Release 0.9.1

Signed-off-by: Tw <wei.tan@intel.com>

.builds/freebsd.yml

@@ -1,7 +1,7 @@
 image: freebsd/latest
 secrets:
   - 51bfddf5-86a6-4e01-8576-358c72a4a0a4
-  - 5cfede76-914e-4071-893e-e5e2e6ae3cea
+  - 512ed797-0927-475a-83fd-bc997792860c
 sources:
   - https://github.com/ziglang/zig
 tasks:

.gitattributes

@@ -9,3 +9,4 @@ lib/libc/** linguist-vendored
 lib/libcxx/** linguist-vendored
 lib/libcxxabi/** linguist-vendored
 lib/libunwind/** linguist-vendored
+lib/tsan/** linguist-vendored

.github/ISSUE_TEMPLATE/bug.yml

@@ -0,0 +1,40 @@
+name: Bug Report
+description: File a bug report
+labels: ["bug"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        A bug is when something works differently than it is expected to.
+        ## Remember to search before filing a new report
+        Please search for this bug in the issue tracker, and use a bug report title that
+        would have made your bug report turn up in the search results for your search query.
+  - type: input
+    id: version
+    attributes:
+      label: Zig Version
+      description: "The output of `zig version`"
+      placeholder: "0.9.0-dev.1275+ac52e0056"
+    validations:
+      required: true
+  - type: textarea
+    id: repro
+    attributes:
+      label: Steps to Reproduce
+      description: What exactly can someone else do, in order to observe the problem that you observed?
+    validations:
+      required: true
+  - type: textarea
+    id: expected
+    attributes:
+      label: Expected Behavior
+      description: What did you expect to happen?
+    validations:
+      required: true
+  - type: textarea
+    id: actual
+    attributes:
+      label: Actual Behavior
+      description: What happened instead? Be sure to include all error messages if any.
+    validations:
+      required: true

.github/ISSUE_TEMPLATE/proposal.yml

@@ -0,0 +1,16 @@
+name: Language Proposal
+description: Propose to improve the Zig language
+labels: ["proposal"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thank you for your interest in improving the Zig language. However, we are
+        not accepting new proposals to change the language at this time.
+  - type: checkboxes
+    id: trash
+    attributes:
+      label: Please do not file a proposal to change the language
+      options:
+        - label: "I understand, thank you. I will not submit a new proposal at this time"
+          required: true

.github/ISSUE_TEMPLATE/question.yml

@@ -0,0 +1,20 @@
+name: Question
+description: Ask a Zig-related question
+labels: ["question"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Welcome! There are a bunch of great places to ask Zig-related questions.
+        Please take a look at
+        [The Community Wiki Page](https://github.com/ziglang/zig/wiki/Community) and
+        find a comfy place to ask questions. You will find plenty of helpful people in
+        these spaces. However, this issue tracker is not for questions. It is for
+        more actionable items such as bug reports and enhancements.
+  - type: checkboxes
+    id: trash
+    attributes:
+      label: Please do not open a question issue on the bug tracker
+      options:
+        - label: "I understand, thank you. I will take my question to one of the community spaces instead"
+          required: true

CMakeLists.txt

@@ -25,8 +25,8 @@ project(zig C CXX)
 set(CMAKE_MODULE_PATH ${CMAKE_SOURCE_DIR}/cmake ${CMAKE_MODULE_PATH})
 
 set(ZIG_VERSION_MAJOR 0)
-set(ZIG_VERSION_MINOR 8)
-set(ZIG_VERSION_PATCH 0)
+set(ZIG_VERSION_MINOR 9)
+set(ZIG_VERSION_PATCH 1)
 set(ZIG_VERSION "" CACHE STRING "Override Zig version string. Default is to find out with git.")
 
 if("${ZIG_VERSION}" STREQUAL "")
@@ -91,7 +91,12 @@ set(ZIG_TARGET_MCPU "baseline" CACHE STRING "-mcpu parameter to output binaries
 set(ZIG_EXECUTABLE "" CACHE STRING "(when cross compiling) path to already-built zig binary")
 set(ZIG_SINGLE_THREADED off CACHE BOOL "limit the zig compiler to use only 1 thread")
 set(ZIG_OMIT_STAGE2 off CACHE BOOL "omit the stage2 backend from stage1")
-set(ZIG_ENABLE_LOGGING off CACHE BOOL "enable logging")
+
+if("${CMAKE_BUILD_TYPE}" STREQUAL "Debug")
+    set(ZIG_ENABLE_LOGGING ON CACHE BOOL "enable logging")
+else()
+    set(ZIG_ENABLE_LOGGING OFF CACHE BOOL "enable logging")
+endif()
 
 if("${ZIG_TARGET_TRIPLE}" STREQUAL "native")
     set(ZIG_USE_LLVM_CONFIG ON CACHE BOOL "use llvm-config to find LLVM libraries")
@@ -99,9 +104,9 @@ else()
     set(ZIG_USE_LLVM_CONFIG OFF CACHE BOOL "use llvm-config to find LLVM libraries")
 endif()
 
-find_package(llvm)
-find_package(clang)
-find_package(lld)
+find_package(llvm 13)
+find_package(clang 13)
+find_package(lld 13)
 
 if(ZIG_STATIC_ZLIB)
     list(REMOVE_ITEM LLVM_LIBRARIES "-lz")
@@ -111,7 +116,10 @@ endif()
 
 if(APPLE AND ZIG_STATIC)
     list(REMOVE_ITEM LLVM_LIBRARIES "-lcurses")
-    find_library(CURSES NAMES libcurses.a curses libcurses libncurses.a ncurses libncurses)
+    find_library(CURSES NAMES libcurses.a libncurses.a
+      PATHS
+        /usr/local/opt/ncurses/lib
+        /opt/homebrew/opt/ncurses/lib)
     list(APPEND LLVM_LIBRARIES "${CURSES}")
 endif()
 
@@ -358,7 +366,9 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/lib/std/crypto/siphash.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/debug.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/dwarf.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/dwarf_bits.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/dwarf/AT.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/dwarf/OP.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/dwarf/TAG.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/elf.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/event.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/event/batch.zig"
@@ -419,18 +429,13 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/lib/std/meta/trait.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/multi_array_list.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux/errno-generic.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux/netlink.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux/prctl.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux/securebits.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/bits/linux/x86_64.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/os/linux.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/os/linux/errno/generic.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/os/linux/x86_64.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/linux.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/linux/io_uring.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/linux/x86_64.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/windows.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/os/windows/bits.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/windows/ntstatus.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/os/windows/win32error.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Progress.zig"
@@ -441,9 +446,11 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/addXf3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/atomics.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/bswap.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/clear_cache.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/clzsi2.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/cmp.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/compareXf2.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/count0bits.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/divdf3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/divsf3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/divtf3.zig"
@@ -488,11 +495,13 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/modti3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/mulXf3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/muldi3.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/mulodi4.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/muloti4.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/mulo.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/multi3.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/negXf2.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/popcountdi2.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/negXi2.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/os_version_check.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/parity.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/popcount.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/shift.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/stack_probe.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/special/compiler_rt/truncXfYf2.zig"
@@ -520,76 +529,85 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/lib/std/target/x86.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Thread.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Thread/AutoResetEvent.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/Thread/Futex.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Thread/Mutex.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Thread/ResetEvent.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/Thread/StaticResetEvent.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/time.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/unicode.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/zig/ast.zig"
-    "${CMAKE_SOURCE_DIR}/lib/std/zig/cross_target.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/zig/Ast.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/zig/CrossTarget.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/parse.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/render.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/string_literal.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/system.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/zig/system/NativePaths.zig"
+    "${CMAKE_SOURCE_DIR}/lib/std/zig/system/NativeTargetInfo.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/system/x86.zig"
     "${CMAKE_SOURCE_DIR}/lib/std/zig/tokenizer.zig"
+    "${CMAKE_SOURCE_DIR}/src/Air.zig"
+    "${CMAKE_SOURCE_DIR}/src/AstGen.zig"
     "${CMAKE_SOURCE_DIR}/src/Cache.zig"
     "${CMAKE_SOURCE_DIR}/src/Compilation.zig"
     "${CMAKE_SOURCE_DIR}/src/DepTokenizer.zig"
+    "${CMAKE_SOURCE_DIR}/src/Liveness.zig"
     "${CMAKE_SOURCE_DIR}/src/Module.zig"
     "${CMAKE_SOURCE_DIR}/src/Package.zig"
     "${CMAKE_SOURCE_DIR}/src/RangeSet.zig"
+    "${CMAKE_SOURCE_DIR}/src/Sema.zig"
     "${CMAKE_SOURCE_DIR}/src/ThreadPool.zig"
     "${CMAKE_SOURCE_DIR}/src/TypedValue.zig"
     "${CMAKE_SOURCE_DIR}/src/WaitGroup.zig"
-    "${CMAKE_SOURCE_DIR}/src/AstGen.zig"
+    "${CMAKE_SOURCE_DIR}/src/Zir.zig"
+    "${CMAKE_SOURCE_DIR}/src/arch/aarch64/bits.zig"
+    "${CMAKE_SOURCE_DIR}/src/arch/arm/bits.zig"
+    "${CMAKE_SOURCE_DIR}/src/arch/riscv64/bits.zig"
+    "${CMAKE_SOURCE_DIR}/src/arch/x86_64/bits.zig"
+    "${CMAKE_SOURCE_DIR}/src/arch/wasm/CodeGen.zig"
     "${CMAKE_SOURCE_DIR}/src/clang.zig"
     "${CMAKE_SOURCE_DIR}/src/clang_options.zig"
     "${CMAKE_SOURCE_DIR}/src/clang_options_data.zig"
     "${CMAKE_SOURCE_DIR}/src/codegen.zig"
-    "${CMAKE_SOURCE_DIR}/src/codegen/aarch64.zig"
-    "${CMAKE_SOURCE_DIR}/src/codegen/arm.zig"
     "${CMAKE_SOURCE_DIR}/src/codegen/c.zig"
     "${CMAKE_SOURCE_DIR}/src/codegen/llvm.zig"
     "${CMAKE_SOURCE_DIR}/src/codegen/llvm/bindings.zig"
-    "${CMAKE_SOURCE_DIR}/src/codegen/riscv64.zig"
-    "${CMAKE_SOURCE_DIR}/src/codegen/wasm.zig"
-    "${CMAKE_SOURCE_DIR}/src/codegen/x86_64.zig"
     "${CMAKE_SOURCE_DIR}/src/glibc.zig"
     "${CMAKE_SOURCE_DIR}/src/introspect.zig"
-    "${CMAKE_SOURCE_DIR}/src/air.zig"
     "${CMAKE_SOURCE_DIR}/src/libc_installation.zig"
     "${CMAKE_SOURCE_DIR}/src/libcxx.zig"
     "${CMAKE_SOURCE_DIR}/src/libtsan.zig"
     "${CMAKE_SOURCE_DIR}/src/libunwind.zig"
     "${CMAKE_SOURCE_DIR}/src/link.zig"
     "${CMAKE_SOURCE_DIR}/src/link/C.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/C/zig.h"
     "${CMAKE_SOURCE_DIR}/src/link/Coff.zig"
     "${CMAKE_SOURCE_DIR}/src/link/Elf.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/Archive.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/MachO/Atom.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/CodeSignature.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/DebugSymbols.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/Dylib.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/Object.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/Symbol.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/Trie.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/Zld.zig"
     "${CMAKE_SOURCE_DIR}/src/link/MachO/bind.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/commands.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/reloc.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/reloc/aarch64.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/MachO/reloc/x86_64.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/Plan9.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/Plan9/aout.zig"
     "${CMAKE_SOURCE_DIR}/src/link/Wasm.zig"
-    "${CMAKE_SOURCE_DIR}/src/link/C/zig.h"
     "${CMAKE_SOURCE_DIR}/src/link/msdos-stub.bin"
-    "${CMAKE_SOURCE_DIR}/src/liveness.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/tapi.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/tapi/Tokenizer.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/tapi/parse.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/tapi/parse/test.zig"
+    "${CMAKE_SOURCE_DIR}/src/link/tapi/yaml.zig"
     "${CMAKE_SOURCE_DIR}/src/main.zig"
     "${CMAKE_SOURCE_DIR}/src/mingw.zig"
     "${CMAKE_SOURCE_DIR}/src/musl.zig"
+    "${CMAKE_SOURCE_DIR}/src/print_air.zig"
     "${CMAKE_SOURCE_DIR}/src/print_env.zig"
     "${CMAKE_SOURCE_DIR}/src/print_targets.zig"
+    "${CMAKE_SOURCE_DIR}/src/print_zir.zig"
     "${CMAKE_SOURCE_DIR}/src/stage1.zig"
     "${CMAKE_SOURCE_DIR}/src/target.zig"
     "${CMAKE_SOURCE_DIR}/src/tracy.zig"
@@ -599,8 +617,6 @@ set(ZIG_STAGE2_SOURCES
     "${CMAKE_SOURCE_DIR}/src/value.zig"
     "${CMAKE_SOURCE_DIR}/src/wasi_libc.zig"
     "${CMAKE_SOURCE_DIR}/src/windows_sdk.zig"
-    "${CMAKE_SOURCE_DIR}/src/Zir.zig"
-    "${CMAKE_SOURCE_DIR}/src/Sema.zig"
 )
 
 if(MSVC)
@@ -771,7 +787,7 @@ else()
     set(ZIG1_RELEASE_ARG -OReleaseFast --strip)
 endif()
 if(ZIG_SINGLE_THREADED)
-  set(ZIG1_SINGLE_THREADED_ARG "--single-threaded")
+  set(ZIG1_SINGLE_THREADED_ARG "-fsingle-threaded")
 else()
   set(ZIG1_SINGLE_THREADED_ARG "")
 endif()
@@ -781,8 +797,9 @@ set(BUILD_ZIG1_ARGS
     -target "${ZIG_TARGET_TRIPLE}"
     "-mcpu=${ZIG_TARGET_MCPU}"
     --name zig1
-    --override-lib-dir "${CMAKE_SOURCE_DIR}/lib"
+    --zig-lib-dir "${CMAKE_SOURCE_DIR}/lib"
     "-femit-bin=${ZIG1_OBJECT}"
+    -fcompiler-rt
     "${ZIG1_RELEASE_ARG}"
     "${ZIG1_SINGLE_THREADED_ARG}"
     -lc
@@ -835,7 +852,7 @@ set(ZIG_SKIP_INSTALL_LIB_FILES off CACHE BOOL
 
 if(NOT ZIG_SKIP_INSTALL_LIB_FILES)
     set(ZIG_INSTALL_ARGS "build"
-        --override-lib-dir "${CMAKE_SOURCE_DIR}/lib"
+        --zig-lib-dir "${CMAKE_SOURCE_DIR}/lib"
         "-Dlib-files-only"
         --prefix "${CMAKE_INSTALL_PREFIX}"
         "-Dconfig_h=${ZIG_CONFIG_H_OUT}"

CODE_OF_CONDUCT.md

@@ -9,7 +9,7 @@ a link. There is no concept of "official" or "unofficial", however, each
 gathering place has its own moderators and rules.
 
 This is Andrew Kelley speaking. At least for now, I'm the moderator of the
-ziglang organization GitHub repositories and the #zig IRC channel on Freenode.
+ziglang organization GitHub repositories and the #zig IRC channel on Libera.chat.
 **This document contains the rules that govern these two spaces only**.
 
 The rules here are strict. This space is for focused, on topic, technical work

CONTRIBUTING.md

@@ -5,7 +5,6 @@
 One of the best ways you can contribute to Zig is to start using it for a
 personal project. Here are some great examples:
 
- * [Oxid](https://github.com/dbandstra/oxid) - arcade style game
  * [TM35-Metronome](https://github.com/TM35-Metronome) - tools for modifying and randomizing Pokémon games
  * [River](https://github.com/ifreund/river/) - a dynamic tiling wayland compositor 
 
@@ -105,9 +104,17 @@ When making changes to the compiler source code, the most helpful test step to
 run is `test-behavior`. When editing documentation it is `docs`. You can find
 this information and more in the `--help` menu.
 
+#### Testing Changes to std lib
+
+To quickly test a change to a file in the standard library, you can run zig test and specify a custom lib directory with the follow command-line argument.
+
+```bash
+./build/zig test lib/std/fmt.zig --zig-lib-dir lib --main-pkg-path lib/std
+```
+
 #### Testing Non-Native Architectures with QEMU
 
-The Linux CI server additionally has qemu installed and sets `-Denable-qemu`.
+The Linux CI server additionally has qemu installed and sets `-fqemu`.
 This provides test coverage for, e.g. aarch64 even on x86_64 machines. It's 
 recommended for Linux users to install qemu and enable this testing option
 when editing the standard library or anything related to a non-native
@@ -116,7 +123,7 @@ architecture.
 ##### glibc
 
 Testing foreign architectures with dynamically linked glibc is one step trickier.
-This requires enabling `-Denable-foreign-glibc=/path/to/glibc/multi/install/glibcs`.
+This requires enabling `--glibc-runtimes /path/to/glibc/multi/install/glibcs`.
 This path is obtained by building glibc for multiple architectures. This
 process for me took an entire day to complete and takes up 65 GiB on my hard
 drive. The CI server does not provide this test coverage. Instructions for
@@ -128,7 +135,7 @@ It's understood that most contributors will not have these tests enabled.
 
 #### Testing Windows from a Linux Machine with Wine
 
-When developing on Linux, another option is available to you: `-Denable-wine`.
+When developing on Linux, another option is available to you: `-fwine`.
 This will enable running behavior tests and std lib tests with Wine. It's
 recommended for Linux users to install Wine and enable this testing option 
 when editing the standard library or anything Windows-related.
@@ -136,7 +143,7 @@ when editing the standard library or anything Windows-related.
 #### Testing WebAssembly using wasmtime
 
 If you have [wasmtime](https://wasmtime.dev/) installed, take advantage of the
-`-Denable-wasmtime` flag which will enable running WASI behavior tests and std
+`-fwasmtime` flag which will enable running WASI behavior tests and std
 lib tests. It's recommended for all users to install wasmtime and enable this
 testing option when editing the standard library and especially anything
 WebAssembly-related.

LICENSE

@@ -1,6 +1,6 @@
 The MIT License (Expat)
 
-Copyright (c) 2015-2021, Zig contributors
+Copyright (c) 2015-2022, Zig contributors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,4 +1,4 @@
-![ZIG](https://ziglang.org/zig-logo.svg)
+![ZIG](https://ziglang.org/img/zig-logo-dynamic.svg)
 
 A general-purpose programming language and toolchain for maintaining
 **robust**, **optimal**, and **reusable** software.

build.zig

@@ -3,7 +3,6 @@ const builtin = std.builtin;
 const Builder = std.build.Builder;
 const tests = @import("test/tests.zig");
 const BufMap = std.BufMap;
-const warn = std.debug.warn;
 const mem = std.mem;
 const ArrayList = std.ArrayList;
 const io = std.io;
@@ -11,14 +10,17 @@ const fs = std.fs;
 const InstallDirectoryOptions = std.build.InstallDirectoryOptions;
 const assert = std.debug.assert;
 
-const zig_version = std.builtin.Version{ .major = 0, .minor = 8, .patch = 0 };
+const zig_version = std.builtin.Version{ .major = 0, .minor = 9, .patch = 1 };
 
 pub fn build(b: *Builder) !void {
     b.setPreferredReleaseMode(.ReleaseFast);
     const mode = b.standardReleaseOptions();
     const target = b.standardTargetOptions(.{});
+    const single_threaded = b.option(bool, "single-threaded", "Build artifacts that run in single threaded mode");
+    const use_zig_libcxx = b.option(bool, "use-zig-libcxx", "If libc++ is needed, use zig's bundled version, don't try to integrate with the system") orelse false;
 
     var docgen_exe = b.addExecutable("docgen", "doc/docgen.zig");
+    docgen_exe.single_threaded = single_threaded;
 
     const rel_zig_exe = try fs.path.relative(b.allocator, b.build_root, b.zig_exe);
     const langref_out_path = fs.path.join(
@@ -40,7 +42,8 @@ pub fn build(b: *Builder) !void {
 
     var test_stage2 = b.addTest("src/test.zig");
     test_stage2.setBuildMode(mode);
-    test_stage2.addPackagePath("stage2_tests", "test/stage2/test.zig");
+    test_stage2.addPackagePath("test_cases", "test/cases.zig");
+    test_stage2.single_threaded = single_threaded;
 
     const fmt_build_zig = b.addFmt(&[_][]const u8{"build.zig"});
 
@@ -61,12 +64,33 @@ pub fn build(b: *Builder) !void {
     const omit_stage2 = b.option(bool, "omit-stage2", "Do not include stage2 behind a feature flag inside stage1") orelse false;
     const static_llvm = b.option(bool, "static-llvm", "Disable integration with system-installed LLVM, Clang, LLD, and libc++") orelse false;
     const enable_llvm = b.option(bool, "enable-llvm", "Build self-hosted compiler with LLVM backend enabled") orelse (is_stage1 or static_llvm);
+    const llvm_has_m68k = b.option(
+        bool,
+        "llvm-has-m68k",
+        "Whether LLVM has the experimental target m68k enabled",
+    ) orelse false;
+    const llvm_has_csky = b.option(
+        bool,
+        "llvm-has-csky",
+        "Whether LLVM has the experimental target csky enabled",
+    ) orelse false;
+    const llvm_has_ve = b.option(
+        bool,
+        "llvm-has-ve",
+        "Whether LLVM has the experimental target ve enabled",
+    ) orelse false;
+    const llvm_has_arc = b.option(
+        bool,
+        "llvm-has-arc",
+        "Whether LLVM has the experimental target arc enabled",
+    ) orelse false;
+    const enable_macos_sdk = b.option(bool, "enable-macos-sdk", "Run tests requiring presence of macOS SDK and frameworks") orelse false;
     const config_h_path_option = b.option([]const u8, "config_h", "Path to the generated config.h");
 
     if (!skip_install_lib_files) {
         b.installDirectory(InstallDirectoryOptions{
             .source_dir = "lib",
-            .install_dir = .Lib,
+            .install_dir = .lib,
             .install_subdir = "zig",
             .exclude_extensions = &[_][]const u8{
                 "README.md",
@@ -85,6 +109,8 @@ pub fn build(b: *Builder) !void {
         return;
 
     const tracy = b.option([]const u8, "tracy", "Enable Tracy integration. Supply path to Tracy source");
+    const tracy_callstack = b.option(bool, "tracy-callstack", "Include callstack information with Tracy data. Does nothing if -Dtracy is not provided") orelse false;
+    const tracy_allocation = b.option(bool, "tracy-allocation", "Include allocation information with Tracy data. Does nothing if -Dtracy is not provided") orelse false;
     const link_libc = b.option(bool, "force-link-libc", "Force self-hosted compiler to link libc") orelse enable_llvm;
     const strip = b.option(bool, "strip", "Omit debug information") orelse false;
 
@@ -101,23 +127,44 @@ pub fn build(b: *Builder) !void {
     exe.install();
     exe.setBuildMode(mode);
     exe.setTarget(target);
-    toolchain_step.dependOn(&exe.step);
+    if (!skip_stage2_tests) {
+        toolchain_step.dependOn(&exe.step);
+    }
     b.default_step.dependOn(&exe.step);
+    exe.single_threaded = single_threaded;
+
+    if (target.isWindows() and target.getAbi() == .gnu) {
+        // LTO is currently broken on mingw, this can be removed when it's fixed.
+        exe.want_lto = false;
+        test_stage2.want_lto = false;
+    }
+
+    const exe_options = b.addOptions();
+    exe.addOptions("build_options", exe_options);
+
+    exe_options.addOption(u32, "mem_leak_frames", mem_leak_frames);
+    exe_options.addOption(bool, "skip_non_native", skip_non_native);
+    exe_options.addOption(bool, "have_llvm", enable_llvm);
+    exe_options.addOption(bool, "llvm_has_m68k", llvm_has_m68k);
+    exe_options.addOption(bool, "llvm_has_csky", llvm_has_csky);
+    exe_options.addOption(bool, "llvm_has_ve", llvm_has_ve);
+    exe_options.addOption(bool, "llvm_has_arc", llvm_has_arc);
 
-    exe.addBuildOption(u32, "mem_leak_frames", mem_leak_frames);
-    exe.addBuildOption(bool, "skip_non_native", skip_non_native);
-    exe.addBuildOption(bool, "have_llvm", enable_llvm);
     if (enable_llvm) {
         const cmake_cfg = if (static_llvm) null else findAndParseConfigH(b, config_h_path_option);
 
         if (is_stage1) {
             exe.addIncludeDir("src");
             exe.addIncludeDir("deps/SoftFloat-3e/source/include");
+
+            test_stage2.addIncludeDir("src");
+            test_stage2.addIncludeDir("deps/SoftFloat-3e/source/include");
             // This is intentionally a dummy path. stage1.zig tries to @import("compiler_rt") in case
             // of being built by cmake. But when built by zig it's gonna get a compiler_rt so that
             // is pointless.
             exe.addPackagePath("compiler_rt", "src/empty.zig");
-            exe.defineCMacro("ZIG_LINK_MODE=Static");
+            exe.defineCMacro("ZIG_LINK_MODE", "Static");
+            test_stage2.defineCMacro("ZIG_LINK_MODE", "Static");
 
             const softfloat = b.addStaticLibrary("softfloat", null);
             softfloat.setBuildMode(.ReleaseFast);
@@ -126,10 +173,16 @@ pub fn build(b: *Builder) !void {
             softfloat.addIncludeDir("deps/SoftFloat-3e/source/8086");
             softfloat.addIncludeDir("deps/SoftFloat-3e/source/include");
             softfloat.addCSourceFiles(&softfloat_sources, &[_][]const u8{ "-std=c99", "-O3" });
+            softfloat.single_threaded = single_threaded;
+
             exe.linkLibrary(softfloat);
+            test_stage2.linkLibrary(softfloat);
 
             exe.addCSourceFiles(&stage1_sources, &exe_cflags);
             exe.addCSourceFiles(&optimized_c_sources, &[_][]const u8{ "-std=c99", "-O3" });
+
+            test_stage2.addCSourceFiles(&stage1_sources, &exe_cflags);
+            test_stage2.addCSourceFiles(&optimized_c_sources, &[_][]const u8{ "-std=c99", "-O3" });
         }
         if (cmake_cfg) |cfg| {
             // Inside this code path, we have to coordinate with system packaged LLVM, Clang, and LLD.
@@ -139,8 +192,8 @@ pub fn build(b: *Builder) !void {
                 b.addSearchPrefix(cfg.cmake_prefix_path);
             }
 
-            try addCmakeCfgOptionsToExe(b, cfg, tracy, exe);
-            try addCmakeCfgOptionsToExe(b, cfg, tracy, test_stage2);
+            try addCmakeCfgOptionsToExe(b, cfg, exe, use_zig_libcxx);
+            try addCmakeCfgOptionsToExe(b, cfg, test_stage2, use_zig_libcxx);
         } else {
             // Here we are -Denable-llvm but no cmake integration.
             try addStaticLlvmOptionsToExe(exe);
@@ -153,6 +206,7 @@ pub fn build(b: *Builder) !void {
     }
 
     const enable_logging = b.option(bool, "log", "Whether to enable logging") orelse false;
+    const enable_link_snapshots = b.option(bool, "link-snapshot", "Whether to enable linker state snapshots") orelse false;
 
     const opt_version_string = b.option([]const u8, "version-string", "Override Zig version string. Default is to find out with git.");
     const version = if (opt_version_string) |version| version else v: {
@@ -168,7 +222,7 @@ pub fn build(b: *Builder) !void {
 
         switch (mem.count(u8, git_describe, "-")) {
             0 => {
-                // Tagged release version (e.g. 0.8.0).
+                // Tagged release version (e.g. 0.9.0).
                 if (!mem.eql(u8, git_describe, version_string)) {
                     std.debug.print("Zig version '{s}' does not match Git tag '{s}'\n", .{ version_string, git_describe });
                     std.process.exit(1);
@@ -176,8 +230,8 @@ pub fn build(b: *Builder) !void {
                 break :v version_string;
             },
             2 => {
-                // Untagged development build (e.g. 0.8.0-684-gbbe2cca1a).
-                var it = mem.split(git_describe, "-");
+                // Untagged development build (e.g. 0.9.0-dev.2025+ecf0050a9).
+                var it = mem.split(u8, git_describe, "-");
                 const tagged_ancestor = it.next() orelse unreachable;
                 const commit_height = it.next() orelse unreachable;
                 const commit_id = it.next() orelse unreachable;
@@ -203,47 +257,68 @@ pub fn build(b: *Builder) !void {
             },
         }
     };
-    exe.addBuildOption([:0]const u8, "version", try b.allocator.dupeZ(u8, version));
+    exe_options.addOption([:0]const u8, "version", try b.allocator.dupeZ(u8, version));
 
     const semver = try std.SemanticVersion.parse(version);
-    exe.addBuildOption(std.SemanticVersion, "semver", semver);
+    exe_options.addOption(std.SemanticVersion, "semver", semver);
 
-    exe.addBuildOption(bool, "enable_logging", enable_logging);
-    exe.addBuildOption(bool, "enable_tracy", tracy != null);
-    exe.addBuildOption(bool, "is_stage1", is_stage1);
-    exe.addBuildOption(bool, "omit_stage2", omit_stage2);
+    exe_options.addOption(bool, "enable_logging", enable_logging);
+    exe_options.addOption(bool, "enable_link_snapshots", enable_link_snapshots);
+    exe_options.addOption(bool, "enable_tracy", tracy != null);
+    exe_options.addOption(bool, "enable_tracy_callstack", tracy_callstack);
+    exe_options.addOption(bool, "enable_tracy_allocation", tracy_allocation);
+    exe_options.addOption(bool, "is_stage1", is_stage1);
+    exe_options.addOption(bool, "omit_stage2", omit_stage2);
     if (tracy) |tracy_path| {
         const client_cpp = fs.path.join(
             b.allocator,
             &[_][]const u8{ tracy_path, "TracyClient.cpp" },
         ) catch unreachable;
+
+        // On mingw, we need to opt into windows 7+ to get some features required by tracy.
+        const tracy_c_flags: []const []const u8 = if (target.isWindows() and target.getAbi() == .gnu)
+            &[_][]const u8{ "-DTRACY_ENABLE=1", "-fno-sanitize=undefined", "-D_WIN32_WINNT=0x601" }
+        else
+            &[_][]const u8{ "-DTRACY_ENABLE=1", "-fno-sanitize=undefined" };
+
         exe.addIncludeDir(tracy_path);
-        exe.addCSourceFile(client_cpp, &[_][]const u8{ "-DTRACY_ENABLE=1", "-fno-sanitize=undefined" });
+        exe.addCSourceFile(client_cpp, tracy_c_flags);
         if (!enable_llvm) {
             exe.linkSystemLibraryName("c++");
         }
         exe.linkLibC();
+
+        if (target.isWindows()) {
+            exe.linkSystemLibrary("dbghelp");
+            exe.linkSystemLibrary("ws2_32");
+        }
     }
 
     const test_filter = b.option([]const u8, "test-filter", "Skip tests that do not match filter");
 
-    const is_wine_enabled = b.option(bool, "enable-wine", "Use Wine to run cross compiled Windows tests") orelse false;
-    const is_qemu_enabled = b.option(bool, "enable-qemu", "Use QEMU to run cross compiled foreign architecture tests") orelse false;
-    const is_wasmtime_enabled = b.option(bool, "enable-wasmtime", "Use Wasmtime to enable and run WASI libstd tests") orelse false;
-    const is_darling_enabled = b.option(bool, "enable-darling", "[Experimental] Use Darling to run cross compiled macOS tests") orelse false;
-    const glibc_multi_dir = b.option([]const u8, "enable-foreign-glibc", "Provide directory with glibc installations to run cross compiled tests that link glibc");
+    const test_stage2_options = b.addOptions();
+    test_stage2.addOptions("build_options", test_stage2_options);
 
-    test_stage2.addBuildOption(bool, "skip_non_native", skip_non_native);
-    test_stage2.addBuildOption(bool, "is_stage1", is_stage1);
-    test_stage2.addBuildOption(bool, "omit_stage2", omit_stage2);
-    test_stage2.addBuildOption(bool, "have_llvm", enable_llvm);
-    test_stage2.addBuildOption(bool, "enable_qemu", is_qemu_enabled);
-    test_stage2.addBuildOption(bool, "enable_wine", is_wine_enabled);
-    test_stage2.addBuildOption(bool, "enable_wasmtime", is_wasmtime_enabled);
-    test_stage2.addBuildOption(u32, "mem_leak_frames", mem_leak_frames * 2);
-    test_stage2.addBuildOption(bool, "enable_darling", is_darling_enabled);
-    test_stage2.addBuildOption(?[]const u8, "glibc_multi_install_dir", glibc_multi_dir);
-    test_stage2.addBuildOption([]const u8, "version", version);
+    test_stage2_options.addOption(bool, "enable_logging", enable_logging);
+    test_stage2_options.addOption(bool, "enable_link_snapshots", enable_link_snapshots);
+    test_stage2_options.addOption(bool, "skip_non_native", skip_non_native);
+    test_stage2_options.addOption(bool, "skip_compile_errors", skip_compile_errors);
+    test_stage2_options.addOption(bool, "is_stage1", is_stage1);
+    test_stage2_options.addOption(bool, "omit_stage2", omit_stage2);
+    test_stage2_options.addOption(bool, "have_llvm", enable_llvm);
+    test_stage2_options.addOption(bool, "llvm_has_m68k", llvm_has_m68k);
+    test_stage2_options.addOption(bool, "llvm_has_csky", llvm_has_csky);
+    test_stage2_options.addOption(bool, "llvm_has_ve", llvm_has_ve);
+    test_stage2_options.addOption(bool, "llvm_has_arc", llvm_has_arc);
+    test_stage2_options.addOption(bool, "enable_qemu", b.enable_qemu);
+    test_stage2_options.addOption(bool, "enable_wine", b.enable_wine);
+    test_stage2_options.addOption(bool, "enable_wasmtime", b.enable_wasmtime);
+    test_stage2_options.addOption(bool, "enable_rosetta", b.enable_rosetta);
+    test_stage2_options.addOption(bool, "enable_darling", b.enable_darling);
+    test_stage2_options.addOption(u32, "mem_leak_frames", mem_leak_frames * 2);
+    test_stage2_options.addOption(?[]const u8, "glibc_runtimes_dir", b.glibc_runtimes_dir);
+    test_stage2_options.addOption([:0]const u8, "version", try b.allocator.dupeZ(u8, version));
+    test_stage2_options.addOption(std.SemanticVersion, "semver", semver);
 
     const test_stage2_step = b.step("test-stage2", "Run the stage2 compiler tests");
     test_stage2_step.dependOn(&test_stage2.step);
@@ -283,14 +358,9 @@ pub fn build(b: *Builder) !void {
         "behavior",
         "Run the behavior tests",
         modes,
-        false,
+        false, // skip_single_threaded
         skip_non_native,
         skip_libc,
-        is_wine_enabled,
-        is_qemu_enabled,
-        is_wasmtime_enabled,
-        is_darling_enabled,
-        glibc_multi_dir,
     ));
 
     toolchain_step.dependOn(tests.addPkgTests(
@@ -300,14 +370,9 @@ pub fn build(b: *Builder) !void {
         "compiler-rt",
         "Run the compiler_rt tests",
         modes,
-        true,
+        true, // skip_single_threaded
         skip_non_native,
-        true,
-        is_wine_enabled,
-        is_qemu_enabled,
-        is_wasmtime_enabled,
-        is_darling_enabled,
-        glibc_multi_dir,
+        true, // skip_libc
     ));
 
     toolchain_step.dependOn(tests.addPkgTests(
@@ -317,18 +382,13 @@ pub fn build(b: *Builder) !void {
         "minilibc",
         "Run the mini libc tests",
         modes,
-        true,
+        true, // skip_single_threaded
         skip_non_native,
-        true,
-        is_wine_enabled,
-        is_qemu_enabled,
-        is_wasmtime_enabled,
-        is_darling_enabled,
-        glibc_multi_dir,
+        true, // skip_libc
     ));
 
     toolchain_step.dependOn(tests.addCompareOutputTests(b, test_filter, modes));
-    toolchain_step.dependOn(tests.addStandaloneTests(b, test_filter, modes, skip_non_native, target));
+    toolchain_step.dependOn(tests.addStandaloneTests(b, test_filter, modes, skip_non_native, enable_macos_sdk, target));
     toolchain_step.dependOn(tests.addStackTraceTests(b, test_filter, modes));
     toolchain_step.dependOn(tests.addCliTests(b, test_filter, modes));
     toolchain_step.dependOn(tests.addAssembleAndLinkTests(b, test_filter, modes));
@@ -339,9 +399,6 @@ pub fn build(b: *Builder) !void {
     }
     // tests for this feature are disabled until we have the self-hosted compiler available
     // toolchain_step.dependOn(tests.addGenHTests(b, test_filter));
-    if (!skip_compile_errors) {
-        toolchain_step.dependOn(tests.addCompileErrorTests(b, test_filter, modes));
-    }
 
     const std_step = tests.addPkgTests(
         b,
@@ -353,11 +410,6 @@ pub fn build(b: *Builder) !void {
         false,
         skip_non_native,
         skip_libc,
-        is_wine_enabled,
-        is_qemu_enabled,
-        is_wasmtime_enabled,
-        is_darling_enabled,
-        glibc_multi_dir,
     );
 
     const test_step = b.step("test", "Run all the tests");
@@ -383,8 +435,8 @@ const exe_cflags = [_][]const u8{
 fn addCmakeCfgOptionsToExe(
     b: *Builder,
     cfg: CMakeConfig,
-    tracy: ?[]const u8,
     exe: *std.build.LibExeObjStep,
+    use_zig_libcxx: bool,
 ) !void {
     exe.addObjectFile(fs.path.join(b.allocator, &[_][]const u8{
         cfg.cmake_binary_dir,
@@ -397,28 +449,32 @@ fn addCmakeCfgOptionsToExe(
     addCMakeLibraryList(exe, cfg.lld_libraries);
     addCMakeLibraryList(exe, cfg.llvm_libraries);
 
-    const need_cpp_includes = tracy != null;
+    if (use_zig_libcxx) {
+        exe.linkLibCpp();
+    } else {
+        const need_cpp_includes = true;
 
-    // System -lc++ must be used because in this code path we are attempting to link
-    // against system-provided LLVM, Clang, LLD.
-    if (exe.target.getOsTag() == .linux) {
-        // First we try to static link against gcc libstdc++. If that doesn't work,
-        // we fall back to -lc++ and cross our fingers.
-        addCxxKnownPath(b, cfg, exe, "libstdc++.a", "", need_cpp_includes) catch |err| switch (err) {
-            error.RequiredLibraryNotFound => {
-                exe.linkSystemLibrary("c++");
-            },
-            else => |e| return e,
-        };
-        exe.linkSystemLibrary("unwind");
-    } else if (exe.target.isFreeBSD()) {
-        try addCxxKnownPath(b, cfg, exe, "libc++.a", null, need_cpp_includes);
-        exe.linkSystemLibrary("pthread");
-    } else if (exe.target.getOsTag() == .openbsd) {
-        try addCxxKnownPath(b, cfg, exe, "libc++.a", null, need_cpp_includes);
-        try addCxxKnownPath(b, cfg, exe, "libc++abi.a", null, need_cpp_includes);
-    } else if (exe.target.isDarwin()) {
-        exe.linkSystemLibrary("c++");
+        // System -lc++ must be used because in this code path we are attempting to link
+        // against system-provided LLVM, Clang, LLD.
+        if (exe.target.getOsTag() == .linux) {
+            // First we try to static link against gcc libstdc++. If that doesn't work,
+            // we fall back to -lc++ and cross our fingers.
+            addCxxKnownPath(b, cfg, exe, "libstdc++.a", "", need_cpp_includes) catch |err| switch (err) {
+                error.RequiredLibraryNotFound => {
+                    exe.linkSystemLibrary("c++");
+                },
+                else => |e| return e,
+            };
+            exe.linkSystemLibrary("unwind");
+        } else if (exe.target.isFreeBSD()) {
+            try addCxxKnownPath(b, cfg, exe, "libc++.a", null, need_cpp_includes);
+            exe.linkSystemLibrary("pthread");
+        } else if (exe.target.getOsTag() == .openbsd) {
+            try addCxxKnownPath(b, cfg, exe, "libc++.a", null, need_cpp_includes);
+            try addCxxKnownPath(b, cfg, exe, "libc++abi.a", null, need_cpp_includes);
+        } else if (exe.target.isDarwin()) {
+            exe.linkSystemLibrary("c++");
+        }
     }
 
     if (cfg.dia_guids_lib.len != 0) {
@@ -449,12 +505,15 @@ fn addStaticLlvmOptionsToExe(
         exe.linkSystemLibrary(lib_name);
     }
 
+    exe.linkSystemLibrary("z");
+
     // This means we rely on clang-or-zig-built LLVM, Clang, LLD libraries.
     exe.linkSystemLibrary("c++");
 
     if (exe.target.getOs().tag == .windows) {
         exe.linkSystemLibrary("version");
         exe.linkSystemLibrary("uuid");
+        exe.linkSystemLibrary("ole32");
     }
 }
 
@@ -470,12 +529,12 @@ fn addCxxKnownPath(
         ctx.cxx_compiler,
         b.fmt("-print-file-name={s}", .{objname}),
     });
-    const path_unpadded = mem.tokenize(path_padded, "\r\n").next().?;
+    const path_unpadded = mem.tokenize(u8, path_padded, "\r\n").next().?;
     if (mem.eql(u8, path_unpadded, objname)) {
         if (errtxt) |msg| {
-            warn("{s}", .{msg});
+            std.debug.print("{s}", .{msg});
         } else {
-            warn("Unable to determine path to {s}\n", .{objname});
+            std.debug.print("Unable to determine path to {s}\n", .{objname});
         }
         return error.RequiredLibraryNotFound;
     }
@@ -486,14 +545,14 @@ fn addCxxKnownPath(
     if (need_cpp_includes) {
         // I used these temporarily for testing something but we obviously need a
         // more general purpose solution here.
-        //exe.addIncludeDir("/nix/store/b3zsk4ihlpiimv3vff86bb5bxghgdzb9-gcc-9.2.0/lib/gcc/x86_64-unknown-linux-gnu/9.2.0/../../../../include/c++/9.2.0");
-        //exe.addIncludeDir("/nix/store/b3zsk4ihlpiimv3vff86bb5bxghgdzb9-gcc-9.2.0/lib/gcc/x86_64-unknown-linux-gnu/9.2.0/../../../../include/c++/9.2.0/x86_64-unknown-linux-gnu");
-        //exe.addIncludeDir("/nix/store/b3zsk4ihlpiimv3vff86bb5bxghgdzb9-gcc-9.2.0/lib/gcc/x86_64-unknown-linux-gnu/9.2.0/../../../../include/c++/9.2.0/backward");
+        //exe.addIncludeDir("/nix/store/fvf3qjqa5qpcjjkq37pb6ypnk1mzhf5h-gcc-9.3.0/lib/gcc/x86_64-unknown-linux-gnu/9.3.0/../../../../include/c++/9.3.0");
+        //exe.addIncludeDir("/nix/store/fvf3qjqa5qpcjjkq37pb6ypnk1mzhf5h-gcc-9.3.0/lib/gcc/x86_64-unknown-linux-gnu/9.3.0/../../../../include/c++/9.3.0/x86_64-unknown-linux-gnu");
+        //exe.addIncludeDir("/nix/store/fvf3qjqa5qpcjjkq37pb6ypnk1mzhf5h-gcc-9.3.0/lib/gcc/x86_64-unknown-linux-gnu/9.3.0/../../../../include/c++/9.3.0/backward");
     }
 }
 
 fn addCMakeLibraryList(exe: *std.build.LibExeObjStep, list: []const u8) void {
-    var it = mem.tokenize(list, ";");
+    var it = mem.tokenize(u8, list, ";");
     while (it.next()) |lib| {
         if (mem.startsWith(u8, lib, "-l")) {
             exe.linkSystemLibrary(lib["-l".len..]);
@@ -587,11 +646,11 @@ fn findAndParseConfigH(b: *Builder, config_h_path_option: ?[]const u8) ?CMakeCon
         },
     };
 
-    var lines_it = mem.tokenize(config_h_text, "\r\n");
+    var lines_it = mem.tokenize(u8, config_h_text, "\r\n");
     while (lines_it.next()) |line| {
         inline for (mappings) |mapping| {
             if (mem.startsWith(u8, line, mapping.prefix)) {
-                var it = mem.split(line, "\"");
+                var it = mem.split(u8, line, "\"");
                 _ = it.next().?; // skip the stuff before the quote
                 const quoted = it.next().?; // the stuff inside the quote
                 @field(ctx, mapping.field) = toNativePathSep(b, quoted);
@@ -602,7 +661,7 @@ fn findAndParseConfigH(b: *Builder, config_h_path_option: ?[]const u8) ?CMakeCon
 }
 
 fn toNativePathSep(b: *Builder, s: []const u8) []u8 {
-    const duplicated = mem.dupe(b.allocator, u8, s) catch unreachable;
+    const duplicated = b.allocator.dupe(u8, s) catch unreachable;
     for (duplicated) |*byte| switch (byte.*) {
         '/' => byte.* = fs.path.sep,
         else => {},
@@ -769,6 +828,7 @@ const zig_cpp_sources = [_][]const u8{
     // These are planned to stay even when we are self-hosted.
     "src/zig_llvm.cpp",
     "src/zig_clang.cpp",
+    "src/zig_llvm-ar.cpp",
     "src/zig_clang_driver.cpp",
     "src/zig_clang_cc1_main.cpp",
     "src/zig_clang_cc1as_main.cpp",
@@ -837,6 +897,7 @@ const llvm_libs = [_][]const u8{
     "LLVMWebAssemblyAsmParser",
     "LLVMWebAssemblyCodeGen",
     "LLVMWebAssemblyDesc",
+    "LLVMWebAssemblyUtils",
     "LLVMWebAssemblyInfo",
     "LLVMSystemZDisassembler",
     "LLVMSystemZAsmParser",
@@ -912,11 +973,12 @@ const llvm_libs = [_][]const u8{
     "LLVMOrcJIT",
     "LLVMMCJIT",
     "LLVMJITLink",
-    "LLVMOrcTargetProcess",
-    "LLVMOrcShared",
     "LLVMInterpreter",
     "LLVMExecutionEngine",
     "LLVMRuntimeDyld",
+    "LLVMOrcTargetProcess",
+    "LLVMOrcShared",
+    "LLVMDWP",
     "LLVMSymbolize",
     "LLVMDebugInfoPDB",
     "LLVMDebugInfoGSYM",
@@ -929,7 +991,6 @@ const llvm_libs = [_][]const u8{
     "LLVMCFGuard",
     "LLVMCoroutines",
     "LLVMObjCARCOpts",
-    "LLVMHelloNew",
     "LLVMipo",
     "LLVMVectorize",
     "LLVMLinker",
@@ -941,6 +1002,7 @@ const llvm_libs = [_][]const u8{
     "LLVMGlobalISel",
     "LLVMMIRParser",
     "LLVMAsmPrinter",
+    "LLVMDebugInfoMSF",
     "LLVMDebugInfoDWARF",
     "LLVMSelectionDAG",
     "LLVMCodeGen",
@@ -962,7 +1024,6 @@ const llvm_libs = [_][]const u8{
     "LLVMMCParser",
     "LLVMMC",
     "LLVMDebugInfoCodeView",
-    "LLVMDebugInfoMSF",
     "LLVMBitReader",
     "LLVMCore",
     "LLVMRemarks",

ci/azure/linux_script

@@ -1,125 +0,0 @@
-#!/bin/sh
-
-set -x
-set -e
-
-sudo apt-get update -q
-sudo apt-get install -y cmake s3cmd tidy
-
-ZIGDIR="$(pwd)"
-ARCH="$(uname -m)"
-TARGET="$ARCH-linux-musl"
-CACHE_BASENAME="zig+llvm+lld+clang-$TARGET-0.8.0-dev.2703+c12704a33"
-PREFIX="$HOME/$CACHE_BASENAME"
-MCPU="baseline"
-JOBS="-j$(nproc)"
-
-rm -rf $PREFIX
-cd $HOME
-
-wget -nv "https://ziglang.org/deps/$CACHE_BASENAME.tar.xz"
-tar xf "$CACHE_BASENAME.tar.xz"
-
-QEMUBASE="qemu-linux-x86_64-5.2.0.1"
-wget -nv "https://ziglang.org/deps/$QEMUBASE.tar.xz"
-tar xf "$QEMUBASE.tar.xz"
-export PATH="$(pwd)/$QEMUBASE/bin:$PATH"
-
-# Bump to v0.23 once this issue is resolved:
-# https://github.com/ziglang/zig/issues/8742
-WASMTIME="wasmtime-v0.22.1-x86_64-linux"
-wget -nv "https://github.com/bytecodealliance/wasmtime/releases/download/v0.22.1/$WASMTIME.tar.xz"
-tar xf "$WASMTIME.tar.xz"
-export PATH="$(pwd)/$WASMTIME:$PATH"
-
-ZIG="$PREFIX/bin/zig"
-export CC="$ZIG cc -target $TARGET -mcpu=$MCPU"
-export CXX="$ZIG c++ -target $TARGET -mcpu=$MCPU"
-
-cd $ZIGDIR
-
-# Make the `zig version` number consistent.
-# This will affect the cmake command below.
-git config core.abbrev 9
-git fetch --unshallow || true
-git fetch --tags
-
-mkdir build
-cd build
-cmake .. \
-  -DCMAKE_INSTALL_PREFIX="$(pwd)/release" \
-  -DCMAKE_PREFIX_PATH="$PREFIX" \
-  -DCMAKE_BUILD_TYPE=Release \
-  -DZIG_TARGET_TRIPLE="$TARGET" \
-  -DZIG_TARGET_MCPU="$MCPU" \
-  -DZIG_STATIC=ON
-
-# Now cmake will use zig as the C/C++ compiler. We reset the environment variables
-# so that installation and testing do not get affected by them.
-unset CC
-unset CXX
-
-make $JOBS install
-
-# Here we rebuild zig but this time using the Zig binary we just now produced to
-# build zig1.o rather than relying on the one built with stage0. See
-# https://github.com/ziglang/zig/issues/6830 for more details.
-cmake .. -DZIG_EXECUTABLE="$(pwd)/release/bin/zig"
-make $JOBS install
-
-for step in test-toolchain test-std docs; do
-  release/bin/zig build $step -Denable-qemu -Denable-wasmtime
-done
-
-# Look for HTML errors.
-tidy -qe ../zig-cache/langref.html
-
-if [ "${BUILD_REASON}" != "PullRequest" ]; then
-  # Produce the experimental std lib documentation.
-  mkdir -p release/docs/std
-  release/bin/zig test ../lib/std/std.zig \
-    --override-lib-dir ../lib \
-    -femit-docs=release/docs/std \
-    -fno-emit-bin
-
-  mv ../LICENSE release/
-  mv ../zig-cache/langref.html release/docs/
-
-  # Remove the unnecessary bin dir in $prefix/bin/zig
-  mv release/bin/zig release/
-  rmdir release/bin
-
-  # Remove the unnecessary zig dir in $prefix/lib/zig/std/std.zig
-  mv release/lib/zig release/lib2
-  rmdir release/lib
-  mv release/lib2 release/lib
-
-  VERSION=$(release/zig version)
-  DIRNAME="zig-linux-$ARCH-$VERSION"
-  TARBALL="$DIRNAME.tar.xz"
-  mv release "$DIRNAME"
-  tar cfJ "$TARBALL" "$DIRNAME"
-
-  mv "$DOWNLOADSECUREFILE_SECUREFILEPATH" "$HOME/.s3cfg"
-  s3cmd put -P --add-header="cache-control: public, max-age=31536000, immutable" "$TARBALL" s3://ziglang.org/builds/
-
-  SHASUM=$(sha256sum $TARBALL | cut '-d ' -f1)
-  BYTESIZE=$(wc -c < $TARBALL)
-
-  JSONFILE="linux-$GITBRANCH.json"
-  touch $JSONFILE
-  echo "{\"tarball\": \"$TARBALL\"," >>$JSONFILE
-  echo "\"shasum\": \"$SHASUM\"," >>$JSONFILE
-  echo "\"size\": \"$BYTESIZE\"}" >>$JSONFILE
-
-  s3cmd put -P --add-header="Cache-Control: max-age=0, must-revalidate" "$JSONFILE" "s3://ziglang.org/builds/$JSONFILE"
-  s3cmd put -P "$JSONFILE" "s3://ziglang.org/builds/$ARCH-linux-$VERSION.json"
-
-  # `set -x` causes these variables to be mangled.
-  # See https://developercommunity.visualstudio.com/content/problem/375679/pipeline-variable-incorrectly-inserts-single-quote.html
-  set +x
-  echo "##vso[task.setvariable variable=tarball;isOutput=true]$TARBALL"
-  echo "##vso[task.setvariable variable=shasum;isOutput=true]$SHASUM"
-  echo "##vso[task.setvariable variable=bytesize;isOutput=true]$BYTESIZE"
-  echo "##vso[task.setvariable variable=version;isOutput=true]$VERSION"
-fi

ci/azure/macos_arm64_script

@@ -3,20 +3,20 @@
 set -x
 set -e
 
-brew update && brew install s3cmd
+brew update && brew install ncurses s3cmd
 
 ZIGDIR="$(pwd)"
 
 HOST_ARCH="x86_64"
 HOST_TARGET="$HOST_ARCH-macos-gnu"
 HOST_MCPU="baseline"
-HOST_CACHE_BASENAME="zig+llvm+lld+clang-$HOST_TARGET-0.8.0-dev.2703+c12704a33"
+HOST_CACHE_BASENAME="zig+llvm+lld+clang-$HOST_TARGET-0.9.0-dev.1249+210ef5af8"
 HOST_PREFIX="$HOME/$HOST_CACHE_BASENAME"
 
 ARCH="aarch64"
 TARGET="$ARCH-macos-gnu"
 MCPU="apple_a14"
-CACHE_BASENAME="zig+llvm+lld+clang-$TARGET-0.8.0-dev.2703+c12704a33"
+CACHE_BASENAME="zig+llvm+lld+clang-$TARGET-0.9.0-dev.1249+210ef5af8"
 PREFIX="$HOME/$CACHE_BASENAME"
 
 JOBS="-j2"
@@ -50,10 +50,11 @@ cd build.host
 cmake .. \
   -DCMAKE_INSTALL_PREFIX="$(pwd)/release" \
   -DCMAKE_PREFIX_PATH="$HOST_PREFIX" \
-  -DCMAKE_BUILD_TYPE=Debug \
+  -DCMAKE_BUILD_TYPE=Release \
   -DZIG_TARGET_TRIPLE="$HOST_TARGET" \
   -DZIG_TARGET_MCPU="$HOST_MCPU" \
-  -DZIG_STATIC=ON
+  -DZIG_STATIC=ON \
+  -DZIG_OMIT_STAGE2=ON
 
 unset CC
 unset CXX
@@ -94,7 +95,7 @@ if [ "${BUILD_REASON}" != "PullRequest" ]; then
   # Produce the experimental std lib documentation.
   mkdir -p release/docs/std
   $ZIG test ../lib/std/std.zig \
-    --override-lib-dir ../lib \
+    --zig-lib-dir ../lib \
     -femit-docs=release/docs/std \
     -fno-emit-bin
 

ci/azure/macos_script

@@ -3,13 +3,13 @@
 set -x
 set -e
 
-brew update && brew install s3cmd
+brew update && brew install ncurses s3cmd
 
 ZIGDIR="$(pwd)"
 ARCH="x86_64"
 TARGET="$ARCH-macos-gnu"
 MCPU="baseline"
-CACHE_BASENAME="zig+llvm+lld+clang-$TARGET-0.8.0-dev.2703+c12704a33"
+CACHE_BASENAME="zig+llvm+lld+clang-$TARGET-0.9.0-dev.1249+210ef5af8"
 PREFIX="$HOME/$CACHE_BASENAME"
 JOBS="-j2"
 
@@ -39,7 +39,8 @@ cmake .. \
   -DCMAKE_BUILD_TYPE=Release \
   -DZIG_TARGET_TRIPLE="$TARGET" \
   -DZIG_TARGET_MCPU="$MCPU" \
-  -DZIG_STATIC=ON
+  -DZIG_STATIC=ON \
+  -DZIG_OMIT_STAGE2=ON
 
 # Now cmake will use zig as the C/C++ compiler. We reset the environment variables
 # so that installation and testing do not get affected by them.
@@ -54,9 +55,12 @@ make $JOBS install
 cmake .. -DZIG_EXECUTABLE="$(pwd)/release/bin/zig"
 make $JOBS install
 
-for step in test-toolchain test-std docs; do
-  release/bin/zig build $step
-done
+# TODO figure out why this causes a segmentation fault
+# release/bin/zig test ../test/behavior.zig -fno-stage1 -fLLVM -I ../test
+
+release/bin/zig build test-toolchain -Denable-macos-sdk
+release/bin/zig build test-std
+release/bin/zig build docs
 
 if [ "${BUILD_REASON}" != "PullRequest" ]; then
   mv ../LICENSE release/
@@ -91,4 +95,5 @@ if [ "${BUILD_REASON}" != "PullRequest" ]; then
   echo "##vso[task.setvariable variable=tarball;isOutput=true]$TARBALL"
   echo "##vso[task.setvariable variable=shasum;isOutput=true]$SHASUM"
   echo "##vso[task.setvariable variable=bytesize;isOutput=true]$BYTESIZE"
+  echo "##vso[task.setvariable variable=version;isOutput=true]$VERSION"
 fi

ci/azure/pipelines.yml

@@ -21,26 +21,14 @@ jobs:
   - script: ci/azure/macos_arm64_script
     name: main
     displayName: 'Build'
-- job: BuildLinux
-  pool:
-    vmImage: 'ubuntu-18.04'
-  timeoutInMinutes: 360
-  steps:
-  - task: DownloadSecureFile@1
-    inputs:
-      secureFile: s3cfg
-  - script: ci/azure/linux_script
-    name: main
-    displayName: 'Build and test'
 - job: BuildWindows
   pool:
     vmImage: 'windows-2019'
   timeoutInMinutes: 360
   steps:
   - powershell: |
-      (New-Object Net.WebClient).DownloadFile("https://github.com/msys2/msys2-installer/releases/download/2021-02-28/msys2-base-x86_64-20210228.sfx.exe", "sfx.exe")
+      (New-Object Net.WebClient).DownloadFile("https://github.com/msys2/msys2-installer/releases/download/2022-01-28/msys2-base-x86_64-20220128.sfx.exe", "sfx.exe")
       .\sfx.exe -y -o\
-      del sfx.exe
     displayName: Download/Extract/Install MSYS2
   - script: |
       @REM install updated filesystem package first without dependency checking
@@ -61,7 +49,6 @@ jobs:
   dependsOn:
   - BuildMacOS
   - BuildMacOS_arm64
-  - BuildLinux
   - BuildWindows
   condition: and(succeeded(), eq(variables['Build.SourceBranch'], 'refs/heads/master'))
   strategy:
@@ -69,7 +56,7 @@ jobs:
   pool:
     vmImage: 'ubuntu-18.04'
   variables:
-    version: $[ dependencies.BuildLinux.outputs['main.version'] ]
+    version: $[ dependencies.BuildMacOS.outputs['main.version'] ]
   steps:
   - task: DownloadSecureFile@1
     inputs:

ci/azure/windows_msvc_install

@@ -6,6 +6,11 @@ set -e
 pacman -Suy --needed --noconfirm
 pacman -S --needed --noconfirm wget p7zip python3-pip tar xz
 
+TARBALL="llvm+clang+lld-13.0.0-x86_64-windows-msvc-release-mt.tar.xz"
+
 pip install s3cmd
-wget -nv "https://ziglang.org/deps/llvm%2bclang%2blld-12.0.1-rc1-x86_64-windows-msvc-release-mt.tar.xz"
-tar xf llvm+clang+lld-12.0.1-rc1-x86_64-windows-msvc-release-mt.tar.xz
+wget -nv "https://ziglang.org/deps/$TARBALL"
+# If the first extraction fails, re-try it once; this can happen if the tarball
+# contains symlinks that are in the table of contents before the files that
+# they point to.
+tar -xf $TARBALL || tar --overwrite -xf $TARBALL

ci/azure/windows_msvc_script.bat

@@ -1,7 +1,7 @@
 @echo on
 SET "SRCROOT=%cd%"
 SET "PREVPATH=%PATH%"
-SET "PREVMSYSEM=%MSYSTEM%"
+SET "PREVMSYSTEM=%MSYSTEM%"
 
 set "PATH=%CD:~0,2%\msys64\usr\bin;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem"
 SET "MSYSTEM=MINGW64"
@@ -11,7 +11,7 @@ SET "MSYSTEM=%PREVMSYSTEM%"
 
 SET "ZIGBUILDDIR=%SRCROOT%\build"
 SET "ZIGINSTALLDIR=%ZIGBUILDDIR%\dist"
-SET "ZIGPREFIXPATH=%SRCROOT%\llvm+clang+lld-12.0.1-rc1-x86_64-windows-msvc-release-mt"
+SET "ZIGPREFIXPATH=%SRCROOT%\llvm+clang+lld-13.0.0-x86_64-windows-msvc-release-mt"
 
 call "C:\Program Files (x86)\Microsoft Visual Studio\2019\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" x64
 
@@ -26,6 +26,10 @@ cd %ZIGBUILDDIR%
 cmake.exe .. -Thost=x64 -G"Visual Studio 16 2019" -A x64 "-DCMAKE_INSTALL_PREFIX=%ZIGINSTALLDIR%" "-DCMAKE_PREFIX_PATH=%ZIGPREFIXPATH%" -DCMAKE_BUILD_TYPE=Release -DZIG_OMIT_STAGE2=ON || exit /b
 msbuild /maxcpucount /p:Configuration=Release INSTALL.vcxproj || exit /b
 
+REM Sadly, stage2 is omitted from this build to save memory on the CI server. Once self-hosted is
+REM built with itself and does not gobble as much memory, we can enable these tests.
+REM "%ZIGINSTALLDIR%\bin\zig.exe" test "..\test\behavior.zig" -fno-stage1 -fLLVM -I "..\test" || exit /b
+
 "%ZIGINSTALLDIR%\bin\zig.exe" build test-toolchain -Dskip-non-native -Dskip-stage2-tests || exit /b
 "%ZIGINSTALLDIR%\bin\zig.exe" build test-std -Dskip-non-native || exit /b
 "%ZIGINSTALLDIR%\bin\zig.exe" build docs || exit /b

ci/azure/windows_upload

@@ -11,6 +11,11 @@ if [ "${BUILD_REASON}" != "PullRequest" ]; then
   mv dist/bin/zig.exe dist/
   rmdir dist/bin
 
+  # Remove the unnecessary zig dir in $prefix/lib/zig/std/std.zig
+  mv dist/lib/zig dist/lib2
+  rmdir dist/lib
+  mv dist/lib2 dist/lib
+
   VERSION=$(dist/zig.exe version)
   DIRNAME="zig-windows-x86_64-$VERSION"
   TARBALL="$DIRNAME.zip"

ci/drone/drone.yml

@@ -7,38 +7,70 @@ platform:
 
 steps:
 - name: build
-  image: ziglang/static-base:llvm12-aarch64-5
+  image: ziglang/static-base:llvm13-aarch64-1
   commands:
   - ./ci/drone/linux_script_build
 
 - name: test-1
   depends_on:
   - build
-  image: ziglang/static-base:llvm12-aarch64-5
+  image: ziglang/static-base:llvm13-aarch64-1
   commands:
   - ./ci/drone/linux_script_test 1
 
 - name: test-2
   depends_on:
   - build
-  image: ziglang/static-base:llvm12-aarch64-5
+  image: ziglang/static-base:llvm13-aarch64-1
   commands:
   - ./ci/drone/linux_script_test 2
 
 - name: test-3
   depends_on:
   - build
-  image: ziglang/static-base:llvm12-aarch64-5
+  image: ziglang/static-base:llvm13-aarch64-1
   commands:
   - ./ci/drone/linux_script_test 3
 
+- name: test-4
+  depends_on:
+  - build
+  image: ziglang/static-base:llvm13-aarch64-1
+  commands:
+  - ./ci/drone/linux_script_test 4
+
+- name: test-5
+  depends_on:
+  - build
+  image: ziglang/static-base:llvm13-aarch64-1
+  commands:
+  - ./ci/drone/linux_script_test 5
+
+- name: test-6
+  depends_on:
+  - build
+  image: ziglang/static-base:llvm13-aarch64-1
+  commands:
+  - ./ci/drone/linux_script_test 6
+
+- name: test-7
+  depends_on:
+  - build
+  image: ziglang/static-base:llvm13-aarch64-1
+  commands:
+  - ./ci/drone/linux_script_test 7
+
 - name: finalize
   depends_on:
   - build
   - test-1
   - test-2
   - test-3
-  image: ziglang/static-base:llvm12-aarch64-5
+  - test-4
+  - test-5
+  - test-6
+  - test-7
+  image: ziglang/static-base:llvm13-aarch64-1
   environment:
     SRHT_OAUTH_TOKEN:
       from_secret: SRHT_OAUTH_TOKEN

ci/drone/linux_script_build

@@ -2,6 +2,13 @@
 
 . ./ci/drone/linux_script_base
 
+# Probe CPU/brand details.
+# TODO: `lscpu` is changing package names in EDGE to `util-linux-misc`
+apk update
+apk add util-linux
+echo "lscpu:"
+lscpu | sed 's,^,  : ,'
+
 PREFIX="/deps/local"
 ZIG="$PREFIX/bin/zig"
 TARGET="$TRIPLEARCH-linux-musl"
@@ -13,7 +20,6 @@ export CXX="$ZIG c++ -target $TARGET -mcpu=$MCPU"
 # The `CMAKE_AR` parameter will consider any spaces to
 # be part of the executable path rather than CLI args, so we have
 # to create wrapper scripts for `zig ar` and zig ranlib`.
-
 cat <<'ENDFILE' >$PREFIX/bin/ar
 #!/bin/sh
 /deps/local/bin/zig ar $@

ci/drone/linux_script_finalize

@@ -13,8 +13,7 @@ pip3 install s3cmd
 cd build
 
 mv ../LICENSE "$DISTDIR/"
-# docs are disabled due to: https://github.com/ziglang/zig/issues/8597
-#mv ../zig-cache/langref.html "$DISTDIR/"
+mv ../zig-cache/langref.html "$DISTDIR/"
 mv "$DISTDIR/bin/zig" "$DISTDIR/"
 rmdir "$DISTDIR/bin"
 

ci/drone/linux_script_test

@@ -2,29 +2,44 @@
 
 . ./ci/drone/linux_script_base
 
-# only release-fast builds of test suite due to: https://github.com/ziglang/zig/issues/8597
-#
-# Some test suite components will be missing because they do not support
-# forcing -OReleaseFast
-#
-# see `zig build --help` for the full list of test-* components
+BUILD_FLAGS="-Dskip-non-native"
+
 case "$1" in
   1)
-    steps="\
-      test-stage2 \
-      test-fmt \
-      test-behavior"
+    ./build/zig build $BUILD_FLAGS test-behavior
+    ./build/zig build $BUILD_FLAGS test-compiler-rt
+    ./build/zig build $BUILD_FLAGS test-fmt
+    ./build/zig build $BUILD_FLAGS docs
     ;;
   2)
-    steps="test-std"
+    # Debug
+    ./build/zig build $BUILD_FLAGS test-std -Dskip-release-safe -Dskip-release-fast -Dskip-release-small
     ;;
   3)
-    steps="\
-      test-compiler-rt \
-      test-minilibc \
-      test-compare-output \
-      test-translate-c \
-      test-run-translated-c"
+    # ReleaseSafe
+    ./build/zig build $BUILD_FLAGS test-std -Dskip-debug -Dskip-release-fast -Dskip-release-small
+    ;;
+  4)
+    # Releasefast
+    ./build/zig build $BUILD_FLAGS test-std -Dskip-debug -Dskip-release-safe -Dskip-release-small
+    ;;
+  5)
+    # ReleaseSmall
+    ./build/zig build $BUILD_FLAGS test-std -Dskip-debug -Dskip-release-safe -Dskip-release-fast
+    ;;
+  6)
+    ./build/zig build $BUILD_FLAGS test-minilibc
+    ./build/zig build $BUILD_FLAGS test-compare-output
+    ./build/zig build $BUILD_FLAGS test-standalone -Dskip-release-safe
+    ./build/zig build $BUILD_FLAGS test-stack-traces
+    ./build/zig build $BUILD_FLAGS test-cli
+    ./build/zig build $BUILD_FLAGS test-asm-link
+    ./build/zig build $BUILD_FLAGS test-runtime-safety
+    ./build/zig build $BUILD_FLAGS test-translate-c
+    ;;
+  7)
+    ./build/zig build $BUILD_FLAGS # test building self-hosted without LLVM
+    ./build/zig build $BUILD_FLAGS test-stage2
     ;;
   '')
     echo "error: expecting test group argument"
@@ -35,12 +50,3 @@ case "$1" in
     exit 1
     ;;
 esac
-
-# only release-fast builds of test suite due to: https://github.com/ziglang/zig/issues/8597
-./build/zig build \
-  -Drelease \
-  -Dskip-debug \
-  -Dskip-release-small \
-  -Dskip-release-safe \
-  -Dskip-non-native \
-  $steps

ci/srht/freebsd_script

@@ -7,7 +7,7 @@ sudo pkg update -fq
 sudo pkg install -y cmake py38-s3cmd wget curl jq samurai
 
 ZIGDIR="$(pwd)"
-CACHE_BASENAME="zig+llvm+lld+clang-x86_64-freebsd-gnu-0.8.0-dev.2703+c12704a33"
+CACHE_BASENAME="zig+llvm+lld+clang-x86_64-freebsd-gnu-0.9.0-dev.1243+456d7e5f5"
 PREFIX="$HOME/$CACHE_BASENAME"
 
 cd $HOME
@@ -38,6 +38,11 @@ cmake .. \
     -GNinja
 samu install
 
+# TODO ld.lld: error: undefined symbol: main
+# >>> referenced by crt1_c.c:75 (/usr/src/lib/csu/amd64/crt1_c.c:75)
+# >>>               /usr/lib/crt1.o:(_start)
+#release/bin/zig test ../test/behavior.zig -fno-stage1 -fLLVM -I ../test
+
 # Here we skip some tests to save time.
 release/bin/zig build test -Dskip-compile-errors -Dskip-non-native
 

ci/srht/index.html

@@ -1,737 +0,0 @@
-<!doctype html>
-<html lang="en">
-  <head>
-    <meta charset="utf-8">
-    <title>Releases ⚡ The Zig Programming Language</title>
-    <link rel="icon" href="/favicon.png">
-    <style>
-      body{
-        font-family: system-ui, -apple-system, Roboto, "Segoe UI", sans-serif;
-        line-height: 1.45;
-        margin-left: 0;
-        margin-right: 0;
-      }
-      p {
-        margin: 0.8em 0;
-      }
-      
-      h1, h2, h3, h4 {
-        margin: 0.5em 0 0.5em;
-        line-height: 1.2;
-        font-weight: bold;
-        color: #666;
-      }
-
-      h1 a, h2 a, h3 a, h4 a {
-        text-decoration: none;
-        color: #666;
-      }
-      
-      a.hdr {
-        visibility: hidden;
-      }
-      h1:hover > a.hdr, h2:hover > a.hdr, h3:hover > a.hdr, h4:hover > a.hdr, h5:hover > a.hdr {
-        visibility: visible;
-      }
-
-      h1 { font-size: 2.0em; }
-      
-      h2 { font-size: 1.5em; }
-      
-      h3 { font-size: 1.25em; }
-
-      h4 { font-size: 1.0em; }
-
-      a {
-        color: #2A6286;
-      }
-      
-      a:not(:hover) {
-        text-decoration: none;
-      }
-      
-      th, td {
-        padding: 0.6em;
-        text-align: left;
-        white-space: nowrap;
-      }
-      
-      td {
-        font-size: 0.96em;
-      }
-      
-      th {
-        border-bottom: 2px solid #f2f3f3;
-      }
-      
-      tr:nth-child(even) {
-        background: #f2f3f3;
-      }
-      
-      .container {
-        margin: 0 auto;
-        position: relative;
-        max-width: 1000px;
-      }
-      
-      #navbar {
-        background-color: #737475;
-        padding: 5px 0;
-        border-top: 4px solid #f7a41d;
-        margin-bottom: 30px;
-      }
-      
-      #navbar .navbar-item, #navbar .navbar-item:visited {
-        color: white;
-        padding-right: 5px;
-        padding-left: 5px;
-      }
-
-      .code {
-        font-family: monospace;
-        font-size: 0.8em;
-      }
-
-      #header-image {
-          background-image: url(https://ziglang.org/img/zig-logo-dark.svg);
-          background-repeat: no-repeat;
-          width: 340px;
-          height: 90px;
-          display: block;
-          padding: 0;
-          margin: 0;
-      }
-
-      @media (prefers-color-scheme: dark) {
-        body{
-          background-color:#111;
-          color: #bbb;
-        }
-        a {
-          color: #88f;
-        }
-        table, th, td {
-          border-color: grey;
-        }
-        tr:nth-child(even) {
-          background: #1e1e1e;
-        }
-        h1 a, h2 a, h3 a, h4 a, h5 a {
-          color: #aaa;
-        }
-        #header-image {
-            background-image: url(https://ziglang.org/img/zig-logo-light.svg);
-        }
-      }
-    </style>
-  </head>
-  <body>
-    <div class="container">
-      <a href="/"><span id="header-image"></span></a>
-    </div>
-    <nav id="navbar">
-      <div class="container">
-        <a href="/download/" class="navbar-item">Download &amp; Documentation</a>
-        <a href="https://github.com/ziglang/zig" class="navbar-item">Source Code</a>
-        <a href="/news/" class="navbar-item">News</a>
-        <a href="https://github.com/ziglang/zig/wiki/Community" class="navbar-item">Join a Community</a>
-        <a href="/zsf/" class="navbar-item">
-          <svg style="color: #ea4aaa; vertical-align: middle;fill: currentColor; margin-right: 5px" viewBox="0 0 12 16" version="1.1" width="12" height="16" aria-hidden="true"><path fill-rule="evenodd" d="M9 2c-.97 0-1.69.42-2.2 1-.51.58-.78.92-.8 1-.02-.08-.28-.42-.8-1-.52-.58-1.17-1-2.2-1-1.632.086-2.954 1.333-3 3 0 .52.09 1.52.67 2.67C1.25 8.82 3.01 10.61 6 13c2.98-2.39 4.77-4.17 5.34-5.33C11.91 6.51 12 5.5 12 5c-.047-1.69-1.342-2.913-3-3z"></path></svg>Sponsor the Zig Software Foundation</a>
-      </div>
-    </nav>
-    <div class="container">
-      <h1>Releases</h1>
-      <p>You can also
-      <a href="https://github.com/ziglang/zig/wiki/Install-Zig-from-a-Package-Manager">install Zig from a package manager</a>.
-      </p>
-      <p>
-      There is a <a href="index.json">JSON version of this page</a>.
-      </p>
-
-      <h2 id="release-master">master</h2>
-      <ul>
-        <li>{{MASTER_DATE}}</li>
-        <li><a href="/documentation/master/">Language Reference</a></li>
-        <li><a href="/documentation/master/std/">Standard Library Documentation</a> (experimental)</li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{SRC_TARBALL}}">{{SRC_TARBALL}}</a></td>
-            <td>Source</td>
-            <td>{{SRC_BYTESIZE}}</td>
-            <td class="code">{{SRC_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{X86_64_LINUX_TARBALL}}">{{X86_64_LINUX_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{X86_64_LINUX_BYTESIZE}}</td>
-            <td class="code">{{X86_64_LINUX_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{AARCH64_LINUX_TARBALL}}">{{AARCH64_LINUX_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{AARCH64_LINUX_BYTESIZE}}</td>
-            <td class="code">{{AARCH64_LINUX_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{X86_64_WINDOWS_TARBALL}}">{{X86_64_WINDOWS_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{X86_64_WINDOWS_BYTESIZE}}</td>
-            <td class="code">{{X86_64_WINDOWS_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{X86_64_MACOS_TARBALL}}">{{X86_64_MACOS_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{X86_64_MACOS_BYTESIZE}}</td>
-            <td class="code">{{X86_64_MACOS_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{AARCH64_MACOS_TARBALL}}">{{AARCH64_MACOS_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{AARCH64_MACOS_BYTESIZE}}</td>
-            <td class="code">{{AARCH64_MACOS_SHASUM}}</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/builds/{{X86_64_FREEBSD_TARBALL}}">{{X86_64_FREEBSD_TARBALL}}</a></td>
-            <td>Binary</td>
-            <td>{{X86_64_FREEBSD_BYTESIZE}}</td>
-            <td class="code">{{X86_64_FREEBSD_SHASUM}}</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.7.1">0.7.1</h2>
-      <ul>
-        <li>2020-12-13</li>
-        <li><a href="0.7.1/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.7.1/">Language Reference</a></li>
-        <li><a href="/documentation/0.7.1/std">Standard Library Documentation</a> (experimental)</li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-0.7.1.tar.xz">zig-0.7.1.tar.xz</a></td>
-            <td>Source</td>
-            <td>11MiB</td>
-            <td class="code">2db3b944ab368d955b48743d9f7c963b8f96de1a441ba5a35e197237cc6dae44</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-bootstrap-0.7.1.tar.xz">zig-bootstrap-0.7.1.tar.xz</a></td>
-            <td>Source</td>
-            <td>39MiB</td>
-            <td class="code">040f27c1fae4b0cac0a2782aecdb691f6a2f8e89db6a6ed35024c31c304fd9b2</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-freebsd-x86_64-0.7.1.tar.xz">zig-freebsd-x86_64-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>38MiB</td>
-            <td class="code">e73c1dca35791a3183fdd5ecde0443ebbe180942efceafe651886034fb8def09</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-linux-aarch64-0.7.1.tar.xz">zig-linux-aarch64-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>33MiB</td>
-            <td class="code">48ec90eba407e4587ddef7eecef25fec7e13587eb98e3b83c5f2f5fff2a5cbe7</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-linux-armv7a-0.7.1.tar.xz">zig-linux-armv7a-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>35MiB</td>
-            <td class="code">5a0662e07b4c4968665e1f97558f8591f6facec45d2e0ff5715e661743107ceb</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-linux-i386-0.7.1.tar.xz">zig-linux-i386-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>38MiB</td>
-            <td class="code">4882e052e5f83690bd0334bb4fc1702b5403cb3a3d2aa63fd7d6043d8afecba3</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-linux-riscv64-0.7.1.tar.xz">zig-linux-riscv64-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36MiB</td>
-            <td class="code">187294bfd35983348c3fe042901b42e67e7e36ab7f77a5f969d21c0051f4d21f</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-linux-x86_64-0.7.1.tar.xz">zig-linux-x86_64-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>37MiB</td>
-            <td class="code">18c7b9b200600f8bcde1cd8d7f1f578cbc3676241ce36d771937ce19a8159b8d</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-macos-x86_64-0.7.1.tar.xz">zig-macos-x86_64-0.7.1.tar.xz</a></td>
-            <td>Binary</td>
-            <td>35MiB</td>
-            <td class="code">845cb17562978af0cf67e3993f4e33330525eaf01ead9386df9105111e3bc519</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-windows-i386-0.7.1.zip">zig-windows-i386-0.7.1.zip</a></td>
-            <td>Binary</td>
-            <td>52MiB</td>
-            <td class="code">a1b9a7421e13153e07fd2e2c93ff29aad64d83105b8fcdafa633dbe689caf1c0</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.1/zig-windows-x86_64-0.7.1.zip">zig-windows-x86_64-0.7.1.zip</a></td>
-            <td>Binary</td>
-            <td>53MiB</td>
-            <td class="code">4818a8a65b4672bc52c0ae7f14d014e0eb8caf10f12c0745176820384cea296a</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.7.0">0.7.0</h2>
-      <ul>
-        <li>2020-11-08</li>
-        <li><a href="0.7.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.7.0/">Language Reference</a></li>
-        <li><a href="/documentation/0.7.0/std">Standard Library Documentation</a> (experimental)</li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-0.7.0.tar.xz">zig-0.7.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>11MiB</td>
-            <td class="code">0efd2cf6c3b05723db80e9cf193bc55150bba84ca41f855a90f53fc756445f83</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-bootstrap-0.7.0.tar.xz">zig-bootstrap-0.7.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>39MiB</td>
-            <td class="code">f073beaf5c53c8c57c0d374cbfcb332ef92ad703173edba0d9e0f2ed28401b72</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-freebsd-x86_64-0.7.0.tar.xz">zig-freebsd-x86_64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>34MiB</td>
-            <td class="code">a0c926272ee4ae720034b4a6a1dc98399d76156dd84182554740f0ca8a41fc99</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-linux-aarch64-0.7.0.tar.xz">zig-linux-aarch64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>32MiB</td>
-            <td class="code">f89933bac87d44be82325754ff88423020c81c7032a6fc41cfeb81e982eeab9b</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-linux-armv7a-0.7.0.tar.xz">zig-linux-armv7a-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>34MiB</td>
-            <td class="code">011c267e25a96ee160505a560c441daa045359a9d50e13ab1bada9d75c95db2d</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-linux-i386-0.7.0.tar.xz">zig-linux-i386-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>37MiB</td>
-            <td class="code">4bb2072cd363bcb1cbeb4872ff5cbc1f683b02d0cc1f90c46e3ea7422ce53222</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-linux-riscv64-0.7.0.tar.xz">zig-linux-riscv64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36MiB</td>
-            <td class="code">40dff81faa6f232ac40abbf88b9371f3cc932b6e09c423b94387c9ea580cb7be</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-linux-x86_64-0.7.0.tar.xz">zig-linux-x86_64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36MiB</td>
-            <td class="code">e619b1c6094c095b932767f527aee2507f847ea981513ff8a08aab0fd730e0ac</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-macos-aarch64-0.7.0.tar.xz">zig-macos-aarch64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>33MiB</td>
-            <td class="code">338238035734db74ea4f30e500a4893bf741d38305c10952d5e39fa05bdb057d</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-macos-x86_64-0.7.0.tar.xz">zig-macos-x86_64-0.7.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>35MiB</td>
-            <td class="code">94063f9a311cbbf7a2e0a12295e09437182cf950f18cb0eb30ea9893f3677f24</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-windows-i386-0.7.0.zip">zig-windows-i386-0.7.0.zip</a></td>
-            <td>Binary</td>
-            <td>51MiB</td>
-            <td class="code">b1e520aacbfbd645ff3521b3eb4d44166d9a0288b8725e4b001f8b50a425eb2e</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.7.0/zig-windows-x86_64-0.7.0.zip">zig-windows-x86_64-0.7.0.zip</a></td>
-            <td>Binary</td>
-            <td>52MiB</td>
-            <td class="code">965f56c0a36f9cda2125e3a348bc654f7f155e2804c3667d231775ec228f8553</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.6.0">0.6.0</h2>
-      <ul>
-        <li>2020-04-13</li>
-        <li><a href="0.6.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.6.0">Language Reference</a></li>
-        <li><a href="/documentation/0.6.0/std">Standard Library Documentation</a> (experimental)</li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-0.6.0.tar.xz">zig-0.6.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>9.9MiB</td>
-            <td class="code">5d167dc19354282dd35dd17b38e99e1763713b9be8a4ba9e9e69284e059e7204</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-bootstrap-0.6.0.tar.xz">zig-bootstrap-0.6.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>36.7MiB</td>
-            <td class="code">5e0e4dc878b3dd0c1852a442b174f0732e8c07869a8fcd226b71a93b89b381ab</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-freebsd-x86_64-0.6.0.tar.xz">zig-freebsd-x86_64-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36MiB</td>
-            <td class="code">190ff79c1eb56805a315d7c7a51082e32f62926250c0702b36760c225e1634a3</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-aarch64-0.6.0.tar.xz">zig-linux-aarch64-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36MiB</td>
-            <td class="code">e7520efd42cfa02be48c2e430d08fe1f3cbb999d21d9f0d3ffd0febb976b2f41</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-armv6kz-0.6.0.tar.xz">zig-linux-armv6kz-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>38MiB</td>
-            <td class="code">36b6493b3fed43eb1f0000e765798ad31a6bb7d7fd3f553ac1c3761dbc919b82</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-armv7a-0.6.0.tar.xz">zig-linux-armv7a-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>38MiB</td>
-            <td class="code">946969abe357def95ca9cbbfcebfcf2d90cf967bcd3f48ee87662e32d91d8f35</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-i386-0.6.0.tar.xz">zig-linux-i386-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>43MiB</td>
-            <td class="code">a97a2f9ae21575743cdd763c1917d49400d83fc562ef64582b18bade43eb24ce</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-riscv64-0.6.0.tar.xz">zig-linux-riscv64-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>41MiB</td>
-            <td class="code">68ddee43f7503c8ae5f26a921f3602c34719a02ed2241f528c0b8b888cc14b38</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-linux-x86_64-0.6.0.tar.xz">zig-linux-x86_64-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>43MiB</td>
-            <td class="code">08fd3c757963630645441c2772362e9c2294020c44f14fce1b89f45de0dc1253</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-macos-x86_64-0.6.0.tar.xz">zig-macos-x86_64-0.6.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>41MiB</td>
-            <td class="code">17270360e87ddc49f737e760047b2fac49f1570a824a306119b1194ac4093895</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-windows-i386-0.6.0.zip">zig-windows-i386-0.6.0.zip</a></td>
-            <td>Binary</td>
-            <td>58MiB</td>
-            <td class="code">3b0a02618743e92175990dc6d1a787bb95ff62c4cda016f1c14c7786f575f8ca</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.6.0/zig-windows-x86_64-0.6.0.zip">zig-windows-x86_64-0.6.0.zip</a></td>
-            <td>Binary</td>
-            <td>47MiB</td>
-            <td class="code">c3b897832523e1026e10b2d8d55d7f895185c0a27a63681f3a23219c3f1c38f4</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.5.0">0.5.0</h2>
-      <ul>
-        <li>2019-09-30</li>
-        <li><a href="0.5.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.5.0">Documentation</a></li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.5.0/zig-0.5.0.tar.xz">zig-0.5.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>10.4MiB</td>
-            <td class="code">55ae16960f152bcb9cf98b4f8570902d0e559a141abf927f0d3555b7cc838a31</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.5.0/zig-linux-x86_64-0.5.0.tar.xz">zig-linux-x86_64-0.5.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>39.0MiB</td>
-            <td class="code">43e8f8a8b8556edd373ddf9c1ef3ca6cf852d4d09fe07d5736d12fefedd2b4f7</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.5.0/zig-windows-x86_64-0.5.0.zip">zig-windows-x86_64-0.5.0.zip</a></td>
-            <td>Binary</td>
-            <td>42.8MiB</td>
-            <td class="code">58141323db8d84a5af62746be5f9140bc161ee760ef33dc91a887bf9ac021976</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.5.0/zig-macos-x86_64-0.5.0.tar.xz">zig-macos-x86_64-0.5.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>36.1MiB</td>
-            <td class="code">28702cc05745c7c0bd450487d5f4091bf0a1ad279b35eb9a640ce3e3a15b300d</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.5.0/zig-freebsd-x86_64-0.5.0.tar.xz">zig-freebsd-x86_64-0.5.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>32.1MiB</td>
-            <td class="code">9e1f4d36c3d584c0aa01f20eb4cd0a0eef3eee5af23e483b8414de55feab6ab6</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.4.0">0.4.0</h2>
-      <ul>
-        <li>2019-04-08</li>
-        <li><a href="0.4.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.4.0">Documentation</a></li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.4.0/zig-0.4.0.tar.xz">zig-0.4.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>5.1MiB</td>
-            <td class="code">fec1f3f6b359a3d942e0a7f9157b3b30cde83927627a0e1ea95c54de3c526cfc</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.4.0/zig-linux-x86_64-0.4.0.tar.xz">zig-linux-x86_64-0.4.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>31.4MiB</td>
-            <td class="code">fb1954e2fb556a01f8079a08130e88f70084e08978ff853bb2b1986d8c39d84e</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.4.0/zig-windows-x86_64-0.4.0.zip">zig-windows-x86_64-0.4.0.zip</a></td>
-            <td>Binary</td>
-            <td>34.1MiB</td>
-            <td class="code">fbc3dd205e064c263063f69f600bedb18e3d0aa2efa747a63ef6cafb6d73f127</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.4.0/zig-macos-x86_64-0.4.0.tar.xz">zig-macos-x86_64-0.4.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>29.4MiB</td>
-            <td class="code">67c932982484d017c5111e54af9f33f15e8e05c6bc5346a55e04052159c964a8</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.4.0/zig-freebsd-x86_64-0.4.0.tar.xz">zig-freebsd-x86_64-0.4.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>26.0MiB</td>
-            <td class="code">3d557c91ac36d8262eb1733bb5f261c95944f9b635e43386e3d00a3272818c30</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.3.0">0.3.0</h2>
-      <ul>
-        <li>2018-09-28</li>
-        <li><a href="0.3.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.3.0">Documentation</a></li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.3.0/zig-0.3.0.tar.xz">zig-0.3.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>2.2MiB</td>
-            <td class="code">d70af604f3a8622f3393d93abb3e056bf60351e32d121e6fa4fe03d8d41e1f5a</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.3.0/zig-linux-x86_64-0.3.0.tar.xz">zig-linux-x86_64-0.3.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>24.0MiB</td>
-            <td class="code">b378d0aae30cb54f28494e7bc4efbc9bfb6326f47bfb302e8b5287af777b2f3c</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.3.0/zig-windows-x86_64-0.3.0.zip">zig-windows-x86_64-0.3.0.zip</a></td>
-            <td>Binary</td>
-            <td>21.5MiB</td>
-            <td class="code">bb568c03950958f8bb3472139c3ab5ed74547c8c694ab50f404c202faf51baf4</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.3.0/zig-macos-x86_64-0.3.0.tar.xz">zig-macos-x86_64-0.3.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>22.6MiB</td>
-            <td class="code">19dec1f1943ab7be26823376d466f7e456143deb34e17502778a949034dc2e7e</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.2.0">0.2.0</h2>
-      <ul>
-        <li>2018-03-15</li>
-        <li><a href="0.2.0/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.2.0">Documentation</a></li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.2.0/zig-0.2.0.tar.xz">zig-0.2.0.tar.xz</a></td>
-            <td>Source</td>
-            <td>1.9MiB</td>
-            <td class="code">29c9beb172737f4d5019b88ceae829ae8bc6512fb4386cfbf895ae2b42aa6965</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.2.0/zig-linux-x86_64-0.2.0.tar.xz">zig-linux-x86_64-0.2.0.tar.xz</a></td>
-            <td>Binary</td>
-            <td>23.5MiB</td>
-            <td class="code">209c6fb745d42474c0a73d6f291c7ae3a38b6a1b6b641eea285a7f840cc1a890</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.2.0/zig-win64-0.2.0.zip">zig-win64-0.2.0.zip</a></td>
-            <td>Binary</td>
-            <td>20.6MiB</td>
-            <td class="code">4f8a2979941a1f081ec8e545cca0b72608c0db1c5a3fd377a94db40649dcd3d4</td>
-          </tr>
-        </tbody>
-      </table>
-
-      <h2 id="release-0.1.1">0.1.1</h2>
-      <ul>
-        <li>2017-10-17</li>
-        <li><a href="0.1.1/release-notes.html">Release Notes</a></li>
-        <li><a href="/documentation/0.1.1">Documentation</a></li>
-      </ul>
-      <table>
-        <colgroup>
-          <col width="40%">
-          <col width="10%">
-          <col width="10%">
-        </colgroup>
-        <thead>
-          <tr>
-            <th>Filename</th>
-            <th>Kind</th>
-            <th>Size</th>
-            <th>Sha256</th>
-          </tr>
-        </thead>
-        <tbody>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.1.1/zig-0.1.1.tar.xz">zig-0.1.1.tar.xz</a></td>
-            <td>Source</td>
-            <td>1.62MiB</td>
-            <td class="code">ffca0cfb263485287e19cc997b08701fcd5f24b700345bcdc3dd8074f5a104e0</td>
-          </tr>
-          <tr>
-            <td><a href="https://ziglang.org/download/0.1.1/zig-win64-0.1.1.zip">zig-win64-0.1.1.zip</a></td>
-            <td>Binary</td>
-            <td>19.3MiB</td>
-            <td class="code">6fc88bef531af7e567fe30bf60da1487b86833cbee84c7a2f3e317030aa5b660</td>
-          </tr>
-        </tbody>
-      </table>
-    </div>
-  </body>
-</html>

ci/srht/index.json

@@ -40,6 +40,138 @@
       "size": "{{AARCH64_LINUX_BYTESIZE}}"
     }
   },
+  "0.8.1": {
+    "date": "2021-09-06",
+    "docs": "https://ziglang.org/documentation/0.8.1/",
+    "stdDocs": "https://ziglang.org/documentation/0.8.1/std/",
+    "notes": "https://ziglang.org/download/0.8.1/release-notes.html",
+    "src": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-0.8.1.tar.xz",
+      "shasum": "8c428e14a0a89cb7a15a6768424a37442292858cdb695e2eb503fa3c7bf47f1a",
+      "size": "12650228"
+    },
+    "bootstrap": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-bootstrap-0.8.1.tar.xz",
+      "shasum": "fa1239247f830ecd51c42537043f5220e4d1dfefdc54356fa419616a0efb3902",
+      "size": "43613464"
+    },
+    "x86_64-freebsd": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-freebsd-x86_64-0.8.1.tar.xz",
+      "shasum": "fc4f6478bcf3a9fce1b8ef677a91694f476dd35be6d6c9c4f44a8b76eedbe176",
+      "size": "39150924"
+    },
+    "aarch64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-linux-aarch64-0.8.1.tar.xz",
+      "shasum": "2166dc9f2d8df387e8b4122883bb979d739281e1ff3f3d5483fec3a23b957510",
+      "size": "37605932"
+    },
+    "armv7a-linux": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-linux-armv7a-0.8.1.tar.xz",
+      "shasum": "5ba58141805e2519f38cf8e715933cbf059f4f3dade92c71838cce341045de05",
+      "size": "39185876"
+    },
+    "i386-linux": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-linux-i386-0.8.1.tar.xz",
+      "shasum": "2f3e84f30492b5f1c5f97cecc0166f07a8a8d50c5f85dbb3a6ef2a4ee6f915e6",
+      "size": "44782932"
+    },
+    "riscv64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-linux-riscv64-0.8.1.tar.xz",
+      "shasum": "4adfaf147b025917c03367462fe5018aaa9edbc6439ef9cd0da2b074ae960554",
+      "size": "41234480"
+    },
+    "x86_64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-linux-x86_64-0.8.1.tar.xz",
+      "shasum": "6c032fc61b5d77a3f3cf781730fa549f8f059ffdb3b3f6ad1c2994d2b2d87983",
+      "size": "41250060"
+    },
+    "aarch64-macos": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-macos-aarch64-0.8.1.tar.xz",
+      "shasum": "5351297e3b8408213514b29c0a938002c5cf9f97eee28c2f32920e1227fd8423",
+      "size": "35340712"
+    },
+    "x86_64-macos": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-macos-x86_64-0.8.1.tar.xz",
+      "shasum": "16b0e1defe4c1807f2e128f72863124bffdd906cefb21043c34b673bf85cd57f",
+      "size": "39946200"
+    },
+    "i386-windows": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-windows-i386-0.8.1.zip",
+      "shasum": "099605051eb0452a947c8eab8fbbc7e43833c8376d267e94e41131c289a1c535",
+      "size": "64152358"
+    },
+    "x86_64-windows": {
+      "tarball": "https://ziglang.org/download/0.8.1/zig-windows-x86_64-0.8.1.zip",
+      "shasum": "43573db14cd238f7111d6bdf37492d363f11ecd1eba802567a172f277d003926",
+      "size": "61897838"
+    }
+  },
+  "0.8.0": {
+    "date": "2021-06-04",
+    "docs": "https://ziglang.org/documentation/0.8.0/",
+    "stdDocs": "https://ziglang.org/documentation/0.8.0/std/",
+    "notes": "https://ziglang.org/download/0.8.0/release-notes.html",
+    "src": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-0.8.0.tar.xz",
+      "shasum": "03a828d00c06b2e3bb8b7ff706997fd76bf32503b08d759756155b6e8c981e77",
+      "size": "12614896"
+    },
+    "bootstrap": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-bootstrap-0.8.0.tar.xz",
+      "shasum": "10600bc9c01f92e343f40d6ecc0ad05d67d27c3e382bce75524c0639cd8ca178",
+      "size": "43574248"
+    },
+    "x86_64-freebsd": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-freebsd-x86_64-0.8.0.tar.xz",
+      "shasum": "0d3ccc436c8c0f50fd55462f72f8492d98723c7218ffc2a8a1831967d81b4bdc",
+      "size": "39125332"
+    },
+    "aarch64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-linux-aarch64-0.8.0.tar.xz",
+      "shasum": "ee204ca2c2037952cf3f8b10c609373a08a291efa4af7b3c73be0f2b27720470",
+      "size": "37575428"
+    },
+    "armv7a-linux": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-linux-armv7a-0.8.0.tar.xz",
+      "shasum": "d00b8bd97b79f45d6f5da956983bafeaa082e6c2ae8c6e1c6d4faa22fa29b320",
+      "size": "38884212"
+    },
+    "i386-linux": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-linux-i386-0.8.0.tar.xz",
+      "shasum": "96e43ee6ed81c3c63401f456bd1c58ee6d42373a43cb324f5cf4974ca0998865",
+      "size": "42136032"
+    },
+    "riscv64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-linux-riscv64-0.8.0.tar.xz",
+      "shasum": "75997527a78cdab64c40c43d9df39c01c4cdb557bb3992a869838371a204cfea",
+      "size": "40016268"
+    },
+    "x86_64-linux": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-linux-x86_64-0.8.0.tar.xz",
+      "shasum": "502625d3da3ae595c5f44a809a87714320b7a40e6dff4a895b5fa7df3391d01e",
+      "size": "41211184"
+    },
+    "aarch64-macos": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-macos-aarch64-0.8.0.tar.xz",
+      "shasum": "b32d13f66d0e1ff740b3326d66a469ee6baddbd7211fa111c066d3bd57683111",
+      "size": "35292180"
+    },
+    "x86_64-macos": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-macos-x86_64-0.8.0.tar.xz",
+      "shasum": "279f9360b5cb23103f0395dc4d3d0d30626e699b1b4be55e98fd985b62bc6fbe",
+      "size": "39969312"
+    },
+    "i386-windows": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-windows-i386-0.8.0.zip",
+      "shasum": "b6ec9aa6cd6f3872fcb30d43ff411802d82008a0c4142ee49e208a09b2c1c5fe",
+      "size": "61507213"
+    },
+    "x86_64-windows": {
+      "tarball": "https://ziglang.org/download/0.8.0/zig-windows-x86_64-0.8.0.zip",
+      "shasum": "8580fbbf3afb72e9b495c7f8aeac752a03475ae0bbcf5d787f3775c7e1f4f807",
+      "size": "61766193"
+    }
+  },
   "0.7.1": {
     "date": "2020-12-13",
     "docs": "https://ziglang.org/documentation/0.7.1/",

ci/srht/netbsd_script

@@ -0,0 +1,97 @@
+#!/bin/sh
+
+set -x
+set -e
+
+sudo pkgin -y update
+sudo pkgin -y upgrade
+sudo pkgin -y install cmake samurai curl jq py27-s3cmd wget clang
+
+ZIGDIR="$(pwd)"
+CACHE_BASENAME="zig+llvm+lld+clang-x86_64-netbsd-gnu-0.8.0"
+PREFIX="$HOME/$CACHE_BASENAME"
+
+cd $HOME
+wget -nv "https://ziglang.org/deps/$CACHE_BASENAME.tar.xz"
+tar xf "$CACHE_BASENAME.tar.xz"
+
+cd $ZIGDIR
+
+# Make the `zig version` number consistent.
+# This will affect the cmake command below.
+git config core.abbrev 9
+git fetch --unshallow || true
+git fetch --tags
+
+# SourceHut reports that it is a terminal that supports escape codes, but it
+# is a filthy liar. Here we tell Zig to not try to send any terminal escape
+# codes to show progress.
+export TERM=dumb
+
+mkdir build
+cd build
+export CC="$PREFIX/bin/zigcc"
+export CXX="$PREFIX/bin/zigcxx"
+echo "#!/bin/sh
+env CC=\"clang\" $PREFIX/bin/zig cc \"\$@\"" > $CC
+echo "#!/bin/sh
+env CC=\"clang\" $PREFIX/bin/zig c++ \"\$@\"" > $CXX
+chmod +x $CC
+chmod +x $CXX
+
+cmake .. \
+    -DCMAKE_BUILD_TYPE=Release \
+    -DCMAKE_PREFIX_PATH=$PREFIX \
+    "-DCMAKE_INSTALL_PREFIX=$(pwd)/release" \
+    -DZIG_STATIC=ON \
+    -DZIG_TARGET_TRIPLE=x86_64-netbsd-none \
+    -GNinja
+samu install
+
+unset CC
+unset CXX
+
+# TODO ld.lld: error: undefined symbol: main
+#>>> referenced by crt0-common.c
+#>>>               /usr/lib/crt0.o:(___start)
+#release/bin/zig test ../test/behavior.zig -fno-stage1 -fLLVM -I ../test
+
+# Here we skip some tests to save time.
+release/bin/zig build test -Dskip-compile-errors -Dskip-non-native
+
+if [ -f ~/.s3cfg ]; then
+  mv ../LICENSE release/
+  mv ../zig-cache/langref.html release/
+  mv release/bin/zig release/
+  rmdir release/bin
+
+  GITBRANCH=$(basename $GITHUB_REF)
+  VERSION=$(release/zig version)
+  DIRNAME="zig-netbsd-x86_64-$VERSION"
+  TARBALL="$DIRNAME.tar.xz"
+  mv release "$DIRNAME"
+  tar cfJ "$TARBALL" "$DIRNAME"
+
+  s3cmd put -P --add-header="cache-control: public, max-age=31536000, immutable" "$TARBALL" s3://ziglang.org/builds/
+
+  SHASUM=$(shasum -a 256 $TARBALL | cut '-d ' -f1)
+  BYTESIZE=$(wc -c < $TARBALL)
+
+  JSONFILE="netbsd-$GITBRANCH.json"
+  touch $JSONFILE
+  echo "{\"tarball\": \"$TARBALL\"," >>$JSONFILE
+  echo "\"shasum\": \"$SHASUM\"," >>$JSONFILE
+  echo "\"size\": \"$BYTESIZE\"}" >>$JSONFILE
+
+  s3cmd put -P --add-header="Cache-Control: max-age=0, must-revalidate" "$JSONFILE" "s3://ziglang.org/builds/$JSONFILE"
+  s3cmd put -P "$JSONFILE" "s3://ziglang.org/builds/x86_64-netbsd-$VERSION.json"
+
+  if [ "$GITBRANCH" = "master" ]; then 
+    # avoid leaking oauth token
+    set +x
+
+    OAUTH_TOKEN="$(cat ~/.oauth_token)"
+    cd "$ZIGDIR"
+    ./ci/srht/on_master_success "$VERSION" "$OAUTH_TOKEN"
+  fi
+fi

ci/srht/on_master_success

@@ -15,6 +15,16 @@ VERSION="$1"
 OAUTH_TOKEN="$2"
 YML_FILE="tmp.yml"
 
+if [ -z "$VERSION" ]; then
+  echo "missing VERSION parameter"
+  exit 1
+fi
+
+if [ -z "$OAUTH_TOKEN" ]; then
+  echo "missing OAUTH_TOKEN parameter"
+  exit 1
+fi
+
 cat <<EOF >"$YML_FILE"
 image: alpine/latest
 packages:

ci/srht/update-download-page.zig

@@ -6,15 +6,10 @@ pub fn main() !void {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
     defer arena.deinit();
 
-    const allocator = &arena.allocator;
+    const allocator = arena.allocator();
 
     const out_dir = "out";
     try std.fs.cwd().makePath(out_dir);
-    {
-        const out_file = out_dir ++ path.sep_str ++ "index.html";
-        const in_file = "index.html";
-        try render(allocator, in_file, out_file, .html);
-    }
     {
         const out_file = out_dir ++ path.sep_str ++ "index.json";
         const in_file = "index.json";
@@ -23,7 +18,7 @@ pub fn main() !void {
 }
 
 fn render(
-    allocator: *mem.Allocator,
+    allocator: mem.Allocator,
     in_file: []const u8,
     out_file: []const u8,
     fmt: enum {
@@ -79,7 +74,7 @@ fn render(
                             try writer.writeAll(trimmed);
                         }
                     } else {
-                        std.debug.warn("line {d}: missing variable: {s}\n", .{ line, var_name });
+                        std.debug.print("line {d}: missing variable: {s}\n", .{ line, var_name });
                         try writer.writeAll("(missing)");
                     }
                     state = State.EndBrace;
@@ -91,7 +86,7 @@ fn render(
                     state = State.Start;
                 },
                 else => {
-                    std.debug.warn("line {d}: invalid byte: '0x{x}'", .{ line, byte });
+                    std.debug.print("line {d}: invalid byte: '0x{x}'", .{ line, byte });
                     std.process.exit(1);
                 },
             },

ci/srht/update_download_page

@@ -15,6 +15,7 @@ X86_64_WINDOWS_JSON_URL="https://ziglang.org/builds/x86_64-windows-$VERSION.json
 AARCH64_MACOS_JSON_URL="https://ziglang.org/builds/aarch64-macos-$VERSION.json"
 X86_64_MACOS_JSON_URL="https://ziglang.org/builds/x86_64-macos-$VERSION.json"
 X86_64_FREEBSD_JSON_URL="https://ziglang.org/builds/x86_64-freebsd-$VERSION.json"
+#X86_64_NETBSD_JSON_URL="https://ziglang.org/builds/x86_64-netbsd-$VERSION.json"
 
 # If any of these fail, it's not really this job failing; rather we have detected
 # that this job will be called again later when other jobs have completed.
@@ -24,6 +25,7 @@ curl --fail -I "$X86_64_WINDOWS_JSON_URL" >/dev/null || exit 0
 curl --fail -I "$AARCH64_MACOS_JSON_URL" >/dev/null || exit 0
 curl --fail -I "$X86_64_MACOS_JSON_URL" >/dev/null || exit 0
 curl --fail -I "$X86_64_FREEBSD_JSON_URL" >/dev/null || exit 0
+#curl --fail -I "$X86_64_NETBSD_JSON_URL" >/dev/null || exit 0
 
 # Without --user, this gave me:
 # ERROR: Could not install packages due to an EnvironmentError: [Errno 13] Permission denied
@@ -84,16 +86,21 @@ export X86_64_FREEBSD_TARBALL="$(echo "$X86_64_FREEBSD_JSON" | jq .tarball -r)"
 export X86_64_FREEBSD_BYTESIZE="$(echo "$X86_64_FREEBSD_JSON" | jq .size -r)"
 export X86_64_FREEBSD_SHASUM="$(echo "$X86_64_FREEBSD_JSON" | jq .shasum -r)"
 
+#X86_64_NETBSD_JSON=$(curl --fail "$X86_64_NETBSD_JSON_URL" || exit 1)
+#export X86_64_NETBSD_TARBALL="$(echo "$X86_64_NETBSD_JSON" | jq .tarball -r)"
+#export X86_64_NETBSD_BYTESIZE="$(echo "$X86_64_NETBSD_JSON" | jq .size -r)"
+#export X86_64_NETBSD_SHASUM="$(echo "$X86_64_NETBSD_JSON" | jq .shasum -r)"
+
 export MASTER_DATE="$(date +%Y-%m-%d)"
 export MASTER_VERSION="$VERSION"
 
+# Create index.json and update the website repo.
 cd "$SRCTARBALLDIR/ci/srht"
 "$ZIG" run update-download-page.zig
 CIDIR="$(pwd)"
 
-# Create index.json and index.html and update the website repo.
 cd "$HOME"
-git clone git@github.com:ziglang/www.ziglang.org.git
+git clone --depth 1 git@github.com:ziglang/www.ziglang.org.git
 cd www.ziglang.org
 WWWDIR="$(pwd)"
 

ci/zinc/drone.yml

@@ -0,0 +1,46 @@
+---
+kind: pipeline
+type: docker
+name: x86_64-linux
+platform:
+  os: linux
+  arch: amd64
+workspace:
+  path: /workspace
+
+steps:
+- name: probe
+  image: ci/debian-amd64:11.1-2
+  commands:
+  - ./ci/zinc/linux_probe.sh
+
+- name: build
+  image: ci/debian-amd64:11.1-2
+  commands:
+  - ./ci/zinc/linux_build.sh
+
+- name: test
+  depends_on:
+  - build
+  image: ci/debian-amd64:11.1-2
+  commands:
+  - ./ci/zinc/linux_test.sh
+
+- name: package
+  depends_on:
+  - test
+  when:
+    branch:
+      - master
+    event:
+      - push
+  image: ci/debian-amd64:11.1-2
+  environment:
+    AWS_ACCESS_KEY_ID:
+      from_secret: AWS_ACCESS_KEY_ID
+    AWS_SECRET_ACCESS_KEY:
+      from_secret: AWS_SECRET_ACCESS_KEY
+    SRHT_OAUTH_TOKEN:
+      from_secret: SRHT_OAUTH_TOKEN
+  commands:
+  - ./ci/zinc/linux_package.sh

ci/zinc/linux_base.sh

@@ -0,0 +1,28 @@
+#!/bin/sh
+
+# https://docs.drone.io/pipeline/docker/syntax/workspace/
+#
+# Drone automatically creates a temporary volume, known as your workspace,
+# where it clones your repository. The workspace is the current working
+# directory for each step in your pipeline.
+#
+# Because the workspace is a volume, filesystem changes are persisted between
+# pipeline steps. In other words, individual steps can communicate and share
+# state using the filesystem.
+#
+# Workspace volumes are ephemeral. They are created when the pipeline starts
+# and destroyed after the pipeline completes.
+
+set -x
+set -e
+
+ARCH="$(uname -m)"
+JOBS="-j$(nproc)"
+
+DEPS_LOCAL="/deps/local"
+WORKSPACE="$DRONE_WORKSPACE"
+
+DEBUG_STAGING="$WORKSPACE/_debug/staging"
+RELEASE_STAGING="$WORKSPACE/_release/staging"
+
+export PATH=$DEPS_LOCAL/bin:$PATH

ci/zinc/linux_build.sh

@@ -0,0 +1,72 @@
+#!/bin/sh
+
+. ./ci/zinc/linux_base.sh
+
+ZIG="$DEPS_LOCAL/bin/zig"
+TARGET="${ARCH}-linux-musl"
+MCPU="baseline"
+
+# Make the `zig version` number consistent.
+# This will affect the cmake command below.
+git config core.abbrev 9
+
+# Build debug zig.
+echo "BUILD debug zig with zig:$($ZIG version)"
+
+export CC="$ZIG cc -target $TARGET -mcpu=$MCPU"
+export CXX="$ZIG c++ -target $TARGET -mcpu=$MCPU"
+
+mkdir _debug
+cd _debug
+cmake .. \
+  -DCMAKE_INSTALL_PREFIX="$DEBUG_STAGING" \
+  -DCMAKE_PREFIX_PATH="$DEPS_LOCAL" \
+  -DCMAKE_BUILD_TYPE=Debug \
+  -DZIG_TARGET_TRIPLE="$TARGET" \
+  -DZIG_TARGET_MCPU="$MCPU" \
+  -DZIG_STATIC=ON \
+  -GNinja
+
+# Now cmake will use zig as the C/C++ compiler. We reset the environment variables
+# so that installation and testing do not get affected by them.
+unset CC
+unset CXX
+
+ninja $JOBS install
+
+ZIG=$DEBUG_STAGING/bin/zig
+
+# Here we rebuild zig but this time using the Zig binary we just now produced to
+# build zig1.o rather than relying on the one built with stage0. See
+# https://github.com/ziglang/zig/issues/6830 for more details.
+cmake .. -DZIG_EXECUTABLE="$ZIG"
+ninja $JOBS install
+
+cd $WORKSPACE
+
+# Build release zig.
+echo "BUILD release zig with zig:$($ZIG version)"
+export CC="$ZIG cc -target $TARGET -mcpu=$MCPU"
+export CXX="$ZIG c++ -target $TARGET -mcpu=$MCPU"
+mkdir _release
+cd _release
+cmake .. \
+  -DCMAKE_INSTALL_PREFIX="$RELEASE_STAGING" \
+  -DCMAKE_PREFIX_PATH="$DEPS_LOCAL" \
+  -DCMAKE_BUILD_TYPE=Release \
+  -DZIG_TARGET_TRIPLE="$TARGET" \
+  -DZIG_TARGET_MCPU="$MCPU" \
+  -DZIG_STATIC=ON \
+  -GNinja
+unset CC
+unset CXX
+ninja $JOBS install
+
+cd $WORKSPACE
+
+# Look for non-conforming code formatting.
+# Formatting errors can be fixed by running `zig fmt` on the files printed here.
+$ZIG fmt --check .
+
+# Explicit exit helps show last command duration.
+exit

ci/zinc/linux_package.sh

@@ -0,0 +1,48 @@
+#!/bin/sh
+
+. ./ci/zinc/linux_base.sh
+
+cp LICENSE $RELEASE_STAGING/
+cp zig-cache/langref.html $RELEASE_STAGING/docs/
+
+# Remove the unnecessary bin dir in $prefix/bin/zig
+mv $RELEASE_STAGING/bin/zig $RELEASE_STAGING/
+rmdir $RELEASE_STAGING/bin
+
+# Remove the unnecessary zig dir in $prefix/lib/zig/std/std.zig
+mv $RELEASE_STAGING/lib/zig $RELEASE_STAGING/lib2
+rmdir $RELEASE_STAGING/lib
+mv $RELEASE_STAGING/lib2 $RELEASE_STAGING/lib
+
+VERSION=$($RELEASE_STAGING/zig version)
+BASENAME="zig-linux-$ARCH-$VERSION"
+TARBALL="$BASENAME.tar.xz"
+mv "$RELEASE_STAGING" "$BASENAME"
+tar cfJ "$TARBALL" "$BASENAME"
+ls -l "$TARBALL"
+
+SHASUM=$(sha256sum $TARBALL | cut '-d ' -f1)
+BYTESIZE=$(wc -c < $TARBALL)
+
+MANIFEST="manifest.json"
+touch $MANIFEST
+echo "{\"tarball\": \"$TARBALL\"," >>$MANIFEST
+echo "\"shasum\": \"$SHASUM\"," >>$MANIFEST
+echo "\"size\": \"$BYTESIZE\"}" >>$MANIFEST
+
+# Publish artifact.
+s3cmd put -P --add-header="cache-control: public, max-age=31536000, immutable" "$TARBALL" s3://ziglang.org/builds/
+
+# Publish manifest.
+s3cmd put -P --add-header="cache-control: max-age=0, must-revalidate" "$MANIFEST" "s3://ziglang.org/builds/$ARCH-linux-$VERSION.json"
+
+# Avoid leaking oauth token.
+set +x
+
+cd $WORKSPACE
+./ci/srht/on_master_success "$VERSION" "$SRHT_OAUTH_TOKEN"
+
+set -x
+
+# Explicit exit helps show last command duration.
+exit

ci/zinc/linux_probe.sh

@@ -0,0 +1,10 @@
+#!/bin/sh
+
+. ./ci/zinc/linux_base.sh
+
+# Probe CPU/brand details.
+echo "lscpu:"
+(lscpu | sed 's,^,  : ,') 1>&2
+
+# Explicit exit helps show last command duration.
+exit

ci/zinc/linux_test.sh

@@ -0,0 +1,40 @@
+#!/bin/sh
+
+. ./ci/zinc/linux_base.sh
+
+ZIG=$DEBUG_STAGING/bin/zig
+
+$ZIG test test/behavior.zig -fno-stage1 -fLLVM -I test
+$ZIG test test/behavior.zig -fno-stage1 -ofmt=c -I test
+$ZIG test test/behavior.zig -fno-stage1 -target wasm32-wasi --test-cmd wasmtime --test-cmd-bin
+
+$ZIG build test-behavior         -fqemu -fwasmtime
+$ZIG build test-compiler-rt      -fqemu -fwasmtime
+$ZIG build test-std              -fqemu -fwasmtime
+$ZIG build test-minilibc         -fqemu -fwasmtime
+$ZIG build test-compare-output   -fqemu -fwasmtime
+$ZIG build test-standalone       -fqemu -fwasmtime
+$ZIG build test-stack-traces     -fqemu -fwasmtime
+$ZIG build test-cli              -fqemu -fwasmtime
+$ZIG build test-asm-link         -fqemu -fwasmtime
+$ZIG build test-runtime-safety   -fqemu -fwasmtime
+$ZIG build test-translate-c      -fqemu -fwasmtime
+$ZIG build test-run-translated-c -fqemu -fwasmtime
+$ZIG build docs                  -fqemu -fwasmtime
+$ZIG build # test building self-hosted without LLVM
+$ZIG build -Dtarget=arm-linux-musleabihf # test building self-hosted for 32-bit arm
+$ZIG build test-fmt              -fqemu -fwasmtime
+$ZIG build test-stage2           -fqemu -fwasmtime
+
+# Produce the experimental std lib documentation.
+mkdir -p $RELEASE_STAGING/docs/std
+$ZIG test lib/std/std.zig \
+  --zig-lib-dir lib \
+  -femit-docs=$RELEASE_STAGING/docs/std \
+  -fno-emit-bin
+
+# Look for HTML errors.
+tidy --drop-empty-elements no -qe zig-cache/langref.html
+
+# Explicit exit helps show last command duration.
+exit

cmake/Findclang.cmake

@@ -9,27 +9,31 @@
 
 find_path(CLANG_INCLUDE_DIRS NAMES clang/Frontend/ASTUnit.h
   PATHS
-    /usr/lib/llvm/12/include
-    /usr/lib/llvm-12/include
-    /usr/lib/llvm-12.0/include
-    /usr/local/llvm120/include
-    /usr/local/llvm12/include
+    /usr/lib/llvm/13/include
+    /usr/lib/llvm-13/include
+    /usr/lib/llvm-13.0/include
+    /usr/local/llvm130/include
+    /usr/local/llvm13/include
+    /usr/local/opt/llvm@13/include
+    /opt/homebrew/opt/llvm@13/include
     /mingw64/include
 )
 
 if(ZIG_PREFER_CLANG_CPP_DYLIB)
   find_library(CLANG_LIBRARIES
     NAMES
-      clang-cpp-12.0
-      clang-cpp120
+      clang-cpp-13.0
+      clang-cpp130
       clang-cpp
     PATHS
       ${CLANG_LIBDIRS}
-      /usr/lib/llvm/12/lib
-      /usr/lib/llvm/12/lib64
-      /usr/lib/llvm-12/lib
-      /usr/local/llvm120/lib
-      /usr/local/llvm12/lib
+      /usr/lib/llvm/13/lib
+      /usr/lib/llvm/13/lib64
+      /usr/lib/llvm-13/lib
+      /usr/local/llvm130/lib
+      /usr/local/llvm13/lib
+      /usr/local/opt/llvm@13/lib
+      /opt/homebrew/opt/llvm@13/lib
   )
 endif()
 
@@ -39,11 +43,13 @@ if(NOT CLANG_LIBRARIES)
     find_library(CLANG_${_prettylibname_}_LIB NAMES ${_libname_}
       PATHS
         ${CLANG_LIBDIRS}
-        /usr/lib/llvm/12/lib
-        /usr/lib/llvm-12/lib
-        /usr/lib/llvm-12.0/lib
-        /usr/local/llvm120/lib
-        /usr/local/llvm12/lib
+        /usr/lib/llvm/13/lib
+        /usr/lib/llvm-13/lib
+        /usr/lib/llvm-13.0/lib
+        /usr/local/llvm130/lib
+        /usr/local/llvm13/lib
+        /usr/local/opt/llvm@13/lib
+        /opt/homebrew/opt/llvm@13/lib
         /mingw64/lib
         /c/msys64/mingw64/lib
         c:\\msys64\\mingw64\\lib

cmake/Findlld.cmake

@@ -8,16 +8,20 @@
 
 find_path(LLD_INCLUDE_DIRS NAMES lld/Common/Driver.h
     PATHS
-        /usr/lib/llvm-12/include
-        /usr/local/llvm120/include
-        /usr/local/llvm12/include
+        /usr/lib/llvm-13/include
+        /usr/local/llvm130/include
+        /usr/local/llvm13/include
+        /usr/local/opt/llvm@13/include
+        /opt/homebrew/opt/llvm@13/include
         /mingw64/include)
 
-find_library(LLD_LIBRARY NAMES lld-12.0 lld120 lld
+find_library(LLD_LIBRARY NAMES lld-13.0 lld130 lld
     PATHS
-        /usr/lib/llvm-12/lib
-        /usr/local/llvm120/lib
-        /usr/local/llvm12/lib
+        /usr/lib/llvm-13/lib
+        /usr/local/llvm130/lib
+        /usr/local/llvm13/lib
+        /usr/local/opt/llvm@13/lib
+        /opt/homebrew/opt/llvm@13/lib
 )
 if(EXISTS ${LLD_LIBRARY})
     set(LLD_LIBRARIES ${LLD_LIBRARY})
@@ -27,9 +31,11 @@ else()
         find_library(LLD_${_prettylibname_}_LIB NAMES ${_libname_}
             PATHS
                 ${LLD_LIBDIRS}
-                /usr/lib/llvm-12/lib
-                /usr/local/llvm120/lib
-                /usr/local/llvm12/lib
+                /usr/lib/llvm-13/lib
+                /usr/local/llvm130/lib
+                /usr/local/llvm13/lib
+                /usr/local/opt/llvm@13/lib
+                /opt/homebrew/opt/llvm@13/lib
                 /mingw64/lib
                 /c/msys64/mingw64/lib
                 c:/msys64/mingw64/lib)
@@ -42,8 +48,8 @@ else()
     FIND_AND_ADD_LLD_LIB(lldMinGW)
     FIND_AND_ADD_LLD_LIB(lldELF)
     FIND_AND_ADD_LLD_LIB(lldCOFF)
-    FIND_AND_ADD_LLD_LIB(lldMachO)
     FIND_AND_ADD_LLD_LIB(lldWasm)
+    FIND_AND_ADD_LLD_LIB(lldMachO)
     FIND_AND_ADD_LLD_LIB(lldReaderWriter)
     FIND_AND_ADD_LLD_LIB(lldCore)
     FIND_AND_ADD_LLD_LIB(lldYAML)

cmake/Findllvm.cmake

@@ -9,37 +9,41 @@
 
 find_path(LLVM_INCLUDE_DIRS NAMES llvm/IR/IRBuilder.h
   PATHS
-    /usr/lib/llvm/12/include
-    /usr/lib/llvm-12/include
-    /usr/lib/llvm-12.0/include
-    /usr/local/llvm12/include
-    /usr/local/llvm120/include
+    /usr/lib/llvm/13/include
+    /usr/lib/llvm-13/include
+    /usr/lib/llvm-13.0/include
+    /usr/local/llvm13/include
+    /usr/local/llvm130/include
+    /usr/local/opt/llvm@13/include
+    /opt/homebrew/opt/llvm@13/include
     /mingw64/include
 )
 
 if(ZIG_PREFER_CLANG_CPP_DYLIB)
   find_library(LLVM_LIBRARIES
     NAMES
-      LLVM-12.0
-      LLVM-12
-      LLVM-120
+      LLVM-13.0
+      LLVM-13
+      LLVM-130
       LLVM
     PATHS
       ${LLVM_LIBDIRS}
-      /usr/lib/llvm/12/lib
-      /usr/lib/llvm/12/lib64
-      /usr/lib/llvm-12/lib
-      /usr/local/llvm12/lib
-      /usr/local/llvm120/lib
+      /usr/lib/llvm/13/lib
+      /usr/lib/llvm/13/lib64
+      /usr/lib/llvm-13/lib
+      /usr/local/llvm13/lib
+      /usr/local/llvm130/lib
+      /usr/local/opt/llvm@13/lib
+      /opt/homebrew/opt/llvm@13/lib
   )
 
   find_program(LLVM_CONFIG_EXE
-      NAMES llvm-config-12 llvm-config-12.0 llvm-config120 llvm-config12 llvm-config
+      NAMES llvm-config-13 llvm-config-13.0 llvm-config130 llvm-config13 llvm-config
       PATHS
           "/mingw64/bin"
           "/c/msys64/mingw64/bin"
           "c:/msys64/mingw64/bin"
-          "C:/Libraries/llvm-12.0.0/bin")
+          "C:/Libraries/llvm-13.0.0/bin")
 
   if ("${LLVM_CONFIG_EXE}" STREQUAL "LLVM_CONFIG_EXE-NOTFOUND")
     message(FATAL_ERROR "unable to find llvm-config")
@@ -54,23 +58,23 @@ if(ZIG_PREFER_CLANG_CPP_DYLIB)
     OUTPUT_VARIABLE LLVM_CONFIG_VERSION
     OUTPUT_STRIP_TRAILING_WHITESPACE)
 
-  if("${LLVM_CONFIG_VERSION}" VERSION_LESS 12)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_LESS 13)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
-  if("${LLVM_CONFIG_VERSION}" VERSION_EQUAL 13)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_EQUAL 14)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
-  if("${LLVM_CONFIG_VERSION}" VERSION_GREATER 13)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_GREATER 14)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
 elseif(ZIG_USE_LLVM_CONFIG)
   find_program(LLVM_CONFIG_EXE
-      NAMES llvm-config-12 llvm-config-12.0 llvm-config120 llvm-config12 llvm-config
+      NAMES llvm-config-13 llvm-config-13.0 llvm-config130 llvm-config13 llvm-config
       PATHS
           "/mingw64/bin"
           "/c/msys64/mingw64/bin"
           "c:/msys64/mingw64/bin"
-          "C:/Libraries/llvm-12.0.0/bin")
+          "C:/Libraries/llvm-13.0.0/bin")
 
   if ("${LLVM_CONFIG_EXE}" STREQUAL "LLVM_CONFIG_EXE-NOTFOUND")
     message(FATAL_ERROR "unable to find llvm-config")
@@ -85,14 +89,14 @@ elseif(ZIG_USE_LLVM_CONFIG)
     OUTPUT_VARIABLE LLVM_CONFIG_VERSION
     OUTPUT_STRIP_TRAILING_WHITESPACE)
 
-  if("${LLVM_CONFIG_VERSION}" VERSION_LESS 12)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_LESS 13)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
-  if("${LLVM_CONFIG_VERSION}" VERSION_EQUAL 13)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_EQUAL 14)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
-  if("${LLVM_CONFIG_VERSION}" VERSION_GREATER 13)
-    message(FATAL_ERROR "expected LLVM 12.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
+  if("${LLVM_CONFIG_VERSION}" VERSION_GREATER 14)
+    message(FATAL_ERROR "expected LLVM 13.x but found ${LLVM_CONFIG_VERSION} using ${LLVM_CONFIG_EXE}")
   endif()
 
   execute_process(
@@ -166,7 +170,7 @@ elseif(ZIG_USE_LLVM_CONFIG)
   set(LLVM_LIBRARIES ${LLVM_LIBRARIES} ${LLVM_SYSTEM_LIBS})
 
   if(NOT LLVM_LIBRARIES)
-    find_library(LLVM_LIBRARIES NAMES LLVM LLVM-12 LLVM-12.0)
+    find_library(LLVM_LIBRARIES NAMES LLVM LLVM-13 LLVM-13.0)
   endif()
 
   link_directories("${CMAKE_PREFIX_PATH}/lib")
@@ -180,11 +184,13 @@ else()
     find_library(LLVM_${_prettylibname_}_LIB NAMES ${_libname_}
       PATHS
       ${LLVM_LIBDIRS}
-      /usr/lib/llvm/12/lib
-      /usr/lib/llvm-12/lib
-      /usr/lib/llvm-12.0/lib
-      /usr/local/llvm120/lib
-      /usr/local/llvm12/lib
+      /usr/lib/llvm/13/lib
+      /usr/lib/llvm-13/lib
+      /usr/lib/llvm-13.0/lib
+      /usr/local/llvm130/lib
+      /usr/local/llvm13/lib
+      /usr/local/opt/llvm@13/lib
+      /opt/homebrew/opt/llvm@13/lib
       /mingw64/lib
       /c/msys64/mingw64/lib
       c:\\msys64\\mingw64\\lib)
@@ -213,6 +219,7 @@ else()
   FIND_AND_ADD_LLVM_LIB(LLVMWebAssemblyAsmParser)
   FIND_AND_ADD_LLVM_LIB(LLVMWebAssemblyCodeGen)
   FIND_AND_ADD_LLVM_LIB(LLVMWebAssemblyDesc)
+  FIND_AND_ADD_LLVM_LIB(LLVMWebAssemblyUtils)
   FIND_AND_ADD_LLVM_LIB(LLVMWebAssemblyInfo)
   FIND_AND_ADD_LLVM_LIB(LLVMSystemZDisassembler)
   FIND_AND_ADD_LLVM_LIB(LLVMSystemZAsmParser)
@@ -288,11 +295,12 @@ else()
   FIND_AND_ADD_LLVM_LIB(LLVMOrcJIT)
   FIND_AND_ADD_LLVM_LIB(LLVMMCJIT)
   FIND_AND_ADD_LLVM_LIB(LLVMJITLink)
-  FIND_AND_ADD_LLVM_LIB(LLVMOrcTargetProcess)
-  FIND_AND_ADD_LLVM_LIB(LLVMOrcShared)
   FIND_AND_ADD_LLVM_LIB(LLVMInterpreter)
   FIND_AND_ADD_LLVM_LIB(LLVMExecutionEngine)
   FIND_AND_ADD_LLVM_LIB(LLVMRuntimeDyld)
+  FIND_AND_ADD_LLVM_LIB(LLVMOrcTargetProcess)
+  FIND_AND_ADD_LLVM_LIB(LLVMOrcShared)
+  FIND_AND_ADD_LLVM_LIB(LLVMDWP)
   FIND_AND_ADD_LLVM_LIB(LLVMSymbolize)
   FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoPDB)
   FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoGSYM)
@@ -305,7 +313,6 @@ else()
   FIND_AND_ADD_LLVM_LIB(LLVMCFGuard)
   FIND_AND_ADD_LLVM_LIB(LLVMCoroutines)
   FIND_AND_ADD_LLVM_LIB(LLVMObjCARCOpts)
-  FIND_AND_ADD_LLVM_LIB(LLVMHelloNew)
   FIND_AND_ADD_LLVM_LIB(LLVMipo)
   FIND_AND_ADD_LLVM_LIB(LLVMVectorize)
   FIND_AND_ADD_LLVM_LIB(LLVMLinker)
@@ -317,6 +324,7 @@ else()
   FIND_AND_ADD_LLVM_LIB(LLVMGlobalISel)
   FIND_AND_ADD_LLVM_LIB(LLVMMIRParser)
   FIND_AND_ADD_LLVM_LIB(LLVMAsmPrinter)
+  FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoMSF)
   FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoDWARF)
   FIND_AND_ADD_LLVM_LIB(LLVMSelectionDAG)
   FIND_AND_ADD_LLVM_LIB(LLVMCodeGen)
@@ -338,7 +346,6 @@ else()
   FIND_AND_ADD_LLVM_LIB(LLVMMCParser)
   FIND_AND_ADD_LLVM_LIB(LLVMMC)
   FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoCodeView)
-  FIND_AND_ADD_LLVM_LIB(LLVMDebugInfoMSF)
   FIND_AND_ADD_LLVM_LIB(LLVMBitReader)
   FIND_AND_ADD_LLVM_LIB(LLVMCore)
   FIND_AND_ADD_LLVM_LIB(LLVMRemarks)

lib/include/__clang_cuda_device_functions.h

@@ -34,10 +34,12 @@ __DEVICE__ unsigned long long __brevll(unsigned long long __a) {
   return __nv_brevll(__a);
 }
 #if defined(__cplusplus)
-__DEVICE__ void __brkpt() { asm volatile("brkpt;"); }
+__DEVICE__ void __brkpt() { __asm__ __volatile__("brkpt;"); }
 __DEVICE__ void __brkpt(int __a) { __brkpt(); }
 #else
-__DEVICE__ void __attribute__((overloadable)) __brkpt(void) { asm volatile("brkpt;"); }
+__DEVICE__ void __attribute__((overloadable)) __brkpt(void) {
+  __asm__ __volatile__("brkpt;");
+}
 __DEVICE__ void __attribute__((overloadable)) __brkpt(int __a) { __brkpt(); }
 #endif
 __DEVICE__ unsigned int __byte_perm(unsigned int __a, unsigned int __b,
@@ -507,7 +509,7 @@ __DEVICE__ float __powf(float __a, float __b) {
 }
 
 // Parameter must have a known integer value.
-#define __prof_trigger(__a) asm __volatile__("pmevent \t%0;" ::"i"(__a))
+#define __prof_trigger(__a) __asm__ __volatile__("pmevent \t%0;" ::"i"(__a))
 __DEVICE__ int __rhadd(int __a, int __b) { return __nv_rhadd(__a, __b); }
 __DEVICE__ unsigned int __sad(int __a, int __b, unsigned int __c) {
   return __nv_sad(__a, __b, __c);
@@ -526,7 +528,7 @@ __DEVICE__ float __tanf(float __a) { return __nv_fast_tanf(__a); }
 __DEVICE__ void __threadfence(void) { __nvvm_membar_gl(); }
 __DEVICE__ void __threadfence_block(void) { __nvvm_membar_cta(); };
 __DEVICE__ void __threadfence_system(void) { __nvvm_membar_sys(); };
-__DEVICE__ void __trap(void) { asm volatile("trap;"); }
+__DEVICE__ void __trap(void) { __asm__ __volatile__("trap;"); }
 __DEVICE__ unsigned int __uAtomicAdd(unsigned int *__p, unsigned int __v) {
   return __nvvm_atom_add_gen_i((int *)__p, __v);
 }
@@ -1051,122 +1053,136 @@ __DEVICE__ unsigned int __bool2mask(unsigned int __a, int shift) {
 }
 __DEVICE__ unsigned int __vabs2(unsigned int __a) {
   unsigned int r;
-  asm("vabsdiff2.s32.s32.s32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(0), "r"(0));
+  __asm__("vabsdiff2.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(0), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabs4(unsigned int __a) {
   unsigned int r;
-  asm("vabsdiff4.s32.s32.s32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(0), "r"(0));
+  __asm__("vabsdiff4.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(0), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabsdiffs2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff2.s32.s32.s32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff2.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 
 __DEVICE__ unsigned int __vabsdiffs4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff4.s32.s32.s32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff4.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabsdiffu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff2.u32.u32.u32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabsdiffu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff4.u32.u32.u32 %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabsss2(unsigned int __a) {
   unsigned int r;
-  asm("vabsdiff2.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(0), "r"(0));
+  __asm__("vabsdiff2.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(0), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vabsss4(unsigned int __a) {
   unsigned int r;
-  asm("vabsdiff4.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(0), "r"(0));
+  __asm__("vabsdiff4.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(0), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vadd2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vadd4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vaddss2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd2.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd2.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vaddss4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd4.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd4.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vaddus2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd2.u32.u32.u32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd2.u32.u32.u32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vaddus4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vadd4.u32.u32.u32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vadd4.u32.u32.u32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vavgs2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vavrg2.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vavrg2.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vavgs4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vavrg4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vavrg4.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vavgu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vavrg2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vavrg2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vavgu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vavrg4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vavrg4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vseteq2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.eq %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.eq %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpeq2(unsigned int __a, unsigned int __b) {
@@ -1174,7 +1190,9 @@ __DEVICE__ unsigned int __vcmpeq2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vseteq4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.eq %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.eq %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpeq4(unsigned int __a, unsigned int __b) {
@@ -1182,7 +1200,9 @@ __DEVICE__ unsigned int __vcmpeq4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetges2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.s32.s32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.s32.s32.ge %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpges2(unsigned int __a, unsigned int __b) {
@@ -1190,7 +1210,9 @@ __DEVICE__ unsigned int __vcmpges2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetges4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.s32.s32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.s32.s32.ge %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpges4(unsigned int __a, unsigned int __b) {
@@ -1198,7 +1220,9 @@ __DEVICE__ unsigned int __vcmpges4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgeu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.ge %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgeu2(unsigned int __a, unsigned int __b) {
@@ -1206,7 +1230,9 @@ __DEVICE__ unsigned int __vcmpgeu2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgeu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.ge %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgeu4(unsigned int __a, unsigned int __b) {
@@ -1214,7 +1240,9 @@ __DEVICE__ unsigned int __vcmpgeu4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgts2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.s32.s32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.s32.s32.gt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgts2(unsigned int __a, unsigned int __b) {
@@ -1222,7 +1250,9 @@ __DEVICE__ unsigned int __vcmpgts2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgts4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.s32.s32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.s32.s32.gt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgts4(unsigned int __a, unsigned int __b) {
@@ -1230,7 +1260,9 @@ __DEVICE__ unsigned int __vcmpgts4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgtu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.gt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgtu2(unsigned int __a, unsigned int __b) {
@@ -1238,7 +1270,9 @@ __DEVICE__ unsigned int __vcmpgtu2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetgtu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.gt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpgtu4(unsigned int __a, unsigned int __b) {
@@ -1246,7 +1280,9 @@ __DEVICE__ unsigned int __vcmpgtu4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetles2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.s32.s32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.s32.s32.le %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmples2(unsigned int __a, unsigned int __b) {
@@ -1254,7 +1290,9 @@ __DEVICE__ unsigned int __vcmples2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetles4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.s32.s32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.s32.s32.le %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmples4(unsigned int __a, unsigned int __b) {
@@ -1262,7 +1300,9 @@ __DEVICE__ unsigned int __vcmples4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetleu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.le %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpleu2(unsigned int __a, unsigned int __b) {
@@ -1270,7 +1310,9 @@ __DEVICE__ unsigned int __vcmpleu2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetleu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.le %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpleu4(unsigned int __a, unsigned int __b) {
@@ -1278,7 +1320,9 @@ __DEVICE__ unsigned int __vcmpleu4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetlts2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.s32.s32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.s32.s32.lt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmplts2(unsigned int __a, unsigned int __b) {
@@ -1286,7 +1330,9 @@ __DEVICE__ unsigned int __vcmplts2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetlts4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.s32.s32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.s32.s32.lt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmplts4(unsigned int __a, unsigned int __b) {
@@ -1294,7 +1340,9 @@ __DEVICE__ unsigned int __vcmplts4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetltu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.lt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpltu2(unsigned int __a, unsigned int __b) {
@@ -1302,7 +1350,9 @@ __DEVICE__ unsigned int __vcmpltu2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetltu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.lt %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpltu4(unsigned int __a, unsigned int __b) {
@@ -1310,7 +1360,9 @@ __DEVICE__ unsigned int __vcmpltu4(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetne2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset2.u32.u32.ne %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset2.u32.u32.ne %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpne2(unsigned int __a, unsigned int __b) {
@@ -1318,7 +1370,9 @@ __DEVICE__ unsigned int __vcmpne2(unsigned int __a, unsigned int __b) {
 }
 __DEVICE__ unsigned int __vsetne4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vset4.u32.u32.ne %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vset4.u32.u32.ne %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vcmpne4(unsigned int __a, unsigned int __b) {
@@ -1345,94 +1399,112 @@ __DEVICE__ unsigned int __vmaxs2(unsigned int __a, unsigned int __b) {
     unsigned mask = __vcmpgts2(__a, __b);
     r = (__a & mask) | (__b & ~mask);
   } else {
-    asm("vmax2.s32.s32.s32 %0,%1,%2,%3;"
-        : "=r"(r)
-        : "r"(__a), "r"(__b), "r"(0));
+    __asm__("vmax2.s32.s32.s32 %0,%1,%2,%3;"
+            : "=r"(r)
+            : "r"(__a), "r"(__b), "r"(0));
   }
   return r;
 }
 __DEVICE__ unsigned int __vmaxs4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmax4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmax4.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vmaxu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmax2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmax2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vmaxu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmax4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmax4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vmins2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmin2.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmin2.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vmins4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmin4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmin4.s32.s32.s32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vminu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmin2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmin2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vminu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vmin4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vmin4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vsads2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff2.s32.s32.s32.add %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff2.s32.s32.s32.add %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vsads4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff4.s32.s32.s32.add %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff4.s32.s32.s32.add %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vsadu2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff2.u32.u32.u32.add %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff2.u32.u32.u32.add %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vsadu4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vabsdiff4.u32.u32.u32.add %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vabsdiff4.u32.u32.u32.add %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 
 __DEVICE__ unsigned int __vsub2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub2.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vneg2(unsigned int __a) { return __vsub2(0, __a); }
 
 __DEVICE__ unsigned int __vsub4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub4.u32.u32.u32 %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vneg4(unsigned int __a) { return __vsub4(0, __a); }
 __DEVICE__ unsigned int __vsubss2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub2.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub2.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vnegss2(unsigned int __a) {
@@ -1440,9 +1512,9 @@ __DEVICE__ unsigned int __vnegss2(unsigned int __a) {
 }
 __DEVICE__ unsigned int __vsubss4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub4.s32.s32.s32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub4.s32.s32.s32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vnegss4(unsigned int __a) {
@@ -1450,16 +1522,16 @@ __DEVICE__ unsigned int __vnegss4(unsigned int __a) {
 }
 __DEVICE__ unsigned int __vsubus2(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub2.u32.u32.u32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub2.u32.u32.u32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 __DEVICE__ unsigned int __vsubus4(unsigned int __a, unsigned int __b) {
   unsigned int r;
-  asm("vsub4.u32.u32.u32.sat %0,%1,%2,%3;"
-      : "=r"(r)
-      : "r"(__a), "r"(__b), "r"(0));
+  __asm__("vsub4.u32.u32.u32.sat %0,%1,%2,%3;"
+          : "=r"(r)
+          : "r"(__a), "r"(__b), "r"(0));
   return r;
 }
 #endif // CUDA_VERSION >= 9020

lib/include/__clang_cuda_math.h

@@ -166,6 +166,8 @@ __DEVICE__ long long llrint(double __a) { return __nv_llrint(__a); }
 __DEVICE__ long long llrintf(float __a) { return __nv_llrintf(__a); }
 __DEVICE__ long long llround(double __a) { return __nv_llround(__a); }
 __DEVICE__ long long llroundf(float __a) { return __nv_llroundf(__a); }
+__DEVICE__ double round(double __a) { return __nv_round(__a); }
+__DEVICE__ float roundf(float __a) { return __nv_roundf(__a); }
 __DEVICE__ double log(double __a) { return __nv_log(__a); }
 __DEVICE__ double log10(double __a) { return __nv_log10(__a); }
 __DEVICE__ float log10f(float __a) { return __nv_log10f(__a); }
@@ -270,8 +272,6 @@ __DEVICE__ float rnorm4df(float __a, float __b, float __c, float __d) {
 __DEVICE__ float rnormf(int __dim, const float *__t) {
   return __nv_rnormf(__dim, __t);
 }
-__DEVICE__ double round(double __a) { return __nv_round(__a); }
-__DEVICE__ float roundf(float __a) { return __nv_roundf(__a); }
 __DEVICE__ double rsqrt(double __a) { return __nv_rsqrt(__a); }
 __DEVICE__ float rsqrtf(float __a) { return __nv_rsqrtf(__a); }
 __DEVICE__ double scalbn(double __a, int __b) { return __nv_scalbn(__a, __b); }

lib/include/__clang_cuda_runtime_wrapper.h

@@ -349,9 +349,14 @@ extern "C" {
 __device__ int vprintf(const char *, const char *);
 __device__ void free(void *) __attribute((nothrow));
 __device__ void *malloc(size_t) __attribute((nothrow)) __attribute__((malloc));
+
+// __assertfail() used to have a `noreturn` attribute. Unfortunately that
+// contributed to triggering the longstanding bug in ptxas when assert was used
+// in sufficiently convoluted code. See
+// https://bugs.llvm.org/show_bug.cgi?id=27738 for the details.
 __device__ void __assertfail(const char *__message, const char *__file,
                              unsigned __line, const char *__function,
-                             size_t __charSize) __attribute__((noreturn));
+                             size_t __charSize);
 
 // In order for standard assert() macro on linux to work we need to
 // provide device-side __assert_fail()

lib/include/__clang_hip_cmath.h

@@ -10,10 +10,11 @@
 #ifndef __CLANG_HIP_CMATH_H__
 #define __CLANG_HIP_CMATH_H__
 
-#if !defined(__HIP__)
+#if !defined(__HIP__) && !defined(__OPENMP_AMDGCN__)
 #error "This file is for HIP and OpenMP AMDGCN device compilation only."
 #endif
 
+#if !defined(__HIPCC_RTC__)
 #if defined(__cplusplus)
 #include <limits>
 #include <type_traits>
@@ -21,102 +22,162 @@
 #endif
 #include <limits.h>
 #include <stdint.h>
+#endif // !defined(__HIPCC_RTC__)
 
 #pragma push_macro("__DEVICE__")
+#pragma push_macro("__CONSTEXPR__")
+#ifdef __OPENMP_AMDGCN__
+#define __DEVICE__ static __attribute__((always_inline, nothrow))
+#define __CONSTEXPR__ constexpr
+#else
 #define __DEVICE__ static __device__ inline __attribute__((always_inline))
+#define __CONSTEXPR__
+#endif // __OPENMP_AMDGCN__
 
 // Start with functions that cannot be defined by DEF macros below.
 #if defined(__cplusplus)
-__DEVICE__ double abs(double __x) { return ::fabs(__x); }
-__DEVICE__ float abs(float __x) { return ::fabsf(__x); }
-__DEVICE__ long long abs(long long __n) { return ::llabs(__n); }
-__DEVICE__ long abs(long __n) { return ::labs(__n); }
-__DEVICE__ float fma(float __x, float __y, float __z) {
+#if defined __OPENMP_AMDGCN__
+__DEVICE__ __CONSTEXPR__ float fabs(float __x) { return ::fabsf(__x); }
+__DEVICE__ __CONSTEXPR__ float sin(float __x) { return ::sinf(__x); }
+__DEVICE__ __CONSTEXPR__ float cos(float __x) { return ::cosf(__x); }
+#endif
+__DEVICE__ __CONSTEXPR__ double abs(double __x) { return ::fabs(__x); }
+__DEVICE__ __CONSTEXPR__ float abs(float __x) { return ::fabsf(__x); }
+__DEVICE__ __CONSTEXPR__ long long abs(long long __n) { return ::llabs(__n); }
+__DEVICE__ __CONSTEXPR__ long abs(long __n) { return ::labs(__n); }
+__DEVICE__ __CONSTEXPR__ float fma(float __x, float __y, float __z) {
   return ::fmaf(__x, __y, __z);
 }
-__DEVICE__ int fpclassify(float __x) {
+#if !defined(__HIPCC_RTC__)
+// The value returned by fpclassify is platform dependent, therefore it is not
+// supported by hipRTC.
+__DEVICE__ __CONSTEXPR__ int fpclassify(float __x) {
   return __builtin_fpclassify(FP_NAN, FP_INFINITE, FP_NORMAL, FP_SUBNORMAL,
                               FP_ZERO, __x);
 }
-__DEVICE__ int fpclassify(double __x) {
+__DEVICE__ __CONSTEXPR__ int fpclassify(double __x) {
   return __builtin_fpclassify(FP_NAN, FP_INFINITE, FP_NORMAL, FP_SUBNORMAL,
                               FP_ZERO, __x);
 }
-__DEVICE__ float frexp(float __arg, int *__exp) {
+#endif // !defined(__HIPCC_RTC__)
+
+__DEVICE__ __CONSTEXPR__ float frexp(float __arg, int *__exp) {
   return ::frexpf(__arg, __exp);
 }
-__DEVICE__ bool isfinite(float __x) { return ::__finitef(__x); }
-__DEVICE__ bool isfinite(double __x) { return ::__finite(__x); }
-__DEVICE__ bool isgreater(float __x, float __y) {
+
+#if defined(__OPENMP_AMDGCN__)
+// For OpenMP we work around some old system headers that have non-conforming
+// `isinf(float)` and `isnan(float)` implementations that return an `int`. We do
+// this by providing two versions of these functions, differing only in the
+// return type. To avoid conflicting definitions we disable implicit base
+// function generation. That means we will end up with two specializations, one
+// per type, but only one has a base function defined by the system header.
+#pragma omp begin declare variant match(                                       \
+    implementation = {extension(disable_implicit_base)})
+
+// FIXME: We lack an extension to customize the mangling of the variants, e.g.,
+//        add a suffix. This means we would clash with the names of the variants
+//        (note that we do not create implicit base functions here). To avoid
+//        this clash we add a new trait to some of them that is always true
+//        (this is LLVM after all ;)). It will only influence the mangled name
+//        of the variants inside the inner region and avoid the clash.
+#pragma omp begin declare variant match(implementation = {vendor(llvm)})
+
+__DEVICE__ __CONSTEXPR__ int isinf(float __x) { return ::__isinff(__x); }
+__DEVICE__ __CONSTEXPR__ int isinf(double __x) { return ::__isinf(__x); }
+__DEVICE__ __CONSTEXPR__ int isfinite(float __x) { return ::__finitef(__x); }
+__DEVICE__ __CONSTEXPR__ int isfinite(double __x) { return ::__finite(__x); }
+__DEVICE__ __CONSTEXPR__ int isnan(float __x) { return ::__isnanf(__x); }
+__DEVICE__ __CONSTEXPR__ int isnan(double __x) { return ::__isnan(__x); }
+
+#pragma omp end declare variant
+#endif // defined(__OPENMP_AMDGCN__)
+
+__DEVICE__ __CONSTEXPR__ bool isinf(float __x) { return ::__isinff(__x); }
+__DEVICE__ __CONSTEXPR__ bool isinf(double __x) { return ::__isinf(__x); }
+__DEVICE__ __CONSTEXPR__ bool isfinite(float __x) { return ::__finitef(__x); }
+__DEVICE__ __CONSTEXPR__ bool isfinite(double __x) { return ::__finite(__x); }
+__DEVICE__ __CONSTEXPR__ bool isnan(float __x) { return ::__isnanf(__x); }
+__DEVICE__ __CONSTEXPR__ bool isnan(double __x) { return ::__isnan(__x); }
+
+#if defined(__OPENMP_AMDGCN__)
+#pragma omp end declare variant
+#endif // defined(__OPENMP_AMDGCN__)
+
+__DEVICE__ __CONSTEXPR__ bool isgreater(float __x, float __y) {
   return __builtin_isgreater(__x, __y);
 }
-__DEVICE__ bool isgreater(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool isgreater(double __x, double __y) {
   return __builtin_isgreater(__x, __y);
 }
-__DEVICE__ bool isgreaterequal(float __x, float __y) {
+__DEVICE__ __CONSTEXPR__ bool isgreaterequal(float __x, float __y) {
   return __builtin_isgreaterequal(__x, __y);
 }
-__DEVICE__ bool isgreaterequal(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool isgreaterequal(double __x, double __y) {
   return __builtin_isgreaterequal(__x, __y);
 }
-__DEVICE__ bool isinf(float __x) { return ::__isinff(__x); }
-__DEVICE__ bool isinf(double __x) { return ::__isinf(__x); }
-__DEVICE__ bool isless(float __x, float __y) {
+__DEVICE__ __CONSTEXPR__ bool isless(float __x, float __y) {
   return __builtin_isless(__x, __y);
 }
-__DEVICE__ bool isless(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool isless(double __x, double __y) {
   return __builtin_isless(__x, __y);
 }
-__DEVICE__ bool islessequal(float __x, float __y) {
+__DEVICE__ __CONSTEXPR__ bool islessequal(float __x, float __y) {
   return __builtin_islessequal(__x, __y);
 }
-__DEVICE__ bool islessequal(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool islessequal(double __x, double __y) {
   return __builtin_islessequal(__x, __y);
 }
-__DEVICE__ bool islessgreater(float __x, float __y) {
+__DEVICE__ __CONSTEXPR__ bool islessgreater(float __x, float __y) {
   return __builtin_islessgreater(__x, __y);
 }
-__DEVICE__ bool islessgreater(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool islessgreater(double __x, double __y) {
   return __builtin_islessgreater(__x, __y);
 }
-__DEVICE__ bool isnan(float __x) { return ::__isnanf(__x); }
-__DEVICE__ bool isnan(double __x) { return ::__isnan(__x); }
-__DEVICE__ bool isnormal(float __x) { return __builtin_isnormal(__x); }
-__DEVICE__ bool isnormal(double __x) { return __builtin_isnormal(__x); }
-__DEVICE__ bool isunordered(float __x, float __y) {
+__DEVICE__ __CONSTEXPR__ bool isnormal(float __x) {
+  return __builtin_isnormal(__x);
+}
+__DEVICE__ __CONSTEXPR__ bool isnormal(double __x) {
+  return __builtin_isnormal(__x);
+}
+__DEVICE__ __CONSTEXPR__ bool isunordered(float __x, float __y) {
   return __builtin_isunordered(__x, __y);
 }
-__DEVICE__ bool isunordered(double __x, double __y) {
+__DEVICE__ __CONSTEXPR__ bool isunordered(double __x, double __y) {
   return __builtin_isunordered(__x, __y);
 }
-__DEVICE__ float modf(float __x, float *__iptr) { return ::modff(__x, __iptr); }
-__DEVICE__ float pow(float __base, int __iexp) {
+__DEVICE__ __CONSTEXPR__ float modf(float __x, float *__iptr) {
+  return ::modff(__x, __iptr);
+}
+__DEVICE__ __CONSTEXPR__ float pow(float __base, int __iexp) {
   return ::powif(__base, __iexp);
 }
-__DEVICE__ double pow(double __base, int __iexp) {
+__DEVICE__ __CONSTEXPR__ double pow(double __base, int __iexp) {
   return ::powi(__base, __iexp);
 }
-__DEVICE__ float remquo(float __x, float __y, int *__quo) {
+__DEVICE__ __CONSTEXPR__ float remquo(float __x, float __y, int *__quo) {
   return ::remquof(__x, __y, __quo);
 }
-__DEVICE__ float scalbln(float __x, long int __n) {
+__DEVICE__ __CONSTEXPR__ float scalbln(float __x, long int __n) {
   return ::scalblnf(__x, __n);
 }
-__DEVICE__ bool signbit(float __x) { return ::__signbitf(__x); }
-__DEVICE__ bool signbit(double __x) { return ::__signbit(__x); }
+__DEVICE__ __CONSTEXPR__ bool signbit(float __x) { return ::__signbitf(__x); }
+__DEVICE__ __CONSTEXPR__ bool signbit(double __x) { return ::__signbit(__x); }
 
 // Notably missing above is nexttoward.  We omit it because
 // ocml doesn't provide an implementation, and we don't want to be in the
 // business of implementing tricky libm functions in this header.
 
 // Other functions.
-__DEVICE__ _Float16 fma(_Float16 __x, _Float16 __y, _Float16 __z) {
+__DEVICE__ __CONSTEXPR__ _Float16 fma(_Float16 __x, _Float16 __y,
+                                      _Float16 __z) {
   return __ocml_fma_f16(__x, __y, __z);
 }
-__DEVICE__ _Float16 pow(_Float16 __base, int __iexp) {
+__DEVICE__ __CONSTEXPR__ _Float16 pow(_Float16 __base, int __iexp) {
   return __ocml_pown_f16(__base, __iexp);
 }
 
+#ifndef __OPENMP_AMDGCN__
 // BEGIN DEF_FUN and HIP_OVERLOAD
 
 // BEGIN DEF_FUN
@@ -127,18 +188,19 @@ __DEVICE__ _Float16 pow(_Float16 __base, int __iexp) {
 
 // Define cmath functions with float argument and returns __retty.
 #define __DEF_FUN1(__retty, __func)                                            \
-  __DEVICE__                                                                   \
-  __retty __func(float __x) { return __func##f(__x); }
+  __DEVICE__ __CONSTEXPR__ __retty __func(float __x) { return __func##f(__x); }
 
 // Define cmath functions with two float arguments and returns __retty.
 #define __DEF_FUN2(__retty, __func)                                            \
-  __DEVICE__                                                                   \
-  __retty __func(float __x, float __y) { return __func##f(__x, __y); }
+  __DEVICE__ __CONSTEXPR__ __retty __func(float __x, float __y) {              \
+    return __func##f(__x, __y);                                                \
+  }
 
 // Define cmath functions with a float and an int argument and returns __retty.
 #define __DEF_FUN2_FI(__retty, __func)                                         \
-  __DEVICE__                                                                   \
-  __retty __func(float __x, int __y) { return __func##f(__x, __y); }
+  __DEVICE__ __CONSTEXPR__ __retty __func(float __x, int __y) {                \
+    return __func##f(__x, __y);                                                \
+  }
 
 __DEF_FUN1(float, acos)
 __DEF_FUN1(float, acosh)
@@ -207,11 +269,117 @@ template <bool __B, class __T = void> struct __hip_enable_if {};
 
 template <class __T> struct __hip_enable_if<true, __T> { typedef __T type; };
 
+namespace __hip {
+template <class _Tp> struct is_integral {
+  enum { value = 0 };
+};
+template <> struct is_integral<bool> {
+  enum { value = 1 };
+};
+template <> struct is_integral<char> {
+  enum { value = 1 };
+};
+template <> struct is_integral<signed char> {
+  enum { value = 1 };
+};
+template <> struct is_integral<unsigned char> {
+  enum { value = 1 };
+};
+template <> struct is_integral<wchar_t> {
+  enum { value = 1 };
+};
+template <> struct is_integral<short> {
+  enum { value = 1 };
+};
+template <> struct is_integral<unsigned short> {
+  enum { value = 1 };
+};
+template <> struct is_integral<int> {
+  enum { value = 1 };
+};
+template <> struct is_integral<unsigned int> {
+  enum { value = 1 };
+};
+template <> struct is_integral<long> {
+  enum { value = 1 };
+};
+template <> struct is_integral<unsigned long> {
+  enum { value = 1 };
+};
+template <> struct is_integral<long long> {
+  enum { value = 1 };
+};
+template <> struct is_integral<unsigned long long> {
+  enum { value = 1 };
+};
+
+// ToDo: specializes is_arithmetic<_Float16>
+template <class _Tp> struct is_arithmetic {
+  enum { value = 0 };
+};
+template <> struct is_arithmetic<bool> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<char> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<signed char> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<unsigned char> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<wchar_t> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<short> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<unsigned short> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<int> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<unsigned int> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<long> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<unsigned long> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<long long> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<unsigned long long> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<float> {
+  enum { value = 1 };
+};
+template <> struct is_arithmetic<double> {
+  enum { value = 1 };
+};
+
+struct true_type {
+  static const __constant__ bool value = true;
+};
+struct false_type {
+  static const __constant__ bool value = false;
+};
+
+template <typename __T, typename __U> struct is_same : public false_type {};
+template <typename __T> struct is_same<__T, __T> : public true_type {};
+
+template <typename __T> struct add_rvalue_reference { typedef __T &&type; };
+
+template <typename __T> typename add_rvalue_reference<__T>::type declval();
+
 // decltype is only available in C++11 and above.
 #if __cplusplus >= 201103L
 // __hip_promote
-namespace __hip {
-
 template <class _Tp> struct __numeric_type {
   static void __test(...);
   static _Float16 __test(_Float16);
@@ -227,8 +395,8 @@ template <class _Tp> struct __numeric_type {
   // No support for long double, use double instead.
   static double __test(long double);
 
-  typedef decltype(__test(std::declval<_Tp>())) type;
-  static const bool value = !std::is_same<type, void>::value;
+  typedef decltype(__test(declval<_Tp>())) type;
+  static const bool value = !is_same<type, void>::value;
 };
 
 template <> struct __numeric_type<void> { static const bool value = true; };
@@ -271,18 +439,17 @@ public:
 
 template <class _A1, class _A2 = void, class _A3 = void>
 class __promote : public __promote_imp<_A1, _A2, _A3> {};
-
-} // namespace __hip
 #endif //__cplusplus >= 201103L
+} // namespace __hip
 
 // __HIP_OVERLOAD1 is used to resolve function calls with integer argument to
 // avoid compilation error due to ambibuity. e.g. floor(5) is resolved with
 // floor(double).
 #define __HIP_OVERLOAD1(__retty, __fn)                                         \
   template <typename __T>                                                      \
-  __DEVICE__ typename __hip_enable_if<std::numeric_limits<__T>::is_integer,    \
-                                      __retty>::type                           \
-  __fn(__T __x) {                                                              \
+  __DEVICE__ __CONSTEXPR__                                                     \
+      typename __hip_enable_if<__hip::is_integral<__T>::value, __retty>::type  \
+      __fn(__T __x) {                                                          \
     return ::__fn((double)__x);                                                \
   }
 
@@ -292,9 +459,8 @@ class __promote : public __promote_imp<_A1, _A2, _A3> {};
 #if __cplusplus >= 201103L
 #define __HIP_OVERLOAD2(__retty, __fn)                                         \
   template <typename __T1, typename __T2>                                      \
-  __DEVICE__ typename __hip_enable_if<                                         \
-      std::numeric_limits<__T1>::is_specialized &&                             \
-          std::numeric_limits<__T2>::is_specialized,                           \
+  __DEVICE__ __CONSTEXPR__ typename __hip_enable_if<                           \
+      __hip::is_arithmetic<__T1>::value && __hip::is_arithmetic<__T2>::value,  \
       typename __hip::__promote<__T1, __T2>::type>::type                       \
   __fn(__T1 __x, __T2 __y) {                                                   \
     typedef typename __hip::__promote<__T1, __T2>::type __result_type;         \
@@ -303,16 +469,15 @@ class __promote : public __promote_imp<_A1, _A2, _A3> {};
 #else
 #define __HIP_OVERLOAD2(__retty, __fn)                                         \
   template <typename __T1, typename __T2>                                      \
-  __DEVICE__                                                                   \
-      typename __hip_enable_if<std::numeric_limits<__T1>::is_specialized &&    \
-                                   std::numeric_limits<__T2>::is_specialized,  \
+  __DEVICE__ __CONSTEXPR__                                                     \
+      typename __hip_enable_if<__hip::is_arithmetic<__T1>::value &&            \
+                                   __hip::is_arithmetic<__T2>::value,          \
                                __retty>::type                                  \
       __fn(__T1 __x, __T2 __y) {                                               \
     return __fn((double)__x, (double)__y);                                     \
   }
 #endif
 
-__HIP_OVERLOAD1(double, abs)
 __HIP_OVERLOAD1(double, acos)
 __HIP_OVERLOAD1(double, acosh)
 __HIP_OVERLOAD1(double, asin)
@@ -336,7 +501,9 @@ __HIP_OVERLOAD1(double, floor)
 __HIP_OVERLOAD2(double, fmax)
 __HIP_OVERLOAD2(double, fmin)
 __HIP_OVERLOAD2(double, fmod)
+#if !defined(__HIPCC_RTC__)
 __HIP_OVERLOAD1(int, fpclassify)
+#endif // !defined(__HIPCC_RTC__)
 __HIP_OVERLOAD2(double, hypot)
 __HIP_OVERLOAD1(int, ilogb)
 __HIP_OVERLOAD1(bool, isfinite)
@@ -381,10 +548,9 @@ __HIP_OVERLOAD2(double, min)
 // Additional Overloads that don't quite match HIP_OVERLOAD.
 #if __cplusplus >= 201103L
 template <typename __T1, typename __T2, typename __T3>
-__DEVICE__ typename __hip_enable_if<
-    std::numeric_limits<__T1>::is_specialized &&
-        std::numeric_limits<__T2>::is_specialized &&
-        std::numeric_limits<__T3>::is_specialized,
+__DEVICE__ __CONSTEXPR__ typename __hip_enable_if<
+    __hip::is_arithmetic<__T1>::value && __hip::is_arithmetic<__T2>::value &&
+        __hip::is_arithmetic<__T3>::value,
     typename __hip::__promote<__T1, __T2, __T3>::type>::type
 fma(__T1 __x, __T2 __y, __T3 __z) {
   typedef typename __hip::__promote<__T1, __T2, __T3>::type __result_type;
@@ -392,10 +558,10 @@ fma(__T1 __x, __T2 __y, __T3 __z) {
 }
 #else
 template <typename __T1, typename __T2, typename __T3>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T1>::is_specialized &&
-                                 std::numeric_limits<__T2>::is_specialized &&
-                                 std::numeric_limits<__T3>::is_specialized,
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_arithmetic<__T1>::value &&
+                                 __hip::is_arithmetic<__T2>::value &&
+                                 __hip::is_arithmetic<__T3>::value,
                              double>::type
     fma(__T1 __x, __T2 __y, __T3 __z) {
   return ::fma((double)__x, (double)__y, (double)__z);
@@ -403,31 +569,31 @@ __DEVICE__
 #endif
 
 template <typename __T>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T>::is_integer, double>::type
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_integral<__T>::value, double>::type
     frexp(__T __x, int *__exp) {
   return ::frexp((double)__x, __exp);
 }
 
 template <typename __T>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T>::is_integer, double>::type
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_integral<__T>::value, double>::type
     ldexp(__T __x, int __exp) {
   return ::ldexp((double)__x, __exp);
 }
 
 template <typename __T>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T>::is_integer, double>::type
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_integral<__T>::value, double>::type
     modf(__T __x, double *__exp) {
   return ::modf((double)__x, __exp);
 }
 
 #if __cplusplus >= 201103L
 template <typename __T1, typename __T2>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T1>::is_specialized &&
-                                 std::numeric_limits<__T2>::is_specialized,
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_arithmetic<__T1>::value &&
+                                 __hip::is_arithmetic<__T2>::value,
                              typename __hip::__promote<__T1, __T2>::type>::type
     remquo(__T1 __x, __T2 __y, int *__quo) {
   typedef typename __hip::__promote<__T1, __T2>::type __result_type;
@@ -435,9 +601,9 @@ __DEVICE__
 }
 #else
 template <typename __T1, typename __T2>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T1>::is_specialized &&
-                                 std::numeric_limits<__T2>::is_specialized,
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_arithmetic<__T1>::value &&
+                                 __hip::is_arithmetic<__T2>::value,
                              double>::type
     remquo(__T1 __x, __T2 __y, int *__quo) {
   return ::remquo((double)__x, (double)__y, __quo);
@@ -445,15 +611,15 @@ __DEVICE__
 #endif
 
 template <typename __T>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T>::is_integer, double>::type
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_integral<__T>::value, double>::type
     scalbln(__T __x, long int __exp) {
   return ::scalbln((double)__x, __exp);
 }
 
 template <typename __T>
-__DEVICE__
-    typename __hip_enable_if<std::numeric_limits<__T>::is_integer, double>::type
+__DEVICE__ __CONSTEXPR__
+    typename __hip_enable_if<__hip::is_integral<__T>::value, double>::type
     scalbn(__T __x, int __exp) {
   return ::scalbn((double)__x, __exp);
 }
@@ -465,17 +631,20 @@ __DEVICE__
 
 // END DEF_FUN and HIP_OVERLOAD
 
+#endif // ifndef __OPENMP_AMDGCN__
 #endif // defined(__cplusplus)
 
+#ifndef __OPENMP_AMDGCN__
 // Define these overloads inside the namespace our standard library uses.
+#if !defined(__HIPCC_RTC__)
 #ifdef _LIBCPP_BEGIN_NAMESPACE_STD
 _LIBCPP_BEGIN_NAMESPACE_STD
 #else
 namespace std {
 #ifdef _GLIBCXX_BEGIN_NAMESPACE_VERSION
 _GLIBCXX_BEGIN_NAMESPACE_VERSION
-#endif
-#endif
+#endif // _GLIBCXX_BEGIN_NAMESPACE_VERSION
+#endif // _LIBCPP_BEGIN_NAMESPACE_STD
 
 // Pull the new overloads we defined above into namespace std.
 // using ::abs; - This may be considered for C++.
@@ -620,11 +789,13 @@ _LIBCPP_END_NAMESPACE_STD
 #else
 #ifdef _GLIBCXX_BEGIN_NAMESPACE_VERSION
 _GLIBCXX_END_NAMESPACE_VERSION
-#endif
+#endif // _GLIBCXX_BEGIN_NAMESPACE_VERSION
 } // namespace std
-#endif
+#endif // _LIBCPP_END_NAMESPACE_STD
+#endif // !defined(__HIPCC_RTC__)
 
 // Define device-side math functions from <ymath.h> on MSVC.
+#if !defined(__HIPCC_RTC__)
 #if defined(_MSC_VER)
 
 // Before VS2019, `<ymath.h>` is also included in `<limits>` and other headers.
@@ -636,29 +807,36 @@ _GLIBCXX_END_NAMESPACE_VERSION
 #if defined(__cplusplus)
 extern "C" {
 #endif // defined(__cplusplus)
-__DEVICE__ __attribute__((overloadable)) double _Cosh(double x, double y) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) double _Cosh(double x,
+                                                                    double y) {
   return cosh(x) * y;
 }
-__DEVICE__ __attribute__((overloadable)) float _FCosh(float x, float y) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) float _FCosh(float x,
+                                                                    float y) {
   return coshf(x) * y;
 }
-__DEVICE__ __attribute__((overloadable)) short _Dtest(double *p) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) short _Dtest(double *p) {
   return fpclassify(*p);
 }
-__DEVICE__ __attribute__((overloadable)) short _FDtest(float *p) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) short _FDtest(float *p) {
   return fpclassify(*p);
 }
-__DEVICE__ __attribute__((overloadable)) double _Sinh(double x, double y) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) double _Sinh(double x,
+                                                                    double y) {
   return sinh(x) * y;
 }
-__DEVICE__ __attribute__((overloadable)) float _FSinh(float x, float y) {
+__DEVICE__ __CONSTEXPR__ __attribute__((overloadable)) float _FSinh(float x,
+                                                                    float y) {
   return sinhf(x) * y;
 }
 #if defined(__cplusplus)
 }
 #endif // defined(__cplusplus)
 #endif // defined(_MSC_VER)
+#endif // !defined(__HIPCC_RTC__)
+#endif // ifndef __OPENMP_AMDGCN__
 
 #pragma pop_macro("__DEVICE__")
+#pragma pop_macro("__CONSTEXPR__")
 
 #endif // __CLANG_HIP_CMATH_H__

lib/include/__clang_hip_libdevice_declares.h

@@ -138,14 +138,22 @@ __device__ __attribute__((const)) float __ocml_fma_rtn_f32(float, float, float);
 __device__ __attribute__((const)) float __ocml_fma_rtp_f32(float, float, float);
 __device__ __attribute__((const)) float __ocml_fma_rtz_f32(float, float, float);
 
-__device__ __attribute__((const)) float
-__llvm_amdgcn_cos_f32(float) __asm("llvm.amdgcn.cos.f32");
-__device__ __attribute__((const)) float
-__llvm_amdgcn_rcp_f32(float) __asm("llvm.amdgcn.rcp.f32");
-__device__ __attribute__((const)) float
-__llvm_amdgcn_rsq_f32(float) __asm("llvm.amdgcn.rsq.f32");
-__device__ __attribute__((const)) float
-__llvm_amdgcn_sin_f32(float) __asm("llvm.amdgcn.sin.f32");
+__device__ inline __attribute__((const)) float
+__llvm_amdgcn_cos_f32(float __x) {
+  return __builtin_amdgcn_cosf(__x);
+}
+__device__ inline __attribute__((const)) float
+__llvm_amdgcn_rcp_f32(float __x) {
+  return __builtin_amdgcn_rcpf(__x);
+}
+__device__ inline __attribute__((const)) float
+__llvm_amdgcn_rsq_f32(float __x) {
+  return __builtin_amdgcn_rsqf(__x);
+}
+__device__ inline __attribute__((const)) float
+__llvm_amdgcn_sin_f32(float __x) {
+  return __builtin_amdgcn_sinf(__x);
+}
 // END INTRINSICS
 // END FLOAT
 
@@ -269,10 +277,14 @@ __device__ __attribute__((const)) double __ocml_fma_rtp_f64(double, double,
 __device__ __attribute__((const)) double __ocml_fma_rtz_f64(double, double,
                                                             double);
 
-__device__ __attribute__((const)) double
-__llvm_amdgcn_rcp_f64(double) __asm("llvm.amdgcn.rcp.f64");
-__device__ __attribute__((const)) double
-__llvm_amdgcn_rsq_f64(double) __asm("llvm.amdgcn.rsq.f64");
+__device__ inline __attribute__((const)) double
+__llvm_amdgcn_rcp_f64(double __x) {
+  return __builtin_amdgcn_rcp(__x);
+}
+__device__ inline __attribute__((const)) double
+__llvm_amdgcn_rsq_f64(double __x) {
+  return __builtin_amdgcn_rsq(__x);
+}
 
 __device__ __attribute__((const)) _Float16 __ocml_ceil_f16(_Float16);
 __device__ _Float16 __ocml_cos_f16(_Float16);

lib/include/__clang_hip_math.h

@@ -9,26 +9,40 @@
 #ifndef __CLANG_HIP_MATH_H__
 #define __CLANG_HIP_MATH_H__
 
-#if !defined(__HIP__)
+#if !defined(__HIP__) && !defined(__OPENMP_AMDGCN__)
 #error "This file is for HIP and OpenMP AMDGCN device compilation only."
 #endif
 
+#if !defined(__HIPCC_RTC__)
 #if defined(__cplusplus)
 #include <algorithm>
 #endif
 #include <limits.h>
 #include <stdint.h>
+#ifdef __OPENMP_AMDGCN__
+#include <omp.h>
+#endif
+#endif // !defined(__HIPCC_RTC__)
 
 #pragma push_macro("__DEVICE__")
+
+#ifdef __OPENMP_AMDGCN__
+#define __DEVICE__ static inline __attribute__((always_inline, nothrow))
+#else
 #define __DEVICE__ static __device__ inline __attribute__((always_inline))
+#endif
 
 // A few functions return bool type starting only in C++11.
 #pragma push_macro("__RETURN_TYPE")
+#ifdef __OPENMP_AMDGCN__
+#define __RETURN_TYPE int
+#else
 #if defined(__cplusplus)
 #define __RETURN_TYPE bool
 #else
 #define __RETURN_TYPE int
 #endif
+#endif // __OPENMP_AMDGCN__
 
 #if defined (__cplusplus) && __cplusplus < 201103L
 // emulate static_assert on type sizes
@@ -36,7 +50,7 @@ template<bool>
 struct __compare_result{};
 template<>
 struct __compare_result<true> {
-  static const bool valid;
+  static const __device__ bool valid;
 };
 
 __DEVICE__
@@ -247,6 +261,9 @@ float fmodf(float __x, float __y) { return __ocml_fmod_f32(__x, __y); }
 __DEVICE__
 float frexpf(float __x, int *__nptr) {
   int __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   float __r =
       __ocml_frexp_f32(__x, (__attribute__((address_space(5))) int *)&__tmp);
   *__nptr = __tmp;
@@ -332,6 +349,9 @@ long int lroundf(float __x) { return __ocml_round_f32(__x); }
 __DEVICE__
 float modff(float __x, float *__iptr) {
   float __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   float __r =
       __ocml_modf_f32(__x, (__attribute__((address_space(5))) float *)&__tmp);
   *__iptr = __tmp;
@@ -412,6 +432,9 @@ float remainderf(float __x, float __y) {
 __DEVICE__
 float remquof(float __x, float __y, int *__quo) {
   int __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   float __r = __ocml_remquo_f32(
       __x, __y, (__attribute__((address_space(5))) int *)&__tmp);
   *__quo = __tmp;
@@ -468,6 +491,9 @@ __RETURN_TYPE __signbitf(float __x) { return __ocml_signbit_f32(__x); }
 __DEVICE__
 void sincosf(float __x, float *__sinptr, float *__cosptr) {
   float __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   *__sinptr =
       __ocml_sincos_f32(__x, (__attribute__((address_space(5))) float *)&__tmp);
   *__cosptr = __tmp;
@@ -476,6 +502,9 @@ void sincosf(float __x, float *__sinptr, float *__cosptr) {
 __DEVICE__
 void sincospif(float __x, float *__sinptr, float *__cosptr) {
   float __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   *__sinptr = __ocml_sincospi_f32(
       __x, (__attribute__((address_space(5))) float *)&__tmp);
   *__cosptr = __tmp;
@@ -788,6 +817,9 @@ double fmod(double __x, double __y) { return __ocml_fmod_f64(__x, __y); }
 __DEVICE__
 double frexp(double __x, int *__nptr) {
   int __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   double __r =
       __ocml_frexp_f64(__x, (__attribute__((address_space(5))) int *)&__tmp);
   *__nptr = __tmp;
@@ -872,6 +904,9 @@ long int lround(double __x) { return __ocml_round_f64(__x); }
 __DEVICE__
 double modf(double __x, double *__iptr) {
   double __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   double __r =
       __ocml_modf_f64(__x, (__attribute__((address_space(5))) double *)&__tmp);
   *__iptr = __tmp;
@@ -960,6 +995,9 @@ double remainder(double __x, double __y) {
 __DEVICE__
 double remquo(double __x, double __y, int *__quo) {
   int __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   double __r = __ocml_remquo_f64(
       __x, __y, (__attribute__((address_space(5))) int *)&__tmp);
   *__quo = __tmp;
@@ -1018,6 +1056,9 @@ double sin(double __x) { return __ocml_sin_f64(__x); }
 __DEVICE__
 void sincos(double __x, double *__sinptr, double *__cosptr) {
   double __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   *__sinptr = __ocml_sincos_f64(
       __x, (__attribute__((address_space(5))) double *)&__tmp);
   *__cosptr = __tmp;
@@ -1026,6 +1067,9 @@ void sincos(double __x, double *__sinptr, double *__cosptr) {
 __DEVICE__
 void sincospi(double __x, double *__sinptr, double *__cosptr) {
   double __tmp;
+#ifdef __OPENMP_AMDGCN__
+#pragma omp allocate(__tmp) allocator(omp_thread_mem_alloc)
+#endif
   *__sinptr = __ocml_sincospi_f64(
       __x, (__attribute__((address_space(5))) double *)&__tmp);
   *__cosptr = __tmp;
@@ -1260,6 +1304,7 @@ float min(float __x, float __y) { return fminf(__x, __y); }
 __DEVICE__
 double min(double __x, double __y) { return fmin(__x, __y); }
 
+#if !defined(__HIPCC_RTC__) && !defined(__OPENMP_AMDGCN__)
 __host__ inline static int min(int __arg1, int __arg2) {
   return std::min(__arg1, __arg2);
 }
@@ -1267,6 +1312,7 @@ __host__ inline static int min(int __arg1, int __arg2) {
 __host__ inline static int max(int __arg1, int __arg2) {
   return std::max(__arg1, __arg2);
 }
+#endif // !defined(__HIPCC_RTC__) && !defined(__OPENMP_AMDGCN__)
 #endif
 
 #pragma pop_macro("__DEVICE__")

lib/include/__clang_hip_runtime_wrapper.h

@@ -18,52 +18,107 @@
 
 #if __HIP__
 
-#include <cmath>
-#include <cstdlib>
-#include <stdlib.h>
-
 #define __host__ __attribute__((host))
 #define __device__ __attribute__((device))
 #define __global__ __attribute__((global))
 #define __shared__ __attribute__((shared))
 #define __constant__ __attribute__((constant))
+#define __managed__ __attribute__((managed))
 
 #if !defined(__cplusplus) || __cplusplus < 201103L
   #define nullptr NULL;
 #endif
 
+#ifdef __cplusplus
+extern "C" {
+  __attribute__((__visibility__("default")))
+  __attribute__((weak))
+  __attribute__((noreturn))
+  __device__ void __cxa_pure_virtual(void) {
+    __builtin_trap();
+  }
+  __attribute__((__visibility__("default")))
+  __attribute__((weak))
+  __attribute__((noreturn))
+  __device__ void __cxa_deleted_virtual(void) {
+    __builtin_trap();
+  }
+}
+#endif //__cplusplus
+
+#if !defined(__HIPCC_RTC__)
+#include <cmath>
+#include <cstdlib>
+#include <stdlib.h>
+#else
+typedef __SIZE_TYPE__ size_t;
+// Define macros which are needed to declare HIP device API's without standard
+// C/C++ headers. This is for readability so that these API's can be written
+// the same way as non-hipRTC use case. These macros need to be popped so that
+// they do not pollute users' name space.
+#pragma push_macro("NULL")
+#pragma push_macro("uint32_t")
+#pragma push_macro("uint64_t")
+#pragma push_macro("CHAR_BIT")
+#pragma push_macro("INT_MAX")
+#define NULL (void *)0
+#define uint32_t __UINT32_TYPE__
+#define uint64_t __UINT64_TYPE__
+#define CHAR_BIT __CHAR_BIT__
+#define INT_MAX __INTMAX_MAX__
+#endif // __HIPCC_RTC__
+
+typedef __SIZE_TYPE__ __hip_size_t;
+
+#ifdef __cplusplus
+extern "C" {
+#endif //__cplusplus
+
 #if __HIP_ENABLE_DEVICE_MALLOC__
-extern "C" __device__ void *__hip_malloc(size_t __size);
-extern "C" __device__ void *__hip_free(void *__ptr);
-static inline __device__ void *malloc(size_t __size) {
+__device__ void *__hip_malloc(__hip_size_t __size);
+__device__ void *__hip_free(void *__ptr);
+__attribute__((weak)) inline __device__ void *malloc(__hip_size_t __size) {
   return __hip_malloc(__size);
 }
-static inline __device__ void *free(void *__ptr) { return __hip_free(__ptr); }
-#else
-static inline __device__ void *malloc(size_t __size) {
-  __builtin_trap();
-  return nullptr;
+__attribute__((weak)) inline __device__ void *free(void *__ptr) {
+  return __hip_free(__ptr);
 }
-static inline __device__ void *free(void *__ptr) {
+#else
+__attribute__((weak)) inline __device__ void *malloc(__hip_size_t __size) {
   __builtin_trap();
-  return nullptr;
+  return (void *)0;
+}
+__attribute__((weak)) inline __device__ void *free(void *__ptr) {
+  __builtin_trap();
+  return (void *)0;
 }
 #endif
 
+#ifdef __cplusplus
+} // extern "C"
+#endif //__cplusplus
+
 #include <__clang_hip_libdevice_declares.h>
 #include <__clang_hip_math.h>
 
-#if !_OPENMP || __HIP_ENABLE_CUDA_WRAPPER_FOR_OPENMP__
+#if defined(__HIPCC_RTC__)
+#include <__clang_hip_cmath.h>
+#else
 #include <__clang_cuda_math_forward_declares.h>
 #include <__clang_hip_cmath.h>
 #include <__clang_cuda_complex_builtins.h>
-
 #include <algorithm>
 #include <complex>
 #include <new>
-#endif // !_OPENMP || __HIP_ENABLE_CUDA_WRAPPER_FOR_OPENMP__
+#endif // __HIPCC_RTC__
 
 #define __CLANG_HIP_RUNTIME_WRAPPER_INCLUDED__ 1
-
+#if defined(__HIPCC_RTC__)
+#pragma pop_macro("NULL")
+#pragma pop_macro("uint32_t")
+#pragma pop_macro("uint64_t")
+#pragma pop_macro("CHAR_BIT")
+#pragma pop_macro("INT_MAX")
+#endif // __HIPCC_RTC__
 #endif // __HIP__
 #endif // __CLANG_HIP_RUNTIME_WRAPPER_H__

lib/include/amxintrin.h

@@ -15,8 +15,13 @@
 #define __AMXINTRIN_H
 #ifdef __x86_64__
 
+/* Define the default attributes for the functions in this file. */
 #define __DEFAULT_FN_ATTRS_TILE                                                \
   __attribute__((__always_inline__, __nodebug__, __target__("amx-tile")))
+#define __DEFAULT_FN_ATTRS_INT8                                                \
+  __attribute__((__always_inline__, __nodebug__, __target__("amx-int8")))
+#define __DEFAULT_FN_ATTRS_BF16                                                \
+  __attribute__((__always_inline__, __nodebug__, __target__("amx-bf16")))
 
 /// Load tile configuration from a 64-byte memory location specified by
 /// "mem_addr". The tile configuration includes the tile type palette, the
@@ -25,7 +30,7 @@
 /// config and the tile data, and the tiles are zeroed. Any invalid
 /// configurations will result in #GP fault.
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> LDTILECFG </c> instruction.
 ///
@@ -41,7 +46,7 @@ _tile_loadconfig(const void *__config) {
 /// palette, the number of bytes per row, and the number of rows. If tiles
 /// are not configured, all zeroes will be stored to memory.
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> STTILECFG </c> instruction.
 ///
@@ -55,7 +60,7 @@ _tile_storeconfig(void *__config) {
 /// Release the tile configuration to return to the init state, which
 /// releases all storage it currently holds.
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TILERELEASE </c> instruction.
 static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
@@ -66,7 +71,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// destination tile "dst" using the tile configuration previously configured
 /// via "_tile_loadconfig".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TILELOADD </c> instruction.
 ///
@@ -86,7 +91,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// that the data will likely not be reused in the near future and the data
 /// caching can be optimized accordingly.
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TILELOADDT1 </c> instruction.
 ///
@@ -104,7 +109,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// "stride" using the tile configuration previously configured via
 /// "_tile_loadconfig".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TILESTORED </c> instruction.
 ///
@@ -119,7 +124,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 
 /// Zero the tile specified by "tdest".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TILEZERO </c> instruction.
 ///
@@ -133,7 +138,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// results. Sum these 4 results with the corresponding 32-bit integer in "dst",
 /// and store the 32-bit result back to tile "dst".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TDPBSSD </c> instruction.
 ///
@@ -152,7 +157,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// 32-bit results. Sum these 4 results with the corresponding 32-bit integer
 /// in "dst", and store the 32-bit result back to tile "dst".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TDPBSUD </c> instruction.
 ///
@@ -171,7 +176,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// results. Sum these 4 results with the corresponding 32-bit integer in "dst",
 /// and store the 32-bit result back to tile "dst".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TDPBUSD </c> instruction.
 ///
@@ -190,7 +195,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// 32-bit results. Sum these 4 results with the corresponding 32-bit integer in
 /// "dst", and store the 32-bit result back to tile "dst".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TDPBUUD </c> instruction.
 ///
@@ -208,7 +213,7 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 /// elements with elements in "dst", and store the 32-bit result back to tile
 /// "dst".
 ///
-/// \headerfile <x86intrin.h>
+/// \headerfile <immintrin.h>
 ///
 /// This intrinsic corresponds to the <c> TDPBF16PS </c> instruction.
 ///
@@ -221,10 +226,12 @@ static __inline__ void __DEFAULT_FN_ATTRS_TILE _tile_release(void) {
 #define _tile_dpbf16ps(dst, src0, src1)                                        \
   __builtin_ia32_tdpbf16ps((dst), (src0), (src1))
 
-#define __DEFAULT_FN_ATTRS_INT8                                                \
-  __attribute__((__always_inline__, __nodebug__, __target__("amx-int8")))
-
+/// AMX tile register size can be configured, the maximum size is 16x64=1024
+/// bytes. Since there is no 2D type in llvm IR, we use vector type to
+/// represent 2D tile and the fixed size is maximum amx tile register size.
 typedef int _tile1024i __attribute__((__vector_size__(1024), __aligned__(64)));
+
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
 static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
 _tile_loadd_internal(unsigned short m, unsigned short n, const void *base,
                      __SIZE_TYPE__ stride) {
@@ -232,12 +239,43 @@ _tile_loadd_internal(unsigned short m, unsigned short n, const void *base,
                                              (__SIZE_TYPE__)(stride));
 }
 
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
+static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
+_tile_loaddt1_internal(unsigned short m, unsigned short n, const void *base,
+                       __SIZE_TYPE__ stride) {
+  return __builtin_ia32_tileloaddt164_internal(m, n, base,
+                                               (__SIZE_TYPE__)(stride));
+}
+
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
 static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
 _tile_dpbssd_internal(unsigned short m, unsigned short n, unsigned short k,
                       _tile1024i dst, _tile1024i src1, _tile1024i src2) {
   return __builtin_ia32_tdpbssd_internal(m, n, k, dst, src1, src2);
 }
 
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
+static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
+_tile_dpbsud_internal(unsigned short m, unsigned short n, unsigned short k,
+                      _tile1024i dst, _tile1024i src1, _tile1024i src2) {
+  return __builtin_ia32_tdpbsud_internal(m, n, k, dst, src1, src2);
+}
+
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
+static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
+_tile_dpbusd_internal(unsigned short m, unsigned short n, unsigned short k,
+                      _tile1024i dst, _tile1024i src1, _tile1024i src2) {
+  return __builtin_ia32_tdpbusd_internal(m, n, k, dst, src1, src2);
+}
+
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
+static __inline__ _tile1024i __DEFAULT_FN_ATTRS_INT8
+_tile_dpbuud_internal(unsigned short m, unsigned short n, unsigned short k,
+                      _tile1024i dst, _tile1024i src1, _tile1024i src2) {
+  return __builtin_ia32_tdpbuud_internal(m, n, k, dst, src1, src2);
+}
+
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
 static __inline__ void __DEFAULT_FN_ATTRS_INT8
 _tile_stored_internal(unsigned short m, unsigned short n, void *base,
                       __SIZE_TYPE__ stride, _tile1024i tile) {
@@ -245,34 +283,211 @@ _tile_stored_internal(unsigned short m, unsigned short n, void *base,
                                               (__SIZE_TYPE__)(stride), tile);
 }
 
+/// This is internal intrinsic. C/C++ user should avoid calling it directly.
+static __inline__ _tile1024i __DEFAULT_FN_ATTRS_BF16
+_tile_dpbf16ps_internal(unsigned short m, unsigned short n, unsigned short k,
+                        _tile1024i dst, _tile1024i src1, _tile1024i src2) {
+  return __builtin_ia32_tdpbf16ps_internal(m, n, k, dst, src1, src2);
+}
+
+/// This struct pack the shape and tile data together for user. We suggest
+/// initializing the struct as early as possible, because compiler depends
+/// on the shape information to do configure. The constant value is preferred
+/// for optimization by compiler.
 typedef struct __tile1024i_str {
   const unsigned short row;
   const unsigned short col;
   _tile1024i tile;
 } __tile1024i;
 
+/// Load tile rows from memory specifieid by "base" address and "stride" into
+/// destination tile "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TILELOADD </c> instruction.
+///
+/// \param dst
+///    A destination tile. Max size is 1024 Bytes.
+/// \param base
+///    A pointer to base address.
+/// \param stride
+///    The stride between the rows' data to be loaded in memory.
 __DEFAULT_FN_ATTRS_TILE
 static void __tile_loadd(__tile1024i *dst, const void *base,
                          __SIZE_TYPE__ stride) {
   dst->tile = _tile_loadd_internal(dst->row, dst->col, base, stride);
 }
 
-__DEFAULT_FN_ATTRS_INT8
-static void __tile_dpbssd(__tile1024i *dst, __tile1024i src1,
-                          __tile1024i src2) {
-  dst->tile = _tile_dpbssd_internal(src1.row, src2.col, src1.col, dst->tile,
-                                    src1.tile, src2.tile);
+/// Load tile rows from memory specifieid by "base" address and "stride" into
+/// destination tile "dst". This intrinsic provides a hint to the implementation
+/// that the data will likely not be reused in the near future and the data
+/// caching can be optimized accordingly.
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TILELOADDT1 </c> instruction.
+///
+/// \param dst
+///    A destination tile. Max size is 1024 Bytes.
+/// \param base
+///    A pointer to base address.
+/// \param stride
+///    The stride between the rows' data to be loaded in memory.
+__DEFAULT_FN_ATTRS_TILE
+static void __tile_stream_loadd(__tile1024i *dst, const void *base,
+                                __SIZE_TYPE__ stride) {
+  dst->tile = _tile_loaddt1_internal(dst->row, dst->col, base, stride);
 }
 
+/// Compute dot-product of bytes in tiles with a source/destination accumulator.
+/// Multiply groups of 4 adjacent pairs of signed 8-bit integers in src0 with
+/// corresponding signed 8-bit integers in src1, producing 4 intermediate 32-bit
+/// results. Sum these 4 results with the corresponding 32-bit integer in "dst",
+/// and store the 32-bit result back to tile "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TDPBSSD </c> instruction.
+///
+/// \param dst
+///    The destination tile. Max size is 1024 Bytes.
+/// \param src0
+///    The 1st source tile. Max size is 1024 Bytes.
+/// \param src1
+///    The 2nd source tile. Max size is 1024 Bytes.
+__DEFAULT_FN_ATTRS_INT8
+static void __tile_dpbssd(__tile1024i *dst, __tile1024i src0,
+                          __tile1024i src1) {
+  dst->tile = _tile_dpbssd_internal(src0.row, src1.col, src0.col, dst->tile,
+                                    src0.tile, src1.tile);
+}
+
+/// Compute dot-product of bytes in tiles with a source/destination accumulator.
+/// Multiply groups of 4 adjacent pairs of signed 8-bit integers in src0 with
+/// corresponding unsigned 8-bit integers in src1, producing 4 intermediate
+/// 32-bit results. Sum these 4 results with the corresponding 32-bit integer
+/// in "dst", and store the 32-bit result back to tile "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TDPBSUD </c> instruction.
+///
+/// \param dst
+///    The destination tile. Max size is 1024 Bytes.
+/// \param src0
+///    The 1st source tile. Max size is 1024 Bytes.
+/// \param src1
+///    The 2nd source tile. Max size is 1024 Bytes.
+__DEFAULT_FN_ATTRS_INT8
+static void __tile_dpbsud(__tile1024i *dst, __tile1024i src0,
+                          __tile1024i src1) {
+  dst->tile = _tile_dpbsud_internal(src0.row, src1.col, src0.col, dst->tile,
+                                    src0.tile, src1.tile);
+}
+
+/// Compute dot-product of bytes in tiles with a source/destination accumulator.
+/// Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in src0 with
+/// corresponding signed 8-bit integers in src1, producing 4 intermediate 32-bit
+/// results. Sum these 4 results with the corresponding 32-bit integer in "dst",
+/// and store the 32-bit result back to tile "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TDPBUSD </c> instruction.
+///
+/// \param dst
+///    The destination tile. Max size is 1024 Bytes.
+/// \param src0
+///    The 1st source tile. Max size is 1024 Bytes.
+/// \param src1
+///    The 2nd source tile. Max size is 1024 Bytes.
+__DEFAULT_FN_ATTRS_INT8
+static void __tile_dpbusd(__tile1024i *dst, __tile1024i src0,
+                          __tile1024i src1) {
+  dst->tile = _tile_dpbusd_internal(src0.row, src1.col, src0.col, dst->tile,
+                                    src0.tile, src1.tile);
+}
+
+/// Compute dot-product of bytes in tiles with a source/destination accumulator.
+/// Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in src0 with
+/// corresponding unsigned 8-bit integers in src1, producing 4 intermediate
+/// 32-bit results. Sum these 4 results with the corresponding 32-bit integer in
+/// "dst", and store the 32-bit result back to tile "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TDPBUUD </c> instruction.
+///
+/// \param dst
+///    The destination tile. Max size is 1024 Bytes.
+/// \param src0
+///    The 1st source tile. Max size is 1024 Bytes.
+/// \param src1
+///    The 2nd source tile. Max size is 1024 Bytes.
+__DEFAULT_FN_ATTRS_INT8
+static void __tile_dpbuud(__tile1024i *dst, __tile1024i src0,
+                          __tile1024i src1) {
+  dst->tile = _tile_dpbuud_internal(src0.row, src1.col, src0.col, dst->tile,
+                                    src0.tile, src1.tile);
+}
+
+/// Store the tile specified by "src" to memory specifieid by "base" address and
+/// "stride".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TILESTORED </c> instruction.
+///
+/// \param dst
+///    A destination tile. Max size is 1024 Bytes.
+/// \param base
+///    A pointer to base address.
+/// \param stride
+///    The stride between the rows' data to be stored in memory.
 __DEFAULT_FN_ATTRS_TILE
 static void __tile_stored(void *base, __SIZE_TYPE__ stride, __tile1024i src) {
   _tile_stored_internal(src.row, src.col, base, stride, src.tile);
 }
 
+/// Zero the tile specified by "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TILEZERO </c> instruction.
+///
+/// \param dst
+///    The destination tile to be zero. Max size is 1024 Bytes.
 __DEFAULT_FN_ATTRS_TILE
 static void __tile_zero(__tile1024i *dst) {
   dst->tile = __builtin_ia32_tilezero_internal(dst->row, dst->col);
 }
 
+/// Compute dot-product of BF16 (16-bit) floating-point pairs in tiles src0 and
+/// src1, accumulating the intermediate single-precision (32-bit) floating-point
+/// elements with elements in "dst", and store the 32-bit result back to tile
+/// "dst".
+///
+/// \headerfile <immintrin.h>
+///
+/// This intrinsic corresponds to the <c> TDPBF16PS </c> instruction.
+///
+/// \param dst
+///    The destination tile. Max size is 1024 Bytes.
+/// \param src0
+///    The 1st source tile. Max size is 1024 Bytes.
+/// \param src1
+///    The 2nd source tile. Max size is 1024 Bytes.
+__DEFAULT_FN_ATTRS_BF16
+static void __tile_dpbf16ps(__tile1024i *dst, __tile1024i src0,
+                            __tile1024i src1) {
+  dst->tile = _tile_dpbf16ps_internal(src0.row, src1.col, src0.col, dst->tile,
+                                      src0.tile, src1.tile);
+}
+
+#undef __DEFAULT_FN_ATTRS_TILE
+#undef __DEFAULT_FN_ATTRS_INT8
+#undef __DEFAULT_FN_ATTRS_BF16
+
 #endif /* __x86_64__ */
 #endif /* __AMXINTRIN_H */

lib/include/arm_acle.h

@@ -639,6 +639,49 @@ __jcvt(double __a) {
 }
 #endif
 
+/* Armv8.5-A FP rounding intrinsics */
+#if __ARM_64BIT_STATE && defined(__ARM_FEATURE_FRINT)
+static __inline__ float __attribute__((__always_inline__, __nodebug__))
+__frint32zf(float __a) {
+  return __builtin_arm_frint32zf(__a);
+}
+
+static __inline__ double __attribute__((__always_inline__, __nodebug__))
+__frint32z(double __a) {
+  return __builtin_arm_frint32z(__a);
+}
+
+static __inline__ float __attribute__((__always_inline__, __nodebug__))
+__frint64zf(float __a) {
+  return __builtin_arm_frint64zf(__a);
+}
+
+static __inline__ double __attribute__((__always_inline__, __nodebug__))
+__frint64z(double __a) {
+  return __builtin_arm_frint64z(__a);
+}
+
+static __inline__ float __attribute__((__always_inline__, __nodebug__))
+__frint32xf(float __a) {
+  return __builtin_arm_frint32xf(__a);
+}
+
+static __inline__ double __attribute__((__always_inline__, __nodebug__))
+__frint32x(double __a) {
+  return __builtin_arm_frint32x(__a);
+}
+
+static __inline__ float __attribute__((__always_inline__, __nodebug__))
+__frint64xf(float __a) {
+  return __builtin_arm_frint64xf(__a);
+}
+
+static __inline__ double __attribute__((__always_inline__, __nodebug__))
+__frint64x(double __a) {
+  return __builtin_arm_frint64x(__a);
+}
+#endif
+
 /* Armv8.7-A load/store 64-byte intrinsics */
 #if __ARM_64BIT_STATE && defined(__ARM_FEATURE_LS64)
 typedef struct {
@@ -709,6 +752,18 @@ __arm_st64bv0(void *__addr, data512_t __value) {
 
 #endif /* __ARM_FEATURE_TME */
 
+/* Armv8.5-A Random number generation intrinsics */
+#if __ARM_64BIT_STATE && defined(__ARM_FEATURE_RNG)
+static __inline__ int __attribute__((__always_inline__, __nodebug__))
+__rndr(uint64_t *__p) {
+  return __builtin_arm_rndr(__p);
+}
+static __inline__ int __attribute__((__always_inline__, __nodebug__))
+__rndrrs(uint64_t *__p) {
+  return __builtin_arm_rndrrs(__p);
+}
+#endif
+
 #if defined(__cplusplus)
 }
 #endif

lib/include/avx512fintrin.h

@@ -9300,8 +9300,11 @@ _mm512_mask_abs_pd(__m512d __W, __mmask8 __K, __m512d __A)
  * computations. In vector-reduction arithmetic, the evaluation order is
  * independent of the order of the input elements of V.
 
- * For floating point types, we always assume the elements are reassociable even
- * if -fast-math is off.
+ * For floating-point intrinsics:
+ * 1. When using fadd/fmul intrinsics, the order of operations within the
+ * vector is unspecified (associative math).
+ * 2. When using fmin/fmax intrinsics, NaN or -0.0 elements within the vector
+ * produce unspecified results.
 
  * Used bisection method. At each step, we partition the vector with previous
  * step in half, and the operation is performed on its two halves.
@@ -9524,75 +9527,49 @@ _mm512_mask_reduce_min_epu32(__mmask16 __M, __m512i __V) {
   return __builtin_ia32_reduce_umin_d512((__v16si)__V);
 }
 
-#define _mm512_mask_reduce_operator(op) \
-  __m256d __t1 = _mm512_extractf64x4_pd(__V, 0); \
-  __m256d __t2 = _mm512_extractf64x4_pd(__V, 1); \
-  __m256d __t3 = _mm256_##op(__t1, __t2); \
-  __m128d __t4 = _mm256_extractf128_pd(__t3, 0); \
-  __m128d __t5 = _mm256_extractf128_pd(__t3, 1); \
-  __m128d __t6 = _mm_##op(__t4, __t5); \
-  __m128d __t7 = __builtin_shufflevector(__t6, __t6, 1, 0); \
-  __m128d __t8 = _mm_##op(__t6, __t7); \
-  return __t8[0]
-
 static __inline__ double __DEFAULT_FN_ATTRS512
 _mm512_reduce_max_pd(__m512d __V) {
-  _mm512_mask_reduce_operator(max_pd);
+  return __builtin_ia32_reduce_fmax_pd512(__V);
 }
 
 static __inline__ double __DEFAULT_FN_ATTRS512
 _mm512_reduce_min_pd(__m512d __V) {
-  _mm512_mask_reduce_operator(min_pd);
+  return __builtin_ia32_reduce_fmin_pd512(__V);
 }
 
 static __inline__ double __DEFAULT_FN_ATTRS512
 _mm512_mask_reduce_max_pd(__mmask8 __M, __m512d __V) {
   __V = _mm512_mask_mov_pd(_mm512_set1_pd(-__builtin_inf()), __M, __V);
-  _mm512_mask_reduce_operator(max_pd);
+  return __builtin_ia32_reduce_fmax_pd512(__V);
 }
 
 static __inline__ double __DEFAULT_FN_ATTRS512
 _mm512_mask_reduce_min_pd(__mmask8 __M, __m512d __V) {
   __V = _mm512_mask_mov_pd(_mm512_set1_pd(__builtin_inf()), __M, __V);
-  _mm512_mask_reduce_operator(min_pd);
+  return __builtin_ia32_reduce_fmin_pd512(__V);
 }
-#undef _mm512_mask_reduce_operator
-
-#define _mm512_mask_reduce_operator(op) \
-  __m256 __t1 = (__m256)_mm512_extractf64x4_pd((__m512d)__V, 0); \
-  __m256 __t2 = (__m256)_mm512_extractf64x4_pd((__m512d)__V, 1); \
-  __m256 __t3 = _mm256_##op(__t1, __t2); \
-  __m128 __t4 = _mm256_extractf128_ps(__t3, 0); \
-  __m128 __t5 = _mm256_extractf128_ps(__t3, 1); \
-  __m128 __t6 = _mm_##op(__t4, __t5); \
-  __m128 __t7 = __builtin_shufflevector(__t6, __t6, 2, 3, 0, 1); \
-  __m128 __t8 = _mm_##op(__t6, __t7); \
-  __m128 __t9 = __builtin_shufflevector(__t8, __t8, 1, 0, 3, 2); \
-  __m128 __t10 = _mm_##op(__t8, __t9); \
-  return __t10[0]
 
 static __inline__ float __DEFAULT_FN_ATTRS512
 _mm512_reduce_max_ps(__m512 __V) {
-  _mm512_mask_reduce_operator(max_ps);
+  return __builtin_ia32_reduce_fmax_ps512(__V);
 }
 
 static __inline__ float __DEFAULT_FN_ATTRS512
 _mm512_reduce_min_ps(__m512 __V) {
-  _mm512_mask_reduce_operator(min_ps);
+  return __builtin_ia32_reduce_fmin_ps512(__V);
 }
 
 static __inline__ float __DEFAULT_FN_ATTRS512
 _mm512_mask_reduce_max_ps(__mmask16 __M, __m512 __V) {
   __V = _mm512_mask_mov_ps(_mm512_set1_ps(-__builtin_inff()), __M, __V);
-  _mm512_mask_reduce_operator(max_ps);
+  return __builtin_ia32_reduce_fmax_ps512(__V);
 }
 
 static __inline__ float __DEFAULT_FN_ATTRS512
 _mm512_mask_reduce_min_ps(__mmask16 __M, __m512 __V) {
   __V = _mm512_mask_mov_ps(_mm512_set1_ps(__builtin_inff()), __M, __V);
-  _mm512_mask_reduce_operator(min_ps);
+  return __builtin_ia32_reduce_fmin_ps512(__V);
 }
-#undef _mm512_mask_reduce_operator
 
 /// Moves the least significant 32 bits of a vector of [16 x i32] to a
 ///    32-bit signed integer value.
@@ -9611,6 +9588,169 @@ _mm512_cvtsi512_si32(__m512i __A) {
   return __b[0];
 }
 
+/// Loads 8 double-precision (64-bit) floating-point elements stored at memory
+/// locations starting at location \a base_addr at packed 32-bit integer indices
+/// stored in the lower half of \a vindex scaled by \a scale them in dst.
+///
+/// This intrinsic corresponds to the <c> VGATHERDPD </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///   dst[i+63:i] := MEM[addr+63:addr]
+/// ENDFOR
+/// dst[MAX:512] := 0
+/// \endoperation
+#define _mm512_i32logather_pd(vindex, base_addr, scale)                        \
+  _mm512_i32gather_pd(_mm512_castsi512_si256(vindex), (base_addr), (scale))
+
+/// Loads 8 double-precision (64-bit) floating-point elements from memory
+/// starting at location \a base_addr at packed 32-bit integer indices stored in
+/// the lower half of \a vindex scaled by \a scale into dst using writemask
+/// \a mask (elements are copied from \a src when the corresponding mask bit is
+/// not set).
+///
+/// This intrinsic corresponds to the <c> VGATHERDPD </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   IF mask[j]
+///     addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///     dst[i+63:i] := MEM[addr+63:addr]
+///   ELSE
+///     dst[i+63:i] := src[i+63:i]
+///   FI
+/// ENDFOR
+/// dst[MAX:512] := 0
+/// \endoperation
+#define _mm512_mask_i32logather_pd(src, mask, vindex, base_addr, scale)        \
+  _mm512_mask_i32gather_pd((src), (mask), _mm512_castsi512_si256(vindex),      \
+                           (base_addr), (scale))
+
+/// Loads 8 64-bit integer elements from memory starting at location \a base_addr
+/// at packed 32-bit integer indices stored in the lower half of \a vindex
+/// scaled by \a scale and stores them in dst.
+///
+/// This intrinsic corresponds to the <c> VPGATHERDQ </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///   dst[i+63:i] := MEM[addr+63:addr]
+/// ENDFOR
+/// dst[MAX:512] := 0
+/// \endoperation
+#define _mm512_i32logather_epi64(vindex, base_addr, scale)                     \
+  _mm512_i32gather_epi64(_mm512_castsi512_si256(vindex), (base_addr), (scale))
+
+/// Loads 8 64-bit integer elements from memory starting at location \a base_addr
+/// at packed 32-bit integer indices stored in the lower half of \a vindex
+/// scaled by \a scale and stores them in dst using writemask \a mask (elements
+/// are copied from \a src when the corresponding mask bit is not set).
+///
+/// This intrinsic corresponds to the <c> VPGATHERDQ </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   IF mask[j]
+///     addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///     dst[i+63:i] := MEM[addr+63:addr]
+///   ELSE
+///     dst[i+63:i] := src[i+63:i]
+///   FI
+/// ENDFOR
+/// dst[MAX:512] := 0
+/// \endoperation
+#define _mm512_mask_i32logather_epi64(src, mask, vindex, base_addr, scale)     \
+  _mm512_mask_i32gather_epi64((src), (mask), _mm512_castsi512_si256(vindex),   \
+                              (base_addr), (scale))
+
+/// Stores 8 packed double-precision (64-bit) floating-point elements in \a v1
+/// and to memory locations starting at location \a base_addr at packed 32-bit
+/// integer indices stored in \a vindex scaled by \a scale.
+///
+/// This intrinsic corresponds to the <c> VSCATTERDPD </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///   MEM[addr+63:addr] := v1[i+63:i]
+/// ENDFOR
+/// \endoperation
+#define _mm512_i32loscatter_pd(base_addr, vindex, v1, scale)                   \
+  _mm512_i32scatter_pd((base_addr), _mm512_castsi512_si256(vindex), (v1), (scale))
+
+/// Stores 8 packed double-precision (64-bit) floating-point elements in \a v1
+/// to memory locations starting at location \a base_addr at packed 32-bit
+/// integer indices stored in \a vindex scaled by \a scale. Only those elements
+/// whose corresponding mask bit is set in writemask \a mask are written to
+/// memory.
+///
+/// This intrinsic corresponds to the <c> VSCATTERDPD </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   IF mask[j]
+///     addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///     MEM[addr+63:addr] := a[i+63:i]
+///   FI
+/// ENDFOR
+/// \endoperation
+#define _mm512_mask_i32loscatter_pd(base_addr, mask, vindex, v1, scale)        \
+  _mm512_mask_i32scatter_pd((base_addr), (mask),                               \
+                            _mm512_castsi512_si256(vindex), (v1), (scale))
+
+/// Stores 8 packed 64-bit integer elements located in \a v1 and stores them in
+/// memory locations starting at location \a base_addr at packed 32-bit integer
+/// indices stored in \a vindex scaled by \a scale.
+///
+/// This intrinsic corresponds to the <c> VPSCATTERDQ </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///   MEM[addr+63:addr] := a[i+63:i]
+/// ENDFOR
+/// \endoperation
+#define _mm512_i32loscatter_epi64(base_addr, vindex, v1, scale)                \
+  _mm512_i32scatter_epi64((base_addr),                                         \
+                          _mm512_castsi512_si256(vindex), (v1), (scale))
+
+/// Stores 8 packed 64-bit integer elements located in a and stores them in
+/// memory locations starting at location \a base_addr at packed 32-bit integer
+/// indices stored in \a vindex scaled by scale using writemask \a mask (elements
+/// whose corresponding mask bit is not set are not written to memory).
+///
+/// This intrinsic corresponds to the <c> VPSCATTERDQ </c> instructions.
+///
+/// \operation
+/// FOR j := 0 to 7
+///   i := j*64
+///   m := j*32
+///   IF mask[j]
+///     addr := base_addr + SignExtend64(vindex[m+31:m]) * ZeroExtend64(scale) * 8
+///     MEM[addr+63:addr] := a[i+63:i]
+///   FI
+/// ENDFOR
+/// \endoperation
+#define _mm512_mask_i32loscatter_epi64(base_addr, mask, vindex, v1, scale)     \
+  _mm512_mask_i32scatter_epi64((base_addr), (mask),                            \
+                               _mm512_castsi512_si256(vindex), (v1), (scale))
+
 #undef __DEFAULT_FN_ATTRS512
 #undef __DEFAULT_FN_ATTRS128
 #undef __DEFAULT_FN_ATTRS

lib/include/builtins.h

@@ -0,0 +1,16 @@
+/*===---- builtins.h - Standard header for extra builtins -----------------===*\
+ *
+ * Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+ * See https://llvm.org/LICENSE.txt for license information.
+ * SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+ *
+\*===----------------------------------------------------------------------===*/
+
+/// Some legacy compilers have builtin definitions in a file named builtins.h.
+/// This header file has been added to allow compatibility with code that was
+/// written for those compilers. Code may have an include line for this file
+/// and to avoid an error an empty file with this name is provided.
+#ifndef __BUILTINS_H
+#define __BUILTINS_H
+
+#endif /* __BUILTINS_H */

lib/include/cuda_wrappers/complex

@@ -72,8 +72,16 @@
 #define _GLIBCXX_USE_C99_COMPLEX 0
 #define _GLIBCXX_USE_C99_COMPLEX_TR1 0
 
+// Work around a compatibility issue with libstdc++ 11.1.0
+// https://bugs.llvm.org/show_bug.cgi?id=50383
+#pragma push_macro("__failed_assertion")
+#if _GLIBCXX_RELEASE == 11
+#define __failed_assertion __cuda_failed_assertion
+#endif
+
 #include_next <complex>
 
+#pragma pop_macro("__failed_assertion")
 #pragma pop_macro("_GLIBCXX_USE_C99_COMPLEX_TR1")
 #pragma pop_macro("_GLIBCXX_USE_C99_COMPLEX")
 

lib/include/hexagon_circ_brev_intrinsics.h

@@ -0,0 +1,298 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef _HEXAGON_CIRC_BREV_INTRINSICS_H_
+#define _HEXAGON_CIRC_BREV_INTRINSICS_H_ 1
+
+#include <hexagon_protos.h>
+#include <stdint.h>
+
+/* Circular Load */
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_load_update_D(Word64 dst, Word64 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_D(dest,ptr,incr,bufsize,K)  \
+    { ptr = (int64_t *) HEXAGON_circ_ldd (ptr, &(dest), ((((K)+1)<<24)|((bufsize)<<3)), ((incr)*8)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_load_update_W(Word32 dst, Word32 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_W(dest,ptr,incr,bufsize,K)  \
+    { ptr = (int *) HEXAGON_circ_ldw (ptr, &(dest), (((K)<<24)|((bufsize)<<2)), ((incr)*4)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_load_update_H(Word16 dst, Word16 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_H(dest,ptr,incr,bufsize,K)  \
+    { ptr = (int16_t *) HEXAGON_circ_ldh (ptr, &(dest), ((((K)-1)<<24)|((bufsize)<<1)), ((incr)*2)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_load_update_UH( UWord16 dst,  UWord16 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_UH(dest,ptr,incr,bufsize,K) \
+    { ptr = (uint16_t *) HEXAGON_circ_lduh (ptr, &(dest), ((((K)-1)<<24)|((bufsize)<<1)), ((incr)*2)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_load_update_B(Word8 dst, Word8 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_B(dest,ptr,incr,bufsize,K)  \
+    { ptr = (int8_t *) HEXAGON_circ_ldb (ptr, &(dest), ((((K)-2)<<24)|(bufsize)), incr); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void  Q6_circ_load_update_UB(UWord8 dst, UWord8 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_load_update_UB(dest,ptr,incr,bufsize,K) \
+    { ptr = (uint8_t *) HEXAGON_circ_ldub (ptr, &(dest), ((((K)-2)<<24)|(bufsize)), incr); }
+
+/* Circular Store */
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_store_update_D(Word64 *src, Word64 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_store_update_D(src,ptr,incr,bufsize,K)  \
+    { ptr = (int64_t *) HEXAGON_circ_std (ptr, src, ((((K)+1)<<24)|((bufsize)<<3)), ((incr)*8)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_store_update_W(Word32 *src, Word32 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_store_update_W(src,ptr,incr,bufsize,K)  \
+    { ptr = (int *) HEXAGON_circ_stw (ptr, src, (((K)<<24)|((bufsize)<<2)), ((incr)*4)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_store_update_HL(Word16 *src, Word16 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_store_update_HL(src,ptr,incr,bufsize,K) \
+    { ptr = (int16_t *) HEXAGON_circ_sth (ptr, src, ((((K)-1)<<24)|((bufsize)<<1)), ((incr)*2)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_store_update_HH(Word16 *src, Word16 *ptr, UWord32 incr, UWord32 bufsize, UWord32 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_store_update_HH(src,ptr,incr,bufsize,K) \
+    { ptr = (int16_t *) HEXAGON_circ_sthhi (ptr, src, ((((K)-1)<<24)|((bufsize)<<1)), ((incr)*2)); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_circ_store_update_B(Word8 *src, Word8 *ptr, UWord32 I4, UWord32 bufsize,  UWord64 K)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_circ_store_update_B(src,ptr,incr,bufsize,K)  \
+    { ptr = (int8_t *) HEXAGON_circ_stb (ptr, src, ((((K)-2)<<24)|(bufsize)), incr); }
+
+
+/* Bit Reverse Load */
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_D(Word64 dst, Word64 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_D(dest,ptr,log2bufsize) \
+    { ptr = (int64_t *) HEXAGON_brev_ldd (ptr, &(dest), (1<<(16-((log2bufsize) + 3)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_W(Word32 dst, Word32 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_W(dest,ptr,log2bufsize) \
+    { ptr = (int *) HEXAGON_brev_ldw (ptr, &(dest), (1<<(16-((log2bufsize) + 2)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_H(Word16 dst, Word16 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_H(dest,ptr,log2bufsize) \
+    { ptr = (int16_t *) HEXAGON_brev_ldh (ptr, &(dest), (1<<(16-((log2bufsize) + 1)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_UH(UWord16 dst,  UWord16 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_UH(dest,ptr,log2bufsize) \
+    { ptr = (uint16_t *) HEXAGON_brev_lduh (ptr, &(dest), (1<<(16-((log2bufsize) + 1)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_B(Word8 dst, Word8 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_B(dest,ptr,log2bufsize) \
+    { ptr = (int8_t *) HEXAGON_brev_ldb (ptr, &(dest), (1<<(16-((log2bufsize))))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_load_update_UB(UWord8 dst, UWord8 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_load_update_UB(dest,ptr,log2bufsize) \
+    { ptr = (uint8_t *) HEXAGON_brev_ldub (ptr, &(dest), (1<<(16-((log2bufsize))))); }
+
+/* Bit Reverse Store */
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_store_update_D(Word64 *src, Word64 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_store_update_D(src,ptr,log2bufsize)   \
+    { ptr = (int64_t *) HEXAGON_brev_std (ptr, src, (1<<(16-((log2bufsize) + 3)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_store_update_W(Word32 *src, Word32 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_store_update_W(src,ptr,log2bufsize)   \
+    { ptr = (int *) HEXAGON_brev_stw (ptr, src, (1<<(16-((log2bufsize) + 2)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_store_update_HL(Word16 *src, Word16 *ptr, Word32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_store_update_HL(src,ptr,log2bufsize)   \
+    { ptr = (int16_t *) HEXAGON_brev_sth (ptr, src, (1<<(16-((log2bufsize) + 1)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_store_update_HH(Word16 *src, Word16 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_store_update_HH(src,ptr,log2bufsize)   \
+    { ptr = (int16_t *) HEXAGON_brev_sthhi (ptr, src, (1<<(16-((log2bufsize) + 1)))); }
+
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: void Q6_bitrev_store_update_B(Word8 *src, Word8 *ptr, UWord32 Iu4)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#define Q6_bitrev_store_update_B(src,ptr,log2bufsize)   \
+    { ptr = (int8_t *) HEXAGON_brev_stb (ptr, src, (1<<(16-((log2bufsize))))); }
+
+
+#define HEXAGON_circ_ldd  __builtin_circ_ldd
+#define HEXAGON_circ_ldw  __builtin_circ_ldw
+#define HEXAGON_circ_ldh  __builtin_circ_ldh
+#define HEXAGON_circ_lduh __builtin_circ_lduh
+#define HEXAGON_circ_ldb  __builtin_circ_ldb
+#define HEXAGON_circ_ldub __builtin_circ_ldub
+
+
+#define HEXAGON_circ_std  __builtin_circ_std
+#define HEXAGON_circ_stw  __builtin_circ_stw
+#define HEXAGON_circ_sth  __builtin_circ_sth
+#define HEXAGON_circ_sthhi __builtin_circ_sthhi
+#define HEXAGON_circ_stb  __builtin_circ_stb
+
+
+#define HEXAGON_brev_ldd  __builtin_brev_ldd
+#define HEXAGON_brev_ldw  __builtin_brev_ldw
+#define HEXAGON_brev_ldh  __builtin_brev_ldh
+#define HEXAGON_brev_lduh __builtin_brev_lduh
+#define HEXAGON_brev_ldb  __builtin_brev_ldb
+#define HEXAGON_brev_ldub __builtin_brev_ldub
+
+#define HEXAGON_brev_std  __builtin_brev_std
+#define HEXAGON_brev_stw  __builtin_brev_stw
+#define HEXAGON_brev_sth  __builtin_brev_sth
+#define HEXAGON_brev_sthhi __builtin_brev_sthhi
+#define HEXAGON_brev_stb  __builtin_brev_stb
+
+#ifdef __HVX__
+/* ==========================================================================
+   Assembly Syntax:       if (Qt) vmem(Rt+#0) = Vs
+   C Intrinsic Prototype: void Q6_vmaskedstoreq_QAV(HVX_VectorPred Qt, HVX_VectorAddress A, HVX_Vector Vs)
+   Instruction Type:      COPROC_VMEM
+   Execution Slots:       SLOT0
+   ========================================================================== */
+
+#define Q6_vmaskedstoreq_QAV __BUILTIN_VECTOR_WRAP(__builtin_HEXAGON_V6_vmaskedstoreq)
+
+/* ==========================================================================
+   Assembly Syntax:       if (!Qt) vmem(Rt+#0) = Vs
+   C Intrinsic Prototype: void Q6_vmaskedstorenq_QAV(HVX_VectorPred Qt, HVX_VectorAddress A, HVX_Vector Vs)
+   Instruction Type:      COPROC_VMEM
+   Execution Slots:       SLOT0
+   ========================================================================== */
+
+#define Q6_vmaskedstorenq_QAV __BUILTIN_VECTOR_WRAP(__builtin_HEXAGON_V6_vmaskedstorenq)
+
+/* ==========================================================================
+   Assembly Syntax:       if (Qt) vmem(Rt+#0):nt = Vs
+   C Intrinsic Prototype: void Q6_vmaskedstorentq_QAV(HVX_VectorPred Qt, HVX_VectorAddress A, HVX_Vector Vs)
+   Instruction Type:      COPROC_VMEM
+   Execution Slots:       SLOT0
+   ========================================================================== */
+
+#define Q6_vmaskedstorentq_QAV __BUILTIN_VECTOR_WRAP(__builtin_HEXAGON_V6_vmaskedstorentq)
+
+/* ==========================================================================
+   Assembly Syntax:       if (!Qt) vmem(Rt+#0):nt = Vs
+   C Intrinsic Prototype: void Q6_vmaskedstorentnq_QAV(HVX_VectorPred Qt, HVX_VectorAddress A, HVX_Vector Vs)
+   Instruction Type:      COPROC_VMEM
+   Execution Slots:       SLOT0
+   ========================================================================== */
+
+#define Q6_vmaskedstorentnq_QAV __BUILTIN_VECTOR_WRAP(__builtin_HEXAGON_V6_vmaskedstorentnq)
+
+#endif
+
+
+#endif  /* #ifndef _HEXAGON_CIRC_BREV_INTRINSICS_H_ */
+
+#ifdef __NOT_DEFINED__
+/*** comment block template  ***/
+/* ==========================================================================
+   Assembly Syntax:       Return=instruction()
+   C Intrinsic Prototype: ReturnType Intrinsic(ParamType Rs, ParamType Rt)
+   Instruction Type:      InstructionType
+   Execution Slots:       SLOT0123
+   ========================================================================== */
+#endif /***  __NOT_DEFINED__  ***/

lib/include/immintrin.h

@@ -72,11 +72,6 @@
 #include <f16cintrin.h>
 #endif
 
-#if !(defined(_MSC_VER) || defined(__SCE__)) || __has_feature(modules) ||      \
-    defined(__VPCLMULQDQ__)
-#include <vpclmulqdqintrin.h>
-#endif
-
 /* No feature check desired due to internal checks */
 #include <bmiintrin.h>
 
@@ -230,6 +225,11 @@
 #include <pkuintrin.h>
 #endif
 
+#if !(defined(_MSC_VER) || defined(__SCE__)) || __has_feature(modules) ||      \
+    defined(__VPCLMULQDQ__)
+#include <vpclmulqdqintrin.h>
+#endif
+
 #if !(defined(_MSC_VER) || defined(__SCE__)) || __has_feature(modules) ||      \
     defined(__VAES__)
 #include <vaesintrin.h>

lib/include/intrin.h

@@ -451,24 +451,47 @@ unsigned char _InterlockedCompareExchange128_rel(__int64 volatile *_Destination,
 static __inline__ void __DEFAULT_FN_ATTRS __movsb(unsigned char *__dst,
                                                   unsigned char const *__src,
                                                   size_t __n) {
-  __asm__ __volatile__("rep movsb" : "+D"(__dst), "+S"(__src), "+c"(__n)
-                       : : "memory");
+#if defined(__x86_64__)
+  __asm__ __volatile__("rep movsb"
+                       : "+D"(__dst), "+S"(__src), "+c"(__n)
+                       :
+                       : "memory");
+#else
+  __asm__ __volatile__("xchg %%esi, %1\nrep movsb\nxchg %%esi, %1"
+                       : "+D"(__dst), "+r"(__src), "+c"(__n)
+                       :
+                       : "memory");
+#endif
 }
 static __inline__ void __DEFAULT_FN_ATTRS __movsd(unsigned long *__dst,
                                                   unsigned long const *__src,
                                                   size_t __n) {
+#if defined(__x86_64__)
   __asm__ __volatile__("rep movsl"
                        : "+D"(__dst), "+S"(__src), "+c"(__n)
                        :
                        : "memory");
+#else
+  __asm__ __volatile__("xchg %%esi, %1\nrep movsl\nxchg %%esi, %1"
+                       : "+D"(__dst), "+r"(__src), "+c"(__n)
+                       :
+                       : "memory");
+#endif
 }
 static __inline__ void __DEFAULT_FN_ATTRS __movsw(unsigned short *__dst,
                                                   unsigned short const *__src,
                                                   size_t __n) {
+#if defined(__x86_64__)
   __asm__ __volatile__("rep movsw"
                        : "+D"(__dst), "+S"(__src), "+c"(__n)
                        :
                        : "memory");
+#else
+  __asm__ __volatile__("xchg %%esi, %1\nrep movsw\nxchg %%esi, %1"
+                       : "+D"(__dst), "+r"(__src), "+c"(__n)
+                       :
+                       : "memory");
+#endif
 }
 static __inline__ void __DEFAULT_FN_ATTRS __stosd(unsigned long *__dst,
                                                   unsigned long __x,
@@ -507,16 +530,26 @@ static __inline__ void __DEFAULT_FN_ATTRS __stosq(unsigned __int64 *__dst,
 |* Misc
 \*----------------------------------------------------------------------------*/
 #if defined(__i386__) || defined(__x86_64__)
+#if defined(__i386__)
+#define __cpuid_count(__leaf, __count, __eax, __ebx, __ecx, __edx)             \
+  __asm("cpuid"                                                                \
+        : "=a"(__eax), "=b"(__ebx), "=c"(__ecx), "=d"(__edx)                   \
+        : "0"(__leaf), "2"(__count))
+#else
+/* x86-64 uses %rbx as the base register, so preserve it. */
+#define __cpuid_count(__leaf, __count, __eax, __ebx, __ecx, __edx)             \
+  __asm("xchgq %%rbx,%q1\n"                                                    \
+        "cpuid\n"                                                              \
+        "xchgq %%rbx,%q1"                                                      \
+        : "=a"(__eax), "=r"(__ebx), "=c"(__ecx), "=d"(__edx)                   \
+        : "0"(__leaf), "2"(__count))
+#endif
 static __inline__ void __DEFAULT_FN_ATTRS __cpuid(int __info[4], int __level) {
-  __asm__("cpuid"
-          : "=a"(__info[0]), "=b"(__info[1]), "=c"(__info[2]), "=d"(__info[3])
-          : "a"(__level), "c"(0));
+  __cpuid_count(__level, 0, __info[0], __info[1], __info[2], __info[3]);
 }
 static __inline__ void __DEFAULT_FN_ATTRS __cpuidex(int __info[4], int __level,
                                                     int __ecx) {
-  __asm__("cpuid"
-          : "=a"(__info[0]), "=b"(__info[1]), "=c"(__info[2]), "=d"(__info[3])
-          : "a"(__level), "c"(__ecx));
+  __cpuid_count(__level, __ecx, __info[0], __info[1], __info[2], __info[3]);
 }
 static __inline__ void __DEFAULT_FN_ATTRS __halt(void) {
   __asm__ volatile("hlt");
@@ -541,6 +574,9 @@ void _WriteStatusReg(int, __int64);
 unsigned short __cdecl _byteswap_ushort(unsigned short val);
 unsigned long __cdecl _byteswap_ulong (unsigned long val);
 unsigned __int64 __cdecl _byteswap_uint64(unsigned __int64 val);
+
+__int64 __mulh(__int64 __a, __int64 __b);
+unsigned __int64 __umulh(unsigned __int64 __a, unsigned __int64 __b);
 #endif
 
 /*----------------------------------------------------------------------------*\

lib/include/keylockerintrin.h

@@ -230,10 +230,12 @@ _mm_aesenc128kl_u8(__m128i* __odata, __m128i __idata, const void *__h) {
 ///                    HandleKeyType (Handle[511:0]) != HANDLE_KEY_TYPE_AES256 )
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   MEM[__odata+127:__odata] := 0
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate512 (Handle[511:0], IWKey)
 ///   IF (Authentic == 0)
 ///     ZF := 1
+///     MEM[__odata+127:__odata] := 0
 ///   ELSE
 ///     MEM[__odata+127:__odata] := AES256Encrypt (__idata[127:0], UnwrappedKey)
 ///     ZF := 0
@@ -267,10 +269,12 @@ _mm_aesenc256kl_u8(__m128i* __odata, __m128i __idata, const void *__h) {
 ///                  HandleKeyType (Handle[383:0]) != HANDLE_KEY_TYPE_AES128)
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   MEM[__odata+127:__odata] := 0
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate384 (Handle[383:0], IWKey)
 ///   IF (Authentic == 0)
 ///     ZF := 1
+///     MEM[__odata+127:__odata] := 0
 ///   ELSE
 ///     MEM[__odata+127:__odata] := AES128Decrypt (__idata[127:0], UnwrappedKey)
 ///     ZF := 0
@@ -304,10 +308,12 @@ _mm_aesdec128kl_u8(__m128i* __odata, __m128i __idata, const void *__h) {
 ///                   HandleKeyType (Handle[511:0]) != HANDLE_KEY_TYPE_AES256)
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   MEM[__odata+127:__odata] := 0
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate512 (Handle[511:0], IWKey)
 ///   IF (Authentic == 0)
 ///     ZF := 1
+///     MEM[__odata+127:__odata] := 0
 ///   ELSE
 ///     MEM[__odata+127:__odata] := AES256Decrypt (__idata[127:0], UnwrappedKey)
 ///     ZF := 0
@@ -354,10 +360,16 @@ _mm_aesdec256kl_u8(__m128i* __odata, __m128i __idata, const void *__h) {
 ///                    HandleKeyType (Handle[383:0]) != HANDLE_KEY_TYPE_AES128 )
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   FOR i := 0 to 7
+///     __odata[i] := 0
+///   ENDFOR
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate384 (Handle[383:0], IWKey)
 ///   IF Authentic == 0
 ///     ZF := 1
+///     FOR i := 0 to 7
+///       __odata[i] := 0
+///     ENDFOR
 ///   ELSE
 ///     FOR i := 0 to 7
 ///       __odata[i] := AES128Encrypt (__idata[i], UnwrappedKey)
@@ -394,10 +406,16 @@ _mm_aesencwide128kl_u8(__m128i __odata[8], const __m128i __idata[8], const void*
 ///                    HandleKeyType (Handle[511:0]) != HANDLE_KEY_TYPE_AES512 )
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   FOR i := 0 to 7
+///     __odata[i] := 0
+///   ENDFOR
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate512 (Handle[511:0], IWKey)
 ///   IF Authentic == 0
 ///     ZF := 1
+///     FOR i := 0 to 7
+///       __odata[i] := 0
+///     ENDFOR
 ///   ELSE
 ///     FOR i := 0 to 7
 ///       __odata[i] := AES256Encrypt (__idata[i], UnwrappedKey)
@@ -434,10 +452,16 @@ _mm_aesencwide256kl_u8(__m128i __odata[8], const __m128i __idata[8], const void*
 ///                    HandleKeyType (Handle) != HANDLE_KEY_TYPE_AES128 )
 /// IF (IllegalHandle)
 ///   ZF := 1
+///   FOR i := 0 to 7
+///     __odata[i] := 0
+///   ENDFOR
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate384 (Handle[383:0], IWKey)
 ///   IF Authentic == 0
 ///     ZF := 1
+///     FOR i := 0 to 7
+///       __odata[i] := 0
+///     ENDFOR
 ///   ELSE
 ///     FOR i := 0 to 7
 ///       __odata[i] := AES128Decrypt (__idata[i], UnwrappedKey)
@@ -474,10 +498,16 @@ _mm_aesdecwide128kl_u8(__m128i __odata[8], const __m128i __idata[8], const void*
 ///                   HandleKeyType (Handle) != HANDLE_KEY_TYPE_AES512 )
 /// If (IllegalHandle)
 ///   ZF := 1
+///   FOR i := 0 to 7
+///     __odata[i] := 0
+///   ENDFOR
 /// ELSE
 ///   (UnwrappedKey, Authentic) := UnwrapKeyAndAuthenticate512 (Handle[511:0], IWKey)
 ///   IF Authentic == 0
 ///     ZF := 1
+///     FOR i := 0 to 7
+///       __odata[i] := 0
+///     ENDFOR
 ///   ELSE
 ///     FOR i := 0 to 7
 ///       __odata[i] := AES256Decrypt (__idata[i], UnwrappedKey)

lib/include/opencl-c-base.h

@@ -21,9 +21,37 @@
 #define cl_khr_subgroup_shuffle 1
 #define cl_khr_subgroup_shuffle_relative 1
 #define cl_khr_subgroup_clustered_reduce 1
+#define cl_khr_extended_bit_ops 1
+#define cl_khr_integer_dot_product 1
+#define __opencl_c_integer_dot_product_input_4x8bit 1
+#define __opencl_c_integer_dot_product_input_4x8bit_packed 1
+
 #endif // defined(__SPIR__)
 #endif // (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200)
 
+// Define feature macros for OpenCL C 2.0
+#if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ == 200)
+#define __opencl_c_pipes 1
+#define __opencl_c_generic_address_space 1
+#define __opencl_c_work_group_collective_functions 1
+#define __opencl_c_atomic_order_acq_rel 1
+#define __opencl_c_atomic_order_seq_cst 1
+#define __opencl_c_atomic_scope_device 1
+#define __opencl_c_atomic_scope_all_devices 1
+#define __opencl_c_device_enqueue 1
+#define __opencl_c_read_write_images 1
+#define __opencl_c_program_scope_global_variables 1
+#define __opencl_c_images 1
+#endif
+
+// Define header-only feature macros for OpenCL C 3.0.
+#if (__OPENCL_C_VERSION__ == 300)
+// For the SPIR target all features are supported.
+#if defined(__SPIR__)
+#define __opencl_c_atomic_scope_all_devices 1
+#endif // defined(__SPIR__)
+#endif // (__OPENCL_C_VERSION__ == 300)
+
 // built-in scalar data types:
 
 /**
@@ -141,7 +169,9 @@ typedef double double8 __attribute__((ext_vector_type(8)));
 typedef double double16 __attribute__((ext_vector_type(16)));
 #endif
 
-#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
+#if defined(__OPENCL_CPP_VERSION__)
+#define NULL nullptr
+#elif defined(__OPENCL_C_VERSION__)
 #define NULL ((void*)0)
 #endif
 
@@ -297,7 +327,12 @@ typedef enum memory_scope {
   memory_scope_work_item = __OPENCL_MEMORY_SCOPE_WORK_ITEM,
   memory_scope_work_group = __OPENCL_MEMORY_SCOPE_WORK_GROUP,
   memory_scope_device = __OPENCL_MEMORY_SCOPE_DEVICE,
+#if defined(__opencl_c_atomic_scope_all_devices)
   memory_scope_all_svm_devices = __OPENCL_MEMORY_SCOPE_ALL_SVM_DEVICES,
+#if (__OPENCL_C_VERSION__ >= CL_VERSION_3_0)
+  memory_scope_all_devices = memory_scope_all_svm_devices,
+#endif // __OPENCL_C_VERSION__ >= CL_VERSION_3_0
+#endif // defined(__opencl_c_atomic_scope_all_devices)
 #if defined(cl_intel_subgroups) || defined(cl_khr_subgroups)
   memory_scope_sub_group = __OPENCL_MEMORY_SCOPE_SUB_GROUP
 #endif
@@ -322,7 +357,9 @@ typedef enum memory_order
   memory_order_acquire = __ATOMIC_ACQUIRE,
   memory_order_release = __ATOMIC_RELEASE,
   memory_order_acq_rel = __ATOMIC_ACQ_REL,
+#if defined(__opencl_c_atomic_order_seq_cst)
   memory_order_seq_cst = __ATOMIC_SEQ_CST
+#endif
 } memory_order;
 
 #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
@@ -445,8 +482,113 @@ typedef struct {
 
 #endif // defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_2_0)
 
+/**
+ * OpenCL v1.1/1.2/2.0 s6.2.4.2 - as_type operators
+ * Reinterprets a data type as another data type of the same size
+ */
+#define as_char(x) __builtin_astype((x), char)
+#define as_char2(x) __builtin_astype((x), char2)
+#define as_char3(x) __builtin_astype((x), char3)
+#define as_char4(x) __builtin_astype((x), char4)
+#define as_char8(x) __builtin_astype((x), char8)
+#define as_char16(x) __builtin_astype((x), char16)
+
+#define as_uchar(x) __builtin_astype((x), uchar)
+#define as_uchar2(x) __builtin_astype((x), uchar2)
+#define as_uchar3(x) __builtin_astype((x), uchar3)
+#define as_uchar4(x) __builtin_astype((x), uchar4)
+#define as_uchar8(x) __builtin_astype((x), uchar8)
+#define as_uchar16(x) __builtin_astype((x), uchar16)
+
+#define as_short(x) __builtin_astype((x), short)
+#define as_short2(x) __builtin_astype((x), short2)
+#define as_short3(x) __builtin_astype((x), short3)
+#define as_short4(x) __builtin_astype((x), short4)
+#define as_short8(x) __builtin_astype((x), short8)
+#define as_short16(x) __builtin_astype((x), short16)
+
+#define as_ushort(x) __builtin_astype((x), ushort)
+#define as_ushort2(x) __builtin_astype((x), ushort2)
+#define as_ushort3(x) __builtin_astype((x), ushort3)
+#define as_ushort4(x) __builtin_astype((x), ushort4)
+#define as_ushort8(x) __builtin_astype((x), ushort8)
+#define as_ushort16(x) __builtin_astype((x), ushort16)
+
+#define as_int(x) __builtin_astype((x), int)
+#define as_int2(x) __builtin_astype((x), int2)
+#define as_int3(x) __builtin_astype((x), int3)
+#define as_int4(x) __builtin_astype((x), int4)
+#define as_int8(x) __builtin_astype((x), int8)
+#define as_int16(x) __builtin_astype((x), int16)
+
+#define as_uint(x) __builtin_astype((x), uint)
+#define as_uint2(x) __builtin_astype((x), uint2)
+#define as_uint3(x) __builtin_astype((x), uint3)
+#define as_uint4(x) __builtin_astype((x), uint4)
+#define as_uint8(x) __builtin_astype((x), uint8)
+#define as_uint16(x) __builtin_astype((x), uint16)
+
+#define as_long(x) __builtin_astype((x), long)
+#define as_long2(x) __builtin_astype((x), long2)
+#define as_long3(x) __builtin_astype((x), long3)
+#define as_long4(x) __builtin_astype((x), long4)
+#define as_long8(x) __builtin_astype((x), long8)
+#define as_long16(x) __builtin_astype((x), long16)
+
+#define as_ulong(x) __builtin_astype((x), ulong)
+#define as_ulong2(x) __builtin_astype((x), ulong2)
+#define as_ulong3(x) __builtin_astype((x), ulong3)
+#define as_ulong4(x) __builtin_astype((x), ulong4)
+#define as_ulong8(x) __builtin_astype((x), ulong8)
+#define as_ulong16(x) __builtin_astype((x), ulong16)
+
+#define as_float(x) __builtin_astype((x), float)
+#define as_float2(x) __builtin_astype((x), float2)
+#define as_float3(x) __builtin_astype((x), float3)
+#define as_float4(x) __builtin_astype((x), float4)
+#define as_float8(x) __builtin_astype((x), float8)
+#define as_float16(x) __builtin_astype((x), float16)
+
+#ifdef cl_khr_fp64
+#define as_double(x) __builtin_astype((x), double)
+#define as_double2(x) __builtin_astype((x), double2)
+#define as_double3(x) __builtin_astype((x), double3)
+#define as_double4(x) __builtin_astype((x), double4)
+#define as_double8(x) __builtin_astype((x), double8)
+#define as_double16(x) __builtin_astype((x), double16)
+#endif // cl_khr_fp64
+
+#ifdef cl_khr_fp16
+#define as_half(x) __builtin_astype((x), half)
+#define as_half2(x) __builtin_astype((x), half2)
+#define as_half3(x) __builtin_astype((x), half3)
+#define as_half4(x) __builtin_astype((x), half4)
+#define as_half8(x) __builtin_astype((x), half8)
+#define as_half16(x) __builtin_astype((x), half16)
+#endif // cl_khr_fp16
+
+#define as_size_t(x) __builtin_astype((x), size_t)
+#define as_ptrdiff_t(x) __builtin_astype((x), ptrdiff_t)
+#define as_intptr_t(x) __builtin_astype((x), intptr_t)
+#define as_uintptr_t(x) __builtin_astype((x), uintptr_t)
+
+// OpenCL v1.1 s6.9, v1.2/2.0 s6.10 - Function qualifiers
+
+#define __kernel_exec(X, typen) __kernel \
+	__attribute__((work_group_size_hint(X, 1, 1))) \
+	__attribute__((vec_type_hint(typen)))
+
+#define kernel_exec(X, typen) __kernel \
+	__attribute__((work_group_size_hint(X, 1, 1))) \
+	__attribute__((vec_type_hint(typen)))
+
+#if defined(__OPENCL_CPP_VERSION__) || (__OPENCL_C_VERSION__ >= CL_VERSION_1_2)
+// OpenCL v1.2 s6.12.13, v2.0 s6.13.13 - printf
+
+int printf(__constant const char* st, ...) __attribute__((format(printf, 1, 2)));
+#endif
+
 #ifdef cl_intel_device_side_avc_motion_estimation
-#pragma OPENCL EXTENSION cl_intel_device_side_avc_motion_estimation : begin
 
 #define CLK_AVC_ME_MAJOR_16x16_INTEL 0x0
 #define CLK_AVC_ME_MAJOR_16x8_INTEL 0x1
@@ -580,7 +722,6 @@ typedef struct {
 #define CLK_AVC_IME_RESULT_DUAL_REFERENCE_STREAMOUT_INITIALIZE_INTEL 0x0
 #define CLK_AVC_IME_RESULT_DUAL_REFERENCE_STREAMIN_INITIALIZE_INTEL 0x0
 
-#pragma OPENCL EXTENSION cl_intel_device_side_avc_motion_estimation : end
 #endif // cl_intel_device_side_avc_motion_estimation
 
 // Disable any extensions we may have enabled previously.

lib/include/openmp_wrappers/__clang_openmp_device_functions.h

@@ -14,13 +14,13 @@
 #error "This file is for OpenMP compilation only."
 #endif
 
-#pragma omp begin declare variant match(                                       \
-    device = {arch(nvptx, nvptx64)}, implementation = {extension(match_any)})
-
 #ifdef __cplusplus
 extern "C" {
 #endif
 
+#pragma omp begin declare variant match(                                       \
+    device = {arch(nvptx, nvptx64)}, implementation = {extension(match_any)})
+
 #define __CUDA__
 #define __OPENMP_NVPTX__
 
@@ -33,10 +33,74 @@ extern "C" {
 #undef __OPENMP_NVPTX__
 #undef __CUDA__
 
+#pragma omp end declare variant
+
+#ifdef __AMDGCN__
+#pragma omp begin declare variant match(device = {arch(amdgcn)})
+
+// Import types which will be used by __clang_hip_libdevice_declares.h
+#ifndef __cplusplus
+#include <stdbool.h>
+#include <stdint.h>
+#endif
+
+#define __OPENMP_AMDGCN__
+#pragma push_macro("__device__")
+#define __device__
+
+/// Include declarations for libdevice functions.
+#include <__clang_hip_libdevice_declares.h>
+
+#pragma pop_macro("__device__")
+#undef __OPENMP_AMDGCN__
+
+#pragma omp end declare variant
+#endif
+
 #ifdef __cplusplus
 } // extern "C"
 #endif
 
-#pragma omp end declare variant
+// Ensure we make `_ZdlPv`, aka. `operator delete(void*)` available without the
+// need to `include <new>` in C++ mode.
+#ifdef __cplusplus
+
+// We require malloc/free.
+#include <cstdlib>
+
+#pragma push_macro("OPENMP_NOEXCEPT")
+#if __cplusplus >= 201103L
+#define OPENMP_NOEXCEPT noexcept
+#else
+#define OPENMP_NOEXCEPT
+#endif
+
+// Device overrides for non-placement new and delete.
+inline void *operator new(__SIZE_TYPE__ size) {
+  if (size == 0)
+    size = 1;
+  return ::malloc(size);
+}
+
+inline void *operator new[](__SIZE_TYPE__ size) { return ::operator new(size); }
+
+inline void operator delete(void *ptr)OPENMP_NOEXCEPT { ::free(ptr); }
+
+inline void operator delete[](void *ptr) OPENMP_NOEXCEPT {
+  ::operator delete(ptr);
+}
+
+// Sized delete, C++14 only.
+#if __cplusplus >= 201402L
+inline void operator delete(void *ptr, __SIZE_TYPE__ size)OPENMP_NOEXCEPT {
+  ::operator delete(ptr);
+}
+inline void operator delete[](void *ptr, __SIZE_TYPE__ size) OPENMP_NOEXCEPT {
+  ::operator delete(ptr);
+}
+#endif
+
+#pragma pop_macro("OPENMP_NOEXCEPT")
+#endif
 
 #endif

lib/include/openmp_wrappers/cmath

@@ -75,4 +75,58 @@ __DEVICE__ float tgamma(float __x) { return ::tgammaf(__x); }
 
 #pragma omp end declare variant
 
+#ifdef __AMDGCN__
+#pragma omp begin declare variant match(device = {arch(amdgcn)})
+
+#pragma push_macro("__constant__")
+#define __constant__ __attribute__((constant))
+#define __OPENMP_AMDGCN__
+
+#include <__clang_hip_cmath.h>
+
+#pragma pop_macro("__constant__")
+#undef __OPENMP_AMDGCN__
+
+// Define overloads otherwise which are absent
+#define __DEVICE__ static constexpr __attribute__((always_inline, nothrow))
+
+__DEVICE__ float acos(float __x) { return ::acosf(__x); }
+__DEVICE__ float acosh(float __x) { return ::acoshf(__x); }
+__DEVICE__ float asin(float __x) { return ::asinf(__x); }
+__DEVICE__ float asinh(float __x) { return ::asinhf(__x); }
+__DEVICE__ float atan(float __x) { return ::atanf(__x); }
+__DEVICE__ float atan2(float __x, float __y) { return ::atan2f(__x, __y); }
+__DEVICE__ float atanh(float __x) { return ::atanhf(__x); }
+__DEVICE__ float cbrt(float __x) { return ::cbrtf(__x); }
+__DEVICE__ float cosh(float __x) { return ::coshf(__x); }
+__DEVICE__ float erf(float __x) { return ::erff(__x); }
+__DEVICE__ float erfc(float __x) { return ::erfcf(__x); }
+__DEVICE__ float exp2(float __x) { return ::exp2f(__x); }
+__DEVICE__ float expm1(float __x) { return ::expm1f(__x); }
+__DEVICE__ float fdim(float __x, float __y) { return ::fdimf(__x, __y); }
+__DEVICE__ float hypot(float __x, float __y) { return ::hypotf(__x, __y); }
+__DEVICE__ int ilogb(float __x) { return ::ilogbf(__x); }
+__DEVICE__ float ldexp(float __arg, int __exp) {
+  return ::ldexpf(__arg, __exp);
+}
+__DEVICE__ float lgamma(float __x) { return ::lgammaf(__x); }
+__DEVICE__ float log1p(float __x) { return ::log1pf(__x); }
+__DEVICE__ float logb(float __x) { return ::logbf(__x); }
+__DEVICE__ float nextafter(float __x, float __y) {
+  return ::nextafterf(__x, __y);
+}
+__DEVICE__ float remainder(float __x, float __y) {
+  return ::remainderf(__x, __y);
+}
+__DEVICE__ float scalbn(float __x, int __y) { return ::scalbnf(__x, __y); }
+__DEVICE__ float sinh(float __x) { return ::sinhf(__x); }
+__DEVICE__ float tan(float __x) { return ::tanf(__x); }
+__DEVICE__ float tanh(float __x) { return ::tanhf(__x); }
+__DEVICE__ float tgamma(float __x) { return ::tgammaf(__x); }
+
+#undef __DEVICE__
+
+#pragma omp end declare variant
+#endif // __AMDGCN__
+
 #endif

lib/include/openmp_wrappers/complex

@@ -17,7 +17,6 @@
 // We require std::math functions in the complex builtins below.
 #include <cmath>
 
-#define __CUDA__
 #define __OPENMP_NVPTX__
 #include <__clang_cuda_complex_builtins.h>
 #undef __OPENMP_NVPTX__
@@ -26,9 +25,6 @@
 // Grab the host header too.
 #include_next <complex>
 
-
-#ifdef __cplusplus
-
 // If we are compiling against libc++, the macro _LIBCPP_STD_VER should be set
 // after including <cmath> above. Since the complex header we use is a
 // simplified version of the libc++, we don't need it in this case. If we
@@ -48,5 +44,3 @@
 #pragma omp end declare variant
 
 #endif
-
-#endif

lib/include/openmp_wrappers/complex.h

@@ -17,7 +17,6 @@
 // We require math functions in the complex builtins below.
 #include <math.h>
 
-#define __CUDA__
 #define __OPENMP_NVPTX__
 #include <__clang_cuda_complex_builtins.h>
 #undef __OPENMP_NVPTX__

lib/include/openmp_wrappers/math.h

@@ -48,4 +48,14 @@
 
 #pragma omp end declare variant
 
+#ifdef __AMDGCN__
+#pragma omp begin declare variant match(device = {arch(amdgcn)})
+
+#define __OPENMP_AMDGCN__
+#include <__clang_hip_math.h>
+#undef __OPENMP_AMDGCN__
+
+#pragma omp end declare variant
+#endif
+
 #endif

lib/include/openmp_wrappers/new

@@ -9,6 +9,8 @@
 #ifndef __CLANG_OPENMP_WRAPPERS_NEW
 #define __CLANG_OPENMP_WRAPPERS_NEW
 
+// We need the system <new> for the std::nothrow_t. The new/delete operators
+// which do not use nothrow_t are provided without the <new> header.
 #include_next <new>
 
 #if defined(__NVPTX__) && defined(_OPENMP)
@@ -22,48 +24,24 @@
 #define OPENMP_NOEXCEPT
 #endif
 
-// Device overrides for non-placement new and delete.
-inline void *operator new(__SIZE_TYPE__ size) {
-  if (size == 0)
-    size = 1;
-  return ::malloc(size);
-}
 inline void *operator new(__SIZE_TYPE__ size,
                           const std::nothrow_t &) OPENMP_NOEXCEPT {
   return ::operator new(size);
 }
 
-inline void *operator new[](__SIZE_TYPE__ size) { return ::operator new(size); }
 inline void *operator new[](__SIZE_TYPE__ size, const std::nothrow_t &) {
   return ::operator new(size);
 }
 
-inline void operator delete(void *ptr)OPENMP_NOEXCEPT {
-  if (ptr)
-    ::free(ptr);
-}
 inline void operator delete(void *ptr, const std::nothrow_t &)OPENMP_NOEXCEPT {
   ::operator delete(ptr);
 }
 
-inline void operator delete[](void *ptr) OPENMP_NOEXCEPT {
-  ::operator delete(ptr);
-}
 inline void operator delete[](void *ptr,
                               const std::nothrow_t &) OPENMP_NOEXCEPT {
   ::operator delete(ptr);
 }
 
-// Sized delete, C++14 only.
-#if __cplusplus >= 201402L
-inline void operator delete(void *ptr, __SIZE_TYPE__ size)OPENMP_NOEXCEPT {
-  ::operator delete(ptr);
-}
-inline void operator delete[](void *ptr, __SIZE_TYPE__ size) OPENMP_NOEXCEPT {
-  ::operator delete(ptr);
-}
-#endif
-
 #pragma pop_macro("OPENMP_NOEXCEPT")
 #endif
 

lib/include/ppc_wrappers/xmmintrin.h

@@ -28,7 +28,7 @@
    Most SSE scalar float intrinsic operations can be performed more
    efficiently as C language float scalar operations or optimized to
    use vector SIMD operations. We recommend this for new applications. */
-#error "Please read comment above.  Use -DNO_WARN_X86_INTRINSICS to disable this error."
+#error "Please read comment above. Use -DNO_WARN_X86_INTRINSICS to disable this error."
 #endif
 
 #ifndef _XMMINTRIN_H_INCLUDED
@@ -62,14 +62,13 @@
 
 /* The Intel API is flexible enough that we must allow aliasing with other
    vector types, and their scalar components.  */
-typedef float __m128 __attribute__ ((__vector_size__ (16), __may_alias__));
+typedef vector float __m128 __attribute__((__may_alias__));
 
 /* Unaligned version of the same type.  */
-typedef float __m128_u __attribute__ ((__vector_size__ (16), __may_alias__,
-				       __aligned__ (1)));
+typedef vector float __m128_u __attribute__((__may_alias__, __aligned__(1)));
 
 /* Internal data types for implementing the intrinsics.  */
-typedef float __v4sf __attribute__ ((__vector_size__ (16)));
+typedef vector float __v4sf;
 
 /* Create an undefined vector.  */
 extern __inline __m128 __attribute__((__gnu_inline__, __always_inline__, __artificial__))

lib/include/uintrintrin.h

@@ -20,6 +20,13 @@
 
 #ifdef __x86_64__
 
+struct __uintr_frame
+{
+  unsigned long long rip;
+  unsigned long long rflags;
+  unsigned long long rsp;
+};
+
 /// Clears the user interrupt flag (UIF). Its effect takes place immediately: a
 ///    user interrupt cannot be delivered on the instruction boundary following
 ///    CLUI. Can be executed only if CR4.UINT = 1, the logical processor is in

lib/include/vaesintrin.h

@@ -28,13 +28,6 @@ static __inline__ __m256i __DEFAULT_FN_ATTRS
               (__v4di) __B);
 }
 
-static __inline__ __m512i __DEFAULT_FN_ATTRS_F
- _mm512_aesenc_epi128(__m512i __A, __m512i __B)
-{
-  return (__m512i) __builtin_ia32_aesenc512((__v8di) __A,
-              (__v8di) __B);
-}
-
 static __inline__ __m256i __DEFAULT_FN_ATTRS
  _mm256_aesdec_epi128(__m256i __A, __m256i __B)
 {
@@ -42,13 +35,6 @@ static __inline__ __m256i __DEFAULT_FN_ATTRS
               (__v4di) __B);
 }
 
-static __inline__ __m512i __DEFAULT_FN_ATTRS_F
- _mm512_aesdec_epi128(__m512i __A, __m512i __B)
-{
-  return (__m512i) __builtin_ia32_aesdec512((__v8di) __A,
-              (__v8di) __B);
-}
-
 static __inline__ __m256i __DEFAULT_FN_ATTRS
  _mm256_aesenclast_epi128(__m256i __A, __m256i __B)
 {
@@ -56,13 +42,6 @@ static __inline__ __m256i __DEFAULT_FN_ATTRS
               (__v4di) __B);
 }
 
-static __inline__ __m512i __DEFAULT_FN_ATTRS_F
- _mm512_aesenclast_epi128(__m512i __A, __m512i __B)
-{
-  return (__m512i) __builtin_ia32_aesenclast512((__v8di) __A,
-              (__v8di) __B);
-}
-
 static __inline__ __m256i __DEFAULT_FN_ATTRS
  _mm256_aesdeclast_epi128(__m256i __A, __m256i __B)
 {
@@ -70,13 +49,35 @@ static __inline__ __m256i __DEFAULT_FN_ATTRS
               (__v4di) __B);
 }
 
+#ifdef __AVX512FINTRIN_H
+static __inline__ __m512i __DEFAULT_FN_ATTRS_F
+ _mm512_aesenc_epi128(__m512i __A, __m512i __B)
+{
+  return (__m512i) __builtin_ia32_aesenc512((__v8di) __A,
+              (__v8di) __B);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS_F
+ _mm512_aesdec_epi128(__m512i __A, __m512i __B)
+{
+  return (__m512i) __builtin_ia32_aesdec512((__v8di) __A,
+              (__v8di) __B);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS_F
+ _mm512_aesenclast_epi128(__m512i __A, __m512i __B)
+{
+  return (__m512i) __builtin_ia32_aesenclast512((__v8di) __A,
+              (__v8di) __B);
+}
+
 static __inline__ __m512i __DEFAULT_FN_ATTRS_F
  _mm512_aesdeclast_epi128(__m512i __A, __m512i __B)
 {
   return (__m512i) __builtin_ia32_aesdeclast512((__v8di) __A,
               (__v8di) __B);
 }
-
+#endif // __AVX512FINTRIN_H
 
 #undef __DEFAULT_FN_ATTRS
 #undef __DEFAULT_FN_ATTRS_F

lib/include/vecintrin.h

@@ -1016,64 +1016,84 @@ vec_scatter_element(__vector double __vec,
 
 static inline __ATTRS_o_ai __vector signed char
 vec_xl(long __offset, const signed char *__ptr) {
-  return *(const __vector signed char *)
-          ((const char *)__ptr + __offset);
+  __vector signed char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed char));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector unsigned char
 vec_xl(long __offset, const unsigned char *__ptr) {
-  return *(const __vector unsigned char *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned char));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector signed short
 vec_xl(long __offset, const signed short *__ptr) {
-  return *(const __vector signed short *)
-          ((const char *)__ptr + __offset);
+  __vector signed short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed short));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector unsigned short
 vec_xl(long __offset, const unsigned short *__ptr) {
-  return *(const __vector unsigned short *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned short));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector signed int
 vec_xl(long __offset, const signed int *__ptr) {
-  return *(const __vector signed int *)
-          ((const char *)__ptr + __offset);
+  __vector signed int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed int));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector unsigned int
 vec_xl(long __offset, const unsigned int *__ptr) {
-  return *(const __vector unsigned int *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned int));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector signed long long
 vec_xl(long __offset, const signed long long *__ptr) {
-  return *(const __vector signed long long *)
-          ((const char *)__ptr + __offset);
+  __vector signed long long V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed long long));
+  return V;
 }
 
 static inline __ATTRS_o_ai __vector unsigned long long
 vec_xl(long __offset, const unsigned long long *__ptr) {
-  return *(const __vector unsigned long long *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned long long V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned long long));
+  return V;
 }
 
 #if __ARCH__ >= 12
 static inline __ATTRS_o_ai __vector float
 vec_xl(long __offset, const float *__ptr) {
-  return *(const __vector float *)
-          ((const char *)__ptr + __offset);
+  __vector float V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector float));
+  return V;
 }
 #endif
 
 static inline __ATTRS_o_ai __vector double
 vec_xl(long __offset, const double *__ptr) {
-  return *(const __vector double *)
-          ((const char *)__ptr + __offset);
+  __vector double V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector double));
+  return V;
 }
 
 /*-- vec_xld2 ---------------------------------------------------------------*/
@@ -1081,64 +1101,82 @@ vec_xl(long __offset, const double *__ptr) {
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed char
 vec_xld2(long __offset, const signed char *__ptr) {
-  return *(const __vector signed char *)
-          ((const char *)__ptr + __offset);
+  __vector signed char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed char));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned char
 vec_xld2(long __offset, const unsigned char *__ptr) {
-  return *(const __vector unsigned char *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned char));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed short
 vec_xld2(long __offset, const signed short *__ptr) {
-  return *(const __vector signed short *)
-          ((const char *)__ptr + __offset);
+  __vector signed short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed short));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned short
 vec_xld2(long __offset, const unsigned short *__ptr) {
-  return *(const __vector unsigned short *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned short));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed int
 vec_xld2(long __offset, const signed int *__ptr) {
-  return *(const __vector signed int *)
-          ((const char *)__ptr + __offset);
+  __vector signed int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed int));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned int
 vec_xld2(long __offset, const unsigned int *__ptr) {
-  return *(const __vector unsigned int *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned int));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed long long
 vec_xld2(long __offset, const signed long long *__ptr) {
-  return *(const __vector signed long long *)
-          ((const char *)__ptr + __offset);
+  __vector signed long long V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed long long));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned long long
 vec_xld2(long __offset, const unsigned long long *__ptr) {
-  return *(const __vector unsigned long long *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned long long V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned long long));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector double
 vec_xld2(long __offset, const double *__ptr) {
-  return *(const __vector double *)
-          ((const char *)__ptr + __offset);
+  __vector double V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector double));
+  return V;
 }
 
 /*-- vec_xlw4 ---------------------------------------------------------------*/
@@ -1146,99 +1184,128 @@ vec_xld2(long __offset, const double *__ptr) {
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed char
 vec_xlw4(long __offset, const signed char *__ptr) {
-  return *(const __vector signed char *)
-          ((const char *)__ptr + __offset);
+  __vector signed char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed char));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned char
 vec_xlw4(long __offset, const unsigned char *__ptr) {
-  return *(const __vector unsigned char *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned char V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned char));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed short
 vec_xlw4(long __offset, const signed short *__ptr) {
-  return *(const __vector signed short *)
-          ((const char *)__ptr + __offset);
+  __vector signed short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed short));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned short
 vec_xlw4(long __offset, const unsigned short *__ptr) {
-  return *(const __vector unsigned short *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned short V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned short));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector signed int
 vec_xlw4(long __offset, const signed int *__ptr) {
-  return *(const __vector signed int *)
-          ((const char *)__ptr + __offset);
+  __vector signed int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector signed int));
+  return V;
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai __vector unsigned int
 vec_xlw4(long __offset, const unsigned int *__ptr) {
-  return *(const __vector unsigned int *)
-          ((const char *)__ptr + __offset);
+  __vector unsigned int V;
+  __builtin_memcpy(&V, ((const char *)__ptr + __offset),
+                   sizeof(__vector unsigned int));
+  return V;
 }
 
 /*-- vec_xst ----------------------------------------------------------------*/
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector signed char __vec, long __offset, signed char *__ptr) {
-  *(__vector signed char *)((char *)__ptr + __offset) = __vec;
+  __vector signed char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed char));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector unsigned char __vec, long __offset, unsigned char *__ptr) {
-  *(__vector unsigned char *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned char));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector signed short __vec, long __offset, signed short *__ptr) {
-  *(__vector signed short *)((char *)__ptr + __offset) = __vec;
+  __vector signed short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed short));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector unsigned short __vec, long __offset, unsigned short *__ptr) {
-  *(__vector unsigned short *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned short));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector signed int __vec, long __offset, signed int *__ptr) {
-  *(__vector signed int *)((char *)__ptr + __offset) = __vec;
+  __vector signed int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector signed int));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector unsigned int __vec, long __offset, unsigned int *__ptr) {
-  *(__vector unsigned int *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned int));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector signed long long __vec, long __offset,
-          signed long long *__ptr) {
-  *(__vector signed long long *)((char *)__ptr + __offset) = __vec;
+        signed long long *__ptr) {
+  __vector signed long long V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed long long));
 }
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector unsigned long long __vec, long __offset,
-          unsigned long long *__ptr) {
-  *(__vector unsigned long long *)((char *)__ptr + __offset) = __vec;
+        unsigned long long *__ptr) {
+  __vector unsigned long long V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned long long));
 }
 
 #if __ARCH__ >= 12
 static inline __ATTRS_o_ai void
 vec_xst(__vector float __vec, long __offset, float *__ptr) {
-  *(__vector float *)((char *)__ptr + __offset) = __vec;
+  __vector float V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector float));
 }
 #endif
 
 static inline __ATTRS_o_ai void
 vec_xst(__vector double __vec, long __offset, double *__ptr) {
-  *(__vector double *)((char *)__ptr + __offset) = __vec;
+  __vector double V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector double));
 }
 
 /*-- vec_xstd2 --------------------------------------------------------------*/
@@ -1246,57 +1313,73 @@ vec_xst(__vector double __vec, long __offset, double *__ptr) {
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector signed char __vec, long __offset, signed char *__ptr) {
-  *(__vector signed char *)((char *)__ptr + __offset) = __vec;
+  __vector signed char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed char));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector unsigned char __vec, long __offset, unsigned char *__ptr) {
-  *(__vector unsigned char *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned char));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector signed short __vec, long __offset, signed short *__ptr) {
-  *(__vector signed short *)((char *)__ptr + __offset) = __vec;
+  __vector signed short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed short));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector unsigned short __vec, long __offset, unsigned short *__ptr) {
-  *(__vector unsigned short *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned short));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector signed int __vec, long __offset, signed int *__ptr) {
-  *(__vector signed int *)((char *)__ptr + __offset) = __vec;
+  __vector signed int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector signed int));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector unsigned int __vec, long __offset, unsigned int *__ptr) {
-  *(__vector unsigned int *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned int));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector signed long long __vec, long __offset,
           signed long long *__ptr) {
-  *(__vector signed long long *)((char *)__ptr + __offset) = __vec;
+  __vector signed long long V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed long long));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector unsigned long long __vec, long __offset,
           unsigned long long *__ptr) {
-  *(__vector unsigned long long *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned long long V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned long long));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstd2(__vector double __vec, long __offset, double *__ptr) {
-  *(__vector double *)((char *)__ptr + __offset) = __vec;
+  __vector double V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector double));
 }
 
 /*-- vec_xstw4 --------------------------------------------------------------*/
@@ -1304,37 +1387,48 @@ vec_xstd2(__vector double __vec, long __offset, double *__ptr) {
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector signed char __vec, long __offset, signed char *__ptr) {
-  *(__vector signed char *)((char *)__ptr + __offset) = __vec;
+  __vector signed char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed char));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector unsigned char __vec, long __offset, unsigned char *__ptr) {
-  *(__vector unsigned char *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned char V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned char));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector signed short __vec, long __offset, signed short *__ptr) {
-  *(__vector signed short *)((char *)__ptr + __offset) = __vec;
+  __vector signed short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector signed short));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector unsigned short __vec, long __offset, unsigned short *__ptr) {
-  *(__vector unsigned short *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned short V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned short));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector signed int __vec, long __offset, signed int *__ptr) {
-  *(__vector signed int *)((char *)__ptr + __offset) = __vec;
+  __vector signed int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V, sizeof(__vector signed int));
 }
 
 // This prototype is deprecated.
 static inline __ATTRS_o_ai void
 vec_xstw4(__vector unsigned int __vec, long __offset, unsigned int *__ptr) {
-  *(__vector unsigned int *)((char *)__ptr + __offset) = __vec;
+  __vector unsigned int V = __vec;
+  __builtin_memcpy(((char *)__ptr + __offset), &V,
+                   sizeof(__vector unsigned int));
 }
 
 /*-- vec_load_bndry ---------------------------------------------------------*/
@@ -9259,6 +9353,41 @@ vec_fp_test_data_class(__vector double __a, int __b, int *__c)
                                    __VEC_CLASS_FP_ZERO | \
                                    __VEC_CLASS_FP_INFINITY)
 
+/*-- vec_extend_to_fp32_hi --------------------------------------------------*/
+
+#if __ARCH__ >= 14
+#define vec_extend_to_fp32_hi(X, W) \
+  ((__vector float)__builtin_s390_vclfnhs((X), (W)));
+#endif
+
+/*-- vec_extend_to_fp32_hi --------------------------------------------------*/
+
+#if __ARCH__ >= 14
+#define vec_extend_to_fp32_lo(X, W) \
+  ((__vector float)__builtin_s390_vclfnls((X), (W)));
+#endif
+
+/*-- vec_round_from_fp32 ----------------------------------------------------*/
+
+#if __ARCH__ >= 14
+#define vec_round_from_fp32(X, Y, W) \
+  ((__vector unsigned short)__builtin_s390_vcrnfs((X), (Y), (W)));
+#endif
+
+/*-- vec_convert_to_fp16 ----------------------------------------------------*/
+
+#if __ARCH__ >= 14
+#define vec_convert_to_fp16(X, W) \
+  ((__vector unsigned short)__builtin_s390_vcfn((X), (W)));
+#endif
+
+/*-- vec_convert_from_fp16 --------------------------------------------------*/
+
+#if __ARCH__ >= 14
+#define vec_convert_from_fp16(X, W) \
+  ((__vector unsigned short)__builtin_s390_vcnf((X), (W)));
+#endif
+
 /*-- vec_cp_until_zero ------------------------------------------------------*/
 
 static inline __ATTRS_o_ai __vector signed char

lib/include/vpclmulqdqintrin.h

@@ -19,10 +19,12 @@
                                        (__v4di)(__m256i)(B),  \
                                        (char)(I))
 
+#ifdef __AVX512FINTRIN_H
 #define _mm512_clmulepi64_epi128(A, B, I) \
   (__m512i)__builtin_ia32_pclmulqdq512((__v8di)(__m512i)(A),  \
                                        (__v8di)(__m512i)(B),  \
                                        (char)(I))
+#endif // __AVX512FINTRIN_H
 
 #endif /* __VPCLMULQDQINTRIN_H */
 

lib/libc/glibc/bits/byteswap.h

@@ -1,5 +1,5 @@
 /* Macros and inline functions to swap the order of bytes in integer values.
-   Copyright (C) 1997-2020 Free Software Foundation, Inc.
+   Copyright (C) 1997-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/floatn-common.h

@@ -1,6 +1,6 @@
 /* Macros to control TS 18661-3 glibc features where the same
    definitions are appropriate for all platforms.
-   Copyright (C) 2017-2020 Free Software Foundation, Inc.
+   Copyright (C) 2017-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/libc-header-start.h

@@ -1,5 +1,5 @@
 /* Handle feature test macros at the start of a header.
-   Copyright (C) 2016-2020 Free Software Foundation, Inc.
+   Copyright (C) 2016-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -44,8 +44,26 @@
 
 /* ISO/IEC TS 18661-1:2014 defines the __STDC_WANT_IEC_60559_BFP_EXT__
    macro.  Most but not all symbols enabled by that macro in TS
-   18661-1 are enabled unconditionally in C2X; the symbols in Annex F
-   still require that macro in C2X.  */
+   18661-1 are enabled unconditionally in C2X.  In C2X, the symbols in
+   Annex F still require a new feature test macro
+   __STDC_WANT_IEC_60559_EXT__ instead (C2X does not define
+   __STDC_WANT_IEC_60559_BFP_EXT__), while a few features from TS
+   18661-1 are not included in C2X (and thus should depend on
+   __STDC_WANT_IEC_60559_BFP_EXT__ even when C2X features are
+   enabled).
+
+   __GLIBC_USE (IEC_60559_BFP_EXT) controls those features from TS
+   18661-1 not included in C2X.
+
+   __GLIBC_USE (IEC_60559_BFP_EXT_C2X) controls those features from TS
+   18661-1 that are also included in C2X (with no feature test macro
+   required in C2X).
+
+   __GLIBC_USE (IEC_60559_EXT) controls those features from TS 18661-1
+   that are included in C2X but conditional on
+   __STDC_WANT_IEC_60559_EXT__.  (There are currently no features
+   conditional on __STDC_WANT_IEC_60559_EXT__ that are not in TS
+   18661-1.)  */
 #undef __GLIBC_USE_IEC_60559_BFP_EXT
 #if defined __USE_GNU || defined __STDC_WANT_IEC_60559_BFP_EXT__
 # define __GLIBC_USE_IEC_60559_BFP_EXT 1
@@ -58,6 +76,12 @@
 #else
 # define __GLIBC_USE_IEC_60559_BFP_EXT_C2X 0
 #endif
+#undef __GLIBC_USE_IEC_60559_EXT
+#if __GLIBC_USE (IEC_60559_BFP_EXT) || defined __STDC_WANT_IEC_60559_EXT__
+# define __GLIBC_USE_IEC_60559_EXT 1
+#else
+# define __GLIBC_USE_IEC_60559_EXT 0
+#endif
 
 /* ISO/IEC TS 18661-4:2015 defines the
    __STDC_WANT_IEC_60559_FUNCS_EXT__ macro.  Other than the reduction

lib/libc/glibc/bits/long-double.h

@@ -1,5 +1,5 @@
 /* Properties of long double type.
-   Copyright (C) 2016-2020 Free Software Foundation, Inc.
+   Copyright (C) 2016-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -37,4 +37,17 @@
 #ifndef __NO_LONG_DOUBLE_MATH
 # define __NO_LONG_DOUBLE_MATH	1
 #endif
-#define __LONG_DOUBLE_USES_FLOAT128 0
+
+/* The macro __LDOUBLE_REDIRECTS_TO_FLOAT128_ABI is used to determine the
+   choice of the underlying ABI of long double.  It will always assume
+   a constant value for each translation unit.
+
+   If the value is non-zero, any API which is parameterized by the long
+   double type (i.e the scanf/printf family of functions or the explicitly
+   parameterized math.h functions) will be redirected to a compatible
+   implementation using _Float128 ABI via symbols suffixed with ieee128.
+
+   The mechanism this macro uses to acquire may be a function
+   of architecture, or target specific options used to invoke the
+   compiler.  */
+#define __LDOUBLE_REDIRECTS_TO_FLOAT128_ABI 0

lib/libc/glibc/bits/select.h

@@ -1,4 +1,4 @@
-/* Copyright (C) 1997-2020 Free Software Foundation, Inc.
+/* Copyright (C) 1997-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/signum-generic.h

@@ -1,5 +1,5 @@
 /* Signal number constants.  Generic template.
-   Copyright (C) 1991-2020 Free Software Foundation, Inc.
+   Copyright (C) 1991-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -57,31 +57,9 @@
 #define	SIGQUIT		3	/* Quit.  */
 #define	SIGTRAP		5	/* Trace/breakpoint trap.  */
 #define	SIGKILL		9	/* Killed.  */
-#define SIGBUS		10	/* Bus error.  */
-#define	SIGSYS		12	/* Bad system call.  */
 #define	SIGPIPE		13	/* Broken pipe.  */
 #define	SIGALRM		14	/* Alarm clock.  */
 
-/* New(er) POSIX signals (1003.1-2008, 1003.1-2013).  */
-#define	SIGURG		16	/* Urgent data is available at a socket.  */
-#define	SIGSTOP		17	/* Stop, unblockable.  */
-#define	SIGTSTP		18	/* Keyboard stop.  */
-#define	SIGCONT		19	/* Continue.  */
-#define	SIGCHLD		20	/* Child terminated or stopped.  */
-#define	SIGTTIN		21	/* Background read from control terminal.  */
-#define	SIGTTOU		22	/* Background write to control terminal.  */
-#define	SIGPOLL		23	/* Pollable event occurred (System V).  */
-#define	SIGXCPU		24	/* CPU time limit exceeded.  */
-#define	SIGXFSZ		25	/* File size limit exceeded.  */
-#define	SIGVTALRM	26	/* Virtual timer expired.  */
-#define	SIGPROF		27	/* Profiling timer expired.  */
-#define	SIGUSR1		30	/* User-defined signal 1.  */
-#define	SIGUSR2		31	/* User-defined signal 2.  */
-
-/* Nonstandard signals found in all modern POSIX systems
-   (including both BSD and Linux).  */
-#define	SIGWINCH	28	/* Window size change (4.3 BSD, Sun).  */
-
 /* Archaic names for compatibility.  */
 #define	SIGIO		SIGPOLL	/* I/O now possible (4.2 BSD).  */
 #define	SIGIOT		SIGABRT	/* IOT instruction, abort() on a PDP-11.  */
@@ -93,8 +71,9 @@
    but some real-time signals may be used internally by glibc.  Do not
    use these constants in application code; use SIGRTMIN and SIGRTMAX
    (defined in signal.h) instead.  */
-#define __SIGRTMIN	32
-#define __SIGRTMAX	__SIGRTMIN
+
+/* Include system specific bits.  */
+#include <bits/signum-arch.h>
 
 /* Biggest signal number + 1 (including real-time signals).  */
 #define _NSIG		(__SIGRTMAX + 1)

lib/libc/glibc/bits/stat.h

@@ -1,4 +1,4 @@
-/* Copyright (C) 1992-2020 Free Software Foundation, Inc.
+/* Copyright (C) 1992-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/stdint-intn.h

@@ -1,5 +1,5 @@
 /* Define intN_t types.
-   Copyright (C) 2017-2020 Free Software Foundation, Inc.
+   Copyright (C) 2017-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/stdlib-bsearch.h

@@ -1,5 +1,5 @@
 /* Perform binary search - inline version.
-   Copyright (C) 1991-2020 Free Software Foundation, Inc.
+   Copyright (C) 1991-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/time64.h

@@ -1,5 +1,5 @@
 /* bits/time64.h -- underlying types for __time64_t.  Generic version.
-   Copyright (C) 2018-2020 Free Software Foundation, Inc.
+   Copyright (C) 2018-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/timesize.h

@@ -1,5 +1,5 @@
 /* Bit size of the time_t type at glibc build time, general case.
-   Copyright (C) 2018-2020 Free Software Foundation, Inc.
+   Copyright (C) 2018-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/types/struct_sched_param.h

@@ -1,5 +1,5 @@
 /* Sched parameter structure.  Generic version.
-   Copyright (C) 1996-2020 Free Software Foundation, Inc.
+   Copyright (C) 1996-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/typesizes.h

@@ -1,5 +1,5 @@
 /* bits/typesizes.h -- underlying types for *_t.  Generic version.
-   Copyright (C) 2002-2020 Free Software Foundation, Inc.
+   Copyright (C) 2002-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -50,6 +50,7 @@
 #define __TIME_T_TYPE		__SLONGWORD_TYPE
 #define __USECONDS_T_TYPE	__U32_TYPE
 #define __SUSECONDS_T_TYPE	__SLONGWORD_TYPE
+#define __SUSECONDS64_T_TYPE	__SQUAD_TYPE
 #define __DADDR_T_TYPE		__S32_TYPE
 #define __KEY_T_TYPE		__S32_TYPE
 #define __CLOCKID_T_TYPE	__S32_TYPE
@@ -75,10 +76,16 @@
 
 /* And for fsblkcnt_t, fsblkcnt64_t, fsfilcnt_t and fsfilcnt64_t.  */
 # define __STATFS_MATCHES_STATFS64  1
+
+/* And for getitimer, setitimer and rusage  */
+# define __KERNEL_OLD_TIMEVAL_MATCHES_TIMEVAL64 1
 #else
 # define __RLIM_T_MATCHES_RLIM64_T	0
 
 # define __STATFS_MATCHES_STATFS64  0
+
+/* And for getitimer, setitimer and rusage  */
+# define __KERNEL_OLD_TIMEVAL_MATCHES_TIMEVAL64 0
 #endif
 
 /* Number of descriptors that can fit in an `fd_set'.  */

lib/libc/glibc/bits/uintn-identity.h

@@ -1,5 +1,5 @@
 /* Inline functions to return unsigned integer values unchanged.
-   Copyright (C) 2017-2020 Free Software Foundation, Inc.
+   Copyright (C) 2017-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/bits/waitflags.h

@@ -1,5 +1,5 @@
 /* Definitions of flag bits for `waitpid' et al.
-   Copyright (C) 1992-2020 Free Software Foundation, Inc.
+   Copyright (C) 1992-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or
@@ -24,3 +24,11 @@
 /* Bits in the third argument to `waitpid'.  */
 #define	WNOHANG		1	/* Don't block waiting.  */
 #define	WUNTRACED	2	/* Report status of stopped children.  */
+
+/* Bits in the fourth argument to `waitid'.  */
+#if defined __USE_XOPEN_EXTENDED || defined __USE_XOPEN2K8
+# define WSTOPPED	WUNTRACED	/* Report stopped child. */
+# define WCONTINUED	4		/* Report continued child.  */
+# define WNOWAIT	8		/* Don't reap, just poll status.  */
+# define WEXITED	16		/* Report dead child.  */
+#endif

lib/libc/glibc/bits/waitstatus.h

@@ -1,5 +1,5 @@
 /* Definitions of status bits for `wait' et al.
-   Copyright (C) 1992-2020 Free Software Foundation, Inc.
+   Copyright (C) 1992-2021 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
 
    The GNU C Library is free software; you can redistribute it and/or

lib/libc/glibc/csu/errno.c

@@ -0,0 +1,35 @@
+/* Definition of `errno' variable.  Canonical version.
+   Copyright (C) 2002-2021 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <https://www.gnu.org/licenses/>.  */
+
+#include <errno.h>
+#include <tls.h>
+#include <dl-sysdep.h>
+#undef errno
+
+#if RTLD_PRIVATE_ERRNO
+
+/* Code compiled for rtld refers only to this name.  */
+int rtld_errno attribute_hidden;
+
+#else
+
+__thread int errno;
+extern __thread int __libc_errno __attribute__ ((alias ("errno")))
+  attribute_hidden;
+
+#endif