Rollup merge of rust-lang#134740 - Flakebi:amdgpu-target, r=workingju…

…bilee Add amdgpu target Add amdgpu target to rustc and enable the LLVM target. Fix compiling `core` with the amdgpu: The amdgpu backend makes heavy use of different address spaces. This leads to situations, where a pointer in one addrspace needs to be casted to a pointer in a different addrspace. `bitcast` is invalid for this case, `addrspacecast` needs to be used. Fix compilation failures that created bitcasts for such cases by creating pointer casts (which creates an `addrspacecast` under the hood) instead. MCP: rust-lang/compiler-team#823 Tracking issue: rust-lang#135024 Kinda related to the original amdgpu tracking issue rust-lang#51575 (though that one has been closed for a while).
matthiaskrgr · Feb 1, 2025 · 72c5602 · 72c5602
2 parents 0141138 + 56795fb
commit 72c5602
Show file tree

Hide file tree

Showing 16 changed files with 220 additions and 10 deletions.
diff --git a/compiler/rustc_target/src/spec/mod.rs b/compiler/rustc_target/src/spec/mod.rs
@@ -1908,6 +1908,8 @@ supported_targets! {
 
     ("nvptx64-nvidia-cuda", nvptx64_nvidia_cuda),
 
+    ("amdgcn-amd-amdhsa", amdgcn_amd_amdhsa),
+
     ("xtensa-esp32-none-elf", xtensa_esp32_none_elf),
     ("xtensa-esp32-espidf", xtensa_esp32_espidf),
     ("xtensa-esp32s2-none-elf", xtensa_esp32s2_none_elf),

diff --git a/compiler/rustc_target/src/spec/targets/amdgcn_amd_amdhsa.rs b/compiler/rustc_target/src/spec/targets/amdgcn_amd_amdhsa.rs
@@ -0,0 +1,51 @@
+use crate::spec::{Cc, LinkerFlavor, Lld, PanicStrategy, Target, TargetOptions};
+
+pub(crate) fn target() -> Target {
+    Target {
+        arch: "amdgpu".into(),
+        data_layout: "e-p:64:64-p1:64:64-p2:32:32-p3:32:32-p4:64:64-p5:32:32-p6:32:32-p7:160:256:256:32-p8:128:128-p9:192:256:256:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-S32-A5-G1-ni:7:8:9".into(),
+        llvm_target: "amdgcn-amd-amdhsa".into(),
+        metadata: crate::spec::TargetMetadata {
+            description: Some("AMD GPU".into()),
+            tier: Some(3),
+            host_tools: Some(false),
+            std: Some(false),
+        },
+        pointer_width: 64,
+
+        options: TargetOptions {
+            os: "amdhsa".into(),
+            vendor: "amd".into(),
+            linker_flavor: LinkerFlavor::Gnu(Cc::No, Lld::Yes),
+            linker: Some("rust-lld".into()),
+
+            // There are many CPUs, one for each hardware generation.
+            // Require to set one explicitly as there is no good default.
+            need_explicit_cpu: true,
+
+            max_atomic_width: Some(64),
+
+            // Unwinding on GPUs is not useful.
+            panic_strategy: PanicStrategy::Abort,
+
+            // amdgpu backend does not support libcalls.
+            no_builtins: true,
+            simd_types_indirect: false,
+
+            // Allow `cdylib` crate type.
+            dynamic_linking: true,
+            only_cdylib: true,
+            executables: false,
+            dll_prefix: "".into(),
+            dll_suffix: ".elf".into(),
+
+            // The LLVM backend does not support stack canaries for this target
+            supports_stack_protector: false,
+
+            // Force LTO, object linking does not yet work with amdgpu.
+            requires_lto: true,
+
+            ..Default::default()
+        },
+    }
+}
diff --git a/config.example.toml b/config.example.toml
@@ -108,7 +108,7 @@
 # the resulting rustc being unable to compile for the disabled architectures.
 #
 # To add support for new targets, see https://rustc-dev-guide.rust-lang.org/building/new-target.html.
-#targets = "AArch64;ARM;BPF;Hexagon;LoongArch;MSP430;Mips;NVPTX;PowerPC;RISCV;Sparc;SystemZ;WebAssembly;X86"
+#targets = "AArch64;AMDGPU;ARM;BPF;Hexagon;LoongArch;MSP430;Mips;NVPTX;PowerPC;RISCV;Sparc;SystemZ;WebAssembly;X86"
 
 # LLVM experimental targets to build support for. These targets are specified in
 # the same format as above, but since these targets are experimental, they are

diff --git a/src/bootstrap/download-ci-llvm-stamp b/src/bootstrap/download-ci-llvm-stamp
@@ -1,4 +1,4 @@
 Change this file to make users of the `download-ci-llvm` configuration download
 a new version of LLVM from CI, even if the LLVM submodule hasn’t changed.
 
-Last change is for: https://github.com/rust-lang/rust/pull/129788
+Last change is for: https://github.com/rust-lang/rust/pull/134740
diff --git a/src/bootstrap/src/core/build_steps/llvm.rs b/src/bootstrap/src/core/build_steps/llvm.rs
@@ -331,7 +331,7 @@ impl Step for Llvm {
         let llvm_targets = match &builder.config.llvm_targets {
             Some(s) => s,
             None => {
-                "AArch64;ARM;BPF;Hexagon;LoongArch;MSP430;Mips;NVPTX;PowerPC;RISCV;\
+                "AArch64;AMDGPU;ARM;BPF;Hexagon;LoongArch;MSP430;Mips;NVPTX;PowerPC;RISCV;\
                      Sparc;SystemZ;WebAssembly;X86"
             }
         };

diff --git a/src/doc/rustc/src/SUMMARY.md b/src/doc/rustc/src/SUMMARY.md
@@ -29,6 +29,7 @@
     - [\*-apple-watchos](platform-support/apple-watchos.md)
     - [\*-apple-visionos](platform-support/apple-visionos.md)
     - [aarch64-nintendo-switch-freestanding](platform-support/aarch64-nintendo-switch-freestanding.md)
+    - [amdgcn-amd-amdhsa](platform-support/amdgcn-amd-amdhsa.md)
     - [armeb-unknown-linux-gnueabi](platform-support/armeb-unknown-linux-gnueabi.md)
     - [arm-none-eabi](platform-support/arm-none-eabi.md)
       - [armv4t-none-eabi](platform-support/armv4t-none-eabi.md)

diff --git a/src/doc/rustc/src/platform-support.md b/src/doc/rustc/src/platform-support.md
@@ -272,6 +272,7 @@ target | std | host | notes
 `aarch64_be-unknown-linux-gnu` | ✓ | ✓ | ARM64 Linux (big-endian)
 `aarch64_be-unknown-linux-gnu_ilp32` | ✓ | ✓ | ARM64 Linux (big-endian, ILP32 ABI)
 [`aarch64_be-unknown-netbsd`](platform-support/netbsd.md) | ✓ | ✓ | ARM64 NetBSD (big-endian)
+[`amdgcn-amd-amdhsa`](platform-support/amdgcn-amd-amdhsa.md) | * |  | `-Ctarget-cpu=gfx...` to specify [the AMD GPU] to compile for
 [`arm64_32-apple-watchos`](platform-support/apple-watchos.md) | ✓ |  | Arm Apple WatchOS 64-bit with 32-bit pointers
 [`arm64e-apple-darwin`](platform-support/arm64e-apple-darwin.md)  | ✓ | ✓ | ARM64e Apple Darwin
 [`arm64e-apple-ios`](platform-support/arm64e-apple-ios.md) | ✓ | | ARM64e Apple iOS
@@ -432,3 +433,4 @@ target | std | host | notes
 [`xtensa-esp32s3-none-elf`](platform-support/xtensa.md) | * |  | Xtensa ESP32-S3
 
 [runs on NVIDIA GPUs]: https://github.com/japaric-archived/nvptx#targets
+[the AMD GPU]: https://llvm.org/docs/AMDGPUUsage.html#processors
diff --git a/src/doc/rustc/src/platform-support/amdgcn-amd-amdhsa.md b/src/doc/rustc/src/platform-support/amdgcn-amd-amdhsa.md
@@ -0,0 +1,111 @@
+# `amdgcn-amd-amdhsa`
+
+**Tier: 3**
+
+AMD GPU target for compute/HSA (Heterogeneous System Architecture).
+
+## Target maintainers
+
+- [@Flakebi](https://github.com/Flakebi)
+
+## Requirements
+
+AMD GPUs can be targeted via cross-compilation.
+Supported GPUs depend on the LLVM version that is used by Rust.
+In general, most GPUs starting from gfx7 (Sea Islands/CI) are supported as compilation targets, though older GPUs are not supported by the latest host runtime.
+Details about supported GPUs can be found in [LLVM’s documentation] and [ROCm documentation].
+
+Binaries can be loaded by [HIP] or by the HSA runtime implemented in [ROCR-Runtime].
+The format of binaries is a linked ELF.
+
+Binaries must be built with no-std.
+They can use `core` and `alloc` (`alloc` only if an allocator is supplied).
+At least one function needs to use the `"gpu-kernel"` calling convention and should be marked with `no_mangle` for simplicity.
+Functions using the `"gpu-kernel"` calling convention are kernel entrypoints and can be used from the host runtime.
+
+## Building the target
+
+The target is included in rustc.
+
+## Building Rust programs
+
+The amdgpu target supports many hardware generations, which need different binaries.
+The generations are exposed as different target-cpus in the backend.
+As there are many, Rust does not ship pre-compiled libraries for this target.
+Therefore, you have to build your own copy of `core` by using `cargo -Zbuild-std=core` or similar.
+
+To build a binary, create a no-std library:
+```rust,ignore (platform-specific)
+// src/lib.rs
+#![feature(abi_gpu_kernel)]
+#![no_std]
+
+#[panic_handler]
+fn panic(_: &core::panic::PanicInfo) -> ! {
+    loop {}
+}
+
+#[no_mangle]
+pub extern "gpu-kernel" fn kernel(/* Arguments */) {
+    // Code
+}
+```
+
+Build the library as `cdylib`:
+```toml
+# Cargo.toml
+[lib]
+crate-type = ["cdylib"]
+
+[profile.dev]
+lto = true # LTO must be explicitly enabled for now
+[profile.release]
+lto = true
+```
+
+The target-cpu must be from the list [supported by LLVM] (or printed with `rustc --target amdgcn-amd-amdhsa --print target-cpus`).
+The GPU version on the current system can be found e.g. with [`rocminfo`].
+
+Example `.cargo/config.toml` file to set the target and GPU generation:
+```toml
+# .cargo/config.toml
+[build]
+target = "amdgcn-amd-amdhsa"
+rustflags = ["-Ctarget-cpu=gfx1100"]
+
+[unstable]
+build-std = ["core"] # Optional: "alloc"
+```
+
+## Running Rust programs
+
+To run a binary on an AMD GPU, a host runtime is needed.
+On Linux and Windows, [HIP] can be used to load and run binaries.
+Example code on how to load a compiled binary and run it is available in [ROCm examples].
+
+On Linux, binaries can also run through the HSA runtime as implemented in [ROCR-Runtime].
+
+<!-- Mention an allocator once a suitable one exists for amdgpu -->
+
+<!--
+## Testing
+
+Does the target support running binaries, or do binaries have varying
+expectations that prevent having a standard way to run them? If users can run
+binaries, can they do so in some common emulator, or do they need native
+hardware? Does the target support running the Rust testsuite?
+
+-->
+
+## Additional information
+
+More information can be found on the [LLVM page for amdgpu].
+
+[LLVM’s documentation]: https://llvm.org/docs/AMDGPUUsage.html#processors
+[ROCm documentation]: https://rocmdocs.amd.com
+[HIP]: https://rocm.docs.amd.com/projects/HIP/
+[ROCR-Runtime]: https://github.com/ROCm/ROCR-Runtime
+[supported by LLVM]: https://llvm.org/docs/AMDGPUUsage.html#processors
+[LLVM page for amdgpu]: https://llvm.org/docs/AMDGPUUsage.html
+[`rocminfo`]: https://github.com/ROCm/rocminfo
+[ROCm examples]: https://github.com/ROCm/rocm-examples/tree/ca8ef5b6f1390176616cd1c18fbc98785cbc73f6/HIP-Basic/module_api
diff --git a/src/tools/build-manifest/src/main.rs b/src/tools/build-manifest/src/main.rs
@@ -67,6 +67,7 @@ static TARGETS: &[&str] = &[
     "aarch64-unknown-none-softfloat",
     "aarch64-unknown-redox",
     "aarch64-unknown-uefi",
+    "amdgcn-amd-amdhsa",
     "arm64e-apple-darwin",
     "arm64e-apple-ios",
     "arm64e-apple-tvos",

diff --git a/tests/assembly/targets/targets-amdgpu.rs b/tests/assembly/targets/targets-amdgpu.rs
@@ -0,0 +1,21 @@
+//@ assembly-output: emit-asm
+// ignore-tidy-linelength
+//@ revisions: amdgcn_amd_amdhsa
+//@ [amdgcn_amd_amdhsa] compile-flags: --target amdgcn-amd-amdhsa -Ctarget-cpu=gfx900
+//@ [amdgcn_amd_amdhsa] needs-llvm-components: amdgpu
+
+// Sanity-check that each target can produce assembly code.
+
+#![feature(no_core, lang_items)]
+#![no_std]
+#![no_core]
+#![crate_type = "lib"]
+
+#[lang = "sized"]
+trait Sized {}
+
+pub fn test() -> u8 {
+    42
+}
+
+// CHECK: .version
diff --git a/tests/ui/amdgpu-require-explicit-cpu.nocpu.stderr b/tests/ui/amdgpu-require-explicit-cpu.nocpu.stderr
@@ -0,0 +1,4 @@
+error: target requires explicitly specifying a cpu with `-C target-cpu`
+
+error: aborting due to 1 previous error
+
diff --git a/tests/ui/amdgpu-require-explicit-cpu.rs b/tests/ui/amdgpu-require-explicit-cpu.rs
@@ -0,0 +1,17 @@
+//@ revisions: nocpu cpu
+//@ no-prefer-dynamic
+//@ compile-flags: --crate-type=cdylib --target=amdgcn-amd-amdhsa
+//@ needs-llvm-components: amdgpu
+//@ needs-rust-lld
+//@[nocpu] error-pattern: target requires explicitly specifying a cpu
+//@[nocpu] build-fail
+//@[cpu] compile-flags: -Ctarget-cpu=gfx900
+//@[cpu] build-pass
+
+#![feature(no_core, lang_items)]
+#![no_core]
+
+#[lang="sized"]
+trait Sized {}
+
+pub fn foo() {}
diff --git a/tests/ui/check-cfg/exhaustive-names-values.empty_cfg.stderr b/tests/ui/check-cfg/exhaustive-names-values.empty_cfg.stderr
@@ -14,7 +14,7 @@ warning: unexpected `cfg` condition value: `value`
 LL | #[cfg(target_vendor = "value")]
    |       ^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_vendor` are: `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
+   = note: expected values for `target_vendor` are: `amd`, `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition name: `feature`

diff --git a/tests/ui/check-cfg/exhaustive-names-values.feature.stderr b/tests/ui/check-cfg/exhaustive-names-values.feature.stderr
@@ -15,7 +15,7 @@ warning: unexpected `cfg` condition value: `value`
 LL | #[cfg(target_vendor = "value")]
    |       ^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_vendor` are: `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
+   = note: expected values for `target_vendor` are: `amd`, `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition value: `unk`

diff --git a/tests/ui/check-cfg/exhaustive-names-values.full.stderr b/tests/ui/check-cfg/exhaustive-names-values.full.stderr
@@ -15,7 +15,7 @@ warning: unexpected `cfg` condition value: `value`
 LL | #[cfg(target_vendor = "value")]
    |       ^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_vendor` are: `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
+   = note: expected values for `target_vendor` are: `amd`, `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition value: `unk`

diff --git a/tests/ui/check-cfg/well-known-values.stderr b/tests/ui/check-cfg/well-known-values.stderr
@@ -138,7 +138,7 @@ warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
 LL |     target_arch = "_UNEXPECTED_VALUE",
    |     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_arch` are: `aarch64`, `arm`, `arm64ec`, `avr`, `bpf`, `csky`, `hexagon`, `loongarch64`, `m68k`, `mips`, `mips32r6`, `mips64`, `mips64r6`, `msp430`, `nvptx64`, `powerpc`, `powerpc64`, `riscv32`, `riscv64`, `s390x`, `sparc`, `sparc64`, `wasm32`, `wasm64`, `x86`, `x86_64`, and `xtensa`
+   = note: expected values for `target_arch` are: `aarch64`, `amdgpu`, `arm`, `arm64ec`, `avr`, `bpf`, `csky`, `hexagon`, `loongarch64`, `m68k`, `mips`, `mips32r6`, `mips64`, `mips64r6`, `msp430`, `nvptx64`, `powerpc`, `powerpc64`, `riscv32`, `riscv64`, `s390x`, `sparc`, `sparc64`, `wasm32`, `wasm64`, `x86`, `x86_64`, and `xtensa`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
@@ -201,7 +201,7 @@ warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
 LL |     target_os = "_UNEXPECTED_VALUE",
    |     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_os` are: `aix`, `android`, `cuda`, `dragonfly`, `emscripten`, `espidf`, `freebsd`, `fuchsia`, `haiku`, `hermit`, `horizon`, `hurd`, `illumos`, `ios`, `l4re`, `linux`, `macos`, `netbsd`, `none`, `nto`, `nuttx`, `openbsd`, `psp`, `psx`, `redox`, `rtems`, `solaris`, `solid_asp3`, `teeos`, `trusty`, `tvos`, `uefi`, `unknown`, `visionos`, `vita`, `vxworks`, `wasi`, `watchos`, `windows`, `xous`, and `zkvm`
+   = note: expected values for `target_os` are: `aix`, `amdhsa`, `android`, `cuda`, `dragonfly`, `emscripten`, `espidf`, `freebsd`, `fuchsia`, `haiku`, `hermit`, `horizon`, `hurd`, `illumos`, `ios`, `l4re`, `linux`, `macos`, `netbsd`, `none`, `nto`, `nuttx`, `openbsd`, `psp`, `psx`, `redox`, `rtems`, `solaris`, `solid_asp3`, `teeos`, `trusty`, `tvos`, `uefi`, `unknown`, `visionos`, `vita`, `vxworks`, `wasi`, `watchos`, `windows`, `xous`, and `zkvm`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
@@ -230,7 +230,7 @@ warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
 LL |     target_vendor = "_UNEXPECTED_VALUE",
    |     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    |
-   = note: expected values for `target_vendor` are: `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
+   = note: expected values for `target_vendor` are: `amd`, `apple`, `espressif`, `fortanix`, `ibm`, `kmc`, `mti`, `nintendo`, `nvidia`, `pc`, `risc0`, `sony`, `sun`, `unikraft`, `unknown`, `uwp`, `win7`, and `wrs`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: unexpected `cfg` condition value: `_UNEXPECTED_VALUE`
@@ -274,7 +274,7 @@ LL | #[cfg(target_os = "linuz")] // testing that we suggest `linux`
    |                   |
    |                   help: there is a expected value with a similar name: `"linux"`
    |
-   = note: expected values for `target_os` are: `aix`, `android`, `cuda`, `dragonfly`, `emscripten`, `espidf`, `freebsd`, `fuchsia`, `haiku`, `hermit`, `horizon`, `hurd`, `illumos`, `ios`, `l4re`, `linux`, `macos`, `netbsd`, `none`, `nto`, `nuttx`, `openbsd`, `psp`, `psx`, `redox`, `rtems`, `solaris`, `solid_asp3`, `teeos`, `trusty`, `tvos`, `uefi`, `unknown`, `visionos`, `vita`, `vxworks`, `wasi`, `watchos`, `windows`, `xous`, and `zkvm`
+   = note: expected values for `target_os` are: `aix`, `amdhsa`, `android`, `cuda`, `dragonfly`, `emscripten`, `espidf`, `freebsd`, `fuchsia`, `haiku`, `hermit`, `horizon`, `hurd`, `illumos`, `ios`, `l4re`, `linux`, `macos`, `netbsd`, `none`, `nto`, `nuttx`, `openbsd`, `psp`, `psx`, `redox`, `rtems`, `solaris`, `solid_asp3`, `teeos`, `trusty`, `tvos`, `uefi`, `unknown`, `visionos`, `vita`, `vxworks`, `wasi`, `watchos`, `windows`, `xous`, and `zkvm`
    = note: see <https://doc.rust-lang.org/nightly/rustc/check-cfg.html> for more information about checking conditional configuration
 
 warning: 28 warnings emitted
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,4 @@
		error: target requires explicitly specifying a cpu with `-C target-cpu`

		error: aborting due to 1 previous error