Rollup merge of rust-lang#90701 - michaelwoerister:more-artifact-sizes, r=davidtwco

Record more artifact sizes during self-profiling.

This PR adds artifact size recording for

- "linked artifacts" (executables, RLIBs, dylibs, static libs)
- object files
- dwo files
- assembly files
- crate metadata
- LLVM bitcode files
- LLVM IR files
- codegen unit size estimates

Currently the identifiers emitted for these are hard-coded as string literals. Is it worth adding constants to https://github.com/rust-lang/measureme/blob/master/measureme/src/rustc.rs instead? We don't do that for query names and the like -- but artifact kinds might be more stable than query names.

Author: matthiaskrgr

compiler/rustc_codegen_llvm/src/back/write.rs

@@ -17,6 +17,7 @@ use rustc_codegen_ssa::back::write::{
 };
 use rustc_codegen_ssa::traits::*;
 use rustc_codegen_ssa::{CompiledModule, ModuleCodegen};
+use rustc_data_structures::profiling::SelfProfilerRef;
 use rustc_data_structures::small_c_str::SmallCStr;
 use rustc_errors::{FatalError, Handler, Level};
 use rustc_fs_util::{link_or_copy, path_to_c_string};
@@ -53,6 +54,7 @@ pub fn write_output_file(
     output: &Path,
     dwo_output: Option<&Path>,
     file_type: llvm::FileType,
+    self_profiler_ref: &SelfProfilerRef,
 ) -> Result<(), FatalError> {
     unsafe {
         let output_c = path_to_c_string(output);
@@ -76,6 +78,19 @@ pub fn write_output_file(
                 file_type,
             )
         };
+
+        // Record artifact sizes for self-profiling
+        if result == llvm::LLVMRustResult::Success {
+            let artifact_kind = match file_type {
+                llvm::FileType::ObjectFile => "object_file",
+                llvm::FileType::AssemblyFile => "assembly_file",
+            };
+            record_artifact_size(self_profiler_ref, artifact_kind, output);
+            if let Some(dwo_file) = dwo_output {
+                record_artifact_size(self_profiler_ref, "dwo_file", dwo_file);
+            }
+        }
+
         result.into_result().map_err(|()| {
             let msg = format!("could not write output to {}", output.display());
             llvm_err(handler, &msg)
@@ -752,6 +767,14 @@ pub(crate) unsafe fn codegen(
             let thin = ThinBuffer::new(llmod);
             let data = thin.data();
 
+            if let Some(bitcode_filename) = bc_out.file_name() {
+                cgcx.prof.artifact_size(
+                    "llvm_bitcode",
+                    bitcode_filename.to_string_lossy(),
+                    data.len() as u64,
+                );
+            }
+
             if config.emit_bc || config.emit_obj == EmitObj::Bitcode {
                 let _timer = cgcx.prof.generic_activity_with_arg(
                     "LLVM_module_codegen_emit_bitcode",
@@ -812,6 +835,11 @@ pub(crate) unsafe fn codegen(
             }
 
             let result = llvm::LLVMRustPrintModule(llmod, out_c.as_ptr(), demangle_callback);
+
+            if result == llvm::LLVMRustResult::Success {
+                record_artifact_size(&cgcx.prof, "llvm_ir", &out);
+            }
+
             result.into_result().map_err(|()| {
                 let msg = format!("failed to write LLVM IR to {}", out.display());
                 llvm_err(diag_handler, &msg)
@@ -842,6 +870,7 @@ pub(crate) unsafe fn codegen(
                     &path,
                     None,
                     llvm::FileType::AssemblyFile,
+                    &cgcx.prof,
                 )
             })?;
         }
@@ -875,6 +904,7 @@ pub(crate) unsafe fn codegen(
                         &obj_out,
                         dwo_out,
                         llvm::FileType::ObjectFile,
+                        &cgcx.prof,
                     )
                 })?;
             }
@@ -1131,3 +1161,19 @@ fn create_msvc_imps(
         symbol_name.starts_with(b"__llvm_profile_")
     }
 }
+
+fn record_artifact_size(
+    self_profiler_ref: &SelfProfilerRef,
+    artifact_kind: &'static str,
+    path: &Path,
+) {
+    // Don't stat the file if we are not going to record its size.
+    if !self_profiler_ref.enabled() {
+        return;
+    }
+
+    if let Some(artifact_name) = path.file_name() {
+        let file_size = std::fs::metadata(path).map(|m| m.len()).unwrap_or(0);
+        self_profiler_ref.artifact_size(artifact_kind, artifact_name.to_string_lossy(), file_size);
+    }
+}

compiler/rustc_codegen_ssa/src/back/link.rs

@@ -121,6 +121,19 @@ pub fn link_binary<'a, B: ArchiveBuilder<'a>>(
             if sess.opts.json_artifact_notifications {
                 sess.parse_sess.span_diagnostic.emit_artifact_notification(&out_filename, "link");
             }
+
+            if sess.prof.enabled() {
+                if let Some(artifact_name) = out_filename.file_name() {
+                    // Record size for self-profiling
+                    let file_size = std::fs::metadata(&out_filename).map(|m| m.len()).unwrap_or(0);
+
+                    sess.prof.artifact_size(
+                        "linked_artifact",
+                        artifact_name.to_string_lossy(),
+                        file_size,
+                    );
+                }
+            }
         }
     }
 

compiler/rustc_metadata/src/rmeta/encoder.rs

@@ -2193,5 +2193,8 @@ fn encode_metadata_impl(tcx: TyCtxt<'_>) -> EncodedMetadata {
     result[header + 2] = (pos >> 8) as u8;
     result[header + 3] = (pos >> 0) as u8;
 
+    // Record metadata size for self-profiling
+    tcx.prof.artifact_size("crate_metadata", "crate_metadata", result.len() as u64);
+
     EncodedMetadata { raw_data: result }
 }

compiler/rustc_monomorphize/src/partitioning/mod.rs

@@ -361,6 +361,17 @@ fn collect_and_partition_mono_items<'tcx>(
         )
     });
 
+    if tcx.prof.enabled() {
+        // Record CGU size estimates for self-profiling.
+        for cgu in codegen_units {
+            tcx.prof.artifact_size(
+                "codegen_unit_size_estimate",
+                &cgu.name().as_str()[..],
+                cgu.size_estimate() as u64,
+            );
+        }
+    }
+
     let mono_items: DefIdSet = items
         .iter()
         .filter_map(|mono_item| match *mono_item {