Update cc crate for bootstrap to v1.0.97

[jfgoog]

Reason:

In order to build the Windows version of the Rust toolchain for the Android platform, the following patch to the cc is crate is required to avoid incorrectly determining that we are building with the Android NDK: rust-lang/cc-rs@57853c4

This patch is present in version 1.0.80 and newer versions of the cc crate. The rustc source distribution currently has 3 different versions of cc in the vendor directory, only one of which has the necessary fix.

We (the Android Rust toolchain) are currently maintaining local patches to upgrade the cc crate dependency versions, which we would like to upstream.

Furthermore, beyond the specific reason, the cc crate in bootstrap is currently pinned at an old version due to problems in the past when trying to update it. It is worthwhile to figure out and resolve these problems so we can keep the dependency up-to-date.

Other fixes:

As of cc v1.0.78, object files are prefixed with a 16-character hash.
Update src/bootstrap/src/core/build_steps/llvm.rs to account for this to
avoid failures when building libunwind and libcrt. Note that while the hash
prefix was introduced in v1.0.78, in order to determine the names of the
object files without scanning the directory, we rely on the compile_intermediates
method, which was introduced in cc v1.0.86

As of cc v1.0.86, compilation on MacOS uses the -mmacosx-version-min flag.
A long-standing bug in the CMake rules for compiler-rt causes compilation
to fail when this flag is specified. So we add a workaround to suppress this
flag.

Updating to cc v1.0.91 and newer requires fixes to bootstrap unit tests.
The unit tests use targets named "A", "B", etc., which fail a validation
check introduced in 1.0.91 of the cc crate.

As of cc v1.0.74, the SDKROOT environment variable is used on Darwin if present,
regardless of whether it is for the correct platform or not. This is fixed in cc v1.0.97.

[rustbot]

r? @albertlarsan68

rustbot has assigned @albertlarsan68.
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

[jfgoog]

As noted in the change description, I'm hoping to get all cc crate dependencies for rustc updated to at least 1.0.80, which has a fix we need for building the Rust toolchain for the Android platform.

I saw the comment in Cargo.toml, and I understand there is a good reason for pinning the version. I am happy to work on any fixes to the bootstrap code that are necessary to land this change.

I set the version to 1.0.90, which is the most recent, but I'd be happy with anything >=1.0.80, if that would be preferable.

[workingjubilee]

@jfgoog Please see my comments in #122498

[workingjubilee]

Bootstrap has its own lockfile, I see. I didn't notice that before. It will need the .lock updated too.

[jfgoog]

Here are all 3 PRs I've sent:

#122498
rust-lang/backtrace-rs#592
#122504

[workingjubilee]

@jfgoog It should just be a matter of using cargo update -p cc in the bootstrap dir to bump the .lock forward.

[jfgoog]

@jfgoog It should just be a matter of using cargo update -p cc in the bootstrap dir to bump the .lock forward.

Done.

[workingjubilee]

Cool.

@albertlarsan68 I will leave this to you but I have to wonder if there is any chance of bootstrap using the workspace cc version? Perhaps we should pursue that conversation separately.

[jfgoog]

The job mingw-check-tidy failed! Check out the build log: (web) (plain)

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

I'm not sure what to do about this failure.

[workingjubilee]

???

@NobodyXu This is a peculiar error message, "Compiler version doesn't include clang or GCC"?

[jfgoog]

???

@NobodyXu This is a peculiar error message, "Compiler version doesn't include clang or GCC"?

I have a suspicion that things like this are the reason that the version is pinned.

[workingjubilee]

Apparently!

[jfgoog]

More conservative version, updating to 1.0.80 instead of all the way to 1.0.90: #122507

Let's see if that fares better in presubmit checks.

[lukas-code]

The Compiler version doesn't include clang or GCC is #122231 / rust-lang/cc-rs#958 and should be fine to ignore.

The actual error is x.py completions were changed; run `x.py run generate-completions` to update them which probably happens, because you also updated clap_complete with the Cargo.lock bump.

[klensy]

This pr accidentally bumped not only cc crate; plus see #119654 where cc bump was attempted and failed.

[workingjubilee]

Ahhh.

[workingjubilee]

@jfgoog Can you back out the other .lock changes and retry the cargo update cc?

[jfgoog]

@jfgoog Can you back out the other .lock changes and retry the cargo update cc?

Done

[jfgoog]

The latest error is not being able to find errno.h in "Building stage0 library artifacts (x86_64-apple-darwin)"

[onur-ozkan]

I am unsure if that's the actual problem.

[jfgoog]

I suspect that I caused this by clearing SDKROOT. There are workarounds in multiple places to ignore SDKROOT if it's clearly set for the wrong platform:

• compiler/rustc_target/src/spec/base/apple/mod.rs
• compiler/rustc_codegen_ssa/src/back/link.rs

Maybe the cc crate needs to do the same thing?

[jfgoog]

I updated #124565 with an explanation of what I think is going on.

[onur-ozkan]

Can we somehow set the correct platform instead of removing SDKROOT?

[jfgoog]

I've been trying to avoid it until now, but I feel like fixing the SDKROOT problem has to be done upstream in the cc crate. I don't see how we can detect every instance where we are cross-compiling, clear SDKROOT, and then presumably restore it afterwards.

[jfgoog]

Can we somehow set the correct platform instead of removing SDKROOT?

I don't think so, for two reasons:

1. The SDKROOT has to be set to a valid MacOSX SDK in order for our build of clang to be able to find headers and libraries.
2. When building for multiple platforms (e.g. iOS and watchOS) , which is common, there is no single correct value of SDKROOT that we can use.

[bors]

☔ The latest upstream changes (presumably #124726) made this pull request unmergeable. Please resolve the merge conflicts.

[jfgoog]

rust-lang/cc-rs#1047 is merged

[jfgoog]

Updated to the latest cc v1.0.97 to pick up the fix to SDKROOT.

[onur-ozkan]

rust-lang/cc-rs#1047 is merged

Updated to the latest cc v1.0.97 to pick up the fix to SDKROOT.

Thanks a lot for working on this!

@bors r+ rollup=never

[bors]

📌 Commit 615b485 has been approved by onur-ozkan

It is now in the queue for this repository.

[bors]

⌛ Testing commit 615b485 with merge e2865db...

[bors]

☀️ Test successful - checks-actions
Approved by: onur-ozkan
Pushing e2865db to master...

[rust-timer]

Finished benchmarking commit (e2865db): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

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

Max RSS (memory usage)

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

Cycles

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)	2.8%	[2.1%, 3.8%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 676.36s -> 675.387s (-0.14%)
Artifact size: 315.83 MiB -> 315.97 MiB (0.04%)

[jfgoog]

Holy crap...it merged!