Skip to content

Commit

Permalink
Rollup merge of #70289 - nnethercote:refactor-codegen, r=eddyb
Browse files Browse the repository at this point in the history 
Refactor `codegen`

`codegen` in `src/librustc_codegen_llvm/back/write.rs` is long and has complex control flow. These commits refactor it and make it easier to understand.
  • Loading branch information
@Centril
Centril authored Mar 24, 2020
2 parents a813cf9 + ca0f997 commit 6e587a4
Show file tree
Hide file tree
Showing 2 changed files with 104 additions and 109 deletions.
174 changes: 82 additions & 92 deletions src/librustc_codegen_llvm/back/write.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ use crate::ModuleLlvm;
use log::debug;
use rustc::bug;
use rustc::ty::TyCtxt;
use rustc_codegen_ssa::back::write::{run_assembler, CodegenContext, ModuleConfig};
use rustc_codegen_ssa::back::write::{run_assembler, CodegenContext, EmbedBitcode, ModuleConfig};
use rustc_codegen_ssa::traits::*;
use rustc_codegen_ssa::{CompiledModule, ModuleCodegen, RLIB_BYTECODE_EXTENSION};
use rustc_data_structures::small_c_str::SmallCStr;
Expand Down Expand Up @@ -634,30 +634,24 @@ pub(crate) unsafe fn codegen(
f(cpm)
}

// If we don't have the integrated assembler, then we need to emit asm
// from LLVM and use `gcc` to create the object file.
let asm_to_obj = config.emit_obj && config.no_integrated_as;

// Change what we write and cleanup based on whether obj files are
// just llvm bitcode. In that case write bitcode, and possibly
// delete the bitcode if it wasn't requested. Don't generate the
// machine code, instead copy the .o file from the .bc
let write_bc = config.emit_bc || config.obj_is_bitcode;
let rm_bc = !config.emit_bc && config.obj_is_bitcode;
let write_obj = config.emit_obj && !config.obj_is_bitcode && !asm_to_obj;
let copy_bc_to_obj = config.emit_obj && config.obj_is_bitcode;
// Two things to note:
// - If object files are just LLVM bitcode we write bitcode, copy it to
// the .o file, and delete the bitcode if it wasn't otherwise
// requested.
// - If we don't have the integrated assembler then we need to emit
// asm from LLVM and use `gcc` to create the object file.

let bc_out = cgcx.output_filenames.temp_path(OutputType::Bitcode, module_name);
let obj_out = cgcx.output_filenames.temp_path(OutputType::Object, module_name);

if write_bc || config.emit_bc_compressed || config.embed_bitcode {
if config.bitcode_needed() {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_make_bitcode", &module.name[..]);
let thin = ThinBuffer::new(llmod);
let data = thin.data();

if write_bc {
if config.emit_bc || config.obj_is_bitcode {
let _timer = cgcx.prof.generic_activity_with_arg(
"LLVM_module_codegen_emit_bitcode",
&module.name[..],
Expand All @@ -668,7 +662,7 @@ pub(crate) unsafe fn codegen(
}
}

if config.embed_bitcode {
if config.embed_bitcode == EmbedBitcode::Full {
let _timer = cgcx.prof.generic_activity_with_arg(
"LLVM_module_codegen_embed_bitcode",
&module.name[..],
Expand All @@ -688,81 +682,75 @@ pub(crate) unsafe fn codegen(
diag_handler.err(&msg);
}
}
} else if config.embed_bitcode_marker {
} else if config.embed_bitcode == EmbedBitcode::Marker {
embed_bitcode(cgcx, llcx, llmod, None);
}

{
if config.emit_ir {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_emit_ir", &module.name[..]);
let out = cgcx.output_filenames.temp_path(OutputType::LlvmAssembly, module_name);
let out_c = path_to_c_string(&out);

extern "C" fn demangle_callback(
input_ptr: *const c_char,
input_len: size_t,
output_ptr: *mut c_char,
output_len: size_t,
) -> size_t {
let input = unsafe {
slice::from_raw_parts(input_ptr as *const u8, input_len as usize)
};

let input = match str::from_utf8(input) {
Ok(s) => s,
Err(_) => return 0,
};

let output = unsafe {
slice::from_raw_parts_mut(output_ptr as *mut u8, output_len as usize)
};
let mut cursor = io::Cursor::new(output);

let demangled = match rustc_demangle::try_demangle(input) {
Ok(d) => d,
Err(_) => return 0,
};

if write!(cursor, "{:#}", demangled).is_err() {
// Possible only if provided buffer is not big enough
return 0;
}

cursor.position() as size_t
if config.emit_ir {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_emit_ir", &module.name[..]);
let out = cgcx.output_filenames.temp_path(OutputType::LlvmAssembly, module_name);
let out_c = path_to_c_string(&out);

extern "C" fn demangle_callback(
input_ptr: *const c_char,
input_len: size_t,
output_ptr: *mut c_char,
output_len: size_t,
) -> size_t {
let input =
unsafe { slice::from_raw_parts(input_ptr as *const u8, input_len as usize) };

let input = match str::from_utf8(input) {
Ok(s) => s,
Err(_) => return 0,
};

let output = unsafe {
slice::from_raw_parts_mut(output_ptr as *mut u8, output_len as usize)
};
let mut cursor = io::Cursor::new(output);

let demangled = match rustc_demangle::try_demangle(input) {
Ok(d) => d,
Err(_) => return 0,
};

if write!(cursor, "{:#}", demangled).is_err() {
// Possible only if provided buffer is not big enough
return 0;
}

let result = llvm::LLVMRustPrintModule(llmod, out_c.as_ptr(), demangle_callback);
result.into_result().map_err(|()| {
let msg = format!("failed to write LLVM IR to {}", out.display());
llvm_err(diag_handler, &msg)
})?;
cursor.position() as size_t
}

if config.emit_asm || asm_to_obj {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_emit_asm", &module.name[..]);
let path = cgcx.output_filenames.temp_path(OutputType::Assembly, module_name);
let result = llvm::LLVMRustPrintModule(llmod, out_c.as_ptr(), demangle_callback);
result.into_result().map_err(|()| {
let msg = format!("failed to write LLVM IR to {}", out.display());
llvm_err(diag_handler, &msg)
})?;
}

// We can't use the same module for asm and binary output, because that triggers
// various errors like invalid IR or broken binaries, so we might have to clone the
// module to produce the asm output
let llmod = if config.emit_obj { llvm::LLVMCloneModule(llmod) } else { llmod };
with_codegen(tm, llmod, config.no_builtins, |cpm| {
write_output_file(
diag_handler,
tm,
cpm,
llmod,
&path,
llvm::FileType::AssemblyFile,
)
})?;
}
let config_emit_normal_obj = config.emit_obj && !config.obj_is_bitcode;

if write_obj {
if config.emit_asm || (config_emit_normal_obj && config.no_integrated_as) {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_emit_asm", &module.name[..]);
let path = cgcx.output_filenames.temp_path(OutputType::Assembly, module_name);

// We can't use the same module for asm and binary output, because that triggers
// various errors like invalid IR or broken binaries, so we might have to clone the
// module to produce the asm output
let llmod = if config.emit_obj { llvm::LLVMCloneModule(llmod) } else { llmod };
with_codegen(tm, llmod, config.no_builtins, |cpm| {
write_output_file(diag_handler, tm, cpm, llmod, &path, llvm::FileType::AssemblyFile)
})?;
}

if config_emit_normal_obj {
if !config.no_integrated_as {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_emit_obj", &module.name[..]);
Expand All @@ -776,7 +764,7 @@ pub(crate) unsafe fn codegen(
llvm::FileType::ObjectFile,
)
})?;
} else if asm_to_obj {
} else {
let _timer = cgcx
.prof
.generic_activity_with_arg("LLVM_module_codegen_asm_to_obj", &module.name[..]);
Expand All @@ -789,17 +777,19 @@ pub(crate) unsafe fn codegen(
}
}

if copy_bc_to_obj {
debug!("copying bitcode {:?} to obj {:?}", bc_out, obj_out);
if let Err(e) = link_or_copy(&bc_out, &obj_out) {
diag_handler.err(&format!("failed to copy bitcode to object file: {}", e));
if config.obj_is_bitcode {
if config.emit_obj {
debug!("copying bitcode {:?} to obj {:?}", bc_out, obj_out);
if let Err(e) = link_or_copy(&bc_out, &obj_out) {
diag_handler.err(&format!("failed to copy bitcode to object file: {}", e));
}
}
}

if rm_bc {
debug!("removing_bitcode {:?}", bc_out);
if let Err(e) = fs::remove_file(&bc_out) {
diag_handler.err(&format!("failed to remove bitcode: {}", e));
if !config.emit_bc {
debug!("removing_bitcode {:?}", bc_out);
if let Err(e) = fs::remove_file(&bc_out) {
diag_handler.err(&format!("failed to remove bitcode: {}", e));
}
}
}

Expand Down
39 changes: 22 additions & 17 deletions src/librustc_codegen_ssa/back/write.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,14 @@ use std::thread;

const PRE_LTO_BC_EXT: &str = "pre-lto.bc";

/// The kind of bitcode to embed in object files.
#[derive(PartialEq)]
pub enum EmbedBitcode {
None,
Marker,
Full,
}

/// Module-specific configuration for `optimize_and_codegen`.
pub struct ModuleConfig {
/// Names of additional optimization passes to run.
Expand All @@ -74,7 +82,6 @@ pub struct ModuleConfig {
pub emit_no_opt_bc: bool,
pub emit_bc: bool,
pub emit_bc_compressed: bool,
pub emit_lto_bc: bool,
pub emit_ir: bool,
pub emit_asm: bool,
pub emit_obj: bool,
Expand All @@ -94,8 +101,7 @@ pub struct ModuleConfig {
// emscripten's ecc compiler, when used as the linker.
pub obj_is_bitcode: bool,
pub no_integrated_as: bool,
pub embed_bitcode: bool,
pub embed_bitcode_marker: bool,
pub embed_bitcode: EmbedBitcode,
}

impl ModuleConfig {
Expand All @@ -116,13 +122,11 @@ impl ModuleConfig {
emit_pre_lto_bc: false,
emit_bc: false,
emit_bc_compressed: false,
emit_lto_bc: false,
emit_ir: false,
emit_asm: false,
emit_obj: false,
obj_is_bitcode: false,
embed_bitcode: false,
embed_bitcode_marker: false,
embed_bitcode: EmbedBitcode::None,
no_integrated_as: false,

verify_llvm_ir: false,
Expand All @@ -145,16 +149,15 @@ impl ModuleConfig {
self.new_llvm_pass_manager = sess.opts.debugging_opts.new_llvm_pass_manager;
self.obj_is_bitcode =
sess.target.target.options.obj_is_bitcode || sess.opts.cg.linker_plugin_lto.enabled();
let embed_bitcode =
sess.target.target.options.embed_bitcode || sess.opts.debugging_opts.embed_bitcode;
if embed_bitcode {
match sess.opts.optimize {
config::OptLevel::No | config::OptLevel::Less => {
self.embed_bitcode_marker = embed_bitcode;
self.embed_bitcode =
if sess.target.target.options.embed_bitcode || sess.opts.debugging_opts.embed_bitcode {
match sess.opts.optimize {
config::OptLevel::No | config::OptLevel::Less => EmbedBitcode::Marker,
_ => EmbedBitcode::Full,
}
_ => self.embed_bitcode = embed_bitcode,
}
}
} else {
EmbedBitcode::None
};

// Copy what clang does by turning on loop vectorization at O2 and
// slp vectorization at O3. Otherwise configure other optimization aspects
Expand Down Expand Up @@ -190,7 +193,10 @@ impl ModuleConfig {
}

pub fn bitcode_needed(&self) -> bool {
self.emit_bc || self.obj_is_bitcode || self.emit_bc_compressed || self.embed_bitcode
self.emit_bc
|| self.obj_is_bitcode
|| self.emit_bc_compressed
|| self.embed_bitcode == EmbedBitcode::Full
}
}

Expand Down Expand Up @@ -379,7 +385,6 @@ pub fn start_async_codegen<B: ExtraBackendMethods>(
modules_config.emit_no_opt_bc = true;
modules_config.emit_pre_lto_bc = true;
modules_config.emit_bc = true;
modules_config.emit_lto_bc = true;
metadata_config.emit_bc = true;
allocator_config.emit_bc = true;
}
Expand Down

0 comments on commit 6e587a4

Please sign in to comment.