Joshua Nelson [Mon, 1 Mar 2021 16:25:03 +0000 (11:25 -0500)]
Rollup merge of #82360 - osa1:issue82317, r=Mark-Simulacrum
config.toml parsing error improvements
Improve error messages for musl-libdir and wasi-root keys. Previously
the parser would panic with `unwrap()`. Now it prints
Target "wasm32-wasi" does not have a "wasi-root" key
(and similar for the `musl-libdir` field, which is used in target that
use musl)
Also update comments around wasi-root field to make it clear that the
field is only valid in wasm32-wasi target and needs to be moved to a
`[target.wasm32-wasi]` section to be valid.
bors [Mon, 1 Mar 2021 11:53:24 +0000 (11:53 +0000)]
Auto merge of #78360 - tmiasko:storage-markers, r=wesleywiser,oli-obk
Remove storage markers if they won't be used during code generation
The storage markers constitute a substantial portion of all MIR
statements. At the same time, for builds without any optimizations,
the storage markers have no further use during and after MIR
optimization phase.
If storage markers are not necessary for code generation, remove them.
bors [Mon, 1 Mar 2021 08:22:06 +0000 (08:22 +0000)]
Auto merge of #82654 - JohnTitor:rollup-nkcdkzp, r=JohnTitor
Rollup of 10 pull requests
Successful merges:
- #82309 (Propagate RUSTDOCFLAGS in the environment when documenting)
- #82403 (rustbuild: print out env vars on verbose rustc invocations)
- #82507 (Rename the `tidy` binary to `rust-tidy`)
- #82531 (Add GUI tests)
- #82532 (Add `build.print_step_rusage` to config.toml)
- #82543 (fix env var name in CI)
- #82622 (Propagate `--test-args` for `x.py test src/tools/cargo`)
- #82628 (Try to clarify GlobalAlloc::realloc documentation comment.)
- #82630 (Fix a typo in the `find_anon_type` doc)
- #82643 (Add more proc-macro attribute tests)
Improve error messages for musl-libdir and wasi-root keys. Previously
the parser would panic with `unwrap()`. Now it prints
Target "wasm32-wasi" does not have a "wasi-root" key
(and similar for the `musl-libdir` field, which is used in target that
use musl)
Also update comments around wasi-root field to make it clear that the
field is only valid in wasm32-wasi target and needs to be moved to a
`[target.wasm32-wasi]` section to be valid.
Yuki Okushi [Mon, 1 Mar 2021 06:07:38 +0000 (15:07 +0900)]
Rollup merge of #82628 - vakaras:realloc-doc, r=Mark-Simulacrum
Try to clarify GlobalAlloc::realloc documentation comment.
This PR tries to improve the documentation of [GlobalAlloc::realloc](https://doc.rust-lang.org/alloc/alloc/trait.GlobalAlloc.html#method.realloc) with two aspects:
1. Explicitly mention that `realloc` preserves the contents of the original memory block.
2. Explicitly mention which layout should be used to deallocate the reallocated block.
Yuki Okushi [Mon, 1 Mar 2021 06:07:35 +0000 (15:07 +0900)]
Rollup merge of #82532 - pnkfelix:rustbuild-print-step-rusage, r=Mark-Simulacrum
Add `build.print_step_rusage` to config.toml
Adds `build.print_step_rusage` to config.toml, which is meant to be an easy way to let compiler developers get feedback on the terminal during bootstrap about resource usage during each step.
The output is piggy-backed on `[PRINT-STEP-TIMINGS]`, mostly because the functionality seemed to naturally fit there in the overall control-flow and output structure (even if very little is shared between the implementations themselves).
Some sample output (from my Linux box, where I believe the `max rss` output to be somewhat trust-worthy...):
Yuki Okushi [Mon, 1 Mar 2021 06:07:34 +0000 (15:07 +0900)]
Rollup merge of #82531 - GuillaumeGomez:gui-tests-start, r=jyn514
Add GUI tests
The start of a lot more of GUI tests! \o/
One test is to ensure that the search input can always be selected in all rustdoc "modes" (mobile, tablet mostly) whereas the second checks the shortcuts.
Yuki Okushi [Mon, 1 Mar 2021 06:07:33 +0000 (15:07 +0900)]
Rollup merge of #82507 - jyn514:tidy-windows, r=Mark-Simulacrum
Rename the `tidy` binary to `rust-tidy`
This avoids naming collisions, particularly on Windows where the
dynamic library variable is PATH and setting it causes the in-tree
`tidy` to take precedence over the HTML tidy used by compiletest.
This doesn't change the x.py interface in any way, it still accepts `x.py test tidy` and prints error messages about `tidy`. It only changes the name of the file on disk.
Yuki Okushi [Mon, 1 Mar 2021 06:07:32 +0000 (15:07 +0900)]
Rollup merge of #82403 - pnkfelix:rustbuild-emit-env-vars-on-verbose-verbose, r=Mark-Simulacrum
rustbuild: print out env vars on verbose rustc invocations
Print out environment variables related to Rust on sufficiently verbose rustc invocations.
Output is filtered via heuristic of only printing environment variables whose keys start with "RUST" or "CARGO." This filtering is mostly motivated by my not caring to see e.g. "PATH" in my own output, though it is also motivated as a way to try to avoid printing out personal secrets like github keys that people might have stored in their environments for better or for worse, especially since build output is often pasted into bug reports or gists.
Fix #38686.
<details>
<summary>Click here to see sample output</summary>
Yuki Okushi [Mon, 1 Mar 2021 06:07:31 +0000 (15:07 +0900)]
Rollup merge of #82309 - jyn514:rustdocflags, r=Mark-Simulacrum
Propagate RUSTDOCFLAGS in the environment when documenting
Previously, RUSTDOCFLAGS would get overriden when bootstrap set
`RUSTDOCFLAGS` itself. Propagate the flag manually, using the same logic
as `RUSTFLAGS`.
bors [Mon, 1 Mar 2021 02:48:29 +0000 (02:48 +0000)]
Auto merge of #81094 - ssomers:btree_drainy_refactor_3, r=Mark-Simulacrum
BTreeMap: split up range_search into two stages
`range_search` expects the caller to pass the same root twice and starts searching a node for both bounds of a range. It's not very clear that in the early iterations, it searches twice in the same node. This PR splits that search up in an initial `find_leaf_edges_spanning_range` that postpones aliasing until the last second, and a second phase for continuing the search for the range in the each subtree independently (`find_lower_bound_edge` & `find_upper_bound_edge`), which greatly helps for use in #81075. It also moves those functions over to the search module.
bors [Mon, 1 Mar 2021 00:07:29 +0000 (00:07 +0000)]
Auto merge of #82601 - terhechte:catalyst-support-m1-fix, r=petrochenkov
Fixed support for macOS Catalyst on ARM64
When I initially added Arm64 Catalyst support in https://github.com/rust-lang/rust/pull/77484 I had access to a DTK. However, while waiting to merge the PR some other changes were merged which caused conflicts in the branch. When fixing those conflicts I had no access to the DTK anymore and didn't try out if the resulting binaries did indeed work on Apple Silicon. I finally have a M1 and I realized that some small changes were necessary to support Apple Silicon. This PR adds the required changes. I've been running binaries generated with this branch for some time now and they work without issues.
bors [Sun, 28 Feb 2021 09:18:27 +0000 (09:18 +0000)]
Auto merge of #82350 - ehuss:test-chapter, r=jyn514
Add a chapter on the test harness.
There isn't really any online documentation on the test harness, so this adds a chapter to the rustc book which provides information on how the harness works and details on the command-line options.
bors [Sun, 28 Feb 2021 06:37:18 +0000 (06:37 +0000)]
Auto merge of #82605 - cuviper:issue81051, r=nikic
Revert LLVM D81803 because it broke Windows 7
This submodule update reverts <https://reviews.llvm.org/D81803>.
While that change is meant to fix a real bug, [LLVM PR42623], it caused
new permission errors on Windows 7 that make it unable to build any
archives. This is probably the same root cause as [LLVM PR48378].
Fixes #81051. We'll file a new Rust issue to track the LLVM resolution.
bors [Sun, 28 Feb 2021 01:15:16 +0000 (01:15 +0000)]
Auto merge of #82611 - Dylan-DPC:rollup-l7xlpks, r=Dylan-DPC
Rollup of 11 pull requests
Successful merges:
- #81856 (Suggest character encoding is incorrect when encountering random null bytes)
- #82395 (Add missing "see its documentation for more" stdio)
- #82401 (Remove a redundant macro)
- #82498 (Use log level to control partitioning debug output)
- #82534 (Link crtbegin/crtend on musl to terminate .eh_frame)
- #82537 (Update measureme dependency to the latest version)
- #82561 (doc: cube root, not cubic root)
- #82563 (Fix intra-doc handling of `Self` in enum)
- #82584 (Add ARIA role to sidebar toggle in Rustdoc)
- #82596 (clarify RW lock's priority gotcha)
- #82607 (Add a getter for Frame.loc)
Tomasz Miąsko [Wed, 13 Jan 2021 00:00:00 +0000 (00:00 +0000)]
compiletest: run mir-opt tests with -C opt-level=1
The storage markers are removed at -C opt-level=0 and as a result output
of mir opt tests vary based on used optimization level. Use opt-level=1
for mir-opt tests to avoid the issue.
Tomasz Miąsko [Wed, 13 Jan 2021 00:00:00 +0000 (00:00 +0000)]
Remove storage markers if they won't be used during code generation
The storage markers constitute a substantial portion of all MIR
statements. At the same time, for builds without any optimizations,
the storage markers have no further use during and after MIR
optimization phase.
If storage markers are not necessary for code generation, remove them.
Dylan DPC [Sat, 27 Feb 2021 20:56:25 +0000 (21:56 +0100)]
Rollup merge of #82607 - bjorn3:frame_loc_getter, r=RalfJung
Add a getter for Frame.loc
This is necessary for Priroda.
For context see https://rust-lang.zulipchat.com/#narrow/stream/146212-t-compiler.2Fconst-eval/topic/Frame.3A.3Aloc.20no.20longer.20public/near/228070266 and oli-obk/priroda#27.
Context: I was completely mystified by a my CI deadlocking on mac ([here](https://github.com/matklad/xshell/pull/7)), until ``@azdavis`` debugged the issue. See a stand-alone reproduciton here: https://github.com/matklad/xshell/pull/15
Dylan DPC [Sat, 27 Feb 2021 20:56:23 +0000 (21:56 +0100)]
Rollup merge of #82584 - tazjin:rustdoc-aria, r=GuillaumeGomez
Add ARIA role to sidebar toggle in Rustdoc
This indicates that the div is an interactive element, and makes the sidebar toggle "clickable" in assistive technologies.
Example of Vimium after this change has been applied (see the issue mentioned below for a screenshot of before):
![Screenshot of Vimium link hints on a Rustdoc page, indicating that the sidebar toggle is clickable](https://user-images.githubusercontent.com/1552853/109384961-ff935400-78f8-11eb-8199-1d35181aeff0.png)
Dylan DPC [Sat, 27 Feb 2021 20:56:20 +0000 (21:56 +0100)]
Rollup merge of #82537 - wesleywiser:update_measureme, r=oli-obk
Update measureme dependency to the latest version
This version adds the ability to use `rdpmc` hardware-based performance
counters instead of wall-clock time for measuring duration. This also
introduces a dependency on the `perf-event-open-sys` crate on Linux
which is used when using hardware counters.
Dylan DPC [Sat, 27 Feb 2021 20:56:19 +0000 (21:56 +0100)]
Rollup merge of #82534 - nikic:musl-crtend, r=nagisa
Link crtbegin/crtend on musl to terminate .eh_frame
For some targets, rustc uses a "CRT fallback", where it links CRT
object files it ships instead of letting the host compiler link
them.
On musl, rustc currently links crt1, crti and crtn (provided by
libc), but does not link crtbegin and crtend (provided by libgcc).
In particular, crtend is responsible for terminating the .eh_frame
section. Lack of terminator may result in segfaults during
unwinding, as reported in #47551 and encountered by the LLVM 12
update in #81451.
This patch links crtbegin and crtend for musl as well, following
the table at the top of crt_objects.rs.
Dylan DPC [Sat, 27 Feb 2021 20:56:15 +0000 (21:56 +0100)]
Rollup merge of #81856 - Smittyvb:utf16-warn, r=matthewjasper
Suggest character encoding is incorrect when encountering random null bytes
This adds a note whenever null bytes are seen at the start of a token unexpectedly, since those tend to come from UTF-16 encoded files without a [BOM](https://en.wikipedia.org/wiki/Byte_order_mark) (if a UTF-16 BOM appears it won't be valid UTF-8, but if there is no BOM it be both valid UTF-16 and valid but garbled UTF-8). This approach was suggested in https://github.com/rust-lang/rust/issues/73979#issuecomment-653976451.
Josh Stone [Sat, 27 Feb 2021 19:09:00 +0000 (11:09 -0800)]
Revert LLVM D81803 because it broke Windows 7
This submodule update reverts <https://reviews.llvm.org/D81803>.
While that change is meant to fix a real bug, [LLVM PR42623], it caused
new permission errors on Windows 7 that make it unable to build any
archives. This is probably the same root cause as [LLVM PR48378].
Fixes #81051. We'll file a new Rust issue to track the LLVM resolution.
bors [Sat, 27 Feb 2021 07:52:11 +0000 (07:52 +0000)]
Auto merge of #82448 - Aaron1011:merge-hastokens-hasattrs, r=petrochenkov
Combine HasAttrs and HasTokens into AstLike
When token-based attribute handling is implemeneted in #80689,
we will need to access tokens from `HasAttrs` (to perform
cfg-stripping), and we will to access attributes from `HasTokens` (to
construct a `PreexpTokenStream`).
This PR merges the `HasAttrs` and `HasTokens` traits into a new
`AstLike` trait. The previous `HasAttrs` impls from `Vec<Attribute>` and `AttrVec`
are removed - they aren't attribute targets, so the impls never really
made sense.
Aaron Hill [Tue, 23 Feb 2021 15:21:20 +0000 (10:21 -0500)]
Combine HasAttrs and HasTokens into AstLike
When token-based attribute handling is implemeneted in #80689,
we will need to access tokens from `HasAttrs` (to perform
cfg-stripping), and we will to access attributes from `HasTokens` (to
construct a `PreexpTokenStream`).
This PR merges the `HasAttrs` and `HasTokens` traits into a new
`AstLike` trait. The previous `HasAttrs` impls from `Vec<Attribute>` and `AttrVec`
are removed - they aren't attribute targets, so the impls never really
made sense.
bors [Sat, 27 Feb 2021 05:06:05 +0000 (05:06 +0000)]
Auto merge of #82511 - jsha:fix-bfcache2, r=GuillaumeGomez
Fix back-forward cache in rustdoc frontend
Rustdoc's frontend set a no-op unload handler, specifically to disable
Firefox's back-forward cache because it caused a bug. It's nice to
allow the back-forward cache because it permits faster navigations.
This change addresses the issues that were caused by back-forward cache.
bors [Sat, 27 Feb 2021 02:18:11 +0000 (02:18 +0000)]
Auto merge of #82577 - Dylan-DPC:rollup-c3si8ju, r=Dylan-DPC
Rollup of 14 pull requests
Successful merges:
- #81794 (update tracking issue for `relaxed_struct_unsize`)
- #82057 (Replace const_cstr with cstr crate)
- #82370 (Improve anonymous lifetime note to indicate the target span)
- #82394 (:arrow_up: rust-analyzer)
- #82396 (Add Future trait for doc_spotlight feature doc)
- #82404 (Test hexagon-enum only when llvm target is present)
- #82419 (expand: Preserve order of inert attributes during expansion)
- #82420 (Enable API documentation for `std::os::wasi`.)
- #82421 (Add a `size()` function to WASI's `MetadataExt`.)
- #82442 (Skip emitting closure diagnostic when closure_kind_origins has no entry)
- #82473 (Use libc::accept4 on Android instead of raw syscall.)
- #82482 (Use small hash set in `mir_inliner_callees`)
- #82490 (Update cargo)
- #82494 (Substitute erased lifetimes on bad placeholder type)
Failed merges:
- #82448 (Combine HasAttrs and HasTokens into AstLike)
Dylan DPC [Sat, 27 Feb 2021 01:34:33 +0000 (02:34 +0100)]
Rollup merge of #82490 - ehuss:update-cargo, r=ehuss
Update cargo
11 commits in bf5a5d5e5d3ae842a63bfce6d070dfd438cf6070..572e201536dc2e4920346e28037b63c0f4d88b3c
2021-02-18 15:49:14 +0000 to 2021-02-24 16:51:20 +0000
- Pass the error message format to rustdoc (rust-lang/cargo#9128)
- Fix test target_in_environment_contains_lower_case (rust-lang/cargo#9203)
- Fix hang on broken stderr. (rust-lang/cargo#9201)
- Make it more clear which module is being tested when running cargo test (rust-lang/cargo#9195)
- Updates to edition handling. (rust-lang/cargo#9184)
- Add --cfg and --rustc-cfg flags to output compiler configuration (rust-lang/cargo#9002)
- Run rustdoc doctests relative to the workspace (rust-lang/cargo#9105)
- Add support for [env] section in .cargo/config.toml (rust-lang/cargo#9175)
- Add schema field and `features2` to the index. (rust-lang/cargo#9161)
- Document the default location where cargo install emitting build artifacts (rust-lang/cargo#9189)
- Do not exit prematurely if anything failed installing. (rust-lang/cargo#9185)
Dylan DPC [Sat, 27 Feb 2021 01:34:31 +0000 (02:34 +0100)]
Rollup merge of #82473 - de-vri-es:android-x86-accept4, r=m-ou-se
Use libc::accept4 on Android instead of raw syscall.
This PR replaces the use of a raw `accept4` syscall with `libc::accept4`. This was originally added (by me) because `std` couldn't update to the latest `libc` with `accept4` support for android. By now, libc is already on 0.2.85, so the workaround can be removed.
Dylan DPC [Sat, 27 Feb 2021 01:34:27 +0000 (02:34 +0100)]
Rollup merge of #82420 - sunfishcode:wasi-docs, r=alexcrichton
Enable API documentation for `std::os::wasi`.
This adds API documentation support for `std::os::wasi` modeled after
how `std::os::unix` works, so that WASI can be documented [here] along
with the other platforms.
- This changes the `AsRawFd` for `io::Stdin` for WASI to return
`libc::STDIN_FILENO` instead of `sys::stdio::Stdin.as_raw_fd()` (and
similar for `Stdout` and `Stderr`), which matches how the `unix`
version works. `STDIN_FILENO` etc. may not always be explicitly
reserved at the WASI level, but as long as we have Rust's `std` and
`libc`, I think it's reasonable to guarantee that we'll always use
`libc::STDIN_FILENO` for stdin.
- This duplicates the `osstr2str` utility function, rather than
trying to share it across all the configurations that need it.
Dylan DPC [Sat, 27 Feb 2021 01:34:22 +0000 (02:34 +0100)]
Rollup merge of #82370 - 0yoyoyo:update-issue-81650-point-anonymous-lifetime, r=estebank
Improve anonymous lifetime note to indicate the target span
Improvement for #81650
Cc #81995
Message after this improvement:
(Improve note in the middle)
```
error[E0311]: the parameter type `T` may not live long enough
--> src/main.rs:25:11
|
24 | fn play_with<T: Animal + Send>(scope: &Scope, animal: T) {
| -- help: consider adding an explicit lifetime bound...: `T: 'a +`
25 | scope.spawn(move |_| {
| ^^^^^
|
note: the parameter type `T` must be valid for the anonymous lifetime defined on the function body at 24:40...
--> src/main.rs:24:40
|
24 | fn play_with<T: Animal + Send>(scope: &Scope, animal: T) {
| ^^^^^
note: ...so that the type `[closure@src/main.rs:25:17: 27:6]` will meet its required lifetime bounds
--> src/main.rs:25:11
|
25 | scope.spawn(move |_| {
| ^^^^^
```
Dylan DPC [Sat, 27 Feb 2021 01:34:21 +0000 (02:34 +0100)]
Rollup merge of #82057 - upsuper-forks:cstr, r=davidtwco,wesleywiser
Replace const_cstr with cstr crate
This PR replaces the `const_cstr` macro inside `rustc_data_structures` with `cstr` macro from [cstr](https://crates.io/crates/cstr) crate.
The two macros basically serve the same purpose, which is to generate `&'static CStr` from a string literal. `cstr` is better because it validates the literal at compile time, while the existing `const_cstr` does it at runtime when `debug_assertions` is enabled. In addition, the value `cstr` generates can be used in constant context (which is seemingly not needed anywhere currently, though).
Guillaume Gomez [Fri, 26 Feb 2021 14:52:36 +0000 (15:52 +0100)]
Rollup merge of #82545 - jsha:woff2, r=GuillaumeGomez
rustdoc: add optional woff2 versions of FiraSans.
For browsers that support woff2 (most modern ones:
https://caniuse.com/woff2), this offers a reduction in download size
for these two fonts from 362k to 257k (32% reduction). It decreases the
total page size for `struct.String.html` (counting all subresources) by
about 2.5%.
If this is interesting, I'm happy to apply the same treatment to the
other fonts, but these two are the biggest.
Guillaume Gomez [Fri, 26 Feb 2021 14:52:35 +0000 (15:52 +0100)]
Rollup merge of #82535 - wesleywiser:wip_codegen_thread_names, r=nagisa
Set codegen thread names
Set names on threads spawned during codegen. Various debugging and profiling tools can take advantage of this to show a more useful identifier for threads.
For example, gdb will show thread names in `info threads`:
```
(gdb) info threads
Id Target Id Frame
1 Thread 0x7fffefa7ec40 (LWP 2905) "rustc" __pthread_clockjoin_ex (threadid=140737214134016, thread_return=0x0, clockid=<optimized out>, abstime=<optimized out>, block=<optimized out>)
at pthread_join_common.c:145
2 Thread 0x7fffefa7b700 (LWP 2957) "rustc" 0x00007ffff125eaa8 in llvm::X86_MC::initLLVMToSEHAndCVRegMapping(llvm::MCRegisterInfo*) ()
from /home/wesley/.rustup/toolchains/stage1/lib/librustc_driver-f866439e29074957.so
3 Thread 0x7fffeef0f700 (LWP 3116) "rustc" futex_wait_cancelable (private=0, expected=0, futex_word=0x7fffe8602ac8) at ../sysdeps/nptl/futex-internal.h:183
* 4 Thread 0x7fffeed0e700 (LWP 3123) "rustc" rustc_codegen_ssa::back::write::spawn_work (cgcx=..., work=...) at /home/wesley/code/rust/rust/compiler/rustc_codegen_ssa/src/back/write.rs:1573
6 Thread 0x7fffe113b700 (LWP 3150) "opt foof.7rcbfp" 0x00007ffff2940e62 in llvm::CallGraph::populateCallGraphNode(llvm::CallGraphNode*) ()
from /home/wesley/.rustup/toolchains/stage1/lib/librustc_driver-f866439e29074957.so
8 Thread 0x7fffe0d39700 (LWP 3158) "opt foof.7rcbfp" 0x00007fffefe8998e in malloc_consolidate (av=av@entry=0x7ffe2c000020) at malloc.c:4492
9 Thread 0x7fffe0f3a700 (LWP 3162) "opt foof.7rcbfp" 0x00007fffefef27c4 in __libc_open64 (file=0x7fffe0f38608 "foof.foof.7rcbfp3g-cgu.6.rcgu.o", oflag=524865) at ../sysdeps/unix/sysv/linux/open64.c:48
(gdb)
```
and Windows Performance Analyzer will also show this information when profiling:
And they try to compile it, they will see an error that looks like this:
```bash
error[E0308]: mismatched types
--> test.rs:4:5
|
1 | fn foo() -> u8 {
| -- expected `u8` because of return type
2 | async fn async_fn() -> () {}
| -- checked the `Output` of this `async fn`, found opaque type
3 |
4 | async_fn()
| ^^^^^^^^^^ expected `u8`, found opaque type
|
= note: while checking the return type of this `async fn`
= note: expected type `u8`
found opaque type `impl Future`
```
Nikita Popov [Fri, 26 Feb 2021 10:01:23 +0000 (11:01 +0100)]
Link crtbegin/crtend on musl to terminate .eh_frame
For some targets, rustc uses a "CRT fallback", where it links CRT
object files it ships instead of letting the host compiler link
them.
On musl, rustc currently links crt1, crti and crtn (provided by
libc), but does not link crtbegin and crtend (provided by libgcc).
In particular, crtend is responsible for terminating the .eh_frame
section. Lack of terminator may result in segfaults during
unwinding, as reported in #47551 and encountered by the LLVM 12
update in #81451.
This patch links crtbegin and crtend for musl as well, following
the table at the top of crt_objects.rs.
For browsers that support woff2 (most modern ones:
https://caniuse.com/woff2), this offers a reduction in download size
for these two fonts from 362k to 257k (32% reduction). It decreases the
total page size for `struct.String.html` (counting all subresources) by
about 2.5%.
If this is interesting, I'm happy to apply the same treatment to the
other fonts, but these two are the biggest.
It seemed pretty peculiar to specifically test lang strings in braces, with all the tokens prefixed by `.`.
I did some digging, and it looks like the test cases were added way back in [this commit from 2014](https://github.com/rust-lang/rust/commit/3fef7a74ca9a) by `@skade.`
It looks like they were added just to make sure that the splitting was permissive, and aren't testing that those strings in particular are accepted.