Overhaul filename handling for cross-compiler consistency by Urgau · Pull Request #149709 · rust-lang/rust

Urgau · 2025-12-06T11:14:33Z

This PR overhauls the way we handle filenames in the compiler and rmeta in order to achieve achieve cross-compiler consistency (ie. having the same path no matter if the filename was created in the current compiler session or is coming from rmeta).

This is required as some parts of the compiler rely on consistent paths for the soundness of generated code (see #148328).

In order to achieved consistency multiple steps are being taken by this PR:

by making RealFileName immutable
by only having SourceMap::to_real_filename create RealFileName
- currently RealFileName can be created from any Path and are remapped afterwards, which creates consistency issue
by also making RealFileName holds it's working directory, embeddable name and the remapped scopes
- this removes the need for a Session, to know the current(!) scopes and cwd, which is invalid as they may not be equal to the scopes used when creating the filename

In order for SourceMap::to_real_filename to know which scopes to apply FilePathMapping now takes the current remapping scopes to apply, which makes FileNameDisplayPreference and company useless and are removed.

This PR is split-up in multiple commits (unfortunately not atomic), but should help review the changes.

Unblocks #147611
Fixes #148328

cc @RalfJung @oli-obk
r? @davidtwco

rustbot · 2025-12-06T11:14:36Z

Some changes occurred in coverage instrumentation.

cc @Zalathar

These commits modify the Cargo.lock file. Unintentional changes to Cargo.lock can be introduced when switching branches and rebasing PRs.

If this was unintentional then you should revert the changes before this PR is merged.
Otherwise, you can ignore this comment.

Some changes occurred to the CTFE machinery

cc @RalfJung, @oli-obk, @lcnr

The Miri subtree was changed

cc @rust-lang/miri

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

Some changes occurred in compiler/rustc_codegen_cranelift

cc @bjorn3

Some changes occurred in compiler/rustc_codegen_gcc

cc @antoyo, @GuillaumeGomez

rustbot · 2025-12-06T12:02:58Z

Some changes occurred in src/tools/rustfmt

cc @rust-lang/rustfmt

RalfJung · 2025-12-06T12:17:53Z

@bors try
@rust-timer queue

Overhaul filename handling for cross-compiler consistency

compiler/rustc_session/src/options.rs

Prefer remapping the relative `library/` and `compiler/` directories This is done to avoid leaking the relative paths to the standard library after the overall of filenames in #149709. Noted that the paths were already leaking before, but to a lesser extent since most (but not all) the paths embedded in the distributed `rlib` were absolute. In general Cargo compiles workspace members with relative paths, so it's better anyway to remap the relative path. In addition to our tests I have manually confirmed that it also works as expected for the printed diagnostics paths. cf. https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/remapping.20of.20the.20standard.20library/near/564093571

Overhaul filename handling for cross-compiler consistency This PR overhauls the way we handle filenames in the compiler and `rmeta` in order to achieve achieve cross-compiler consistency (ie. having the same path no matter if the filename was created in the current compiler session or is coming from `rmeta`). This is required as some parts of the compiler rely on consistent paths for the soundness of generated code (see rust-lang#148328). In order to achieved consistency multiple steps are being taken by this PR: - by making `RealFileName` immutable - by only having `SourceMap::to_real_filename` create `RealFileName` - currently `RealFileName` can be created from any `Path` and are remapped afterwards, which creates consistency issue - by also making `RealFileName` holds it's working directory, embeddable name and the remapped scopes - this removes the need for a `Session`, to know the current(!) scopes and cwd, which is invalid as they may not be equal to the scopes used when creating the filename In order for `SourceMap::to_real_filename` to know which scopes to apply `FilePathMapping` now takes the current remapping scopes to apply, which makes `FileNameDisplayPreference` and company useless and are removed. This PR is split-up in multiple commits (unfortunately not atomic), but should help review the changes. Unblocks rust-lang#147611 Fixes rust-lang#148328

Prefer remapping the relative `library/` and `compiler/` directories This is done to avoid leaking the relative paths to the standard library after the overall of filenames in #149709. Noted that the paths were already leaking before, but to a lesser extent since most (but not all) the paths embedded in the distributed `rlib` were absolute. In general Cargo compiles workspace members with relative paths, so it's better anyway to remap the relative path. In addition to our tests I have manually confirmed that it also works as expected for the printed diagnostics paths. cf. https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/remapping.20of.20the.20standard.20library/near/564093571

… r=davidtwco Improve filenames encoding and misc This PR is a follow-up to rust-lang#149709, it aims at preventing a double encoding when there are no remapping, as well as making some small improvements to the code. Best reviewed commit by commit.

Prefer remapping the relative `library/` and `compiler/` directories This is done to avoid leaking the relative paths to the standard library after the overall of filenames in rust-lang/rust#149709. Noted that the paths were already leaking before, but to a lesser extent since most (but not all) the paths embedded in the distributed `rlib` were absolute. In general Cargo compiles workspace members with relative paths, so it's better anyway to remap the relative path. In addition to our tests I have manually confirmed that it also works as expected for the printed diagnostics paths. cf. https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/remapping.20of.20the.20standard.20library/near/564093571

Rollup merge of #149989 - Urgau:filenames-post-improvements, r=davidtwco Improve filenames encoding and misc This PR is a follow-up to #149709, it aims at preventing a double encoding when there are no remapping, as well as making some small improvements to the code. Best reviewed commit by commit.

…aumeGomez Handle remapped paths correctly when generating "Source" links Fixes rust-lang#150100. This PR fixes a regression introduced by [rust-lang#149709](rust-lang#149709), I was overzealous in my changes (https://github.com/rust-lang/rust/pull/149709/changes#diff-e1cf7ef2fb411d24980cd4cbea1e867cc36029e9496e1ceca64cfb6a0e3510f6) and accidentally changed the behavior of `rustdoc` in the presence of remapped, to simply reject them instead of handling them. With this PR remapped paths are handled correctly, in a similar way as it was before. ~~I added a run-make test to make sure we don't regress it again, a simple `rustdoc` test in not sufficient as `rustdoc` is not called on the auxiliary crate. It's not pretty but it works.~~ rust-lang#150172 (comment) rustdoc doesn't have any handling for `--remap-path-scope`, so I used the `MACRO` scope (it was already used [elsewhere](https://github.com/rust-lang/rust/blob/b889870082dd0b0e3594bbfbebb4545d54710829/src/librustdoc/clean/types.rs#L154). cf. https://rust-lang.zulipchat.com/#narrow/channel/266220-t-rustdoc/topic/source.20links.20missing.20for.20nightly.20std.20docs

Rollup merge of #150172 - Urgau:rustdoc-remap-fixes, r=GuillaumeGomez Handle remapped paths correctly when generating "Source" links Fixes #150100. This PR fixes a regression introduced by [#149709](#149709), I was overzealous in my changes (https://github.com/rust-lang/rust/pull/149709/changes#diff-e1cf7ef2fb411d24980cd4cbea1e867cc36029e9496e1ceca64cfb6a0e3510f6) and accidentally changed the behavior of `rustdoc` in the presence of remapped, to simply reject them instead of handling them. With this PR remapped paths are handled correctly, in a similar way as it was before. ~~I added a run-make test to make sure we don't regress it again, a simple `rustdoc` test in not sufficient as `rustdoc` is not called on the auxiliary crate. It's not pretty but it works.~~ #150172 (comment) rustdoc doesn't have any handling for `--remap-path-scope`, so I used the `MACRO` scope (it was already used [elsewhere](https://github.com/rust-lang/rust/blob/b889870082dd0b0e3594bbfbebb4545d54710829/src/librustdoc/clean/types.rs#L154). cf. https://rust-lang.zulipchat.com/#narrow/channel/266220-t-rustdoc/topic/source.20links.20missing.20for.20nightly.20std.20docs

Overhaul filename handling for cross-compiler consistency This PR overhauls the way we handle filenames in the compiler and `rmeta` in order to achieve achieve cross-compiler consistency (ie. having the same path no matter if the filename was created in the current compiler session or is coming from `rmeta`). This is required as some parts of the compiler rely on consistent paths for the soundness of generated code (see rust-lang#148328). In order to achieved consistency multiple steps are being taken by this PR: - by making `RealFileName` immutable - by only having `SourceMap::to_real_filename` create `RealFileName` - currently `RealFileName` can be created from any `Path` and are remapped afterwards, which creates consistency issue - by also making `RealFileName` holds it's working directory, embeddable name and the remapped scopes - this removes the need for a `Session`, to know the current(!) scopes and cwd, which is invalid as they may not be equal to the scopes used when creating the filename In order for `SourceMap::to_real_filename` to know which scopes to apply `FilePathMapping` now takes the current remapping scopes to apply, which makes `FileNameDisplayPreference` and company useless and are removed. This PR is split-up in multiple commits (unfortunately not atomic), but should help review the changes. Unblocks rust-lang#147611 Fixes rust-lang#148328

Improve filenames encoding and misc This PR is a follow-up to rust-lang/rust#149709, it aims at preventing a double encoding when there are no remapping, as well as making some small improvements to the code. Best reviewed commit by commit.

Handle remapped paths correctly when generating "Source" links Fixes rust-lang/rust#150100. This PR fixes a regression introduced by [#149709](rust-lang/rust#149709), I was overzealous in my changes (https://github.com/rust-lang/rust/pull/149709/changes#diff-e1cf7ef2fb411d24980cd4cbea1e867cc36029e9496e1ceca64cfb6a0e3510f6) and accidentally changed the behavior of `rustdoc` in the presence of remapped, to simply reject them instead of handling them. With this PR remapped paths are handled correctly, in a similar way as it was before. ~~I added a run-make test to make sure we don't regress it again, a simple `rustdoc` test in not sufficient as `rustdoc` is not called on the auxiliary crate. It's not pretty but it works.~~ rust-lang/rust#150172 (comment) rustdoc doesn't have any handling for `--remap-path-scope`, so I used the `MACRO` scope (it was already used [elsewhere](https://github.com/rust-lang/rust/blob/b889870082dd0b0e3594bbfbebb4545d54710829/src/librustdoc/clean/types.rs#L154). cf. https://rust-lang.zulipchat.com/#narrow/channel/266220-t-rustdoc/topic/source.20links.20missing.20for.20nightly.20std.20docs

Prefer remapping the relative `library/` and `compiler/` directories This is done to avoid leaking the relative paths to the standard library after the overall of filenames in rust-lang/rust#149709. Noted that the paths were already leaking before, but to a lesser extent since most (but not all) the paths embedded in the distributed `rlib` were absolute. In general Cargo compiles workspace members with relative paths, so it's better anyway to remap the relative path. In addition to our tests I have manually confirmed that it also works as expected for the printed diagnostics paths. cf. https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/remapping.20of.20the.20standard.20library/near/564093571

Overhaul filename handling for cross-compiler consistency This PR overhauls the way we handle filenames in the compiler and `rmeta` in order to achieve achieve cross-compiler consistency (ie. having the same path no matter if the filename was created in the current compiler session or is coming from `rmeta`). This is required as some parts of the compiler rely on consistent paths for the soundness of generated code (see rust-lang#148328). In order to achieved consistency multiple steps are being taken by this PR: - by making `RealFileName` immutable - by only having `SourceMap::to_real_filename` create `RealFileName` - currently `RealFileName` can be created from any `Path` and are remapped afterwards, which creates consistency issue - by also making `RealFileName` holds it's working directory, embeddable name and the remapped scopes - this removes the need for a `Session`, to know the current(!) scopes and cwd, which is invalid as they may not be equal to the scopes used when creating the filename In order for `SourceMap::to_real_filename` to know which scopes to apply `FilePathMapping` now takes the current remapping scopes to apply, which makes `FileNameDisplayPreference` and company useless and are removed. This PR is split-up in multiple commits (unfortunately not atomic), but should help review the changes. Unblocks rust-lang#147611 Fixes rust-lang#148328

Prefer remapping the relative `library/` and `compiler/` directories This is done to avoid leaking the relative paths to the standard library after the overall of filenames in rust-lang/rust#149709. Noted that the paths were already leaking before, but to a lesser extent since most (but not all) the paths embedded in the distributed `rlib` were absolute. In general Cargo compiles workspace members with relative paths, so it's better anyway to remap the relative path. In addition to our tests I have manually confirmed that it also works as expected for the printed diagnostics paths. cf. https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/remapping.20of.20the.20standard.20library/near/564093571

Improve filenames encoding and misc This PR is a follow-up to rust-lang/rust#149709, it aims at preventing a double encoding when there are no remapping, as well as making some small improvements to the code. Best reviewed commit by commit.

rustbot assigned davidtwco Dec 6, 2025