Test if reserved filenames are allowed in Windows #10322

eholk · 2022-01-24T19:11:27Z

Recent versions of Windows have removed the limitation on filenames like aux or con. This change allows the package::reserved_windows_name to still pass by first trying to create a file with a reserved name to see if Windows supports it. If so, it skips the rest of the test. Otherwise, we keep the same behavior as before.

rust-highfive · 2022-01-24T19:11:30Z

r? @ehuss

(rust-highfive has picked a reviewer for you, use r? to override)

weihanglo

Hi. If I want to verify the behavior of this patch. What environment should I prepare? Specifically, can you provide some official doc about this change? Thanks!

src/cargo/util/restricted_names.rs

eholk · 2022-01-27T18:10:59Z

Hi. If I want to verify the behavior of this patch. What environment should I prepare? Specifically, can you provide some official doc about this change? Thanks!

I'm on the Windows 11 Insiders Beta channel, and winver says it's version 21H2 (OS Build 22000.466).

Unfortunately, I haven't been able to find official documentation about the change, but I did confirm the behavior with the Windows dev who made the change.

joshtriplett · 2022-03-01T16:34:55Z

Could you post a transcript of the test failure (including with nocapture so we can see the actual test output)?

eholk · 2022-03-02T23:34:37Z

Sure, here's the output from cargo test -- package::reserved_windows_name --nocapture on master:

    Finished test [unoptimized + debuginfo] target(s) in 0.27s
     Running unittests src/cargo/lib.rs (target\debug\deps\cargo-9040197b200da57a.exe)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 34 filtered out; finished in 0.00s

     Running tests\build-std\main.rs (target\debug\deps\build_std-9cc2dc147da3e21f.exe)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 3 filtered out; finished in 0.00s

     Running tests\internal.rs (target\debug\deps\internal-870b738d2af5cbc1.exe)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 2 filtered out; finished in 0.00s

     Running tests\testsuite\main.rs (target\debug\deps\testsuite-508cfd0591301aff.exe)

running 1 test
running `C:\Users\ericholk\repo\cargo\target\debug\cargo.exe package`
thread 'package::reserved_windows_name' panicked at '
test failed running `C:\Users\ericholk\repo\cargo\target\debug\cargo.exe package`
error: process exited with code 0 (expected 101)
--- stdout

--- stderr
warning: manifest has no documentation, homepage or repository.
See https://doc.rust-lang.org/cargo/reference/manifest.html#package-metadata for more info.
   Packaging foo v0.0.1 (C:\Users\ericholk\repo\cargo\target\tmp\cit\t0\foo)
    Updating `dummy-registry` index
   Verifying foo v0.0.1 (C:\Users\ericholk\repo\cargo\target\tmp\cit\t0\foo)
 Downloading crates ...
  Downloaded bar v1.0.0 (registry `dummy-registry`)
   Compiling bar v1.0.0
   Compiling foo v0.0.1 (C:\Users\ericholk\repo\cargo\target\tmp\cit\t0\foo\target\package\foo-0.0.1)
    Finished dev [unoptimized + debuginfo] target(s) in 0.56s
', tests\testsuite\package.rs:1939:10
stack backtrace:
   0: std::panicking::begin_panic_handler
             at /rustc/f0c4da49983aa699f715caf681e3154b445fb60b\/library\std\src\panicking.rs:584
   1: core::panicking::panic_fmt
             at /rustc/f0c4da49983aa699f715caf681e3154b445fb60b\/library\core\src\panicking.rs:143
   2: cargo_test_support::panic_error::pe
             at .\crates\cargo-test-support\src\lib.rs:48
   3: cargo_test_support::panic_error<anyhow::Error>
             at .\crates\cargo-test-support\src\lib.rs:40
   4: cargo_test_support::Execs::run
             at .\crates\cargo-test-support\src\lib.rs:802
   5: testsuite::package::reserved_windows_name
             at .\tests\testsuite\package.rs:1915
   6: testsuite::package::reserved_windows_name::closure$0
             at .\tests\testsuite\package.rs:1890
   7: core::ops::function::FnOnce::call_once<testsuite::package::reserved_windows_name::closure_env$0,tuple$<> >
             at /rustc/f0c4da49983aa699f715caf681e3154b445fb60b\library\core\src\ops\function.rs:227
   8: core::ops::function::FnOnce::call_once
             at /rustc/f0c4da49983aa699f715caf681e3154b445fb60b\library\core\src\ops\function.rs:227
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
test package::reserved_windows_name ... FAILED

failures:

failures:
    package::reserved_windows_name

test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 2345 filtered out; finished in 1.02s

error: test failed, to rerun pass '--test testsuite'

joshtriplett · 2022-03-02T23:42:41Z

So, reading that test case, I don't think that's a test that should pass in general, even if the platform doesn't actually reserve such filenames. The logic that checks for aux.rs and avoids unpacking the file should not depend on what Windows does, at least not until we stop supporting all versions of Windows with that limitation.

I think, rather than changing the test so its behavior depends on the OS, this should instead change the implementation of the code that this test tests to ensure that it catches reserved filenames on all versions of Windows.

eholk · 2022-03-02T23:48:04Z

That seems reasonable. If I remember right, the code for publishing packages tries to prevent you from publishing packages that couldn't be unpacked on all OSes supported by Cargo, so it makes sense to have similar behavior for unpacking packages.

In that case though, I think we should also remove the #[cfg(windows)] on this test.

joshtriplett · 2022-03-02T23:54:36Z

If I remember right, the code for publishing packages tries to prevent you from publishing packages that couldn't be unpacked on all OSes supported by Cargo, so it makes sense to have similar behavior for unpacking packages.

Exactly.

In that case though, I think we should also remove the #[cfg(windows)] on this test.

Agreed.

eholk · 2022-03-04T23:45:14Z

I decided to close this PR in favor of #10461 which cleanly does what was decided in the discussion here.

eholk · 2022-03-07T18:14:54Z

Reopening this since there's disagreement about whether we want this change or the one in #10461. We only need one of these to land though.

joshtriplett · 2022-03-07T20:34:57Z

Per discussion on Zulip, this approach seems fine to me, as long as it doesn't actually use the same code in both this check and the code being tested (which would be circular).

In that spirit, and in the spirit of not providing this as a public API of Cargo, could you move this function into the test code rather than the util module?

eholk · 2022-03-07T22:09:35Z

In that spirit, and in the spirit of not providing this as a public API of Cargo, could you move this function into the test code rather than the util module?

Yup, that's done now! I moved it to the cargo_test_support crate, which required adding winapi as a dependency there too. Hopefully that's okay!

joshtriplett · 2022-03-07T22:18:24Z

Seems reasonable to me. Does anyone see an issue with adding that dependency to the test support crate?

ehuss

I'm sorry for the super long delay. I didn't have a Windows 11 system handy. I would be happy to get this merged with a small change.

crates/cargo-test-support/src/paths.rs

Recent versions of Windows have removed the limitation on filenames like `aux` or `con`. This change allows the `package::reserved_windows_name` to still pass by first trying to create a file with a reserved name to see if Windows supports it. If so, it skips the rest of the test. Otherwise, we keep the same behavior as before.

The previous commit tests whether the current machine supports Windows restricted names by creating a file and checking whether that succeeds. This commit reworks this testto use GetFullPathNameW, which can be done without having to create and remove new files.

ehuss · 2022-08-04T00:32:44Z

Thanks!

@bors r+

bors · 2022-08-04T00:32:45Z

📌 Commit 8de9adf has been approved by ehuss

It is now in the queue for this repository.

bors · 2022-08-04T00:32:52Z

⌛ Testing commit 8de9adf with merge 7259757...

bors · 2022-08-04T01:34:19Z

☀️ Test successful - checks-actions
Approved by: ehuss
Pushing 7259757 to master...

Update cargo 7 commits in 4fd148c47e733770c537efac5220744945d572ef..ce40690a5e4e315d3dab0aae1eae69d0252c52ac 2022-08-03 15:03:52 +0000 to 2022-08-09 22:32:17 +0000 - Make the `rust-version` error recommend `cargo update --precise -p crate@ver` (rust-lang/cargo#10891) - resolver docs: link to version requirements syntax full explanation (rust-lang/cargo#10946) - Bump os_info to 3.5.0 (rust-lang/cargo#10943) - Mark --timings=html unstable in the document (rust-lang/cargo#10941) - Mention that aliases are recursive (rust-lang/cargo#10935) - Test if reserved filenames are allowed in Windows (rust-lang/cargo#10322) - improve error message for `no such subcommand` (rust-lang/cargo#10924)

rust-highfive assigned ehuss Jan 24, 2022

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jan 24, 2022

weihanglo reviewed Jan 27, 2022

View reviewed changes

src/cargo/util/restricted_names.rs Outdated Show resolved Hide resolved

eholk mentioned this pull request Mar 4, 2022

Handle Windows-reserved paths consistently on all platforms #10461

Closed

eholk closed this Mar 4, 2022

eholk reopened this Mar 7, 2022

eholk force-pushed the reserved-windows-name branch from 46cbe37 to a1e0981 Compare March 7, 2022 22:08

ChrisDenton mentioned this pull request Mar 15, 2022

Improve Windows path prefix parsing rust-lang/rust#94887

Merged

ehuss reviewed Jul 29, 2022

View reviewed changes

crates/cargo-test-support/src/paths.rs Show resolved Hide resolved

eholk added 4 commits August 3, 2022 15:59

Update comment

d03530f

Move Windows path test function to test code

e9c7544

eholk force-pushed the reserved-windows-name branch from a1e0981 to e9c7544 Compare August 3, 2022 23:03

Use old Windows path handling on CI

8de9adf

bors removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Aug 4, 2022

bors added the S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. label Aug 4, 2022

bors merged commit 7259757 into rust-lang:master Aug 4, 2022

eholk mentioned this pull request Aug 8, 2022

Add min_$TARGET_API_version cfg predicates rust-lang/rfcs#3036

Closed

ehuss mentioned this pull request Aug 10, 2022

Update cargo rust-lang/rust#100352

Merged

ehuss added this to the 1.65.0 milestone Nov 6, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Test if reserved filenames are allowed in Windows #10322

Test if reserved filenames are allowed in Windows #10322

eholk commented Jan 24, 2022

rust-highfive commented Jan 24, 2022

weihanglo left a comment

eholk commented Jan 27, 2022

joshtriplett commented Mar 1, 2022

eholk commented Mar 2, 2022 •

edited

Loading

joshtriplett commented Mar 2, 2022 •

edited

Loading

eholk commented Mar 2, 2022 •

edited

Loading

joshtriplett commented Mar 2, 2022

eholk commented Mar 4, 2022

eholk commented Mar 7, 2022

joshtriplett commented Mar 7, 2022

eholk commented Mar 7, 2022

joshtriplett commented Mar 7, 2022

ehuss left a comment

ehuss commented Aug 4, 2022

bors commented Aug 4, 2022

bors commented Aug 4, 2022

bors commented Aug 4, 2022

Test if reserved filenames are allowed in Windows #10322

Test if reserved filenames are allowed in Windows #10322

Conversation

eholk commented Jan 24, 2022

rust-highfive commented Jan 24, 2022

weihanglo left a comment

Choose a reason for hiding this comment

eholk commented Jan 27, 2022

joshtriplett commented Mar 1, 2022

eholk commented Mar 2, 2022 • edited Loading

joshtriplett commented Mar 2, 2022 • edited Loading

eholk commented Mar 2, 2022 • edited Loading

joshtriplett commented Mar 2, 2022

eholk commented Mar 4, 2022

eholk commented Mar 7, 2022

joshtriplett commented Mar 7, 2022

eholk commented Mar 7, 2022

joshtriplett commented Mar 7, 2022

ehuss left a comment

Choose a reason for hiding this comment

ehuss commented Aug 4, 2022

bors commented Aug 4, 2022

bors commented Aug 4, 2022

bors commented Aug 4, 2022

eholk commented Mar 2, 2022 •

edited

Loading

joshtriplett commented Mar 2, 2022 •

edited

Loading

eholk commented Mar 2, 2022 •

edited

Loading