about summary refs log tree commit diff
path: root/compiler/rustc_llvm/llvm-wrapper
AgeCommit message (Collapse)AuthorLines
2024-04-09Set target-abi module flag for RISC-V targetskxxt-1/+10
Fixes cross-language LTO on RISC-V targets (Fixes #121924)
2024-04-08Rollup merge of #123591 - Zalathar:useless-cast, r=cuviperMatthias Krüger-2/+2
Remove unnecessary cast from `LLVMRustGetInstrProfIncrementIntrinsic` (Noticed while reviewing #123409.) This particular cast appears to have been copied over from clang, but there are plenty of other call sites in clang that don't bother with a cast here, and it works fine without one. For context, `llvm::Intrinsic::ID` is a typedef for `unsigned`, and `llvm::Intrinsic::instrprof_increment` is a member of `enum IndependentIntrinsics : unsigned`. --- The formatting change in `unwrap(M)` is the result of manually running `clang-format` on this file, and then reverting all changes other than the ones affecting these lines.
2024-04-08Rollup merge of #122807 - danielhuang:fix-1, r=davidtwcoMatthias Krüger-1/+1
Add consistency with phrases "meantime" and "mean time" "mean time" is used in a few places while "meantime" is used everywhere else; this would make usage consistent throughout the codebase.
2024-04-07Remove unnecessary cast from `LLVMRustGetInstrProfIncrementIntrinsic`Zalathar-2/+2
This particular cast appears to have been copied over from clang, but there are plenty of other call sites in clang that don't bother with a cast here, and it works fine without one. For context, `llvm::Intrinsic::ID` is a typedef for `unsigned`, and `llvm::Intrinsic::instrprof_increment` is a member of `enum IndependentIntrinsics : unsigned`.
2024-04-04Rollup merge of #123437 - Zalathar:clang-format, r=cuviperJacob Pratt-28/+20
Manually run `clang-format` on `CoverageMappingWrapper.cpp` In the current version of #123409, there are several unrelated changes to `CoverageMappingWrapper.cpp` that seem to be the result of running `clang-format` on that file. Instead of asking for those changes to be undone, I figure it's easier to just make them myself as a separate PR, since I was vaguely intending to do that at some point anyway. In a few cases I've strategically added comments to make the grouping of parameters a little nicer, but mostly it doesn't matter much.
2024-04-04Manually run `clang-format` on `CoverageMappingWrapper.cpp`Zalathar-28/+20
2024-04-03update messagesDan-1/+1
2024-04-03coverage: Correctly report and check LLVM's coverage mapping versionZalathar-1/+4
2024-03-26RustWrapper: update call for ↵Augie Fackler-0/+4
llvm/llvm-project@44d037cc258dcf179d2c48c93996bb406ecd0fae Easy change. @rustbot label: +llvm-main
2024-03-17Update the minimum external LLVM to 17Josh Stone-52/+4
2024-03-15Install the bad-alloc handler before fatal errorsJosh Stone-1/+1
The bad-alloc installer was incorrectly asserting that the other handler isn't set yet, instead of checking its own, but we can avoid that by changing the order we install them. Ref: https://github.com/llvm/llvm-project/issues/83040
2024-03-15Aggressively ignore write errors during bad-allocJosh Stone-3/+3
2024-03-15Register LLVM handlers for bad-alloc / OOMJosh Stone-1/+24
LLVM's default bad-alloc handler may throw if exceptions are enabled, and `operator new` isn't hooked at all by default. Now we register our own handler that prints a message similar to fatal errors, then aborts. We also call the function that registers the C++ `std::new_handler`.
2024-03-12llvm-wrapper: adapt for LLVM API changesKrasimir Georgiev-2/+7
Adapts rust for https://github.com/llvm/llvm-project/commit/9997e0397156ff7e01aecbd17bdeb7bfe5fb15b0.
2024-03-10Fix 32-bit overflows in LLVM composite constantserer1243-8/+34
2024-03-08Rollup merge of #122143 - durin42:llvm-19-compression-options, r=workingjubileeMatthias Krüger-0/+16
PassWrapper: update for llvm/llvm-project@a3319371970b ``@rustbot`` label: +llvm-main
2024-03-08Rollup merge of #119365 - nbdd0121:asm-goto, r=AmanieuMatthias Krüger-0/+25
Add asm goto support to `asm!` Tracking issue: #119364 This PR implements asm-goto support, using the syntax described in "future possibilities" section of [RFC2873](https://rust-lang.github.io/rfcs/2873-inline-asm.html#asm-goto). Currently I have only implemented the `label` part, not the `fallthrough` part (i.e. fallthrough is implicit). This doesn't reduce the expressive though, since you can use label-break to get arbitrary control flow or simply set a value and rely on jump threading optimisation to get the desired control flow. I can add that later if deemed necessary. r? ``@Amanieu`` cc ``@ojeda``
2024-03-07PassWrapper: update for llvm/llvm-project@a3319371970bAugie Fackler-0/+16
@rustbot label: +llvm-main
2024-03-07Rollup merge of #122062 - workingjubilee:initialize-my-fist, r=cuviperMatthias Krüger-52/+53
Explicitly assign constructed C++ classes C++ style guides I am aware of recommend specifically preferring = syntax for any classes with fairly obvious constructors[^0] that do not perform any complicated logic in their constructor. I contend that all constructors that the `rustc_llvm` code uses qualify. This has only become more common since C++ 17 guaranteed many cases of copy initialization elision. The other detail is that I tried to ask another contributor with infinitely more C++ experience than me (i.e. any) what this constructor syntax was, and they thought it was a macro. I know of no other language that has adopted this same syntax. As the rustc codebase features many contributors experienced in many other languages, using a less... unique... style has many other benefits in making this code more lucid and maintainable, which is something it direly needs. [^0]: e.g. https://abseil.io/tips/88
2024-03-05Explicitly assign constructed C++ classesJubilee Young-52/+53
C++ style guides I am aware of recommend specifically preferring = syntax for any classes with fairly obvious constructors[^0] that do not perform any complicated logic in their constructor. I contend that all constructors that the `rustc_llvm` code uses qualify. This has only become more common since C++ 17 guaranteed many cases of copy initialization elision. The other detail is that I tried to ask another contributor with infinitely more C++ experience than me (i.e. any) what this constructor syntax was, and they thought it was a macro. I know of no other language that has adopted this same syntax. As the rustc codebase features many contributors experienced in many other languages, using a less... unique... style has many other benefits in making this code more lucid and maintainable, which is something it direly needs. [^0]: e.g. https://abseil.io/tips/88
2024-03-05Clarify FatalErrorHandlerJubilee Young-2/+12
Clarify the FatalErrorHandler API that we use: - Identify rustc's LLVM ERRORs by prefixing them - Comment heavily on its interior, while we are here
2024-03-01Add initial support for DataFlowSanitizerRamon de C Valle-0/+16
Adds initial support for DataFlowSanitizer to the Rust compiler. It currently supports `-Zsanitizer-dataflow-abilist`. Additional options for it can be passed to LLVM command line argument processor via LLVM arguments using `llvm-args` codegen option (e.g., `-Cllvm-args=-dfsan-combine-pointer-labels-on-load=false`).
2024-02-26Rollup merge of #121389 - klensy:llvm-warn-fix, r=nikicMatthias Krüger-2/+2
llvm-wrapper: fix few warnings Two fixes: first one is simple unsigned -> uint64_t, but how second one is more subtile, see commit description.
2024-02-24Add callbr support to LLVM wrapperGary Guo-0/+25
2024-02-21remove simd_reduce_{min,max}_nanlessRalf Jung-7/+6
2024-02-21make simd_reduce_{mul,add}_unordered use only the 'reassoc' flag, not all ↵Ralf Jung-0/+15
fast-math flags
2024-02-21llvm-wrapper: fix warning C4305klensy-1/+1
llvm-wrapper/ArchiveWrapper.cpp(70): warning C4305: 'argument': truncation from 'int' to 'bool' while in llvm 12 signature was static ErrorOr<std::unique_ptr<MemoryBuffer>> getFile(const Twine &Filename, int64_t FileSize = -1, bool RequiresNullTerminator = true, bool IsVolatile = false); https://github.com/llvm/llvm-project/blame/fed41342a82f5a3a9201819a82bf7a48313e296b/llvm/include/llvm/Support/MemoryBuffer.h#L85-L87 in llvm 13 and later it was changed to static ErrorOr<std::unique_ptr<MemoryBuffer>> getFile(const Twine &Filename, bool IsText = false, bool RequiresNullTerminator = true, bool IsVolatile = false); https://github.com/llvm/llvm-project/blame/75e33f71c2dae584b13a7d1186ae0a038ba98838/llvm/include/llvm/Support/MemoryBuffer.h#L86-L88 so code was interpreted as MemoryBuffer::getFile(Path, /*IsText*/true, /*RequiresNullTerminator=*/false), but now will be MemoryBuffer::getFile(Path, /*IsText*/false, /*RequiresNullTerminator=*/false). How that worked before?
2024-02-21llvm-wrapper: fix warning C4244klensy-1/+1
llvm-wrapper/RustWrapper.cpp(1234): warning C4244: '=': conversion from 'uint64_t' to 'unsigned int', possible loss of data nice consistency: uint64_t https://github.com/llvm/llvm-project/blob/6009708b4367171ccdbf4b5905cb6a803753fe18/llvm/include/llvm/IR/DiagnosticInfo.h#L172 but unsigned https://github.com/llvm/llvm-project/blob/6009708b4367171ccdbf4b5905cb6a803753fe18/llvm/include/llvm/IR/DiagnosticInfo.h#L1091
2024-02-20Add "algebraic" versions of the fast-math intrinsicsBen Kimock-1/+24
2024-02-13Auto merge of #121036 - matthiaskrgr:rollup-ul05q8e, r=matthiaskrgrbors-1/+1
Rollup of 8 pull requests Successful merges: - #114877 (unstable-book: add quick-edit link) - #120548 (rustdoc: Fix handling of doc_auto_cfg feature for cfg attributes on glob reexport) - #120549 (modify alias-relate to also normalize ambiguous opaques) - #120959 (Remove good path delayed bugs) - #120978 (match lowering: simplify block creation) - #121019 (coverage: Simplify some parts of the coverage span refiner) - #121021 (Extend intra-doc link chapter in the rustdoc book) - #121031 (RustWrapper: adapt for coverage mapping API changes) Failed merges: - #121014 (Remove `force_print_diagnostic`) r? `@ghost` `@rustbot` modify labels: rollup
2024-02-13Auto merge of #120055 - nikic:llvm-18, r=cuviperbors-4/+4
Update to LLVM 18 LLVM 18 final is planned to be released on Mar 5th. Rust 1.78 is planned to be released on May 2nd. Tested images: dist-x86_64-linux, dist-s390x-linux, dist-aarch64-linux, dist-riscv64-linux, dist-loongarch64-linux, dist-x86_64-freebsd, dist-x86_64-illumos, dist-x86_64-musl, x86_64-linux-integration, test-various, armhf-gnu, i686-msvc, x86_64-msvc, i686-mingw, x86_64-mingw, x86_64-apple-1, x86_64-apple-2, dist-aarch64-apple r? `@ghost`
2024-02-13RustWrapper: adapt for coverage mapping API changesTim Neumann-1/+1
2024-02-13Use MCSubtargetInfo::getAllProcessorFeatures()Nikita Popov-4/+4
This method is now available in upstream LLVM \o/
2024-02-13Rollup merge of #120995 - durin42:llvm-19-pgo-coldfuncopt, r=cuviperMatthias Krüger-0/+15
PassWrapper: adapt for llvm/llvm-project@93cdd1b5cfa3735c Should be no functional change. `@rustbot` label: +llvm-main
2024-02-12PassWrapper: adapt for ↵Augie Fackler-0/+15
llvm/llvm-project@93cdd1b5cfa3735c599949b77e24dbfbe570441a Should be no functional change. @rustbot label: +llvm-main
2024-02-12llvm-wrapper: adapt for LLVM API change: Add support for EXPORTAS name typesHans Wennborg-0/+3
Adapt for llvm/llvm-project@8f23464.
2024-02-06Rollup merge of #120502 - clubby789:remove-ffi-returns-twice, r=compiler-errorsMatthias Krüger-3/+0
Remove `ffi_returns_twice` feature The [tracking issue](https://github.com/rust-lang/rust/issues/58314) and [RFC](https://github.com/rust-lang/rfcs/pull/2633) have been closed for a couple of years. There is also an attribute gate in R-A which should be removed if this lands.
2024-02-06reviewklensy-6/+2
2024-02-06llvm-wrapper: remove llvm 12 hackklensy-6/+1
effectively reverts https://github.com/rust-lang/rust/commit/9a8acea78355b604dbeb29bc38bd4dbf7bfce95f
2024-01-30Remove `ffi_returns_twice` featureclubby789-3/+0
2024-01-12Revert "Auto merge of #113923 - DianQK:restore-no-builtins-lto, r=pnkfelix"DianQK-2/+7
This reverts commit 8c2b57721728233e074db69d93517614de338055, reversing changes made to 9cf18e98f82d85fa41141391d54485b8747da46f.
2024-01-05Pass LLVM error message back to pass wrapper.Ao Li-1/+3
2023-12-18Rollup merge of #118941 - krasimirgg:llvm-cov, r=nikicMatthias Krüger-0/+3
llvm-wrapper: adapt for LLVM API changes Adapt for https://github.com/llvm/llvm-project/commit/8ecbb0404d740d1ab173554e47cef39cd5e3ef8c. r? `@nikic` `@rustbot` label: +llvm-main
2023-12-18llvm-wrapper: adapt for LLVM API changesKrasimir Georgiev-0/+3
Adapt for https://github.com/llvm/llvm-project/commit/8ecbb0404d740d1ab173554e47cef39cd5e3ef8c.
2023-12-16Auto merge of #110494 - majaha:noTrapAfterNoreturn, r=nikicbors-0/+8
Use the LLVM option NoTrapAfterNoreturn Use this LLVM option: https://llvm.org/doxygen/classllvm_1_1TargetOptions.html#acd83fce25de1ac9f6c975135a8235c22 when TrapUnreachable is enabled. This prevents codegenning unnecessary double-traps in some situations. See further discussion here: https://github.com/rust-lang/compiler-team/issues/618
2023-12-12llvm-wrapper: adapt for LLVM API changeKrasimir Georgiev-1/+3
LLVM commit https://github.com/llvm/llvm-project/commit/f09cf34d00625e57dea5317a3ac0412c07292148 moved some functions to a different header: https://buildkite.com/llvm-project/rust-llvm-integrate-prototype/builds/24416#018c5de6-b9c9-4b22-9473-6070d99dcfa7/233-537
2023-12-11Auto merge of #117962 - weihanglo:debug-name-table, r=wesleywiserbors-2/+23
fix: stop emitting `.debug_pubnames` and `.debug_pubtypes` A continuation of #94181. Fixes #48762 MCP can be found in <https://github.com/rust-lang/compiler-team/issues/688>. `.debug_pubnames` and `.debug_pubtypes` are poorly designed and people seldom use them. However, they take a considerable portion of size in the final binary. This tells LLVM stop emitting those sections on DWARFv4 or lower. DWARFv5 use `.debug_names` which is more concise in size and performant for name lookup. Some other no-really-useful personal notes: <details><summary>Details</summary> <p> ## Pepole saying they are not useful * https://github.com/rust-lang/rust/issues/48762 * https://rust-lang.zulipchat.com/#narrow/stream/317568-t-compiler.2Fwg-debugging/topic/investigating.20debuginfo.20size/near/342713604 * `DwarfCompileUnit::hasDwarfPubSections()` — https://github.com/llvm/llvm-project/blob/f633f325a1b808d33ca9653ed373353549ddcde6/llvm/lib/CodeGen/AsmPrinter/DwarfCompileUnit.cpp#L1477-L1494 * clang default to no debug name table when no option provided — https://github.com/llvm/llvm-project/blob/f633f325a1b808d33ca9653ed373353549ddcde6/clang/lib/Frontend/CompilerInvocation.cpp#L1819-L1824 * GCC explicitly says GDB doesn't use pub sections (`TARGET_WANT_DEBUG_PUB_SECTIONS` only be true on Darwin) — https://github.com/gcc-mirror/gcc/blob/5d2a360f0a541646abb11efdbabc33c6a04de7ee/gcc/target.def#L6985-L6990 and https://github.com/bminor/binutils-gdb/blob/319b460545dc79280e2904dcc280057cf71fb753/gold/dwarf_reader.h#L424-L427 * Probably the only place that makes use of pub section in lldb — https://github.com/llvm/llvm-project/blob/725115d7bba2faf3d0c21442f4661dea77b8a77c/lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp#L2117-L2135 * "The -gsplit-dwarf option requires -ggnu-pubnames." — https://github.com/gcc-mirror/gcc/blob/5d2a360f0a541646abb11efdbabc33c6a04de7ee/gcc/opts.cc#L1205 * LLVM: Always emit `.debug_names` with dwarf 5 for Apple platforms — https://reviews.llvm.org/D118754 </p> </details>
2023-12-11fix: stop emitting `.debug_pubnames` and `.debug_pubtypes`Weihang Lo-2/+23
`.debug_pubnames` and `.debug_pubtypes` are poorly designed and people seldom use them. However, they take a considerable portion of size in the final binary. This tells LLVM stop emitting those sections on DWARFv4 or lower. DWARFv5 use `.debug_names` which is more concise in size and performant for name lookup.
2023-12-11llvm-wrapper: adapt for LLVM API changeKrasimir Georgiev-0/+4
LLVM commit https://github.com/llvm/llvm-project/commit/1d608fc755a3e15d0020f61c9535c9b730ab9dec renamed the pass.
2023-12-09Auto merge of #117873 - quininer:android-emutls, r=Amanieubors-7/+3
Add emulated TLS support This is a reopen of https://github.com/rust-lang/rust/pull/96317 . many android devices still only use 128 pthread keys, so using emutls can be helpful. Currently LLVM uses emutls by default for some targets (such as android, openbsd), but rust does not use it, because `has_thread_local` is false. This commit has some changes to allow users to enable emutls: 1. add `-Zhas-thread-local` flag to specify that std uses `#[thread_local]` instead of pthread key. 2. when using emutls, decorate symbol names to find thread local symbol correctly. 3. change `-Zforce-emulated-tls` to `-Ztls-model=emulated` to explicitly specify whether to generate emutls. r? `@Amanieu`
generated by cgit-pink 1.4.1-3-g733a5 (git 2.37.1) at 2026-07-02 17:51:06 +0000