run-make regression test for issue #70924.

[pnkfelix]

Sometime after my PR #72767 (to fix issue #70924) landed, I realized that I could make a local regression test, thanks to rustc --print sysroot: I can make a fresh "copy" (really mostly symlinks) of the sysroot, and then modify it to recreate the terms of this bug.

[rust-highfive]

r? @nikomatsakis

(rust_highfive has picked a reviewer for you, use r? to override)

[pnkfelix]

(okay it looks like this test won't break the windows build. I'll remove the test commit pnkfelix@56d185f now )

[nikomatsakis]

@bors r+ rollup

[bors]

📌 Commit 94605b9 has been approved by nikomatsakis

[mati865]

Failed in rollup: https://dev.azure.com/rust-lang/rust/_build/results?buildId=31133&view=logs&j=a6a2dd6f-5140-5fa9-b933-dfe57be2bdd5&t=53c32ceb-c4c2-54e8-ee5b-d47a390af72f

make[1]: Entering directory '/d/a/1/s/src/test/run-make-fulldeps/incr-add-rust-src-component'
mkdir /d/a/1/s/build/x86_64-pc-windows-msvc/test/run-make-fulldeps/incr-add-rust-src-component/incr-add-rust-src-component/fakeroot
ln -s C:\MORE_SPACE\x86_64-pc-windows-msvc\stage2/* /d/a/1/s/build/x86_64-pc-windows-msvc/test/run-make-fulldeps/incr-add-rust-src-component/incr-add-rust-src-component/fakeroot
make[1]: Leaving directory '/d/a/1/s/src/test/run-make-fulldeps/incr-add-rust-src-component'

------------------------------------------
stderr:
------------------------------------------
ln: failed to create symbolic link '/d/a/1/s/build/x86_64-pc-windows-msvc/test/run-make-fulldeps/incr-add-rust-src-component/incr-add-rust-src-component/fakeroot/*': No such file or directory
make[1]: *** [Makefile:25: all] Error 1

[Dylan-DPC-zz]

@bors r-

[pnkfelix]

argh, I specifically thought I tested windows-msvc in pnkfelix@56d185f

I'll see what I can do about making this test only run on compatible platforms. And also check with the infrastructure team about what I did wrong in my attempt to do a test run.

[pnkfelix]

@bors r=nikomatsakis rollup

[bors]

📌 Commit a9d5dff has been approved by nikomatsakis

[pnkfelix]

argh, I specifically thought I tested windows-msvc in pnkfelix@56d185f

I'll see what I can do about making this test only run on compatible platforms. And also check with the infrastructure team about what I did wrong in my attempt to do a test run.

(the answer to what I did wrong: I added my test commit, but I never actually asked bors to do a try run! 🤦)

[RalfJung]

This is one of only a few PRs that are in the intersection of 4 rollups that all failed the same way: #72952, #72977, #73007.

The failure is on 32bit Windows, a bunch of tests show

------------------------------------------
stderr:
------------------------------------------
thread 'rustc' panicked at 'assertion failed: self.ptr <= self.end', D:\a\1\s\src\libstd\macros.rs:13:23
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

The line number is utterly useless, it points to the macro definition (how is that possible?!?).

Affected tests:

failures:
    [ui] ui\fn\expr-fn-panic.rs
    [ui] ui\imports\glob-use-std.rs
    [ui] ui\macros\die-macro-2.rs
    [ui] ui\macros\die-macro-pure.rs
    [ui] ui\macros\unimplemented-macro-panic.rs
    [ui] ui\macros\unreachable-macro-panic.rs
    [ui] ui\macros\unreachable.rs
    [ui] ui\meta-revision-ok.rs#bar
    [ui] ui\meta-revision-ok.rs#foo
    [ui] ui\mir\mir_codegen_calls_diverging.rs
    [ui] ui\panics\doublepanic.rs
    [ui] ui\panics\explicit-panic.rs
    [ui] ui\panics\main-panic.rs
    [ui] ui\panics\panic-macro-explicit.rs
    [ui] ui\panics\panic-macro-owned.rs
    [ui] ui\panics\panic-macro-static.rs
    [ui] ui\panics\panic-main.rs
    [ui] ui\rfc-1937-termination-trait\termination-trait-for-never.rs

All 3 PRs in the intersection seem impossible to be the cause here, so I'd appreciate help with figuring out what is going on.^^

[Dylan-DPC-zz]

I am going to put 2 of them in a rollup so we can bisect (or trisect) which PR is causing the failure

[Dylan-DPC-zz]

@bors rollup=maybe