NVPTX target specification

.travis.yml

@@ -186,6 +186,8 @@ matrix:
       if: branch = auto
     - env: IMAGE=mingw-check
       if: type = pull_request OR branch = auto
+    - env: IMAGE=nvptx-cuda
+      if: branch = auto
 
     - stage: publish toolstate
       if: branch = master AND type = push

src/bootstrap/lib.rs

@@ -831,6 +831,7 @@ impl Build {
                   !target.contains("msvc") &&
                   !target.contains("emscripten") &&
                   !target.contains("wasm32") &&
+                  !target.contains("nvptx") &&
                   !target.contains("fuchsia") {
             Some(self.cc(target))
         } else {

src/bootstrap/sanity.rs

@@ -156,7 +156,7 @@ pub fn check(build: &mut Build) {
             panic!("the iOS target is only supported on macOS");
         }
 
-        if target.contains("-none-") {
+        if target.contains("-none-") || target.contains("nvptx") {
             if build.no_std(*target).is_none() {
                 let target = build.config.target_config.entry(target.clone())
                     .or_default();
@@ -165,7 +165,7 @@ pub fn check(build: &mut Build) {
             }
 
             if build.no_std(*target) == Some(false) {
-                panic!("All the *-none-* targets are no-std targets")
+                panic!("All the *-none-* and nvptx* targets are no-std targets")
             }
         }
 

src/ci/docker/dist-various-2/Dockerfile

@@ -70,6 +70,7 @@ ENV TARGETS=$TARGETS,x86_64-sun-solaris
 ENV TARGETS=$TARGETS,x86_64-unknown-linux-gnux32
 ENV TARGETS=$TARGETS,x86_64-unknown-cloudabi
 ENV TARGETS=$TARGETS,x86_64-fortanix-unknown-sgx
+ENV TARGETS=$TARGETS,nvptx64-nvidia-cuda
 
 ENV X86_FORTANIX_SGX_LIBS="/x86_64-fortanix-unknown-sgx/lib/"
 

src/ci/docker/nvptx-cuda/Dockerfile

@@ -0,0 +1,18 @@
+FROM ubuntu:18.04
+
+RUN apt-get update
+RUN apt-get install -y --no-install-recommends \
+    g++ make file curl ca-certificates python git \
+    cmake sudo gdb
+
+# TODO(denzp): setup `ptx-linker` CI for auttomatic binary releases.
+RUN curl -sL https://github.com/denzp/rust-ptx-linker/releases/download/v0.9.0-alpha/rust-ptx-linker.linux64.tar.gz | \
+    tar -xzvC /usr/bin
+
+COPY scripts/sccache.sh /scripts/
+RUN sh /scripts/sccache.sh
+
+ENV TARGETS=nvptx64-nvidia-cuda
+
+ENV SCRIPT python2.7 /checkout/x.py test --target $TARGETS \
+    src/test/run-make

src/librustc/ty/context.rs

@@ -1675,6 +1675,12 @@ impl<'a, 'gcx, 'tcx> TyCtxt<'a, 'gcx, 'tcx> {
         }
         false
     }
+
+    /// Determine whether identifiers in the assembly have strict naming rules.
+    /// Currently, only NVPTX* targets need it.
+    pub fn has_strict_asm_symbol_naming(&self) -> bool {
+        self.gcx.sess.target.target.arch.contains("nvptx")
+    }
 }
 
 impl<'a, 'tcx> TyCtxt<'a, 'tcx, 'tcx> {

src/librustc_codegen_ssa/back/link.rs

@@ -149,6 +149,7 @@ pub fn linker_and_flavor(sess: &Session) -> (PathBuf, LinkerFlavor) {
                 LinkerFlavor::Ld => "ld",
                 LinkerFlavor::Msvc => "link.exe",
                 LinkerFlavor::Lld(_) => "lld",
+                LinkerFlavor::PtxLinker => "rust-ptx-linker",
             }), flavor)),
             (Some(linker), None) => {
                 let stem = if linker.extension().and_then(|ext| ext.to_str()) == Some("exe") {

src/librustc_codegen_ssa/back/linker.rs

@@ -83,6 +83,10 @@ impl LinkerInfo {
             LinkerFlavor::Lld(LldFlavor::Wasm) => {
                 Box::new(WasmLd::new(cmd, sess, self)) as Box<dyn Linker>
             }
+
+            LinkerFlavor::PtxLinker => {
+                Box::new(PtxLinker { cmd, sess }) as Box<dyn Linker>
+            }
         }
     }
 }
@@ -1080,3 +1084,124 @@ fn exported_symbols(tcx: TyCtxt, crate_type: CrateType) -> Vec<String> {
 
     symbols
 }
+
+/// Much simplified and explicit CLI for the NVPTX linker. The linker operates
+/// with bitcode and uses LLVM backend to generate a PTX assembly.
+pub struct PtxLinker<'a> {
+    cmd: Command,
+    sess: &'a Session,
+}
+
+impl<'a> Linker for PtxLinker<'a> {
+    fn link_rlib(&mut self, path: &Path) {
+        self.cmd.arg("--rlib").arg(path);
+    }
+
+    fn link_whole_rlib(&mut self, path: &Path) {
+        self.cmd.arg("--rlib").arg(path);
+    }
+
+    fn include_path(&mut self, path: &Path) {
+        self.cmd.arg("-L").arg(path);
+    }
+
+    fn debuginfo(&mut self) {
+        self.cmd.arg("--debug");
+    }
+
+    fn add_object(&mut self, path: &Path) {
+        self.cmd.arg("--bitcode").arg(path);
+    }
+
+    fn args(&mut self, args: &[String]) {
+        self.cmd.args(args);
+    }
+
+    fn optimize(&mut self) {
+        self.cmd.arg(match self.sess.opts.optimize {
+            OptLevel::No => "-O0",
+            OptLevel::Less => "-O1",
+            OptLevel::Default => "-O2",
+            OptLevel::Aggressive => "-O3",
+            OptLevel::Size => "-Os",
+            OptLevel::SizeMin => "-Os"
+        });
+    }
+
+    fn output_filename(&mut self, path: &Path) {
+        self.cmd.arg("-o").arg(path);
+    }
+
+    fn finalize(&mut self) -> Command {
+        ::std::mem::replace(&mut self.cmd, Command::new(""))
+    }
+
+    fn link_dylib(&mut self, _lib: &str) {
+        panic!("external dylibs not supported")
+    }
+
+    fn link_rust_dylib(&mut self, _lib: &str, _path: &Path) {
+        panic!("external dylibs not supported")
+    }
+
+    fn link_staticlib(&mut self, _lib: &str) {
+        panic!("staticlibs not supported")
+    }
+
+    fn link_whole_staticlib(&mut self, _lib: &str, _search_path: &[PathBuf]) {
+        panic!("staticlibs not supported")
+    }
+
+    fn framework_path(&mut self, _path: &Path) {
+        panic!("frameworks not supported")
+    }
+
+    fn link_framework(&mut self, _framework: &str) {
+        panic!("frameworks not supported")
+    }
+
+    fn position_independent_executable(&mut self) {
+    }
+
+    fn full_relro(&mut self) {
+    }
+
+    fn partial_relro(&mut self) {
+    }
+
+    fn no_relro(&mut self) {
+    }
+
+    fn build_static_executable(&mut self) {
+    }
+
+    fn gc_sections(&mut self, _keep_metadata: bool) {
+    }
+
+    fn pgo_gen(&mut self) {
+    }
+
+    fn no_default_libraries(&mut self) {
+    }
+
+    fn build_dylib(&mut self, _out_filename: &Path) {
+    }
+
+    fn export_symbols(&mut self, _tmpdir: &Path, _crate_type: CrateType) {
+    }
+
+    fn subsystem(&mut self, _subsystem: &str) {
+    }
+
+    fn no_position_independent_executable(&mut self) {
+    }
+
+    fn group_start(&mut self) {
+    }
+
+    fn group_end(&mut self) {
+    }
+
+    fn cross_lang_lto(&mut self) {
+    }
+}

src/librustc_codegen_utils/lib.rs

@@ -12,6 +12,7 @@
 #![feature(nll)]
 #![allow(unused_attributes)]
 #![feature(rustc_diagnostic_macros)]
+#![feature(in_band_lifetimes)]
 
 #![recursion_limit="256"]
 

src/librustc_codegen_utils/symbol_names.rs

@@ -103,7 +103,7 @@ use rustc_mir::monomorphize::Instance;
 
 use syntax_pos::symbol::Symbol;
 
-use std::fmt::Write;
+use std::fmt::{self, Write};
 use std::mem::discriminant;
 
 pub fn provide(providers: &mut Providers) {
@@ -221,7 +221,7 @@ fn get_symbol_hash<'a, 'tcx>(
 }
 
 fn def_symbol_name<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, def_id: DefId) -> ty::SymbolName {
-    let mut buffer = SymbolPathBuffer::new();
+    let mut buffer = SymbolPathBuffer::new(tcx);
     item_path::with_forced_absolute_paths(|| {
         tcx.push_item_path(&mut buffer, def_id, false);
     });
@@ -317,7 +317,7 @@ fn compute_symbol_name<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, instance: Instance
 
     let hash = get_symbol_hash(tcx, def_id, instance, instance_ty, substs);
 
-    let mut buf = SymbolPathBuffer::from_interned(tcx.def_symbol_name(def_id));
+    let mut buf = SymbolPathBuffer::from_interned(tcx.def_symbol_name(def_id), tcx);
 
     if instance.is_vtable_shim() {
         buf.push("{{vtable-shim}}");
@@ -339,26 +339,28 @@ fn compute_symbol_name<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, instance: Instance
 //
 // To be able to work on all platforms and get *some* reasonable output, we
 // use C++ name-mangling.
-#[derive(Debug)]
-struct SymbolPathBuffer {
+struct SymbolPathBuffer<'a, 'tcx> {
+    tcx: TyCtxt<'a, 'tcx, 'tcx>,
     result: String,
     temp_buf: String,
 }
 
-impl SymbolPathBuffer {
-    fn new() -> Self {
+impl SymbolPathBuffer<'a, 'tcx> {
+    fn new(tcx: TyCtxt<'a, 'tcx, 'tcx>) -> Self {
         let mut result = SymbolPathBuffer {
             result: String::with_capacity(64),
             temp_buf: String::with_capacity(16),
+            tcx,
         };
         result.result.push_str("_ZN"); // _Z == Begin name-sequence, N == nested
         result
     }
 
-    fn from_interned(symbol: ty::SymbolName) -> Self {
+    fn from_interned(symbol: ty::SymbolName, tcx: TyCtxt<'a, 'tcx, 'tcx>) -> Self {
         let mut result = SymbolPathBuffer {
             result: String::with_capacity(64),
             temp_buf: String::with_capacity(16),
+            tcx,
         };
         result.result.push_str(&symbol.as_str());
         result
@@ -377,15 +379,15 @@ impl SymbolPathBuffer {
     }
 }
 
-impl ItemPathBuffer for SymbolPathBuffer {
+impl ItemPathBuffer for SymbolPathBuffer<'a, 'tcx> {
     fn root_mode(&self) -> &RootMode {
         const ABSOLUTE: &RootMode = &RootMode::Absolute;
         ABSOLUTE
     }
 
     fn push(&mut self, text: &str) {
         self.temp_buf.clear();
-        let need_underscore = sanitize(&mut self.temp_buf, text);
+        let need_underscore = sanitize(&mut self.temp_buf, text, self.tcx);
         let _ = write!(
             self.result,
             "{}",
@@ -398,12 +400,24 @@ impl ItemPathBuffer for SymbolPathBuffer {
     }
 }
 
+// Manual Debug implementation to omit non-Debug `tcx` field.
+impl fmt::Debug for SymbolPathBuffer<'_, '_> {
+    fn fmt(&self, fmt: &mut fmt::Formatter) -> fmt::Result {
+        fmt.debug_struct("SymbolPathBuffer")
+            .field("result", &self.result)
+            .field("temp_buf", &self.temp_buf)
+            .finish()
+    }
+}
+
 // Name sanitation. LLVM will happily accept identifiers with weird names, but
 // gas doesn't!
 // gas accepts the following characters in symbols: a-z, A-Z, 0-9, ., _, $
+// NVPTX assembly has more strict naming rules than gas, so additionally, dots
+// are replaced with '$' there.
 //
 // returns true if an underscore must be added at the start
-pub fn sanitize(result: &mut String, s: &str) -> bool {
+pub fn sanitize(result: &mut String, s: &str, tcx: TyCtxt<'a, 'tcx, 'tcx>) -> bool {
     for c in s.chars() {
         match c {
             // Escape these with $ sequences
@@ -416,12 +430,25 @@ pub fn sanitize(result: &mut String, s: &str) -> bool {
             ')' => result.push_str("$RP$"),
             ',' => result.push_str("$C$"),
 
-            // '.' doesn't occur in types and functions, so reuse it
-            // for ':' and '-'
-            '-' | ':' => result.push('.'),
+            '-' | ':' => if tcx.has_strict_asm_symbol_naming() {
+                // NVPTX doesn't support these characters in symbol names.
+                result.push('$')
+            }
+            else {
+                // '.' doesn't occur in types and functions, so reuse it
+                // for ':' and '-'
+                result.push('.')
+            },
+
+            '.' => if tcx.has_strict_asm_symbol_naming() {
+                result.push('$')
+            }
+            else {
+                result.push('.')
+            },
 
             // These are legal symbols
-            'a'..='z' | 'A'..='Z' | '0'..='9' | '_' | '.' | '$' => result.push(c),
+            'a'..='z' | 'A'..='Z' | '0'..='9' | '_' | '$' => result.push(c),
 
             _ => {
                 result.push('$');

src/librustc_target/spec/mod.rs

@@ -75,6 +75,7 @@ pub enum LinkerFlavor {
     Ld,
     Msvc,
     Lld(LldFlavor),
+    PtxLinker,
 }
 
 #[derive(Clone, Copy, Debug, Eq, Ord, PartialEq, PartialOrd, Hash,
@@ -143,6 +144,7 @@ flavor_mappings! {
     ((LinkerFlavor::Gcc), "gcc"),
     ((LinkerFlavor::Ld), "ld"),
     ((LinkerFlavor::Msvc), "msvc"),
+    ((LinkerFlavor::PtxLinker), "ptx-linker"),
     ((LinkerFlavor::Lld(LldFlavor::Wasm)), "wasm-ld"),
     ((LinkerFlavor::Lld(LldFlavor::Ld64)), "ld64.lld"),
     ((LinkerFlavor::Lld(LldFlavor::Ld)), "ld.lld"),
@@ -455,6 +457,8 @@ supported_targets! {
     ("x86_64-fortanix-unknown-sgx", x86_64_fortanix_unknown_sgx),
 
     ("x86_64-unknown-uefi", x86_64_unknown_uefi),
+
+    ("nvptx64-nvidia-cuda", nvptx64_nvidia_cuda),
 }
 
 /// Everything `rustc` knows about how to compile for a specific target.