Replace all uses of sysroot_candidates with get_or_default_sysroot #142089

bjorn3 · 2025-06-05T16:59:43Z

Before this change we had two different ways to attempt to locate the sysroot which are inconsistently used:

get_or_default_sysroot which tries to locate based on the 0th cli argument and if that doesn't work falls back to locating it using the librustc_driver.so location and returns a single path.,
sysroot_candidates which takes the former and additionally does another attempt at locating using librustc_driver.so except without linux multiarch handling and then returns both paths.,

The latter was originally introduced to be able to locate the codegen backend back when cg_llvm was dynamically linked even for a custom driver when the --sysroot passed in does not contain a copy of cg_llvm. Back then get_or_default_sysroot did not attempt to locate the sysroot based on the location of librustc_driver.so yet. Because that is now done, the only case where removing sysroot_candidates can break things is if you have a custom driver inside what looks like a sysroot including the lib/rustlib directory, but which is missing some parts of the full sysroot like eg rust-lld.

Follow up to #138404

And consistently use try_canonicalize rather than canonicalize.

Before this change we had two different ways to attempt to locate the sysroot which are inconsistently used: * get_or_default_sysroot which tries to locate based on the 0th cli argument and if that doesn't work falls back to locating it using the librustc_driver.so location and returns a single path., * sysroot_candidates which takes the former and additionally does another attempt at locating using librustc_driver.so except without linux multiarch handling and then returns both paths., The latter was originally introduced to be able to locate the codegen backend back when cg_llvm was dynamically linked even for a custom driver when the --sysroot passed in does not contain a copy of cg_llvm. Back then get_or_default_sysroot did not attempt to locate the sysroot based on the location of librustc_driver.so yet. Because that is now done, the only case where removing sysroot_candidates can break things is if you have a custom driver inside what looks like a sysroot including the lib/rustlib directory, but which is missing some parts of the full sysroot like eg rust-lld.

rustbot · 2025-06-05T16:59:48Z

r? @BoxyUwU

rustbot has assigned @BoxyUwU.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

rustbot · 2025-06-05T16:59:50Z

rustc_error_messages was changed

cc @davidtwco, @compiler-errors, @TaKO8Ki

compiler/rustc_error_messages/src/lib.rs

compiler/rustc_interface/src/util.rs

petrochenkov · 2025-06-06T13:29:27Z

compiler/rustc_session/src/filesearch.rs

-                );
-            }
-        }
+pub fn sysroot_with_fallback(sysroot: &Path) -> SmallVec<[PathBuf; 2]> {


This is an improvement, but it still feels a bit sub-optimal.
The passed sysroot here can only be --sysroot or get_or_default_sysroot(), so perhaps it makes sense to make a structure like

struct Sysroot { explicit: Option<PathBuf>, default: PathBuf, }

with two methods for accessing either the main path or both, and then pass that structure around.

The Session only stores the result of materialize_sysroot(), which does maybe_sysroot.unwrap_or_else(|| get_or_default_sysroot()).

rustbot · 2025-06-06T14:20:55Z

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.

bjorn3 · 2025-06-07T17:10:27Z

@rustbot ready

petrochenkov · 2025-06-08T09:55:09Z

@bors r+

bors · 2025-06-08T09:55:12Z

📌 Commit e4c4c4c has been approved by petrochenkov

It is now in the queue for this repository.

…r=petrochenkov Replace all uses of sysroot_candidates with get_or_default_sysroot Before this change we had two different ways to attempt to locate the sysroot which are inconsistently used: * `get_or_default_sysroot` which tries to locate based on the 0th cli argument and if that doesn't work falls back to locating it using the librustc_driver.so location and returns a single path., * `sysroot_candidates` which takes the former and additionally does another attempt at locating using `librustc_driver.so` except without linux multiarch handling and then returns both paths., The latter was originally introduced to be able to locate the codegen backend back when cg_llvm was dynamically linked even for a custom driver when the `--sysroot` passed in does not contain a copy of cg_llvm. Back then `get_or_default_sysroot` did not attempt to locate the sysroot based on the location of librustc_driver.so yet. Because that is now done, the only case where removing `sysroot_candidates` can break things is if you have a custom driver inside what looks like a sysroot including the `lib/rustlib` directory, but which is missing some parts of the full sysroot like eg rust-lld. Follow up to rust-lang#138404

Rollup of 11 pull requests Successful merges: - #140774 (Affirm `-Cforce-frame-pointers=off` does not override) - #141001 (Make NonZero<char> possible) - #141700 (Atomic intrinsics : use const generic ordering, part 2) - #142008 (const-eval error: always say in which item the error occurred) - #142053 (Add new Tier-3 targets: `loongarch32-unknown-none*`) - #142089 (Replace all uses of sysroot_candidates with get_or_default_sysroot) - #142108 (compiler: Add track_caller to AbiMapping::unwrap) - #142132 (`tests/ui`: A New Order [6/N]) - #142162 (UnsafePinned: update get() docs and signature to allow shared mutation) - #142171 (`tests/ui`: A New Order [7/N]) - #142179 (store `target.min_global_align` as an `Align`) r? `@ghost` `@rustbot` modify labels: rollup

…r=petrochenkov Replace all uses of sysroot_candidates with get_or_default_sysroot Before this change we had two different ways to attempt to locate the sysroot which are inconsistently used: * `get_or_default_sysroot` which tries to locate based on the 0th cli argument and if that doesn't work falls back to locating it using the librustc_driver.so location and returns a single path., * `sysroot_candidates` which takes the former and additionally does another attempt at locating using `librustc_driver.so` except without linux multiarch handling and then returns both paths., The latter was originally introduced to be able to locate the codegen backend back when cg_llvm was dynamically linked even for a custom driver when the `--sysroot` passed in does not contain a copy of cg_llvm. Back then `get_or_default_sysroot` did not attempt to locate the sysroot based on the location of librustc_driver.so yet. Because that is now done, the only case where removing `sysroot_candidates` can break things is if you have a custom driver inside what looks like a sysroot including the `lib/rustlib` directory, but which is missing some parts of the full sysroot like eg rust-lld. Follow up to rust-lang#138404

Rollup of 12 pull requests Successful merges: - #141803 (Remove rustc's notion of "preferred" alignment AKA `__alignof`) - #142053 (Add new Tier-3 targets: `loongarch32-unknown-none*`) - #142089 (Replace all uses of sysroot_candidates with get_or_default_sysroot) - #142108 (compiler: Add track_caller to AbiMapping::unwrap) - #142132 (`tests/ui`: A New Order [6/N]) - #142162 (UnsafePinned: update get() docs and signature to allow shared mutation) - #142171 (`tests/ui`: A New Order [7/N]) - #142179 (store `target.min_global_align` as an `Align`) - #142183 (Added test for 30904) - #142194 (Remove all unused feature gates from the compiler) - #142199 (Do not free disk space in the `mingw-check-tidy` job) - #142210 (Run `mingw-check-tidy` on auto builds) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of #142089 - bjorn3:sysroot_handling_cleanup3, r=petrochenkov Replace all uses of sysroot_candidates with get_or_default_sysroot Before this change we had two different ways to attempt to locate the sysroot which are inconsistently used: * `get_or_default_sysroot` which tries to locate based on the 0th cli argument and if that doesn't work falls back to locating it using the librustc_driver.so location and returns a single path., * `sysroot_candidates` which takes the former and additionally does another attempt at locating using `librustc_driver.so` except without linux multiarch handling and then returns both paths., The latter was originally introduced to be able to locate the codegen backend back when cg_llvm was dynamically linked even for a custom driver when the `--sysroot` passed in does not contain a copy of cg_llvm. Back then `get_or_default_sysroot` did not attempt to locate the sysroot based on the location of librustc_driver.so yet. Because that is now done, the only case where removing `sysroot_candidates` can break things is if you have a custom driver inside what looks like a sysroot including the `lib/rustlib` directory, but which is missing some parts of the full sysroot like eg rust-lld. Follow up to #138404

bjorn3 added 2 commits June 5, 2025 16:53

Move canonicalization into current_dll_path

387dae9

And consistently use try_canonicalize rather than canonicalize.

rustbot assigned BoxyUwU Jun 5, 2025

petrochenkov self-assigned this Jun 6, 2025

petrochenkov reviewed Jun 6, 2025

View reviewed changes

petrochenkov added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jun 6, 2025

Fix review comments

e4c4c4c

BoxyUwU removed their assignment Jun 6, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Jun 7, 2025

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jun 8, 2025

workingjubilee mentioned this pull request Jun 8, 2025

Rollup of 11 pull requests #142195

Closed

This was referenced Jun 8, 2025

Rollup of 13 pull requests #142213

Closed

Rollup of 12 pull requests #142220

Merged

bors merged commit 41bc5d7 into rust-lang:master Jun 9, 2025
10 checks passed

rustbot added this to the 1.89.0 milestone Jun 9, 2025

bjorn3 deleted the sysroot_handling_cleanup3 branch June 9, 2025 07:39

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Replace all uses of sysroot_candidates with get_or_default_sysroot #142089

Replace all uses of sysroot_candidates with get_or_default_sysroot #142089

Uh oh!

bjorn3 commented Jun 5, 2025

Uh oh!

rustbot commented Jun 5, 2025

Uh oh!

rustbot commented Jun 5, 2025

Uh oh!

Uh oh!

Uh oh!

petrochenkov Jun 6, 2025

Uh oh!

bjorn3 Jun 6, 2025

Uh oh!

rustbot commented Jun 6, 2025

Uh oh!

bjorn3 commented Jun 7, 2025

Uh oh!

petrochenkov commented Jun 8, 2025

Uh oh!

bors commented Jun 8, 2025

Uh oh!

Uh oh!

Uh oh!

Replace all uses of sysroot_candidates with get_or_default_sysroot #142089

Replace all uses of sysroot_candidates with get_or_default_sysroot #142089

Uh oh!

Conversation

bjorn3 commented Jun 5, 2025

Uh oh!

rustbot commented Jun 5, 2025

Uh oh!

rustbot commented Jun 5, 2025

Uh oh!

Uh oh!

Uh oh!

petrochenkov Jun 6, 2025

Choose a reason for hiding this comment

Uh oh!

bjorn3 Jun 6, 2025

Choose a reason for hiding this comment

Uh oh!

rustbot commented Jun 6, 2025

Uh oh!

bjorn3 commented Jun 7, 2025

Uh oh!

petrochenkov commented Jun 8, 2025

Uh oh!

bors commented Jun 8, 2025

Uh oh!

Uh oh!

Uh oh!