Auto merge of #128936 - bjorn3:fix_thin_archive_reading, r=jieyouxu

Support reading thin archives in ArArchiveBuilder

And switch to using ArArchiveBuilder with the LLVM backend too now that all regressions are fixed.

Fixes #107407
Fixes #107162
#107495 has been fixed in a previous PR already.

compiler/rustc_codegen_llvm/src/back/archive.rs

@@ -106,9 +106,11 @@ pub struct LlvmArchiveBuilderBuilder;
 
 impl ArchiveBuilderBuilder for LlvmArchiveBuilderBuilder {
     fn new_archive_builder<'a>(&self, sess: &'a Session) -> Box<dyn ArchiveBuilder + 'a> {
-        // FIXME use ArArchiveBuilder on most targets again once reading thin archives is
-        // implemented
-        if true {
+        // Keeping LlvmArchiveBuilder around in case of a regression caused by using
+        // ArArchiveBuilder.
+        // FIXME(#128955) remove a couple of months after #128936 gets merged in case
+        // no regression is found.
+        if false {
             Box::new(LlvmArchiveBuilder { sess, additions: Vec::new() })
         } else {
             Box::new(ArArchiveBuilder::new(sess, &LLVM_OBJECT_READER))
@@ -198,25 +200,11 @@ static LLVM_OBJECT_READER: ObjectReader = ObjectReader {
     get_xcoff_member_alignment: DEFAULT_OBJECT_READER.get_xcoff_member_alignment,
 };
 
-fn should_use_llvm_reader(buf: &[u8]) -> bool {
-    let is_bitcode = unsafe { llvm::LLVMRustIsBitcode(buf.as_ptr(), buf.len()) };
-
-    // COFF bigobj file, msvc LTO file or import library. See
-    // https://github.com/llvm/llvm-project/blob/453f27bc9/llvm/lib/BinaryFormat/Magic.cpp#L38-L51
-    let is_unsupported_windows_obj_file = buf.get(0..4) == Some(b"\0\0\xFF\xFF");
-
-    is_bitcode || is_unsupported_windows_obj_file
-}
-
 #[deny(unsafe_op_in_unsafe_fn)]
 fn get_llvm_object_symbols(
     buf: &[u8],
     f: &mut dyn FnMut(&[u8]) -> io::Result<()>,
 ) -> io::Result<bool> {
-    if !should_use_llvm_reader(buf) {
-        return (DEFAULT_OBJECT_READER.get_symbols)(buf, f);
-    }
-
     let mut state = Box::new(f);
 
     let err = unsafe {
@@ -253,18 +241,10 @@ fn get_llvm_object_symbols(
 }
 
 fn llvm_is_64_bit_object_file(buf: &[u8]) -> bool {
-    if !should_use_llvm_reader(buf) {
-        return (DEFAULT_OBJECT_READER.is_64_bit_object_file)(buf);
-    }
-
     unsafe { llvm::LLVMRustIs64BitSymbolicFile(buf.as_ptr(), buf.len()) }
 }
 
 fn llvm_is_ec_object_file(buf: &[u8]) -> bool {
-    if !should_use_llvm_reader(buf) {
-        return (DEFAULT_OBJECT_READER.is_ec_object_file)(buf);
-    }
-
     unsafe { llvm::LLVMRustIsECObject(buf.as_ptr(), buf.len()) }
 }
 

compiler/rustc_codegen_ssa/src/back/archive.rs

@@ -307,10 +307,15 @@ impl<'a> ArchiveBuilder for ArArchiveBuilder<'a> {
             let file_name = String::from_utf8(entry.name().to_vec())
                 .map_err(|err| io::Error::new(io::ErrorKind::InvalidData, err))?;
             if !skip(&file_name) {
-                self.entries.push((
-                    file_name.into_bytes(),
-                    ArchiveEntry::FromArchive { archive_index, file_range: entry.file_range() },
-                ));
+                if entry.is_thin() {
+                    let member_path = archive_path.parent().unwrap().join(Path::new(&file_name));
+                    self.entries.push((file_name.into_bytes(), ArchiveEntry::File(member_path)));
+                } else {
+                    self.entries.push((
+                        file_name.into_bytes(),
+                        ArchiveEntry::FromArchive { archive_index, file_range: entry.file_range() },
+                    ));
+                }
             }
         }
 

compiler/rustc_llvm/llvm-wrapper/SymbolWrapper.cpp

@@ -77,22 +77,18 @@ LLVMRustGetSymbols(char *BufPtr, size_t BufLen, void *State,
   Expected<std::unique_ptr<object::SymbolicFile>> ObjOrErr =
       getSymbolicFile(Buf->getMemBufferRef(), Context);
   if (!ObjOrErr) {
-    Error E = ObjOrErr.takeError();
-    SmallString<0> ErrorBuf;
-    auto Error = raw_svector_ostream(ErrorBuf);
-    Error << E << '\0';
-    return ErrorCallback(Error.str().data());
+    return ErrorCallback(toString(ObjOrErr.takeError()).c_str());
   }
   std::unique_ptr<object::SymbolicFile> Obj = std::move(*ObjOrErr);
+  if (Obj == nullptr) {
+    return 0;
+  }
 
   for (const object::BasicSymbolRef &S : Obj->symbols()) {
     if (!isArchiveSymbol(S))
       continue;
     if (Error E = S.printName(SymName)) {
-      SmallString<0> ErrorBuf;
-      auto Error = raw_svector_ostream(ErrorBuf);
-      Error << E << '\0';
-      return ErrorCallback(Error.str().data());
+      return ErrorCallback(toString(std::move(E)).c_str());
     }
     SymName << '\0';
     if (void *E = Callback(State, SymNameBuf.str().data())) {

src/tools/run-make-support/src/external_deps/llvm.rs

@@ -285,6 +285,12 @@ impl LlvmAr {
         self
     }
 
+    /// Like `obj_to_ar` except creating a thin archive.
+    pub fn obj_to_thin_ar(&mut self) -> &mut Self {
+        self.cmd.arg("rcus").arg("--thin");
+        self
+    }
+
     /// Extract archive members back to files.
     pub fn extract(&mut self) -> &mut Self {
         self.cmd.arg("x");

tests/run-make/staticlib-thin-archive/bin.rs

@@ -0,0 +1,5 @@
+fn main() {
+    unsafe {
+        rust_lib::simple_fn();
+    }
+}

tests/run-make/staticlib-thin-archive/rmake.rs

@@ -0,0 +1,23 @@
+// Regression test for https://github.com/rust-lang/rust/issues/107407 which
+// checks that rustc can read thin archive. Before the object crate added thin
+// archive support rustc would add emit object files to the staticlib and after
+// the object crate added thin archive support it would previously crash the
+// compiler due to a missing special case for thin archive members.
+use run_make_support::{llvm_ar, path, rfs, rust_lib_name, rustc, static_lib_name};
+
+fn main() {
+    rfs::create_dir("archive");
+
+    // Build a thin archive
+    rustc().input("simple_obj.rs").emit("obj").output("archive/simple_obj.o").run();
+    llvm_ar()
+        .obj_to_thin_ar()
+        .output_input(path("archive").join(static_lib_name("thin_archive")), "archive/simple_obj.o")
+        .run();
+
+    // Build an rlib which includes the members of this thin archive
+    rustc().input("rust_lib.rs").library_search_path("archive").run();
+
+    // Build a binary which requires a symbol from the thin archive
+    rustc().input("bin.rs").extern_("rust_lib", rust_lib_name("rust_lib")).run();
+}

tests/run-make/staticlib-thin-archive/rust_lib.rs

@@ -0,0 +1,6 @@
+#![crate_type = "rlib"]
+
+#[link(name = "thin_archive", kind = "static")]
+extern "C" {
+    pub fn simple_fn();
+}

tests/run-make/staticlib-thin-archive/simple_obj.rs

@@ -0,0 +1,4 @@
+#![crate_type = "staticlib"]
+
+#[no_mangle]
+extern "C" fn simple_fn() {}