[WIP] Build a new incr comp session dir from scratch every time

Rather than copying the old incr comp dir and then modifying it. This
saves a copy/hardlink for files that are modified. And it removes the
need for accurate work product tracking to avoid accumulating cruft,
which is non-trivial. We don't accurately track the pre-LTO bitcode
files for ThinLTO for example.

FIXME cleanup

Author: bjorn3

compiler/rustc_codegen_cranelift/src/driver/aot.rs

@@ -140,7 +140,6 @@ fn emit_module(
         bytecode: None,
         assembly: None,
         llvm_ir: None,
-        links_from_incr_cache: Vec::new(),
     })
 }
 

compiler/rustc_codegen_llvm/src/back/lto.rs

@@ -463,23 +463,33 @@ fn thin_lto(
 
         info!("thin LTO data created");
 
-        let (key_map_path, prev_key_map, curr_key_map) = if let Some(ref incr_comp_session_dir) =
-            cgcx.incr_comp_session_dir
+        let new_key_map_path = cgcx
+            .new_incr_comp_session_dir
+            .as_ref()
+            .map(|dir| dir.join(THIN_LTO_KEYS_INCR_COMP_FILE_NAME));
+
+        let (prev_key_map, curr_key_map) = if let Some(ref old_incr_comp_session_dir) =
+            cgcx.old_incr_comp_session_dir
         {
-            let path = incr_comp_session_dir.join(THIN_LTO_KEYS_INCR_COMP_FILE_NAME);
+            let old_path = old_incr_comp_session_dir.join(THIN_LTO_KEYS_INCR_COMP_FILE_NAME);
+
             // If the previous file was deleted, or we get an IO error
             // reading the file, then we'll just use `None` as the
             // prev_key_map, which will force the code to be recompiled.
-            let prev =
-                if path.exists() { ThinLTOKeysMap::load_from_file(&path).ok() } else { None };
+            let prev = if old_path.exists() {
+                ThinLTOKeysMap::load_from_file(&old_path).ok()
+            } else {
+                None
+            };
             let curr = ThinLTOKeysMap::from_thin_lto_modules(&data, &thin_modules, &module_names);
-            (Some(path), prev, curr)
+
+            (prev, curr)
         } else {
             // If we don't compile incrementally, we don't need to load the
             // import data from LLVM.
             assert!(green_modules.is_empty());
             let curr = ThinLTOKeysMap::default();
-            (None, None, curr)
+            (None, curr)
         };
         info!("thin LTO cache key map loaded");
         info!("prev_key_map: {:#?}", prev_key_map);
@@ -500,15 +510,15 @@ fn thin_lto(
             if let (Some(prev_key_map), true) =
                 (prev_key_map.as_ref(), green_modules.contains_key(module_name))
             {
-                assert!(cgcx.incr_comp_session_dir.is_some());
+                assert!(cgcx.old_incr_comp_session_dir.is_some());
+                assert!(cgcx.new_incr_comp_session_dir.is_some());
 
                 // If a module exists in both the current and the previous session,
                 // and has the same LTO cache key in both sessions, then we can re-use it
                 if prev_key_map.keys.get(module_name) == curr_key_map.keys.get(module_name) {
                     let work_product = green_modules[module_name].clone();
                     copy_jobs.push(work_product);
                     info!(" - {}: re-used", module_name);
-                    assert!(cgcx.incr_comp_session_dir.is_some());
                     continue;
                 }
             }
@@ -518,8 +528,8 @@ fn thin_lto(
         }
 
         // Save the current ThinLTO import information for the next compilation
-        // session, overwriting the previous serialized data (if any).
-        if let Some(path) = key_map_path
+        // session.
+        if let Some(path) = new_key_map_path
             && let Err(err) = curr_key_map.save_to_file(&path)
         {
             write::llvm_err(dcx, LlvmError::WriteThinLtoKey { err });

compiler/rustc_codegen_ssa/src/back/write.rs

@@ -15,7 +15,9 @@ use rustc_errors::{
 };
 use rustc_fs_util::link_or_copy;
 use rustc_hir::find_attr;
-use rustc_incremental::{copy_cgu_workproduct_to_incr_comp_cache_dir, in_incr_comp_dir_sess};
+use rustc_incremental::{
+    copy_cgu_workproduct_to_incr_comp_cache_dir, in_incr_comp_dir_sess, in_old_incr_comp_dir_sess,
+};
 use rustc_macros::{Decodable, Encodable};
 use rustc_metadata::fs::copy_to_stdout;
 use rustc_middle::bug;
@@ -349,9 +351,12 @@ pub struct CodegenContext {
     /// Directory into which should the LLVM optimization remarks be written.
     /// If `None`, they will be written to stderr.
     pub remark_dir: Option<PathBuf>,
+    /// The previous incremental compilation session directory, or None if we
+    /// are not compiling incrementally or there is no previous session.
+    pub old_incr_comp_session_dir: Option<PathBuf>,
     /// The incremental compilation session directory, or None if we are not
     /// compiling incrementally
-    pub incr_comp_session_dir: Option<PathBuf>,
+    pub new_incr_comp_session_dir: Option<PathBuf>,
     /// `true` if the codegen should be run in parallel.
     ///
     /// Depends on [`WriteBackendMethods::supports_parallel()`] and `-Zno_parallel_backend`.
@@ -498,7 +503,6 @@ fn copy_all_cgu_workproducts_to_incr_comp_cache_dir(
             incr_comp_session.unwrap(),
             &module.name,
             files.as_slice(),
-            &module.links_from_incr_cache,
         );
         work_products.insert(id, product);
     }
@@ -840,7 +844,7 @@ fn execute_optimize_work_item<B: WriteBackendMethods>(
     // save our module to disk first.
     let bitcode = if cgcx.module_config.emit_pre_lto_bc {
         let filename = pre_lto_bitcode_filename(&module.name);
-        cgcx.incr_comp_session_dir.as_ref().map(|path| path.join(&filename))
+        cgcx.new_incr_comp_session_dir.as_ref().map(|path| path.join(&filename))
     } else {
         None
     };
@@ -887,11 +891,9 @@ fn execute_copy_from_cache_work_item(
     let dcx = DiagCtxt::new(Box::new(shared_emitter));
     let dcx = dcx.handle();
 
-    let incr_comp_session_dir = cgcx.incr_comp_session_dir.as_ref().unwrap();
-
-    let mut links_from_incr_cache = Vec::new();
+    let incr_comp_session_dir = cgcx.old_incr_comp_session_dir.as_ref().unwrap();
 
-    let mut load_from_incr_comp_dir = |output_path: PathBuf, saved_path: &str| {
+    let load_from_incr_comp_dir = |output_path: PathBuf, saved_path: &str| {
         let source_file_in_incr_comp_dir = incr_comp_session_dir.join(saved_path);
         debug!(
             "copying preexisting module `{}` from {:?} to {}",
@@ -900,10 +902,7 @@ fn execute_copy_from_cache_work_item(
             output_path.display()
         );
         match link_or_copy(&source_file_in_incr_comp_dir, &output_path) {
-            Ok(_) => {
-                links_from_incr_cache.push(source_file_in_incr_comp_dir);
-                Some(output_path)
-            }
+            Ok(_) => Some(output_path),
             Err(error) => {
                 dcx.emit_err(diagnostics::CopyPathBuf {
                     source_file: source_file_in_incr_comp_dir,
@@ -926,7 +925,7 @@ fn execute_copy_from_cache_work_item(
             load_from_incr_comp_dir(dwarf_obj_out, saved_dwarf_object_file)
         });
 
-    let mut load_from_incr_cache = |perform, output_type: OutputType| {
+    let load_from_incr_cache = |perform, output_type: OutputType| {
         if perform {
             let saved_file = module.source.saved_files.get(output_type.extension())?;
             let output_path = cgcx.output_filenames.temp_path_for_cgu(output_type, &module.name);
@@ -954,7 +953,6 @@ fn execute_copy_from_cache_work_item(
     }
 
     CompiledModule {
-        links_from_incr_cache,
         kind: ModuleKind::Regular,
         name: module.name,
         object,
@@ -1288,10 +1286,14 @@ fn start_executing_work<B: WriteBackendMethods>(
         time_trace: sess.opts.unstable_opts.llvm_time_trace,
         remark: sess.opts.cg.remark.clone(),
         remark_dir,
-        incr_comp_session_dir: tcx
+        old_incr_comp_session_dir: tcx
+            .incr_comp_session
+            .as_ref()
+            .and_then(|incr_comp_session| incr_comp_session.old_session_directory.clone()),
+        new_incr_comp_session_dir: tcx
             .incr_comp_session
             .as_ref()
-            .map(|incr_comp_session| incr_comp_session.session_directory.clone()),
+            .map(|incr_comp_session| incr_comp_session.new_session_directory.clone()),
         output_filenames: Arc::clone(tcx.output_filenames(())),
         module_config: regular_config,
         opt_level,
@@ -2262,13 +2264,28 @@ pub(crate) fn submit_pre_lto_module_to_llvm<B: WriteBackendMethods>(
     module: CachedModuleCodegen,
 ) {
     let filename = pre_lto_bitcode_filename(&module.name);
+    let old_bitcode_path =
+        in_old_incr_comp_dir_sess(tcx.incr_comp_session.unwrap(), &filename).unwrap();
     let bitcode_path = in_incr_comp_dir_sess(tcx.incr_comp_session.unwrap(), &filename);
+
+    // FIXME is this necessary?
+    // match link_or_copy(&old_bitcode_path, &bitcode_path) {
+    //     Ok(_) => {}
+    //     Err(error) => {
+    //         tcx.sess.dcx().emit_err(diagnostics::CopyPathBuf {
+    //             source_file: old_bitcode_path,
+    //             output_path: bitcode_path,
+    //             error,
+    //         });
+    //         return;
+    //     }
+    // }
+
     // Schedule the module to be loaded
-    drop(
-        coordinator
-            .sender
-            .send(Message::AddImportOnlyModule::<B> { bitcode_path, work_product: module.source }),
-    );
+    drop(coordinator.sender.send(Message::AddImportOnlyModule::<B> {
+        bitcode_path: old_bitcode_path,
+        work_product: module.source,
+    }));
 }
 
 fn pre_lto_bitcode_filename(module_name: &str) -> String {

compiler/rustc_codegen_ssa/src/lib.rs

@@ -115,7 +115,6 @@ impl<M> ModuleCodegen<M> {
             bytecode,
             assembly,
             llvm_ir,
-            links_from_incr_cache: Vec::new(),
         }
     }
 }
@@ -130,7 +129,6 @@ pub struct CompiledModule {
     pub bytecode: Option<PathBuf>,
     pub assembly: Option<PathBuf>, // --emit=asm
     pub llvm_ir: Option<PathBuf>,  // --emit=llvm-ir, llvm-bc is in bytecode
-    pub links_from_incr_cache: Vec<PathBuf>,
 }
 
 impl CompiledModule {

compiler/rustc_incremental/src/diagnostics.rs

@@ -169,21 +169,6 @@ pub(crate) struct DeleteLock<'a> {
     pub err: std::io::Error,
 }
 
-#[derive(Diagnostic)]
-#[diag(
-    "hard linking files in the incremental compilation cache failed. copying files instead. consider moving the cache directory to a file system which supports hard linking in session dir `{$path}`"
-)]
-pub(crate) struct HardLinkFailed<'a> {
-    pub path: &'a Path,
-}
-
-#[derive(Diagnostic)]
-#[diag("failed to delete partly initialized session dir `{$path}`: {$err}")]
-pub(crate) struct DeletePartial<'a> {
-    pub path: &'a Path,
-    pub err: std::io::Error,
-}
-
 #[derive(Diagnostic)]
 #[diag("did not finalize incremental compilation session directory `{$path}`: {$err}")]
 #[help("the next build will not be able to reuse work from this compilation")]

compiler/rustc_incremental/src/lib.rs

@@ -11,7 +11,7 @@ mod persist;
 
 pub use persist::{
     copy_cgu_workproduct_to_incr_comp_cache_dir, finalize_session_directory, in_incr_comp_dir_sess,
-    load_query_result_cache, save_work_product_index, setup_dep_graph,
+    in_old_incr_comp_dir_sess, load_query_result_cache, save_work_product_index, setup_dep_graph,
 };
 use rustc_middle::util::Providers;