm-project/pull/200533.\n\nGitOrigin-RevId: 0b109532693fec0dce3a8d275b716f5a5fb6e45c\n" }, { "commit": "e87bdc67e7b5d09284c01d199f913e4f0b54cea2", "tree": "c075e78066d2b2d55ef255bca7a7a611503fbd7a", "parents": [ "5ca5bdc73ed704da63e8b7fe9f96d50ebdc8723d" ], "author": { "name": "Med Ismail Bennani", "email": "ismail@bennani.ma", "time": "Wed Jun 03 08:06:59 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 08:09:08 2026 -0700" }, "message": "[lldb/test] Fix TestExpeditedThreadPCs on remote-darwin targets (#201275)\n\nGitOrigin-RevId: e97e67681d0d7de9c6ab8b293852068ea5591066\n" }, { "commit": "5ca5bdc73ed704da63e8b7fe9f96d50ebdc8723d", "tree": "34bad94748d7200fa2c20083540eb771cc148d8f", "parents": [ "5ca802bc4325536ba2a8040eefae6f99f5e3152a" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Wed Jun 03 10:03:20 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Wed Jun 03 01:06:15 2026 -0700" }, "message": "[lldb][test] Always call quit when tearing down pexpect tests (#201100)\n\nRight now we manually have to call quit at the end of each pexpect test.\nThis patches makes this call automatic.\n\nThis also makes tests that missed the call and where previously waiting\nfor a timeout faster. For example, TestClangREPL.py now only takes about\n10 seconds to run instead of 1 minute.\n\nGitOrigin-RevId: edcafdb6af16350da3444ac6bde10897a4c4ceee\n" }, { "commit": "5ca802bc4325536ba2a8040eefae6f99f5e3152a", "tree": "eddf250eabbe84b11064007745421bc84ef8faa6", "parents": [ "d5fe011dc7100f395d641ebf8e9765597f0da062" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Tue Jun 02 19:41:22 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 19:45:42 2026 -0700" }, "message": "[lldb] Enable MyST colon_fence and deflist extensions (NFC) (#201250)\n\nEnable the colon_fence and deflist MyST parser extensions in the LLDB\ndocs configuration. This is a preparatory step for converting the\nremaining reStructuredText documentation pages to Markdown, where these\ntwo extensions are needed to translate RST admonition directives\n(:::{note}) and definition lists.\n\nContext:\nhttps://discourse.llvm.org/t/rfc-make-myst-markdown-the-llvm-docs-format-rip-rest/\nGitOrigin-RevId: 45c4ebb92ae393a541105c120e0dba2992237c03\n" }, { "commit": "d5fe011dc7100f395d641ebf8e9765597f0da062", "tree": "55fec9bb48ff0b4d9158ae6519028f90a313ebb6", "parents": [ "8506ffd1c3ec13f783ffb329205613e2b22ff362" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Tue Jun 02 16:47:54 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 16:50:55 2026 -0700" }, "message": "[lldb] Stop hard-linking libpython into the dynamic Python plugin (#200530)\n\nDrops ${Python3_LIBRARIES} from the SHARED build of\nlldbPluginScriptInterpreterPython and lets undefined Python symbols\nthrough at link time (`-undefined dynamic_lookup` on Darwin,\n`--allow-shlib-undefined` on Linux; Windows keeps its existing\ndelay-load + import lib).\n\nSystemInitializerFull::Initialize resolves the Python runtime loader\nvia ScriptInterpreterRuntimeLoader::Get(eScriptLanguagePython) and\ncalls Load() before initializing any plugin, so libpython is mapped\ninto the process before either entry point that references it: the\nstatic script interpreter\u0027s Initialize() (which invokes Python via\nthe LLDB_PLUGIN_INITIALIZE loop) and the dynamic plugin\u0027s dlopen\n(whose undefined references resolve against the in-process\nlibpython). This covers both LLDB_ENABLE_DYNAMIC_SCRIPTINTERPRETERS\n\u003dON and \u003dOFF, and keeps Windows working in static builds where the\ndelay-load thunks live in liblldb itself. The loader is\nonce_flag-cached, and errors propagate out via the existing Expected\nreturn.\n\n`import lldb` from a running Python works because libpython is\nalready mapped into the process. The Python runtime loader probes\nfor stable-ABI symbols (Py_IsInitialized, Py_InitializeFromConfig)\nvia dlsym(RTLD_DEFAULT) / GetProcAddress and, finding them, returns\nsuccess without dlopen\u0027ing a second libpython on top.\n\nThe dynamic plugins are added as test-suite dependencies so\ncheck-lldb-shell and check-lldb-api wait for them; they can\u0027t be\nbuild-order deps of liblldb itself because the plugins link against\nliblldb (CMake SHARED cycle).\n\nThe unit-test static variant lldbStaticScriptInterpreterPython keeps\nhard-linking libpython so link-time symbol checking is preserved.\n\nGitOrigin-RevId: 3eb13f8db39ed42827122489c830c414cb6660e3\n" }, { "commit": "8506ffd1c3ec13f783ffb329205613e2b22ff362", "tree": "5f4e8489e5a5895a88ecf5e0f27b94794022286e", "parents": [ "e07b6af6def194e6e7e48da8601e385b6c3d7e30" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Tue Jun 02 16:17:47 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 16:20:41 2026 -0700" }, "message": "[lldb] Add PythonRuntimeLoader for runtime libpython lookup (NFC) (#200524)\n\nGeneralizes the Windows-only Python lookup in PythonPathSetup into a\ncross-platform abstraction. Adds an abstract ScriptInterpreterRuntimeLoader\nwith a per-language factory. The Python implementation dynamically loads Python\nlibrary into the current process.\n\nThe loader no-ops when Python is already in the process, then walks\nLLDB_PYTHON_LIBRARY env override, the build-time Python\n(LLDB_PYTHON_RUNTIME_LIBRARY_BUILD_PATH) and finally a platform candidate list:\n\n- Darwin: DEVELOPER_DIR, the bundled Xcode.app, and Command Line Tools joined\n against Python3.framework. Then python.org, /opt/homebrew, and /usr/local\n joined against Python.framework. Then xcrun -f python3 and if that fails,\n libpython3.dylib as a last resort.\n- Linux: libpython3.so plus descending stable-ABI SONAMEs.\n- Windows: the LLDB_PYTHON_RUNTIME_LIBRARY_FILENAME bare name (resolved via the\n loader\u0027s default search list) and the exe-relative\n LLDB_PYTHON_DLL_RELATIVE_PATH fallback (built off GetModuleFileNameW).\n Pre-mapping python3xx.dll lets the script interpreter plugin\u0027s delay-load\n thunks resolve against the already-loaded module by base name on first use.\n\nThis commit only introduces the abstraction. No existing call site is changed,\nand the script interpreter plugin still hard-links Python, which are part of\ntwo follow-up PRs.\n\nGitOrigin-RevId: 142ad481b6254104a51da7d636ad9e3f30518a32\n" }, { "commit": "e07b6af6def194e6e7e48da8601e385b6c3d7e30", "tree": "103bd3296384a43287245bb92daf0b80cb235bc8", "parents": [ "bddf8e4730b40181ea4b8417d5cc1de897de6536" ], "author": { "name": "Jason Molenda", "email": "jmolenda@apple.com", "time": "Tue Jun 02 16:01:02 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 16:05:34 2026 -0700" }, "message": "[lldb][debugserver] Arguments to kill(2) are reversed (#201226)\n\nThis codepath is only executed as an attempt to clean up during a failed\nlaunch, so the reversed arguments were rarely actually used.\n\nrdar://175507620\n\nGitOrigin-RevId: 2a936ce9316042cd7e45e72b528ad1d55aa2fe3c\n" }, { "commit": "bddf8e4730b40181ea4b8417d5cc1de897de6536", "tree": "61a1c0d9cfc6a155d6b48b76f6575df3cd0e2667", "parents": [ "bc06d8a890305f132e1a823d41b33110fdc46ca0" ], "author": { "name": "Jason Molenda", "email": "jmolenda@apple.com", "time": "Tue Jun 02 15:54:47 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 15:57:20 2026 -0700" }, "message": "[lldb] Have TestRunLocker run both styles of launch (#200978)\n\nWhile debugging flakey behavior with TestRunLocker, I noticed that is\nintended to run its test once with a stop at the entry function (and\nthen Continues) and once where we launch to the main() loop. But we were\nnever exercising the stop-at-entry codepath.\n\nThis doesn\u0027t fix the flakey behavior, although that only happens with\nthe launch-directly-into-main() codepath; I don\u0027t get failures when I\nstop at the entry point and then continue.\n\nGitOrigin-RevId: 72a05d59bc8cba3728ad1e3f60bdcc8504f7a708\n" }, { "commit": "bc06d8a890305f132e1a823d41b33110fdc46ca0", "tree": "307033a7c8ebf37a7d5f319a5178a69629bdba2a", "parents": [ "da9e2cb41f74f728544e076d7479f4301c9d90af" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Tue Jun 02 15:43:34 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 15:45:42 2026 -0700" }, "message": "[lldb] Add lldb-mcp to LLVM_DISTRIBUTION_COMPONENTS (#201225)\n\nAdd lldb-mcp to LLVM_DISTRIBUTION_COMPONENTS\n\nGitOrigin-RevId: 155c6e4f887a9b4fb8900ec655124303e7308b4b\n" }, { "commit": "da9e2cb41f74f728544e076d7479f4301c9d90af", "tree": "bada9250e1bfd4481a8892a587724c419a64f566", "parents": [ "cd12a3a0779763f487f2c281ad532ea2f766d40f" ], "author": { "name": "Jonas Devlieghere", "email": "jonas@devlieghere.com", "time": "Tue Jun 02 14:45:39 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 14:50:15 2026 -0700" }, "message": "[lldb][test] Make delayed-definition-die-searching CU-layout agnostic (#201206)\n\nThe second `ParseTypeFromDWARF` for t1 (after `p v2`) only fires when\nt1\u0027s definition lives in a separate CU from its forward declaration:\nLLDB parses the forward-decl DIE during `p v1` and a distinct definition\nDIE during `p v2`. dsymutil\u0027s parallel linker collapses both into a\nsingle DIE in the artificial type unit, so t1 is parsed once during `p\nv1` and only re-resolved during `p v2`.\n\nDrop the second-parse CHECK so the test no longer presumes a per-CU type\nlayout. The remaining `\u0027t1\u0027 resolving forward declaration...` CHECK\nafter `p v2` still verifies what the test was designed to catch: t1\u0027s\ncomplete-type resolution is deferred until v2 is evaluated. If LLDB\nregressed to eager resolution during `p v1`, that log line would move\nand the test would fail. Add a `(t1) (x \u003d 0)` CHECK at the end to cover\nthe end-to-end value.\n\nGitOrigin-RevId: f2a780bbe2d8f66bcffb76e9d5f008d581be7917\n" }, { "commit": "cd12a3a0779763f487f2c281ad532ea2f766d40f", "tree": "23a3e5546a05444a10ac9cdfc94fc4118443c07a", "parents": [ "fb0ee8efdc0852e4a8898d4063a1a94a62b1b527" ], "author": { "name": "Adrian Prantl", "email": "aprantl@apple.com", "time": "Tue Jun 02 12:49:49 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 12:50:51 2026 -0700" }, "message": "[LLDB] Add always-on logging for \"settings set\" (#201185)\n\nWhen triaging bugreports it can be very useful to understand what LLDB\ndefaults have been changed. Adding settings to the always-on logging\nchannel helps with that.\n\nrdar://176482205\n\nAssisted-by: claude\nGitOrigin-RevId: 7d2dce084307356948a3ef78c902fd68558afe2e\n" }, { "commit": "fb0ee8efdc0852e4a8898d4063a1a94a62b1b527", "tree": "06ed57b2d799b0842072a56b8d3ad08a83e8a872", "parents": [ "404bf19d36917f6c76b970408032442051075eaf" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Tue Jun 02 10:35:07 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 12:50:41 2026 -0700" }, "message": "[lldb] Skip TestQemuLaunch.py on arm64e (#200969)\n\nThis test is set up to use a python-based qemu-like stub instead of the\nactual qemu (which makes it far more portable). When trying to run the\ntest as arm64e, LLDB will attempt to launch an arm64e-based qemu.\nBecause the fake qemu is a python script, LLDB will try to launch python\nas arm64e. Neither the python that ships with Xcode nor the python from\npython.org have an arm64e slice, so this test will not work for arm64e.\n\nGitOrigin-RevId: 2511bdd645ee7b2ce01064b5fc31524c5e532480\n" }, { "commit": "404bf19d36917f6c76b970408032442051075eaf", "tree": "bcdd40fb216552643f90a7d52d0931097cb84bcc", "parents": [ "7f832218adfec848b7591d2e2950121a86299856" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Tue Jun 02 10:29:46 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 10:35:36 2026 -0700" }, "message": "[lldb] Skip simulator platform tests on arm64e (#200505)\n\nSimulators do not support arm64e.\n\nGitOrigin-RevId: 13e532562f13ce94c7b7ffc8621252de92a29a42\n" }, { "commit": "7f832218adfec848b7591d2e2950121a86299856", "tree": "fd97f713049880ec1296ace25faebe2ff533b393", "parents": [ "b3be0c15bd14cfa5b5d41d8845e741c538b2d8f7" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Tue Jun 02 18:20:15 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 10:22:06 2026 -0700" }, "message": "[lldb][windows] return false in default branch of RegisterContextWindows_*::WriteRegister (#200885)\n\nGitOrigin-RevId: 55af104f1c9eeb93ca127104b810bc2520100f9c\n" }, { "commit": "b3be0c15bd14cfa5b5d41d8845e741c538b2d8f7", "tree": "f331c2a86547a391889ea6e2cb8739833fa932e6", "parents": [ "20a6023226adf844cb407c795d53633de9909d03" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Tue Jun 02 09:29:21 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 09:31:27 2026 -0700" }, "message": "[lldb] Fix Symbol static_assert for 32 bit Windows (#201130)\n\nFollow up to\nhttps://github.com/llvm/llvm-project/pull/200919#issuecomment-4600914496\n\nGitOrigin-RevId: 7b199bdb2fe3d852684fab77bfdd3141dbc987f9\n" }, { "commit": "20a6023226adf844cb407c795d53633de9909d03", "tree": "99f7b4a309e408e3beb2719ff597834e4e09156c", "parents": [ "e25b6f94af56ebacf4e3b8e4698dba4c3469d370" ], "author": { "name": "Lukacma", "email": "Marian.Lukac@arm.com", "time": "Tue Jun 02 16:44:37 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 08:45:29 2026 -0700" }, "message": "Revert \"[AArch64] Fix definition of system register move instructions\" (#201137)\n\nReverts llvm/llvm-project#185709\n\nGitOrigin-RevId: 9083fb24c709238b1f18f0486bcf014e490e1997\n" }, { "commit": "e25b6f94af56ebacf4e3b8e4698dba4c3469d370", "tree": "629fb74a2edaed03f9df40724e98a21e0ed9ce90", "parents": [ "be6217d7fdccccfe1f0f1cd2c20e91f820a405ad" ], "author": { "name": "Felipe de Azevedo Piovezan", "email": "fpiovezan@apple.com", "time": "Tue Jun 02 16:02:37 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 08:05:38 2026 -0700" }, "message": "[lldb][NFC] Factor out helper code from MemoryCache (#201120)\n\nThis will be reused in a subsequent commit.\n\nGitOrigin-RevId: 8dd51399f8a9f13cbe35fb2060f05bd30acfb6f8\n" }, { "commit": "be6217d7fdccccfe1f0f1cd2c20e91f820a405ad", "tree": "7982382332a342b46410e0d5c5be707dd752fa2d", "parents": [ "79449ae87a80304973591960d582096a3631323e" ], "author": { "name": "Ilia Kuklin", "email": "ikuklin@accesssoftek.com", "time": "Tue Jun 02 19:20:35 2026 +0500" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 07:25:34 2026 -0700" }, "message": "[lldb] Add bitwise shift operators and fix literals\u0027 sign in DIL (#192506)\n\nThis patch add bitwise shift operators and fixes the Scalar\u0027s sign\ninside of the literals with signed types. The only way to test this fix,\nfor now, is to use these literals in an arithmetic right shift that does\nsign extension, so this is included in this patch.\n\nGitOrigin-RevId: d25f10e8b4a37c2759925883fb59a9da4ea0eb94\n" }, { "commit": "79449ae87a80304973591960d582096a3631323e", "tree": "3d044b6bda25c92b130a84b584d7cd710f86acc2", "parents": [ "6fb4dae2c33dd3fe806c2689839481580255df51" ], "author": { "name": "Nerixyz", "email": "nerixdev@outlook.de", "time": "Tue Jun 02 16:06:41 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 07:10:25 2026 -0700" }, "message": "[lldb][Windows] Use LazyImport for WaitForDebugEventEx (#201118)\n\nFrom\nhttps://github.com/llvm/llvm-project/pull/196395#pullrequestreview-4409209050.\n\nThis uses `LazyImport` for `WaitForDebugEventEx` to simplify the import.\n\nGitOrigin-RevId: 023248a62673ce82045fa284022673882f209677\n" }, { "commit": "6fb4dae2c33dd3fe806c2689839481580255df51", "tree": "877a90cdc6af83a20b29ed1d376cd3e97967325b", "parents": [ "c1aa527b98dc7ea3978f6fc8b87fd4bc506f2e14" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Tue Jun 02 14:12:26 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 06:15:21 2026 -0700" }, "message": "[lldb][Windows] Register MSVCRTCFrameRecognizer from DynamicLoaderWindowsDYLD (#201097)\n\nGitOrigin-RevId: 6180a4899ff913e0a0e2338e4d019b683f8f06da\n" }, { "commit": "c1aa527b98dc7ea3978f6fc8b87fd4bc506f2e14", "tree": "68b8dd04ff33477ba4ff1323a8dbbd2970b9b662", "parents": [ "70baaf293b045017e22ab647af0078d560387c3e" ], "author": { "name": "Nerixyz", "email": "nerixdev@outlook.de", "time": "Tue Jun 02 15:07:56 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 06:10:40 2026 -0700" }, "message": "[lldb][Windows] Support OutputDebugString (#196395)\n\nThis picks https://reviews.llvm.org/D128541 back up. It implements\nsupport for `OutputDebugStringA/W` on Windows. It\u0027s used by some logging\nsystems.\n\nThe main changes since the original patch:\n- Use `WaitForDebugEventEx` for getting debug events over\n`WaitForDebugEvent`. According to the\n[docs](https://learn.microsoft.com/en-us/windows/win32/api/debugapi/nf-debugapi-waitfordebugeventex),\nthe only difference is that the -Ex version correctly outputs Unicode\nstrings.\n- Support strings longer than 64 KiB. I set an arbitrary limit of 1 MiB\nfor the strings we read. The debugger interface only tells us the length\nmodulo 64 KiB which is a bit awkward. Libraries like Qt already chunk\ncalls to `OutputDebugString`, so strings shouldn\u0027t be too big in\npractice.\n- Output to stdout instead of a log channel, so the output is always\nenabled. I don\u0027t know if this should go to stdout or stderr.\n\nTested that it works with lldb-dap and shows up in the debug console in\nall modes.\n\nCloses #185891.\n\n---------\n\nCo-authored-by: Alvin Wong \u003calvin@alvinhc.com\u003e\nGitOrigin-RevId: 2a7524b4e33c8dfd667fa28341ec7c6c529d9c50\n" }, { "commit": "70baaf293b045017e22ab647af0078d560387c3e", "tree": "0ab38c0ca7ad679788c7b337a0e28fc1329a1424", "parents": [ "cab92c8794820fe382ac69779eed37f9e4d3bd9a" ], "author": { "name": "Hans Wennborg", "email": "hans@hanshq.net", "time": "Tue Jun 02 13:43:23 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 04:45:42 2026 -0700" }, "message": "Revert \"[clang] fix getTemplateInstantiationArgs\" (#201093)\n\nThis caused bootstrap builds on macOS to fail with\n\n error: definition with same mangled name ... as another definition\n\nin JSONGenerator.cpp. See comments on the original PR.\n\nReverts llvm/llvm-project#199528\n\nGitOrigin-RevId: b29352f7ea1d46fa7d90900d7a279851b6de9f74\n" }, { "commit": "cab92c8794820fe382ac69779eed37f9e4d3bd9a", "tree": "8b9da56529dc7410d58024ee3aa85510ccf845ac", "parents": [ "cc9614d6ffd9ec69c3b0b4ec2c65d518e447335f" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Tue Jun 02 11:57:43 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 04:00:27 2026 -0700" }, "message": "[lldb][windows] copy vcpkg runtime dlls (#200416)\n\nGitOrigin-RevId: 6e1f2f513c84759e0d0727925dbb8c3a3e8b5199\n" }, { "commit": "cc9614d6ffd9ec69c3b0b4ec2c65d518e447335f", "tree": "a4aebf52dd05936c3dc8b03bf41364e79fc07b6e", "parents": [ "5b31eb86c6cd9beaacf88522a4a73d3509bdac4e" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Tue Jun 02 11:45:37 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 03:51:07 2026 -0700" }, "message": "[lldb][Windows] Default-initialize PseudoConsole Kernel32 members (#200880)\n\nGitOrigin-RevId: 3c0daa26db4ba76359ca3285af6b709d8199c1cc\n" }, { "commit": "5b31eb86c6cd9beaacf88522a4a73d3509bdac4e", "tree": "b5ac9347ff80f03df91cc2485ba20d680587fc77", "parents": [ "ac0d425f9d87fc6730776003e268e2c3d27c8329" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Tue Jun 02 11:45:35 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 03:50:56 2026 -0700" }, "message": "[lldb][windows] Close pipe HANDLE if _open_osfhandle fails (#200882)\n\nGitOrigin-RevId: c51296345de89f80fbaeac85e92d72d0df06c0cb\n" }, { "commit": "ac0d425f9d87fc6730776003e268e2c3d27c8329", "tree": "6cf5c0491ed0c9809a7ba757069edf436a548262", "parents": [ "b4de424aa7bfd40ac05cb851096150187aa037aa" ], "author": { "name": "David Spickett", "email": "david.spickett@arm.com", "time": "Tue Jun 02 10:33:10 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 02:35:27 2026 -0700" }, "message": "[lldb][test] Disable part of TestThreadJump.py on WoA (#201074)\n\nSee https://github.com/llvm/llvm-project/issues/201068.\n\nGitOrigin-RevId: 7d62e753bb9ef0c21d1be9d155e4f6456391a406\n" }, { "commit": "b4de424aa7bfd40ac05cb851096150187aa037aa", "tree": "35e03e5b061e8cae8f3b5de77815545d68555fb6", "parents": [ "88fa41f6fe364a19d2bd22bf385880353bfd3f08" ], "author": { "name": "Felipe de Azevedo Piovezan", "email": "fpiovezan@apple.com", "time": "Tue Jun 02 08:35:22 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Tue Jun 02 00:40:34 2026 -0700" }, "message": "[lldb][NFCI] Cleanup AppleObjCClassDescriptorV2::ivar_t API (#201042)\n\nGitOrigin-RevId: 114e135010ba92afd87462dd120f865b83f966b3\n" }, { "commit": "88fa41f6fe364a19d2bd22bf385880353bfd3f08", "tree": "50b1a4ddc3522db3ba69dd2447520e5b27050afb", "parents": [ "c74847d6edd7265602c15051bd6be0a071771a1f" ], "author": { "name": "Matheus Izvekov", "email": "mizvekov@gmail.com", "time": "Mon Jun 01 22:07:01 2026 -0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 18:10:31 2026 -0700" }, "message": "[clang] fix getTemplateInstantiationArgs (#199528)\n\nThis implements a new strategy for collecting the template arguments, by\nrelying on the qualifiers and template parameter lists to navigate the\ntemplate context of out-of-line definitions.\n\nThis greatly simplifies the signature of that function, by removing a\nbunch of workarounds, and simpliffying a couple that weren\u0027t removed\nyet.\n\nSince this now relies on qualifiers and template parameter lists, this\npatch expends most of its effort making sure these are placed,\ntransformed and propagated to template instantiations.\n\nAlso makes the explicit specialization AST nodes stop abusing the\ntemplate parameter lists by storing it\u0027s own template parameter list,\ncreating a dedicated field for them, similar to partial specializations.\n\nFixes #101330\n\nGitOrigin-RevId: ccd2606d0c30c58554cc11038a2160672ef529d5\n" }, { "commit": "c74847d6edd7265602c15051bd6be0a071771a1f", "tree": "8b012fe1d516e8f11c8981639e0ba659ff5762cb", "parents": [ "ed2f09a4ec0292afaa0d777cae22b3036849533c" ], "author": { "name": "Med Ismail Bennani", "email": "ismail@bennani.ma", "time": "Mon Jun 01 15:54:06 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 15:55:53 2026 -0700" }, "message": "[lldb] Skip libc++ category tests on Darwin when no in-tree libc++ is built (#199262)\n\n`canRunLibcxxTests()` previously short-circuited with \"libc++ always\npresent\" for all Darwin targets, meaning the \"libc++\" test category was\nnever skipped on macOS — even when `LLDB_HAS_LIBCXX` is `OFF` and no\n`--libcxx-include-dir` / `--libcxx-library-dir` are passed to dotest.\nThe tests would silently run against the system libc++ instead of an\nin-tree build, producing results inconsistent with what the suite is\ndesigned to validate.\n\nThis fixes `canRunLibcxxTests()` to apply the same `libcxx_include_dir`\n/ `libcxx_library_dir` guard on `Darwin` that `Linux` already uses. When\nthose dirs are absent (i.e. no in-tree libc++ was built), the function\nreturns `False` and `checkLibcxxSupport()` appends `libc++` to\n`skip_categories` — skipping those tests exactly as Linux does.\n\nOn the CMake side, the `SEND_ERROR` for `LLDB_HAS_LIBCXX\u003dOFF` is\ndowngraded to a `WARNING` so downstreams that intentionally skip the\nruntimes build can keep `LLDB_INCLUDE_TESTS\u003dON` for the tests they\nactually want to run. The warning is also gated on the new\n`LLDB_ENABLE_LIBCXX_TESTS` option (default `ON`): setting it to OFF\nacknowledges the deliberate choice and silences the warning without\nrequiring either an in-tree libc++ build or disabling all tests.\n\nSigned-off-by: Med Ismail Bennani \u003cismail@bennani.ma\u003e\nGitOrigin-RevId: 0597087c44e493610d3f96608bef46e721201af6\n" }, { "commit": "ed2f09a4ec0292afaa0d777cae22b3036849533c", "tree": "192e373a2325aaf0d53c8f6b10c66030151d8175", "parents": [ "a11a79a527e0ddb47a1fb1e93fc7185537c6cb07" ], "author": { "name": "Med Ismail Bennani", "email": "ismail@bennani.ma", "time": "Mon Jun 01 15:53:53 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 15:55:43 2026 -0700" }, "message": "[lldb/test] Fix variant double-expansion in LLDBTestCaseFactory (#200943)\n\nLLDBTestCaseFactory generates the actual test methods that get run: for\nevery `test*` method on a TestBase subclass, it stamps out one copy per\ndebug-info format (dwarf, dsym, ...) and, if a variant like\nswift_clang\u0027s clang/noclang is registered, one copy per variant value on\ntop. When two test methods in the same class have names that share a\nprefix and the longer one is declared first in the file, the shorter\nmethod\u0027s expansion ends up re-stamping every copy already produced for\nthe longer one.\n\nThe variant-expansion helper decides what to copy by name-prefix match:\nwhen processing `test_expr`, it grabs `test_expr`, `test_expr_dwarf`,\n`test_expr_dsym`, ... and produces a clang/noclang suffix for each one.\nThe factory was handing it the full running dict of already-synthesized\nmethods, so by the time `test_expr`\u0027s turn came, the dict also held\n`test_expr_stripped_dwarf` and `test_expr_stripped_dsym` from\n`test_expr_stripped`. Those names start with `test_expr_` too, so they\npick up a second clang/noclang suffix and inherit `test_expr`\u0027s\nxfail/skip predicates.\n\nHiddenIvarsTestCase trips this twice in a row (test_expr_stripped before\ntest_expr; test_frame_variable_stripped before test_frame_variable).\nUpstream `_test_variants` is empty so the bug is latent on mainline, but\nany registered variant exposes it.\n\nWith swift_clang (values clang/noclang) registered, test_expr_stripped +\ntest_expr should produce eight methods (4 dwarf/dsym × 2 clang/noclang\nper original). Before this patch the factory emits twelve, four of them\njunk:\n\n```\n test_expr_stripped_dsym_clang_clang\n test_expr_stripped_dsym_clang_noclang\n test_expr_stripped_dsym_noclang_clang\n test_expr_stripped_dsym_noclang_noclang\n test_expr_stripped_dwarf_clang_clang\n test_expr_stripped_dwarf_clang_noclang\n test_expr_stripped_dwarf_noclang_clang\n test_expr_stripped_dwarf_noclang_noclang\n test_expr_dsym_clang\n test_expr_dsym_noclang\n test_expr_dwarf_clang\n test_expr_dwarf_noclang\n```\n\nTrack each method\u0027s expansion in a local dict and merge it back into the\nshared dict only once that method is fully processed, so the variant\nhelper never sees another method\u0027s copies.\n\nGitOrigin-RevId: 94337fb6b2d5935cdaa9abecdba8274924119dcc\n" }, { "commit": "a11a79a527e0ddb47a1fb1e93fc7185537c6cb07", "tree": "dfbb06cd2df5801fc09551b3b2982247db08374c", "parents": [ "8ff905902953283118e967018503b0ecc8e2cf67" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Mon Jun 01 14:22:43 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 14:25:38 2026 -0700" }, "message": "[lldb] Add static_assert to catch increases to size of Symbol (#200919)\n\nGitOrigin-RevId: 333bc302b49f37da1987e15ae6f66a3027f615f7\n" }, { "commit": "8ff905902953283118e967018503b0ecc8e2cf67", "tree": "ea214067d0f81b79d315cd65fdeee353dfd3aff8", "parents": [ "77607bda46b67613d1f1dc04c2ccdbe25e2e8b2c" ], "author": { "name": "Fangrui Song", "email": "i@maskray.me", "time": "Mon Jun 01 14:07:15 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 14:10:44 2026 -0700" }, "message": "[lldb] Remove unused DenseMapInfo::getTombstoneKey (#200635)\n\n#200595 changed DenseMap to no longer create tombstone buckets, so\nDenseMapInfo\u003cT\u003e::getTombstoneKey() is never called. Remove dead\ndefinitions and dead tombstone branches.\n\nGitOrigin-RevId: 54796f26bf917bd68434f4c13c2a9bb16506cba9\n" }, { "commit": "77607bda46b67613d1f1dc04c2ccdbe25e2e8b2c", "tree": "f0682e060c4b1db587b40215b6d243723cde7df2", "parents": [ "40d3b5e971956735da6d700e463a8e33c35a4700" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Mon Jun 01 13:37:55 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 13:40:31 2026 -0700" }, "message": "[lldb] Strip objc superclass pointer in trampoline handler (#200490)\n\nThe pointer needs to be stripped before being handed off to any objc\nruntime functions. Otherwise the utility expression will hit a PAC\nexception and the thread plan will fail to execute correctly.\n\nThis fixes TestObjCStepping.py on arm64e.\n\nGitOrigin-RevId: 4d3513075df51a2cb579e3d82e9593aac42a4bce\n" }, { "commit": "40d3b5e971956735da6d700e463a8e33c35a4700", "tree": "f5015e812c4ef88b3e4634922a6b2ee6e74910f0", "parents": [ "588820d71a4d9ea2eeb362da532f78db6731883e" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Mon Jun 01 13:27:28 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 13:30:14 2026 -0700" }, "message": "[lldb] Fix TestBranchIslands.py for arm64e (#200498)\n\nNeed to pass CFLAGS to clang when building the asm files, otherwise the\ntriple isn\u0027t used and they\u0027re automatically compiled for the host\nplatform.\n\nGitOrigin-RevId: 385ba8821c75a13ca45a38047e69670921ba9a06\n" }, { "commit": "588820d71a4d9ea2eeb362da532f78db6731883e", "tree": "3fd8f311f8320151073114a3c2d03963a85ee2f5", "parents": [ "7cf09c5a75be6800eb4f6f0d62c2fc4f0bdf7c32" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Mon Jun 01 13:01:29 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 13:05:45 2026 -0700" }, "message": "[lldb] Skip minidump case in TestDynamicValue.py for arm64e (#200047)\n\nThe minidump format does not currently have a way to distinguish arm64e\nfrom arm64.\n\nGitOrigin-RevId: 7a9fdaa0b1a0c003d9c278bd40647b357952ca93\n" }, { "commit": "7cf09c5a75be6800eb4f6f0d62c2fc4f0bdf7c32", "tree": "4ddc0b19572986cb0a015d6af2ea70c6534c1728", "parents": [ "5f33d03f97e85d17429567398b1c08ca2f03c479" ], "author": { "name": "Sergei Druzhkov", "email": "serzhdruzhok@gmail.com", "time": "Mon Jun 01 22:36:09 2026 +0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 12:41:14 2026 -0700" }, "message": "[lldb-dap] Mark source deemphasize if path doesn\u0027t exist (#194702)\n\nLLDB-DAP has a problem with sanitizers in GCC. When we stop in\nsanitizer\u0027s code, lldb-dap sends stack frames with path (sanitizer\u0027s\nbuild dir path) that doesn\u0027t exist on machine. It leads to problems in\nVS Code UI (see issue below).\n\nFixes #184789\n\nGitOrigin-RevId: 659c3fbd732771ad30052a9cd2c5896cfe00737e\n" }, { "commit": "5f33d03f97e85d17429567398b1c08ca2f03c479", "tree": "05b3b4469280e3bd2c7c015c03eba9a250f6920e", "parents": [ "5b5926f7c6ddf52ba44a1e999223b35ca7826519" ], "author": { "name": "Sergei Druzhkov", "email": "serzhdruzhok@gmail.com", "time": "Mon Jun 01 19:45:50 2026 +0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 09:50:26 2026 -0700" }, "message": "[lldb-dap] Use SetTarget for launch and attach commands (#200133)\n\nWithout this patch event listener registration was skipped, as a result\n`Modules` view in UI was not displayed in case of launching target via\n`launchCommands` or `attachCommands`.\n\nGitOrigin-RevId: 09f30ef6105e3f9695ad876fcb5add2e7917712b\n" }, { "commit": "5b5926f7c6ddf52ba44a1e999223b35ca7826519", "tree": "b8337b025c9dc56e40aca34a10f6330663c4d826", "parents": [ "18737d7e314da14dfbd661df9df920ad8c56a171" ], "author": { "name": "Sergei Druzhkov", "email": "serzhdruzhok@gmail.com", "time": "Mon Jun 01 19:34:46 2026 +0300" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 09:40:33 2026 -0700" }, "message": "[lldb-dap] Cleanup InstructionBreakpoint (#200228)\n\nAdded mutex like in other breakpoints\n([PR](https://github.com/llvm/llvm-project/pull/134030)). Also removed\nunused `m_offset` field.\n\nGitOrigin-RevId: 8b901cc2a0a3310b2b25d01f31d4bec665b714ef\n" }, { "commit": "18737d7e314da14dfbd661df9df920ad8c56a171", "tree": "45eb1be61d65d335b2d8643058edef71eabfa5b1", "parents": [ "55622556527a1d8c80e1e05be2ddbb67e0b41a1f" ], "author": { "name": "Adrian Prantl", "email": "aprantl@apple.com", "time": "Mon Jun 01 09:19:17 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 09:20:44 2026 -0700" }, "message": "[LLDB] Detect cycles during Type resolution (#200304)\n\nI got LLDB crash reports from the Swift plugin where (presumably\nmalformed) debug info sends lldb_private::Type into an infite recursion.\nMost likely this is a bug in the DWARF parser, however, even malformed\ninputs shouldn\u0027t crash LLDB so this patch adds cycle detection.\n\nrdar://177856769\n\nAssisted-by: claude\nGitOrigin-RevId: 9e0d8a65db700435bda0ca8d63dadb07e6e18370\n" }, { "commit": "55622556527a1d8c80e1e05be2ddbb67e0b41a1f", "tree": "d56442c8a7b691833ba82f608b304e2d5d648402", "parents": [ "356426c25e052e714b9e0bc25b29fc2d44d4359c" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 16:41:15 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 08:45:37 2026 -0700" }, "message": "[lldb][Windows] Use captured error in ConnectionGenericFile::Read (#200803)\n\nUse the captured value on both branches so the reported error matches\nthe one that was tested against.\n\nGitOrigin-RevId: 45660cb7d894741cfaeb5f94058662c81f9a07ca\n" }, { "commit": "356426c25e052e714b9e0bc25b29fc2d44d4359c", "tree": "c22672c28056f2f2fc53685ba368b46189b09d78", "parents": [ "8b66d4b7fafdccddfb863575e69750779f916c4b" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 16:26:23 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 08:30:49 2026 -0700" }, "message": "[NFC][lldb][windows] cleanup ProcessLauncherWindows result check (#200804)\n\nGitOrigin-RevId: 5fa7b5547d2a70003a681ecf2823170c69c44bec\n" }, { "commit": "8b66d4b7fafdccddfb863575e69750779f916c4b", "tree": "2d67868598aae8029ea4d891a1c44659a3d86ec2", "parents": [ "4ef1a622a68e3b7e678a25ee543c70990e12e5ca" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 16:25:37 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 08:30:39 2026 -0700" }, "message": "[NFC][lldb][windows] break down NativeProcessWindows::OnDebugException (#200832)\n\nThis patch breaks down `NativeProcessWindows::OnDebugException` into 3\ndifferent handlers (one for each exception) for readability.\n\nGitOrigin-RevId: e68990b844ee4bfbb29a187b32232a1edb04a873\n" }, { "commit": "4ef1a622a68e3b7e678a25ee543c70990e12e5ca", "tree": "53a8c4d6699cdb012bed86460624f6b4d84d88fa", "parents": [ "c4425307a1fa63edf361d4220b4f3e2d43d9751d" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:52:56 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:55:38 2026 -0700" }, "message": "[lldb][windows] refactor null handling in DynamicLoaderWindowsDYLD (#200821)\n\nThis patch thightens the null guards.\n\n---------\n\nCo-authored-by: Nerixyz \u003cnero.9@hotmail.de\u003e\nGitOrigin-RevId: 25ebed5fc05af9263300e381c684b3202fec3b2f\n" }, { "commit": "c4425307a1fa63edf361d4220b4f3e2d43d9751d", "tree": "2d8bc1a2d4956ec904338f325387ae85964fc4c9", "parents": [ "92c6ca868975203bf1ce104e8b99f7a4539df40f" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:27:51 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:32:09 2026 -0700" }, "message": "[lldb][Windows] Don\u0027t kill lldb when a thread fails to stop (#200805)\n\nReplace `exit(1)` in `NativeProcessWindows::StopThread` with logging so\na single thread suspend failure no longer terminates lldb itself.\n\nGitOrigin-RevId: b05c5e4d8238bfe34f3b4df72377f714f2e60d2f\n" }, { "commit": "92c6ca868975203bf1ce104e8b99f7a4539df40f", "tree": "139dcdc87a65015c09917ad5e7ebdc4faffc5cf6", "parents": [ "c7b44298d9ce94935cfa606c95922bf1b76c04b4" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:25:53 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:32:01 2026 -0700" }, "message": "[NFC][lldb][windows] follow styling guidelines for hProcess (#200812)\n\n`hProcess` does not follow the LLVM styling guidelines. Rename it to\n`process`.\n\nGitOrigin-RevId: d2bea40d8822e51f58c2655996f27d159179d120\n" }, { "commit": "c7b44298d9ce94935cfa606c95922bf1b76c04b4", "tree": "889e73fbb6a31794d9a487cc3be163bbf1fe4e4a", "parents": [ "c767fada1b965a8636e779dddab3a661d4d6530f" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:25:47 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:31:53 2026 -0700" }, "message": "[NFC][lldb][windows] simplify NativeProcessWindows::FindSoftwareBreakpoint (#200820)\n\nGitOrigin-RevId: cedd8d914432253879a1b830c0550c0fb7f01776\n" }, { "commit": "c767fada1b965a8636e779dddab3a661d4d6530f", "tree": "25528d0ac19fd4fc79f35b37e20a4cd7afb0c8d5", "parents": [ "f9fee4a4f9a2875a08741add333b528cf4b0bc52" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:25:34 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:31:44 2026 -0700" }, "message": "[NFC][lldb][windows] Simplify PlatformWindows::ConnectRemote (#200823)\n\nGitOrigin-RevId: 6a80387760c17b7e72e77b1b24ac1808bdf0929e\n" }, { "commit": "f9fee4a4f9a2875a08741add333b528cf4b0bc52", "tree": "c64a5d6a567bbaf617c7ab0ef880f01c12b1ba2b", "parents": [ "3783b4b660457bdb5ea446675901661c8efffd96" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 15:24:42 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 07:31:31 2026 -0700" }, "message": "[lldb][windows] fix late null check (#200822)\n\n`process_sp` should be null checked before calling the\n`HijackProcessEvents` method. This patch also removes 2 unused\nvariables.\n\nGitOrigin-RevId: fc8911a78e3618adbf422d7856ed0cb4115eada1\n" }, { "commit": "3783b4b660457bdb5ea446675901661c8efffd96", "tree": "49c154abe329ecbd7a5e0c4c37788f1851c560f9", "parents": [ "f6ffe9ae7ec37a4b032722d666fa16e179cc1e4d" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Mon Jun 01 14:12:50 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 06:16:31 2026 -0700" }, "message": "[lldb][windows] bound ReadProcessMemory (#200230)\n\nGitOrigin-RevId: b99a6fc497a014c80091ab7582e331b3d03090e9\n" }, { "commit": "f6ffe9ae7ec37a4b032722d666fa16e179cc1e4d", "tree": "38be0551a7b495495682334bd0a92d6d50aaa11b", "parents": [ "c0ce1de0f7f45e4a1daefb92288ee8cf2dec9ecb" ], "author": { "name": "Raphael Isemann", "email": "rise@apple.com", "time": "Mon Jun 01 12:27:08 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 03:30:50 2026 -0700" }, "message": "[lldb] Reduce size of Mangled class (#200181)\n\nThe Mangled class is used in several places in LLDB, most notably as a\ndirect member of Symbol. This makes this class one of the most\nfrequently long-lived allocations in LLDB.\n\nIn commit a2672250be871bdac18c1a955265a98704434218 , this class got a\n(large) cache that stores information about demangled data. This cache\nis stored in a std::optional member, which means the memory for the\nclass is allocated within our Mangled object. It should be noted that\nthis cache is only used when we actually demangle the name, which\ndoesn\u0027t happen for every mangled name we encounter.\n\nThe additional cache member caused that the size of Mangled went from\n16B to 152B by default (that is, even if the Mangled name was never\ndemangled).\n\nThis patch replaces the std::optional with a unique_ptr which stores the\ncache on first use in a separate heap allocation. This changes decreases\nthe amount of allocated memory when debugging a relatively small\nObjective-C project from 1.57GiB to 1.18GiB (-400MiB).\n\nGitOrigin-RevId: bd1b3d47462acf4f854f593bdd77b3f127adea46\n" }, { "commit": "c0ce1de0f7f45e4a1daefb92288ee8cf2dec9ecb", "tree": "7db9226d80ed4df132988bf0db2813dc69a53bda", "parents": [ "17140b02b2d8ef261884dc3286428ca86f63732e" ], "author": { "name": "David Spickett", "email": "david.spickett@arm.com", "time": "Mon Jun 01 09:42:54 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 01:45:17 2026 -0700" }, "message": "[lldb][lldb-server][Darwin] Error when gdbserver mode is requested (#199654)\n\nFixes #199035\n\nlldb-server\u0027s platform mode works on Apple platforms but the gdbserver\nmode does not. Users should use debugserver instead, and platform mode\nknows to spawn debugserver instead of lldb-server.\n\nI\u0027m adding an error to state this, because until now it would maybe\nappear to work, or crash in strange ways. None of which can be fixed by\na user and are a waste of our time dealing with the bug reports.\n\nGitOrigin-RevId: 683c367c21d8595a45a97e0439540d3efcd29b25\n" }, { "commit": "17140b02b2d8ef261884dc3286428ca86f63732e", "tree": "1c77dc07e748f965690572b285ee6b289751a899", "parents": [ "52ff58d806966a05a19fa1b7c7852c074c5bbc26" ], "author": { "name": "Felipe de Azevedo Piovezan", "email": "fpiovezan@apple.com", "time": "Mon Jun 01 09:21:59 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 01:25:51 2026 -0700" }, "message": "[lldb] Strip pointer metadata in ReadMemoryRanges (#200398)\n\nThe Process base class is generally responsible for fixing pointer\nmetadata before delegating memory reads to concrete Process\nspecializations. However, ReadMemoryRanges was a direct path into the\nderived classes, which made it so that pointer metadata was never\nstripped.\n\nThis commit creates a non-virtual ReadMemoryRanges in Process, which\nclears pointer metadata, before delegating to the new virtual method\nDoReadMemoryRanges. This also allows, in the future, to plug into the\nmemory cache system.\n\nGitOrigin-RevId: fb8cb1b3d0edd14ff0b40562200b827aca72598f\n" }, { "commit": "52ff58d806966a05a19fa1b7c7852c074c5bbc26", "tree": "455c10031710f7dc6120188f3a93177e865514ea", "parents": [ "91abcc73af0ad40f71365e5a955aff5e3faa9b13" ], "author": { "name": "Pavel Labath", "email": "pavel@labath.sk", "time": "Mon Jun 01 09:31:27 2026 +0200" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Mon Jun 01 00:35:32 2026 -0700" }, "message": "[lldb] Fix CommandInterpreter::GetExecutionContext (#199922)\n\nFollowing up Jonas\u0027s comment in #198949, this patch changes the function\nto not ignore the override context in the adopt_dummy_target\u003dfalse case.\n\nHowever, I don\u0027t implement Jonas\u0027s suggestion exactly either. Instead of\n\"bypassing\" the override context if it contains the dummy target, this\npatch returns an empty context instead. I think this makes more sense,\nas the intention of the user may very well have been to run the command\nin the context of the dummy target. The test suite has no opinion either\nway, but this is sufficient to fix the regression that #198949 was\ntrying to fix.\n\nI also delete the test added in that PR, as the tests in this patch\ncover more cases, and the test has the potential of interfering with\nanother process/test running on the same system.\n\nGitOrigin-RevId: e5ab4f8a1f766febdc65ce89c00dec85393cfd68\n" }, { "commit": "91abcc73af0ad40f71365e5a955aff5e3faa9b13", "tree": "69b909633fca9b3ac55c8020cdcbe057a37b5f82", "parents": [ "0e93544d90cba4aa4ca087d25e1318660aab8307" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Sun May 31 09:54:09 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Sun May 31 09:55:31 2026 -0700" }, "message": "[lldb] Fix up comment placement (NFC) (#200671)\n\nFollow up to https://github.com/llvm/llvm-project/pull/200084\n\nGitOrigin-RevId: c42f1884730fb51a5f802ad621c0cc7f68715424\n" }, { "commit": "0e93544d90cba4aa4ca087d25e1318660aab8307", "tree": "800eafcc5d882b8637d5fa971b299f0cc3319604", "parents": [ "06bfbf5a711792e84efcbb0f936545bc0ba8b333" ], "author": { "name": "Ayush Sahay", "email": "asahay@qti.qualcomm.com", "time": "Sat May 30 12:00:43 2026 +0530" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 23:35:27 2026 -0700" }, "message": "[lldb][Windows] Cache thread context in NativeRegisterContextWindows_arm64 (#197385)\n\nCache thread context in _NativeRegisterContextWindows_arm64_ to improve\nread performance. Previously, the thread context was retrieved for every\nread or write operation.\n\nThis change intends to lay the groundwork for provisioning debug support\nfor SVE on WoA.\n\nAssisted-by: Claude Sonnet 4.6\nGitOrigin-RevId: 3c193d4d88c8c05157b61da3f3cf925eef8d6c5d\n" }, { "commit": "06bfbf5a711792e84efcbb0f936545bc0ba8b333", "tree": "21da834e61eb708fa4bf7ed778dcf4fd91b71e5d", "parents": [ "27e54a32da36f29ae40a384c99cb55b984d4a62d" ], "author": { "name": "Med Ismail Bennani", "email": "ismail@bennani.ma", "time": "Fri May 29 21:01:24 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 21:05:46 2026 -0700" }, "message": "[lldb/test] Trim @expectedFailureAll(remote\u003dTrue) decorators (#200529)\n\nThese three tests pass when run against a remote-darwin platform backed\nby lldb-platform on device. Update each decorator to reflect where it\u0027s\nstill expected to fail rather than blanket-XFAILing every remote run.\n\n- `TestAssertMessages.test_createTestTarget`: was XFAIL on\noslist\u003dno_match([\"linux\"]) + remote\u003dTrue. Add darwin_all to the no_match\nlist so the XFAIL stays only on remote-windows / remote-freebsd /\nremote-netbsd / remote-android.\n- `TestDebuggerAPI.test_CreateTarget_platform`: scope to non-Darwin\nremotes (bug llvm/llvm-project#92419 still tracks the underlying issue\non those platforms).\n- `TestObjcOptimized`: drop @expectedFailureAll(remote\u003dTrue) from the\ntest method and put @skipUnlessDarwin on the class. The Makefile depends\non `-framework Foundation` and `-lobjc`, so the test cannot build on\nnon-Darwin platforms — skip it there outright instead of pretending it\ncould XFAIL.\n\nSigned-off-by: Med Ismail Bennani \u003cismail@bennani.ma\u003e\nGitOrigin-RevId: 524734d88b3cdd874f00853ea6309e70bbb1cd4e\n" }, { "commit": "27e54a32da36f29ae40a384c99cb55b984d4a62d", "tree": "a19043f56e9a438d131e6a77ba25698e675f582c", "parents": [ "616a251209a86bc92465d19942533f96fd1bd53b" ], "author": { "name": "Adrian Prantl", "email": "aprantl@apple.com", "time": "Fri May 29 17:32:36 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 17:35:31 2026 -0700" }, "message": "[LLDB] Simplify the hint when po-ing an object with no object descrip… (#200499)\n\n…tion\n\nThe current wording of the hint is so long that the output obscures the\noutput of the command, which can be confusing. By shortening the message\nthe command output hopefully comes back into the center of attention.\n\nGitOrigin-RevId: f561d59856c1a2a1f7adb9604d47e9e1e940c7a0\n" }, { "commit": "616a251209a86bc92465d19942533f96fd1bd53b", "tree": "750b9cf164923bc0d5859bc8770a51c96fef824e", "parents": [ "55964c5946595ac5ade6a0c4ca1639557988aece" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Fri May 29 16:54:11 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 16:55:32 2026 -0700" }, "message": "[lldb] Fix copy-paste typo in Symbol::operator\u003d (#200528)\n\nGitOrigin-RevId: 20f117f386e53fe348967c355d07842190c001ee\n" }, { "commit": "55964c5946595ac5ade6a0c4ca1639557988aece", "tree": "66cd0c9ff8661fcf6c909b32ae0cd17fd47f741b", "parents": [ "f92f7e3289fa2272878fed39408aa14b1eccfcb1" ], "author": { "name": "Jason Molenda", "email": "jmolenda@apple.com", "time": "Fri May 29 15:59:49 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 16:05:31 2026 -0700" }, "message": "[lldb][Darwin] Correct jGetLoadedDynamicLibrariesInfos typeo \u0026 fallback (#200515)\n\nJonas caught that I had a typeo in checking for the\n`sizeof_mh_and_loadcmds` key in the `jGetLoadedDynamicLibrariesInfos`\nresponse from debugserver in DyanmicLoaderDarwin. Fix that.\n\nAlso I originally picked a fallback value for the mach header + load\ncommands as a guess. I\u0027ve sinced looked at a large UI app\u0027s binaries and\nbased on the size of their actual mh+load commands, picked a default\nthat will read all the data needed in the majority of cases.\n\nrdar://178283767\n\nGitOrigin-RevId: 9b77b22a0d14b08fb4d3b217f323449911ef8213\n" }, { "commit": "f92f7e3289fa2272878fed39408aa14b1eccfcb1", "tree": "d06113be3c8a4e7af536c05f460fad4a015cca4e", "parents": [ "01c7ad622af6ed8348d07b216f02a9ca5f9ea907" ], "author": { "name": "satyanarayana reddy janga", "email": "satyajanga@fb.com", "time": "Fri May 29 17:33:55 2026 -0500" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 15:36:05 2026 -0700" }, "message": "[lldb-server] Add accelerator plugin infrastructure for debugging hardware accelerators like gpus (#198907)\n\nThis is the first patch of many related to\nhttps://discourse.llvm.org/t/upstreaming-basic-support-for-accelerators/89827/6\n\n### What this patch adds\n\n - **`LLDBServerAcceleratorPlugin`** base class in\n`source/Plugins/Process/gdb-remote/` so accelerator can implement the\nown plugin\n - **`accelerator-plugins+`** feature in `qSupported` response, only\n advertised when at least one plugin is installed\n- **`jAcceleratorPluginInitialize`** GDB remote packet and its\nimplementation in handlers, request and response.\n- **`AcceleratorActions`** struct so every plugin can return the actions\nthat needs to be on the initilaize. in the future we will extend this\ninstall breakpoints etc.\n - **Mock accelerator plugin** for testing, gated by CMake option\n `LLDB_ENABLE_MOCK_ACCELERATOR_PLUGIN` (default OFF)\n - **Tests** that connect to a real lldb-server, verify\n `accelerator-plugins+` in `qSupported`, send\n `jAcceleratorPluginInitialize`, and validate the JSON response\n\n ### Design decisions\n - CMake option defaults to OFF so normal builds are unaffected\n - Tests skip automatically when the plugin is not compiled in\n\nGitOrigin-RevId: 3d24f9acc92bc3288efa8e17eee6ae41d694677f\n" }, { "commit": "01c7ad622af6ed8348d07b216f02a9ca5f9ea907", "tree": "39af90c9e69db14043fb3b60535e6a82a7196d45", "parents": [ "1e436bc88f951bda11941cb11a85308563ba2e6b" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Fri May 29 15:25:52 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 15:30:24 2026 -0700" }, "message": "[lldb] Halve the size of indexes in DemangledNameInfo (NFC) (#200506)\n\nMangled should not be more than 4GB. This will halve the size of\n`DemangledNameInfo` from 128 to 64 bytes.\n\nGitOrigin-RevId: 9b5d0af42f8d5f10f6ca566eeebf162dc22abe7b\n" }, { "commit": "1e436bc88f951bda11941cb11a85308563ba2e6b", "tree": "483bcf9ee4c87e6ef6366a874e4c6c39d88c8f85", "parents": [ "6184d330a223ecf632ecfd0c268d2a067f923095" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Fri May 29 11:49:00 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 11:50:18 2026 -0700" }, "message": "[lldb][test] Link test binaries with -random_uuid (#199385)\n\nLink using `-random_uuid` on macOS to avoid accidental UUID matching in tests.\n\nGitOrigin-RevId: 352fd0bd9b8ee5f89ec1fe508d5aa35525cd94e9\n" }, { "commit": "6184d330a223ecf632ecfd0c268d2a067f923095", "tree": "8e4113cf0315d7836a2c14b622746f4538000383", "parents": [ "13d158f4604e21cba03b0a667ab181b2a84f4dd4" ], "author": { "name": "Alex Langford", "email": "alangford@apple.com", "time": "Fri May 29 10:11:15 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 10:15:19 2026 -0700" }, "message": "[lldb] Tighten TestObjcMethods test case for arm64e (#200262)\n\nThis test explicitly checks for spurious DWARF lookups by looking for\nthe string `$__lldb` in the dwarf lookup logs. The intent here is to\nvalidate that any calls to `FindFunctions` or `FindTypes` won\u0027t have\n`$__lldb` in its lookup target. On arm64, this is trivial because `expr\nself` goes through the IRInterpreter.\n\nHowever, on arm64e, `expr self` is JIT compiled and executed in the\ninferior process. LLDB installs some utility functions to check\ninvariants and those functions are usually prefixed with `$__lldb`. When\nthose utility functions appear in the logs, the test incorrectly fails.\n\nGitOrigin-RevId: 8a3ed55e56dc24cfa7d78875b89f0911a791dc55\n" }, { "commit": "13d158f4604e21cba03b0a667ab181b2a84f4dd4", "tree": "af4e95bade7c6fc0b9bcb26a3ce5a181979d879e", "parents": [ "bb6671ce346a5a34422d0c06dfd8992824a5e80a" ], "author": { "name": "Adrian Prantl", "email": "aprantl@apple.com", "time": "Fri May 29 09:50:50 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 09:55:41 2026 -0700" }, "message": "[LLDB] Skip test with older versions of clang (#200456)\n\nGitOrigin-RevId: 4472e935678ae8aba762b2464867f9e3fe81ca15\n" }, { "commit": "bb6671ce346a5a34422d0c06dfd8992824a5e80a", "tree": "42dfbca75421b3564dbcbfe69a552a414b584b69", "parents": [ "5dfdc059f6d4178b35e5c5fb01c40c8f912d87fb" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri May 29 17:08:21 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 09:11:16 2026 -0700" }, "message": "[lldb][windows] fix scope_exit warning (#200439)\n\nUsing the constructor rather than `make_scope_exit` is preferred.\n\nGitOrigin-RevId: 7dcce4c16da12365709adc685679df70572ac49e\n" }, { "commit": "5dfdc059f6d4178b35e5c5fb01c40c8f912d87fb", "tree": "8f4892b007157cc1260eb8e2d5b04c82775788eb", "parents": [ "160caaff28b9211c5481a8b663bee3c2d581e8a5" ], "author": { "name": "Dave Lee", "email": "davelee.com@gmail.com", "time": "Fri May 29 09:02:33 2026 -0700" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 09:05:19 2026 -0700" }, "message": "[lldb] Avoid trie prefix recompuation (NFC) (#200026)\n\nAvoid O(depth) string concatenations for every symbol in the Mach-O\nexports trie.\n\nThis change uses a string to track the current trie prefix, instead of a\nvector of substrings. Instead of pushing and popping on the vector, the\nstring is appended and truncated.\n\nAssisted-by: claude\nGitOrigin-RevId: d043ec846b79e30b8de88e601e8c9e36eee771eb\n" }, { "commit": "160caaff28b9211c5481a8b663bee3c2d581e8a5", "tree": "68d771de5a177d19b879d6b61624fd15965d3e2b", "parents": [ "e6ed12f67c92ae5f714220b6949b13d9339ff039" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri May 29 14:09:10 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 06:10:20 2026 -0700" }, "message": "[lldb][windows] enumerate volumes via FindFirstVolumeW (#200231)\n\nGitOrigin-RevId: 4b703fb5fbb9e517eb205f310a129472f3cac655\n" }, { "commit": "e6ed12f67c92ae5f714220b6949b13d9339ff039", "tree": "68bfcc1721bc0b6915e49499432e9a4e56d355b2", "parents": [ "0a7286137cc7c7ca3c7726bd7827e563e5209d4a" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri May 29 14:02:19 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 06:05:03 2026 -0700" }, "message": "[lldb] Make PTY initialization idempotent (#197717)\n\nGitOrigin-RevId: f545a3b3f558a29e2d7c2dc18748e8bedc3ce6c6\n" }, { "commit": "0a7286137cc7c7ca3c7726bd7827e563e5209d4a", "tree": "f2fb760da826536f224a41662b86ee8e06aedafc", "parents": [ "200913217b5b3ecafd1e7d08b0d4515dad6794c3" ], "author": { "name": "Charles Zablit", "email": "c_zablit@apple.com", "time": "Fri May 29 13:47:20 2026 +0100" }, "committer": { "name": "Copybara-Service", "email": "copybara-worker@google.com", "time": "Fri May 29 05:50:33 2026 -0700" }, "message": "[lldb][windows] fix command source hitting EOF (#194950)\n\nGitOrigin-RevId: a182b4bed2c443f383258fe7dcd886b39624526f\n" } ], "next": "200913217b5b3ecafd1e7d08b0d4515dad6794c3" }