Restrict linker version script of proc-macro crates to just its two symbols

[pnkfelix]

Restrict linker version script of proc-macro crates to just the two symbols of each proc-macro crate.

The main known effect of doing this is to stop including #[no_mangle] symbols in the linker version script.

Background:

The combination of a proc-macro crate with an import of another crate that itself exports a no_mangle function was broken for a period of time, because:

• In PR Limit symbols exported from proc macros #99944 we stopped exporting no_mangle symbols from proc-macro crates; proc-macro crates have a very limited interface and are meant to be treated as a blackbox to everything except rustc itself. However: he constructed linker version script still referred to them, but resolving that discrepancy was left as a FIXME in the code, tagged with issue Don't export no_mangle symbols from proc macros #99978.
• In PR Add --no-undefined-version link flag and fix associated breakage #108017 we started telling the linker to check (via the--no-undefined-version linker invocation flag) that every symbol referenced in the "linker version script" is provided as linker input. So the unresolved discrepancy from Don't export no_mangle symbols from proc macros #99978 started surfacing as a compile-time error (e.g. regression - nightly 2023-03-30 -> 2023-03-31 causes linker error when importing wasm_bindgen::JsValue in a proc_macro #111888).

Fix #111888
Fix #99978.

[rustbot]

r? @davidtwco

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

[pnkfelix]

r? @bjorn3

[bjorn3]

@bors r+

[bors]

📌 Commit a2058dd has been approved by bjorn3

It is now in the queue for this repository.

[wesleywiser]

@bors p=1

Bumping priority as this fixes a P-high regression and is beta-nominated.

[bors]

⌛ Testing commit a2058dd with merge 0c68940e5658fa62941986c021cc6b447f48e281...

[bors]

💔 Test failed - checks-actions

[pnkfelix]

To only update this specific test, also pass `--test-args proc-macro/no-mangle-in-proc-macro-issue-111888.rs`

error: 1 errors occurred comparing output.
status: exit status: 0
command: RUSTC_ICE="0" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "/checkout/tests/ui/proc-macro/no-mangle-in-proc-macro-issue-111888.rs" "-Zthreads=1" "-Zsimulate-remapped-rust-src-base=/rustc/FAKE_PREFIX" "-Ztranslate-remapped-path-to-local-path=no" "--sysroot" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2" "--target=wasm32-unknown-unknown" "--error-format" "json" "--json" "future-incompat" "-Ccodegen-units=1" "-Zui-testing" "-Zdeduplicate-diagnostics=no" "-Zwrite-long-types-to-disk=no" "-Cstrip=debuginfo" "--out-dir" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/proc-macro/no-mangle-in-proc-macro-issue-111888" "-A" "unused" "-A" "internal_features" "-Crpath" "-Cdebuginfo=0" "-Lnative=/checkout/obj/build/wasm32-unknown-unknown/native/rust-test-helpers" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/proc-macro/no-mangle-in-proc-macro-issue-111888/auxiliary"
--- stderr -------------------------------
warning: dropping unsupported crate type `proc-macro` for target `wasm32-unknown-unknown`

warning: 1 warning emitted

Whoops, I guess I need to add wasm32-unknown-unknown to the targets I include in my local testing.

I'm skimming and I'm not seeing a clear pattern in how the other proc-macro crates in this directory avoid hitting this error. I'm guessing most of them are handling it via // force-host (and others via // ignore-wasm32), but it is strange that not all of the crate_type = "proc-macro" include at least one of those two header entries...

(oh, maybe the ones that don't need either of those two header entries are all compile-fail tests anyway... but derive-test.rs is still an exception to all of those conditions... I'm guessing --test is what sidesteps it in that case, I have no idea what --test will do on a proc-macro crate...)

[pnkfelix]

@bors r=bjorn3

[bors]

📌 Commit a2a7f27 has been approved by bjorn3

It is now in the queue for this repository.

[bjorn3]

I'm guessing --test is what sidesteps it in that case, I have no idea what --test will do on a proc-macro crate...

--test will force a crate type of bin. See collect_crate_types in rustc_interface::util.

[bors]

⌛ Testing commit a2a7f27 with merge b442d5cfa218c1af27110b73fe6e86e4dcb4ffe5...

[bors]

💔 Test failed - checks-actions

[rust-log-analyzer]

The job armhf-gnu failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[bjorn3]

thread 'main' panicked at 'client.read_exact(&mut header) failed with Connection reset by peer (os error 104)', src/tools/remote-test-client/src/main.rs:310:9

@bors retry spurious connection error for remote test client

[bors]

⌛ Testing commit a2a7f27 with merge a946c1e...

[bors]

☀️ Test successful - checks-actions
Approved by: bjorn3
Pushing a946c1e to master...

[rust-timer]

Finished benchmarking commit (a946c1e): comparison URL.

Overall result: ❌ regressions - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	1.0%	[0.3%, 1.6%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.6%	[2.5%, 2.6%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.1%	[-3.1%, -3.1%]	1
All ❌✅ (primary)	-	-	0

Cycles

This benchmark run did not return any relevant results for this metric.

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 631.766s -> 633.338s (0.25%)

[apiraino]

Beta backport approved as per compiler team on Zulip

@rustbot label +beta-accepted