Auto merge of #132514 - Zalathar:print-target-cpus, r=jieyouxu

Port most of `--print=target-cpus` to Rust

The logic and formatting needed by `--print=target-cpus` has historically been carried out in C++ code. Originally it used `printf` to write directly to the console, but later it switched over to writing to a `std::ostringstream` and then passing its buffer to a callback function pointer.

This PR replaces that C++ code with a very simple function that writes a list of CPU names to a `&RustString`, with the rest of the logic and formatting being handled by ordinary safe Rust code.

compiler/rustc_codegen_llvm/src/llvm/ffi.rs

@@ -2190,12 +2190,8 @@ unsafe extern "C" {
 
     pub fn LLVMRustHasFeature(T: &TargetMachine, s: *const c_char) -> bool;
 
-    pub fn LLVMRustPrintTargetCPUs(
-        T: &TargetMachine,
-        cpu: *const c_char,
-        print: unsafe extern "C" fn(out: *mut c_void, string: *const c_char, len: usize),
-        out: *mut c_void,
-    );
+    #[allow(improper_ctypes)]
+    pub(crate) fn LLVMRustPrintTargetCPUs(TM: &TargetMachine, OutStr: &RustString);
     pub fn LLVMRustGetTargetFeaturesCount(T: &TargetMachine) -> size_t;
     pub fn LLVMRustGetTargetFeature(
         T: &TargetMachine,
@@ -2204,7 +2200,7 @@ unsafe extern "C" {
         Desc: &mut *const c_char,
     );
 
-    pub fn LLVMRustGetHostCPUName(len: *mut usize) -> *const c_char;
+    pub fn LLVMRustGetHostCPUName(LenOut: &mut size_t) -> *const u8;
 
     // This function makes copies of pointed to data, so the data's lifetime may end after this
     // function returns.

compiler/rustc_codegen_llvm/src/llvm_util.rs

@@ -1,4 +1,5 @@
-use std::ffi::{CStr, CString, c_char, c_void};
+use std::collections::VecDeque;
+use std::ffi::{CStr, CString};
 use std::fmt::Write;
 use std::path::Path;
 use std::sync::Once;
@@ -387,7 +388,65 @@ fn llvm_target_features(tm: &llvm::TargetMachine) -> Vec<(&str, &str)> {
     ret
 }
 
-fn print_target_features(out: &mut String, sess: &Session, tm: &llvm::TargetMachine) {
+pub(crate) fn print(req: &PrintRequest, out: &mut String, sess: &Session) {
+    require_inited();
+    let tm = create_informational_target_machine(sess, false);
+    match req.kind {
+        PrintKind::TargetCPUs => print_target_cpus(sess, &tm, out),
+        PrintKind::TargetFeatures => print_target_features(sess, &tm, out),
+        _ => bug!("rustc_codegen_llvm can't handle print request: {:?}", req),
+    }
+}
+
+fn print_target_cpus(sess: &Session, tm: &llvm::TargetMachine, out: &mut String) {
+    let cpu_names = llvm::build_string(|s| unsafe {
+        llvm::LLVMRustPrintTargetCPUs(&tm, s);
+    })
+    .unwrap();
+
+    struct Cpu<'a> {
+        cpu_name: &'a str,
+        remark: String,
+    }
+    // Compare CPU against current target to label the default.
+    let target_cpu = handle_native(&sess.target.cpu);
+    let make_remark = |cpu_name| {
+        if cpu_name == target_cpu {
+            // FIXME(#132514): This prints the LLVM target string, which can be
+            // different from the Rust target string. Is that intended?
+            let target = &sess.target.llvm_target;
+            format!(
+                " - This is the default target CPU for the current build target (currently {target})."
+            )
+        } else {
+            "".to_owned()
+        }
+    };
+    let mut cpus = cpu_names
+        .lines()
+        .map(|cpu_name| Cpu { cpu_name, remark: make_remark(cpu_name) })
+        .collect::<VecDeque<_>>();
+
+    // Only print the "native" entry when host and target are the same arch,
+    // since otherwise it could be wrong or misleading.
+    if sess.host.arch == sess.target.arch {
+        let host = get_host_cpu_name();
+        cpus.push_front(Cpu {
+            cpu_name: "native",
+            remark: format!(" - Select the CPU of the current host (currently {host})."),
+        });
+    }
+
+    let max_name_width = cpus.iter().map(|cpu| cpu.cpu_name.len()).max().unwrap_or(0);
+    writeln!(out, "Available CPUs for this target:").unwrap();
+    for Cpu { cpu_name, remark } in cpus {
+        // Only pad the CPU name if there's a remark to print after it.
+        let width = if remark.is_empty() { 0 } else { max_name_width };
+        writeln!(out, "    {cpu_name:<width$}{remark}").unwrap();
+    }
+}
+
+fn print_target_features(sess: &Session, tm: &llvm::TargetMachine, out: &mut String) {
     let mut llvm_target_features = llvm_target_features(tm);
     let mut known_llvm_target_features = FxHashSet::<&'static str>::default();
     let mut rustc_target_features = sess
@@ -447,52 +506,31 @@ fn print_target_features(out: &mut String, sess: &Session, tm: &llvm::TargetMach
     writeln!(out, "and may be renamed or removed in a future version of LLVM or rustc.\n").unwrap();
 }
 
-pub(crate) fn print(req: &PrintRequest, mut out: &mut String, sess: &Session) {
-    require_inited();
-    let tm = create_informational_target_machine(sess, false);
-    match req.kind {
-        PrintKind::TargetCPUs => {
-            // SAFETY generate a C compatible string from a byte slice to pass
-            // the target CPU name into LLVM, the lifetime of the reference is
-            // at least as long as the C function
-            let cpu_cstring = CString::new(handle_native(sess.target.cpu.as_ref()))
-                .unwrap_or_else(|e| bug!("failed to convert to cstring: {}", e));
-            unsafe extern "C" fn callback(out: *mut c_void, string: *const c_char, len: usize) {
-                let out = unsafe { &mut *(out as *mut &mut String) };
-                let bytes = unsafe { slice::from_raw_parts(string as *const u8, len) };
-                write!(out, "{}", String::from_utf8_lossy(bytes)).unwrap();
-            }
-            unsafe {
-                llvm::LLVMRustPrintTargetCPUs(
-                    &tm,
-                    cpu_cstring.as_ptr(),
-                    callback,
-                    (&raw mut out) as *mut c_void,
-                );
-            }
-        }
-        PrintKind::TargetFeatures => print_target_features(out, sess, &tm),
-        _ => bug!("rustc_codegen_llvm can't handle print request: {:?}", req),
-    }
-}
-
-fn handle_native(name: &str) -> &str {
-    if name != "native" {
-        return name;
-    }
-
-    unsafe {
+/// Returns the host CPU name, according to LLVM.
+fn get_host_cpu_name() -> &'static str {
     let mut len = 0;
+    // SAFETY: The underlying C++ global function returns a `StringRef` that
+    // isn't tied to any particular backing buffer, so it must be 'static.
+    let slice: &'static [u8] = unsafe {
         let ptr = llvm::LLVMRustGetHostCPUName(&mut len);
-        str::from_utf8(slice::from_raw_parts(ptr as *const u8, len)).unwrap()
+        assert!(!ptr.is_null());
+        slice::from_raw_parts(ptr, len)
+    };
+    str::from_utf8(slice).expect("host CPU name should be UTF-8")
+}
+
+/// If the given string is `"native"`, returns the host CPU name according to
+/// LLVM. Otherwise, the string is returned as-is.
+fn handle_native(cpu_name: &str) -> &str {
+    match cpu_name {
+        "native" => get_host_cpu_name(),
+        _ => cpu_name,
     }
 }
 
 pub(crate) fn target_cpu(sess: &Session) -> &str {
-    match sess.opts.cg.target_cpu {
-        Some(ref name) => handle_native(name),
-        None => handle_native(sess.target.cpu.as_ref()),
-    }
+    let cpu_name = sess.opts.cg.target_cpu.as_deref().unwrap_or_else(|| &sess.target.cpu);
+    handle_native(cpu_name)
 }
 
 /// The list of LLVM features computed from CLI flags (`-Ctarget-cpu`, `-Ctarget-feature`,

compiler/rustc_llvm/llvm-wrapper/PassWrapper.cpp

@@ -317,49 +317,17 @@ template <typename KV> static size_t getLongestEntryLength(ArrayRef<KV> Table) {
   return MaxLen;
 }
 
-using PrintBackendInfo = void(void *, const char *Data, size_t Len);
-
 extern "C" void LLVMRustPrintTargetCPUs(LLVMTargetMachineRef TM,
-                                        const char *TargetCPU,
-                                        PrintBackendInfo Print, void *Out) {
-  const TargetMachine *Target = unwrap(TM);
-  const Triple::ArchType HostArch =
-      Triple(sys::getDefaultTargetTriple()).getArch();
-  const Triple::ArchType TargetArch = Target->getTargetTriple().getArch();
+                                        RustStringRef OutStr) {
+  ArrayRef<SubtargetSubTypeKV> CPUTable =
+      unwrap(TM)->getMCSubtargetInfo()->getAllProcessorDescriptions();
+  auto OS = RawRustStringOstream(OutStr);
 
-  std::ostringstream Buf;
-
-  const MCSubtargetInfo *MCInfo = Target->getMCSubtargetInfo();
-  const ArrayRef<SubtargetSubTypeKV> CPUTable =
-      MCInfo->getAllProcessorDescriptions();
-  unsigned MaxCPULen = getLongestEntryLength(CPUTable);
-
-  Buf << "Available CPUs for this target:\n";
-  // Don't print the "native" entry when the user specifies --target with a
-  // different arch since that could be wrong or misleading.
-  if (HostArch == TargetArch) {
-    MaxCPULen = std::max(MaxCPULen, (unsigned)std::strlen("native"));
-    const StringRef HostCPU = sys::getHostCPUName();
-    Buf << "    " << std::left << std::setw(MaxCPULen) << "native"
-        << " - Select the CPU of the current host "
-           "(currently "
-        << HostCPU.str() << ").\n";
-  }
+  // Just print a bare list of target CPU names, and let Rust-side code handle
+  // the full formatting of `--print=target-cpus`.
   for (auto &CPU : CPUTable) {
-    // Compare cpu against current target to label the default
-    if (strcmp(CPU.Key, TargetCPU) == 0) {
-      Buf << "    " << std::left << std::setw(MaxCPULen) << CPU.Key
-          << " - This is the default target CPU for the current build target "
-             "(currently "
-          << Target->getTargetTriple().str() << ").";
-    } else {
-      Buf << "    " << CPU.Key;
+    OS << CPU.Key << "\n";
   }
-    Buf << "\n";
-  }
-
-  const auto &BufString = Buf.str();
-  Print(Out, BufString.data(), BufString.size());
 }
 
 extern "C" size_t LLVMRustGetTargetFeaturesCount(LLVMTargetMachineRef TM) {
@@ -382,9 +350,9 @@ extern "C" void LLVMRustGetTargetFeature(LLVMTargetMachineRef TM, size_t Index,
   *Desc = Feat.Desc;
 }
 
-extern "C" const char *LLVMRustGetHostCPUName(size_t *len) {
+extern "C" const char *LLVMRustGetHostCPUName(size_t *OutLen) {
   StringRef Name = sys::getHostCPUName();
-  *len = Name.size();
+  *OutLen = Name.size();
   return Name.data();
 }
 

tests/run-make/print-target-cpus-native/rmake.rs

@@ -0,0 +1,39 @@
+//@ ignore-cross-compile
+//@ needs-llvm-components: aarch64 x86
+// FIXME(#132514): Is needs-llvm-components actually necessary for this test?
+
+use run_make_support::{assert_contains_regex, rfs, rustc, target};
+
+// Test that when querying `--print=target-cpus` for a target with the same
+// architecture as the host, the first CPU is "native" with a suitable remark.
+
+fn main() {
+    let expected = r"^Available CPUs for this target:
+    native +- Select the CPU of the current host \(currently [^ )]+\)\.
+";
+
+    // Without an explicit target.
+    rustc().print("target-cpus").run().assert_stdout_contains_regex(expected);
+
+    // With an explicit target that happens to be the host.
+    let host = target(); // Because of ignore-cross-compile, assume host == target.
+    rustc().print("target-cpus").target(host).run().assert_stdout_contains_regex(expected);
+
+    // With an explicit output path.
+    rustc().print("target-cpus=./xyzzy.txt").run().assert_stdout_equals("");
+    assert_contains_regex(rfs::read_to_string("./xyzzy.txt"), expected);
+
+    // Now try some cross-target queries with the same arch as the host.
+    // (Specify multiple targets so that at least one of them is not the host.)
+    let cross_targets: &[&str] = if cfg!(target_arch = "aarch64") {
+        &["aarch64-unknown-linux-gnu", "aarch64-apple-darwin"]
+    } else if cfg!(target_arch = "x86_64") {
+        &["x86_64-unknown-linux-gnu", "x86_64-apple-darwin"]
+    } else {
+        &[]
+    };
+    for target in cross_targets {
+        println!("Trying target: {target}");
+        rustc().print("target-cpus").target(target).run().assert_stdout_contains_regex(expected);
+    }
+}