Revert "Clean up LLVM module naming (just use CodegenUnit names)."

michaelwoerister · michaelwoerister · commit 0830cc92bd4c · 2018-07-16T08:58:29.000+02:00
This reverts commit f6894eb.
diff --git a/src/librustc_codegen_llvm/back/link.rs b/src/librustc_codegen_llvm/back/link.rs
@@ -45,6 +45,13 @@ use std::process::{Output, Stdio};
 use std::str;
 use syntax::attr;
 
+/// The LLVM module name containing crate-metadata. This includes a `.` on
+/// purpose, so it cannot clash with the name of a user-defined module.
+pub const METADATA_MODULE_NAME: &'static str = "crate.metadata";
+
+// same as for metadata above, but for allocator shim
+pub const ALLOCATOR_MODULE_NAME: &'static str = "crate.allocator";
+
 pub use rustc_codegen_utils::link::{find_crate_name, filename_for_input, default_output_for_target,
                                   invalid_output_for_target, build_link_meta, out_filename,
                                   check_file_is_writeable};
diff --git a/src/librustc_codegen_llvm/back/lto.rs b/src/librustc_codegen_llvm/back/lto.rs
@@ -238,7 +238,7 @@ fn fat_lto(cgcx: &CodegenContext,
         .expect("must be codegen'ing at least one module");
     let module = modules.remove(costliest_module);
     let llmod = module.llvm().expect("can't lto pre-codegened modules").llmod;
-    info!("using {:?} as a base module", module.name);
+    info!("using {:?} as a base module", module.llmod_id);
 
     // For all other modules we codegened we'll need to link them into our own
     // bitcode. All modules were codegened in their own LLVM context, however,
@@ -248,7 +248,7 @@ fn fat_lto(cgcx: &CodegenContext,
     for module in modules {
         let llvm = module.llvm().expect("can't lto pre-codegened modules");
         let buffer = ModuleBuffer::new(llvm.llmod);
-        let llmod_id = CString::new(&module.name[..]).unwrap();
+        let llmod_id = CString::new(&module.llmod_id[..]).unwrap();
         serialized_modules.push((SerializedModule::Local(buffer), llmod_id));
     }
 
@@ -376,9 +376,9 @@ fn thin_lto(cgcx: &CodegenContext,
         //        the most expensive portion of this small bit of global
         //        analysis!
         for (i, module) in modules.iter().enumerate() {
-            info!("local module: {} - {}", i, module.name);
+            info!("local module: {} - {}", i, module.llmod_id);
             let llvm = module.llvm().expect("can't lto precodegened module");
-            let name = CString::new(module.name.clone()).unwrap();
+            let name = CString::new(module.llmod_id.clone()).unwrap();
             let buffer = ThinBuffer::new(llvm.llmod);
             thin_modules.push(llvm::ThinLTOModule {
                 identifier: name.as_ptr(),
@@ -387,7 +387,7 @@ fn thin_lto(cgcx: &CodegenContext,
             });
             thin_buffers.push(buffer);
             module_names.push(name);
-            timeline.record(&module.name);
+            timeline.record(&module.llmod_id);
         }
 
         // FIXME: All upstream crates are deserialized internally in the
@@ -676,6 +676,7 @@ impl ThinModule {
                 llcx,
                 tm,
             }),
+            llmod_id: self.name().to_string(),
             name: self.name().to_string(),
             kind: ModuleKind::Regular,
         };
diff --git a/src/librustc_codegen_llvm/back/write.rs b/src/librustc_codegen_llvm/back/write.rs
@@ -694,7 +694,7 @@ unsafe fn codegen(cgcx: &CodegenContext,
 
         if config.emit_bc_compressed {
             let dst = bc_out.with_extension(RLIB_BYTECODE_EXTENSION);
-            let data = bytecode::encode(&module.name, data);
+            let data = bytecode::encode(&module.llmod_id, data);
             if let Err(e) = fs::write(&dst, data) {
                 diag_handler.err(&format!("failed to write bytecode: {}", e));
             }
@@ -1306,6 +1306,7 @@ fn execute_work_item(cgcx: &CodegenContext,
         assert_eq!(bytecode_compressed.is_some(), config.emit_bc_compressed);
 
         Ok(WorkItemResult::Compiled(CompiledModule {
+            llmod_id: module.llmod_id.clone(),
             name: module_name,
             kind: ModuleKind::Regular,
             pre_existing: true,
diff --git a/src/librustc_codegen_llvm/base.rs b/src/librustc_codegen_llvm/base.rs
@@ -739,18 +739,15 @@ pub fn codegen_crate<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
     let link_meta = link::build_link_meta(crate_hash);
 
     // Codegen the metadata.
-    let metadata_cgu_name = CodegenUnit::build_cgu_name(tcx,
-                                                        LOCAL_CRATE,
-                                                        &["crate"],
-                                                        Some("metadata")).as_str()
-                                                                         .to_string();
+    let llmod_id = "metadata";
     let (metadata_llcx, metadata_llmod, metadata) =
         time(tcx.sess, "write metadata", || {
-            write_metadata(tcx, &metadata_cgu_name, &link_meta)
+            write_metadata(tcx, llmod_id, &link_meta)
         });
 
     let metadata_module = ModuleCodegen {
-        name: metadata_cgu_name,
+        name: link::METADATA_MODULE_NAME.to_string(),
+        llmod_id: llmod_id.to_string(),
         source: ModuleSource::Codegened(ModuleLlvm {
             llcx: metadata_llcx,
             llmod: metadata_llmod,
@@ -813,30 +810,26 @@ pub fn codegen_crate<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
 
     // Codegen an allocator shim, if any
     let allocator_module = if let Some(kind) = *tcx.sess.allocator_kind.get() {
-        let llmod_id = CodegenUnit::build_cgu_name(tcx,
-                                                   LOCAL_CRATE,
-                                                   &["crate"],
-                                                   Some("allocator")).as_str()
-                                                                     .to_string();
-        let (llcx, llmod) = unsafe {
-            context::create_context_and_module(tcx.sess, &llmod_id)
-        };
-        let modules = ModuleLlvm {
-            llmod,
-            llcx,
-            tm: create_target_machine(tcx.sess, false),
-        };
-        time(tcx.sess, "write allocator module", || {
-            unsafe {
+        unsafe {
+            let llmod_id = "allocator";
+            let (llcx, llmod) =
+                context::create_context_and_module(tcx.sess, llmod_id);
+            let modules = ModuleLlvm {
+                llmod,
+                llcx,
+                tm: create_target_machine(tcx.sess, false),
+            };
+            time(tcx.sess, "write allocator module", || {
                 allocator::codegen(tcx, &modules, kind)
-            }
-        });
+            });
 
-        Some(ModuleCodegen {
-            name: llmod_id,
-            source: ModuleSource::Codegened(modules),
-            kind: ModuleKind::Allocator,
-        })
+            Some(ModuleCodegen {
+                name: link::ALLOCATOR_MODULE_NAME.to_string(),
+                llmod_id: llmod_id.to_string(),
+                source: ModuleSource::Codegened(modules),
+                kind: ModuleKind::Allocator,
+            })
+        }
     } else {
         None
     };
@@ -879,10 +872,21 @@ pub fn codegen_crate<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
                 // succeed it means that none of the dependencies has changed
                 // and we can safely re-use.
                 if let Some(dep_node_index) = tcx.dep_graph.try_mark_green(tcx, dep_node) {
+                    // Append ".rs" to LLVM module identifier.
+                    //
+                    // LLVM code generator emits a ".file filename" directive
+                    // for ELF backends. Value of the "filename" is set as the
+                    // LLVM module identifier.  Due to a LLVM MC bug[1], LLVM
+                    // crashes if the module identifier is same as other symbols
+                    // such as a function name in the module.
+                    // 1. http://llvm.org/bugs/show_bug.cgi?id=11479
+                    let llmod_id = format!("{}.rs", cgu.name());
+
                     let module = ModuleCodegen {
                         name: cgu.name().to_string(),
                         source: ModuleSource::Preexisting(buf),
                         kind: ModuleKind::Regular,
+                        llmod_id,
                     };
                     tcx.dep_graph.mark_loaded_from_cache(dep_node_index, true);
                     write::submit_codegened_module_to_llvm(tcx, module, 0);
@@ -1191,8 +1195,21 @@ fn compile_codegen_unit<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
     {
         let cgu_name = cgu.name().to_string();
 
+        // Append ".rs" to LLVM module identifier.
+        //
+        // LLVM code generator emits a ".file filename" directive
+        // for ELF backends. Value of the "filename" is set as the
+        // LLVM module identifier.  Due to a LLVM MC bug[1], LLVM
+        // crashes if the module identifier is same as other symbols
+        // such as a function name in the module.
+        // 1. http://llvm.org/bugs/show_bug.cgi?id=11479
+        let llmod_id = format!("{}-{}.rs",
+                               cgu.name(),
+                               tcx.crate_disambiguator(LOCAL_CRATE)
+                                   .to_fingerprint().to_hex());
+
         // Instantiate monomorphizations without filling out definitions yet...
-        let cx = CodegenCx::new(tcx, cgu);
+        let cx = CodegenCx::new(tcx, cgu, &llmod_id);
         let module = {
             let mono_items = cx.codegen_unit
                                  .items_in_deterministic_order(cx.tcx);
@@ -1250,6 +1267,7 @@ fn compile_codegen_unit<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>,
                 name: cgu_name,
                 source: ModuleSource::Codegened(llvm_module),
                 kind: ModuleKind::Regular,
+                llmod_id,
             }
         };
 
diff --git a/src/librustc_codegen_llvm/context.rs b/src/librustc_codegen_llvm/context.rs
@@ -214,7 +214,8 @@ pub unsafe fn create_context_and_module(sess: &Session, mod_name: &str) -> (Cont
 
 impl<'a, 'tcx> CodegenCx<'a, 'tcx> {
     pub fn new(tcx: TyCtxt<'a, 'tcx, 'tcx>,
-               codegen_unit: Arc<CodegenUnit<'tcx>>)
+               codegen_unit: Arc<CodegenUnit<'tcx>>,
+               llmod_id: &str)
                -> CodegenCx<'a, 'tcx> {
         // An interesting part of Windows which MSVC forces our hand on (and
         // apparently MinGW didn't) is the usage of `dllimport` and `dllexport`
@@ -267,7 +268,7 @@ impl<'a, 'tcx> CodegenCx<'a, 'tcx> {
 
         unsafe {
             let (llcx, llmod) = create_context_and_module(&tcx.sess,
-                                                          &codegen_unit.name().as_str());
+                                                          &llmod_id[..]);
 
             let dbg_cx = if tcx.sess.opts.debuginfo != NoDebugInfo {
                 let dctx = debuginfo::CrateDebugContext::new(llmod);
diff --git a/src/librustc_codegen_llvm/lib.rs b/src/librustc_codegen_llvm/lib.rs
@@ -258,8 +258,8 @@ struct ModuleCodegen {
     /// unique amongst **all** crates.  Therefore, it should contain
     /// something unique to this crate (e.g., a module path) as well
     /// as the crate name and disambiguator.
-    /// We currently generate these names via CodegenUnit::build_cgu_name().
     name: String,
+    llmod_id: String,
     source: ModuleSource,
     kind: ModuleKind,
 }
@@ -306,6 +306,7 @@ impl ModuleCodegen {
         };
 
         CompiledModule {
+            llmod_id: self.llmod_id,
             name: self.name.clone(),
             kind: self.kind,
             pre_existing,
@@ -319,6 +320,7 @@ impl ModuleCodegen {
 #[derive(Debug)]
 struct CompiledModule {
     name: String,
+    llmod_id: String,
     kind: ModuleKind,
     pre_existing: bool,
     object: Option<PathBuf>,
