Auto merge of #52266 - michaelwoerister:incr-thinlto-preliminaries, r…

…=alexcrichton

Preliminary work for incremental ThinLTO.

Since implementing incremental ThinLTO is a bit more involved than I initially thought, I'm splitting out some of the things that already work. This PR (1) adds a way accessing some ThinLTO information in `rustc` and (2) does some cleanup around CGU/object file naming (which makes things quite a bit nicer).

This is probably best reviewed one commit at a time.

src/librustc/mir/mono.rs

@@ -8,15 +8,16 @@
 // option. This file may not be copied, modified, or distributed
 // except according to those terms.
 
-use hir::def_id::DefId;
+use hir::def_id::{DefId, CrateNum};
 use syntax::ast::NodeId;
-use syntax::symbol::InternedString;
+use syntax::symbol::{Symbol, InternedString};
 use ty::{Instance, TyCtxt};
 use util::nodemap::FxHashMap;
 use rustc_data_structures::base_n;
 use rustc_data_structures::stable_hasher::{HashStable, StableHasherResult,
                                            StableHasher};
 use ich::{Fingerprint, StableHashingContext, NodeIdHashingMode};
+use std::fmt;
 use std::hash::Hash;
 
 #[derive(PartialEq, Eq, Clone, Copy, Debug, Hash)]
@@ -173,6 +174,80 @@ impl<'tcx> CodegenUnit<'tcx> {
             self.size_estimate = Some(size_estimate + delta);
         }
     }
+
+    /// CGU names should fulfill the following requirements:
+    /// - They should be able to act as a file name on any kind of file system
+    /// - They should not collide with other CGU names, even for different versions
+    ///   of the same crate.
+    ///
+    /// Consequently, we don't use special characters except for '.' and '-' and we
+    /// prefix each name with the crate-name and crate-disambiguator.
+    ///
+    /// This function will build CGU names of the form:
+    ///
+    /// ```
+    /// <crate-name>.<crate-disambiguator>(-<component>)*[.<special-suffix>]
+    /// ```
+    ///
+    /// The '.' before `<special-suffix>` makes sure that names with a special
+    /// suffix can never collide with a name built out of regular Rust
+    /// identifiers (e.g. module paths).
+    pub fn build_cgu_name<I, C, S>(tcx: TyCtxt,
+                                   cnum: CrateNum,
+                                   components: I,
+                                   special_suffix: Option<S>)
+                                   -> InternedString
+        where I: IntoIterator<Item=C>,
+              C: fmt::Display,
+              S: fmt::Display,
+    {
+        let cgu_name = CodegenUnit::build_cgu_name_no_mangle(tcx,
+                                                             cnum,
+                                                             components,
+                                                             special_suffix);
+
+        if tcx.sess.opts.debugging_opts.human_readable_cgu_names {
+            cgu_name
+        } else {
+            let cgu_name = &cgu_name.as_str()[..];
+            Symbol::intern(&CodegenUnit::mangle_name(cgu_name)).as_interned_str()
+        }
+    }
+
+    /// Same as `CodegenUnit::build_cgu_name()` but will never mangle the
+    /// resulting name.
+    pub fn build_cgu_name_no_mangle<I, C, S>(tcx: TyCtxt,
+                                             cnum: CrateNum,
+                                             components: I,
+                                             special_suffix: Option<S>)
+                                             -> InternedString
+        where I: IntoIterator<Item=C>,
+              C: fmt::Display,
+              S: fmt::Display,
+    {
+        use std::fmt::Write;
+
+        let mut cgu_name = String::with_capacity(64);
+
+        // Start out with the crate name and disambiguator
+        write!(cgu_name,
+               "{}.{}",
+               tcx.crate_name(cnum),
+               tcx.crate_disambiguator(cnum)).unwrap();
+
+        // Add the components
+        for component in components {
+            write!(cgu_name, "-{}", component).unwrap();
+        }
+
+        if let Some(special_suffix) = special_suffix {
+            // We add a dot in here so it cannot clash with anything in a regular
+            // Rust identifier
+            write!(cgu_name, ".{}", special_suffix).unwrap();
+        }
+
+        Symbol::intern(&cgu_name[..]).as_interned_str()
+    }
 }
 
 impl<'a, 'tcx> HashStable<StableHashingContext<'a>> for CodegenUnit<'tcx> {

src/librustc/session/mod.rs

@@ -26,6 +26,7 @@ use util::nodemap::{FxHashMap, FxHashSet};
 use util::common::{duration_to_secs_str, ErrorReported};
 use util::common::ProfileQueriesMsg;
 
+use rustc_data_structures::base_n;
 use rustc_data_structures::sync::{self, Lrc, Lock, LockCell, OneThread, Once, RwLock};
 
 use syntax::ast::NodeId;
@@ -1185,6 +1186,14 @@ impl CrateDisambiguator {
     }
 }
 
+impl fmt::Display for CrateDisambiguator {
+    fn fmt(&self, f: &mut fmt::Formatter) -> Result<(), fmt::Error> {
+        let (a, b) = self.0.as_value();
+        let as_u128 = a as u128 | ((b as u128) << 64);
+        f.write_str(&base_n::encode(as_u128, base_n::CASE_INSENSITIVE))
+    }
+}
+
 impl From<Fingerprint> for CrateDisambiguator {
     fn from(fingerprint: Fingerprint) -> CrateDisambiguator {
         CrateDisambiguator(fingerprint)

src/librustc_codegen_llvm/back/link.rs

@@ -45,13 +45,6 @@ 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};

src/librustc_codegen_llvm/back/lto.rs

@@ -20,16 +20,23 @@ use rustc::hir::def_id::LOCAL_CRATE;
 use rustc::middle::exported_symbols::SymbolExportLevel;
 use rustc::session::config::{self, Lto};
 use rustc::util::common::time_ext;
+use rustc_data_structures::fx::FxHashMap;
 use time_graph::Timeline;
 use {ModuleCodegen, ModuleLlvm, ModuleKind, ModuleSource};
 
 use libc;
 
-use std::ffi::CString;
+use std::ffi::{CString, CStr};
+use std::fs::File;
+use std::io;
+use std::mem;
+use std::path::Path;
 use std::ptr;
 use std::slice;
 use std::sync::Arc;
 
+pub const THIN_LTO_IMPORTS_INCR_COMP_FILE_NAME: &str = "thin-lto-imports.bin";
+
 pub fn crate_type_allows_lto(crate_type: config::CrateType) -> bool {
     match crate_type {
         config::CrateTypeExecutable |
@@ -193,7 +200,7 @@ pub(crate) fn run(cgcx: &CodegenContext,
         }
         Lto::Thin |
         Lto::ThinLocal => {
-            thin_lto(&diag_handler, modules, upstream_modules, &arr, timeline)
+            thin_lto(cgcx, &diag_handler, modules, upstream_modules, &arr, timeline)
         }
         Lto::No => unreachable!(),
     }
@@ -231,7 +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.llmod_id);
+    info!("using {:?} as a base module", module.name);
 
     // 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,
@@ -241,7 +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.llmod_id[..]).unwrap();
+        let llmod_id = CString::new(&module.name[..]).unwrap();
         serialized_modules.push((SerializedModule::Local(buffer), llmod_id));
     }
 
@@ -346,7 +353,8 @@ impl Drop for Linker {
 /// calculating the *index* for ThinLTO. This index will then be shared amongst
 /// all of the `LtoModuleCodegen` units returned below and destroyed once
 /// they all go out of scope.
-fn thin_lto(diag_handler: &Handler,
+fn thin_lto(cgcx: &CodegenContext,
+            diag_handler: &Handler,
             modules: Vec<ModuleCodegen>,
             serialized_modules: Vec<(SerializedModule, CString)>,
             symbol_white_list: &[*const libc::c_char],
@@ -368,9 +376,9 @@ fn thin_lto(diag_handler: &Handler,
         //        the most expensive portion of this small bit of global
         //        analysis!
         for (i, module) in modules.iter().enumerate() {
-            info!("local module: {} - {}", i, module.llmod_id);
+            info!("local module: {} - {}", i, module.name);
             let llvm = module.llvm().expect("can't lto precodegened module");
-            let name = CString::new(module.llmod_id.clone()).unwrap();
+            let name = CString::new(module.name.clone()).unwrap();
             let buffer = ThinBuffer::new(llvm.llmod);
             thin_modules.push(llvm::ThinLTOModule {
                 identifier: name.as_ptr(),
@@ -379,7 +387,7 @@ fn thin_lto(diag_handler: &Handler,
             });
             thin_buffers.push(buffer);
             module_names.push(name);
-            timeline.record(&module.llmod_id);
+            timeline.record(&module.name);
         }
 
         // FIXME: All upstream crates are deserialized internally in the
@@ -424,6 +432,18 @@ fn thin_lto(diag_handler: &Handler,
             let msg = format!("failed to prepare thin LTO context");
             return Err(write::llvm_err(&diag_handler, msg))
         }
+
+        // Save the ThinLTO import information for incremental compilation.
+        if let Some(ref incr_comp_session_dir) = cgcx.incr_comp_session_dir {
+            let path = incr_comp_session_dir.join(THIN_LTO_IMPORTS_INCR_COMP_FILE_NAME);
+            let imports = ThinLTOImports::from_thin_lto_data(data);
+            if let Err(err) = imports.save_to_file(&path) {
+                let msg = format!("Error while writing ThinLTO import data: {}",
+                                  err);
+                return Err(write::llvm_err(&diag_handler, msg));
+            }
+        }
+
         let data = ThinData(data);
         info!("thin LTO data created");
         timeline.record("data");
@@ -656,7 +676,6 @@ impl ThinModule {
                 llcx,
                 tm,
             }),
-            llmod_id: self.name().to_string(),
             name: self.name().to_string(),
             kind: ModuleKind::Regular,
         };
@@ -776,3 +795,117 @@ impl ThinModule {
         Ok(module)
     }
 }
+
+
+#[derive(Debug)]
+pub struct ThinLTOImports {
+    // key = llvm name of importing module, value = list of modules it imports from
+    imports: FxHashMap<String, Vec<String>>,
+}
+
+impl ThinLTOImports {
+
+    pub fn new() -> ThinLTOImports {
+        ThinLTOImports {
+            imports: FxHashMap(),
+        }
+    }
+
+    /// Load the ThinLTO import map from ThinLTOData.
+    unsafe fn from_thin_lto_data(data: *const llvm::ThinLTOData) -> ThinLTOImports {
+
+        fn module_name_to_str(c_str: &CStr) -> &str {
+            match c_str.to_str() {
+                Ok(s) => s,
+                Err(e) => {
+                    bug!("Encountered non-utf8 LLVM module name `{}`: {}",
+                        c_str.to_string_lossy(),
+                        e)
+                }
+            }
+        }
+
+        unsafe extern "C" fn imported_module_callback(payload: *mut libc::c_void,
+                                                      importing_module_name: *const libc::c_char,
+                                                      imported_module_name: *const libc::c_char) {
+            let map = &mut* (payload as *mut ThinLTOImports);
+
+            let importing_module_name = CStr::from_ptr(importing_module_name);
+            let importing_module_name = module_name_to_str(&importing_module_name);
+            let imported_module_name = CStr::from_ptr(imported_module_name);
+            let imported_module_name = module_name_to_str(&imported_module_name);
+
+            if !map.imports.contains_key(importing_module_name) {
+                map.imports.insert(importing_module_name.to_owned(), vec![]);
+            }
+
+            map.imports
+               .get_mut(importing_module_name)
+               .unwrap()
+               .push(imported_module_name.to_owned());
+        }
+
+        let mut map = ThinLTOImports {
+            imports: FxHashMap(),
+        };
+
+        llvm::LLVMRustGetThinLTOModuleImports(data,
+                                              imported_module_callback,
+                                              &mut map as *mut _ as *mut libc::c_void);
+        map
+    }
+
+    pub fn save_to_file(&self, path: &Path) -> io::Result<()> {
+        use std::io::Write;
+
+        let file = File::create(path)?;
+        let mut writer = io::BufWriter::new(file);
+
+        for (importing_module_name, imported_modules) in &self.imports {
+            writeln!(writer, "{}", importing_module_name)?;
+
+            for imported_module in imported_modules {
+                writeln!(writer, "  {}", imported_module)?;
+            }
+
+            writeln!(writer)?;
+        }
+
+        Ok(())
+    }
+
+    pub fn load_from_file(path: &Path) -> io::Result<ThinLTOImports> {
+        use std::io::BufRead;
+
+        let mut imports = FxHashMap();
+        let mut current_module = None;
+        let mut current_imports = vec![];
+
+        let file = File::open(path)?;
+
+        for line in io::BufReader::new(file).lines() {
+            let line = line?;
+
+            if line.is_empty() {
+                let importing_module = current_module
+                    .take()
+                    .expect("Importing module not set");
+
+                imports.insert(importing_module,
+                               mem::replace(&mut current_imports, vec![]));
+            } else if line.starts_with(" ") {
+                // This is an imported module
+                assert_ne!(current_module, None);
+                current_imports.push(line.trim().to_string());
+            } else {
+                // This is the beginning of a new module
+                assert_eq!(current_module, None);
+                current_module = Some(line.trim().to_string());
+            }
+        }
+
+        Ok(ThinLTOImports {
+            imports
+        })
+    }
+}

src/librustc_codegen_llvm/back/write.rs

@@ -696,7 +696,7 @@ unsafe fn codegen(cgcx: &CodegenContext,
 
         if config.emit_bc_compressed {
             let dst = bc_out.with_extension(RLIB_BYTECODE_EXTENSION);
-            let data = bytecode::encode(&module.llmod_id, data);
+            let data = bytecode::encode(&module.name, data);
             if let Err(e) = fs::write(&dst, data) {
                 diag_handler.err(&format!("failed to write bytecode: {}", e));
             }
@@ -1308,7 +1308,6 @@ 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,