Commit Graph

174 Commits

Author SHA1 Message Date
Erik Desjardins
a19138f1f9 cleanup: remove unused ability to have LLVM null-terminate const strings 2022-03-06 12:28:46 -05:00
Amanieu d'Antras
6fcfc3d877 Add -Z oom={panic,abort} command-line option 2022-03-03 12:58:38 +00:00
cuishuang
8238b91402 all: fix some typos
Signed-off-by: cuishuang <imcusg@gmail.com>
2022-03-03 19:47:23 +08:00
mark
081bf75aea rename ErrorReported -> ErrorGuaranteed 2022-03-02 09:45:25 -06:00
bors
7df92e44c0 Auto merge of #87402 - nagisa:nagisa/request-feature-requests-for-features, r=estebank
Direct users towards using Rust target feature names in CLI

This PR consists of a couple of changes on how we handle target features.

In particular there is a bug-fix wherein we avoid passing through features that aren't prefixed by `+` or `-` to LLVM. These appear to be causing LLVM to assert, which is pretty poor a behaviour (and also makes it pretty clear we expect feature names to be prefixed).

The other commit, I anticipate to be somewhat more controversial is outputting a warning when users specify a LLVM-specific, or otherwise unknown, feature name on the CLI. In those situations we request users to either replace it with a known Rust feature name (e.g. `bmi` -> `bmi1`) or file a feature request. I've a couple motivations for this: first of all, if users are specifying these features on the command line, I'm pretty confident there is also a need for these features to be usable via `#[cfg(target_feature)]` machinery.  And second, we're growing a fair number of backends recently and having ability to provide some sort of unified-ish interface in this place seems pretty useful to me.

Sponsored by: standard.ai
2022-03-02 03:03:22 +00:00
Simonas Kazlauskas
4f132cb340 Querify global_backend_features
At the very least this serves to deduplicate the diagnostics that are
output about unknown target features provided via CLI.
2022-03-01 01:57:25 +02:00
Erik Desjardins
962b8cf38b Revert "Auto merge of #92419 - erikdesjardins:coldland, r=nagisa"
This reverts commit 4f49627c6f, reversing
changes made to 028c6f1454.
2022-02-27 23:11:03 -05:00
bors
648d038c86 Auto merge of #94123 - bjorn3:cg_ssa_singleton_builder, r=tmiasko
Partially move cg_ssa towards using a single builder

Not all codegen backends can handle hopping between blocks well. For example Cranelift requires blocks to be terminated before switching to building a new block. Rust-gpu requires a `RefCell` to allow hopping between blocks and cg_gcc currently has a buggy implementation of hopping between blocks. This PR reduces the amount of cases where cg_ssa switches between blocks before they are finished and mostly fixes the block hopping in cg_gcc. (~~only `scalar_to_backend` doesn't handle it correctly yet in cg_gcc~~ fixed that one.)

`@antoyo` please review the cg_gcc changes.
2022-02-24 12:28:19 +00:00
bjorn3
18c34e41a4 Introduce Bx::switch_to_block 2022-02-24 12:18:21 +01:00
bjorn3
56b7080bdb Remove build_sibling_block 2022-02-20 13:38:15 +01:00
Mark Rousskov
e855e2d15c Move ty::print methods to Drop-based scope guards 2022-02-16 17:24:23 -05:00
bjorn3
40d30ceb43 Unconditionally update symbols
All paths to an ArchiveBuilder::build call update_symbols first.
2022-02-10 18:27:18 +01:00
bors
22a30689ed Auto merge of #93511 - cjgillot:query-copy, r=oli-obk
Ensure that queries only return Copy types.

This should pervent the perf footgun of returning a result with an expensive `Clone` impl (like a `Vec` of a hash map).

I went for the stupid solution of allocating on an arena everything that was not `Copy`. Some query results could be made Copy easily, but I did not really investigate.
2022-02-10 09:37:07 +00:00
Camille GILLOT
1b2337c19d Make FnAbiError Copy. 2022-02-09 20:11:29 +01:00
Camille GILLOT
96340e5668 Ensure that queries only return Copy types. 2022-02-09 20:07:38 +01:00
cynecx
20506e31ab #[used(linker)] attribute (https://github.com/dtolnay/linkme/issues/41) 2022-02-06 20:23:23 +01:00
bjorn3
4bdc78436b Merge landing_pad and set_cleanup into cleanup_landing_pad 2022-01-24 14:10:05 +01:00
bjorn3
2eaf0bc20a Merge add_handler into catch_switch
Some codegen backends may require all handlers to be immediately known
2022-01-24 14:10:05 +01:00
bjorn3
81597f2721 Remove unused return values from resume and cleanup_ret
Given that these instructions are diverging, not every codegen backend
may be able to produce a return value for them.
2022-01-24 13:48:09 +01:00
bjorn3
ba419539d5 Reorder unwinding related builder methods to differentiate between dwarf and msvc instructions 2022-01-24 13:45:34 +01:00
William D. Jones
94dc3753aa Add preliminary support for inline assembly for msp430. 2022-01-22 23:42:46 -05:00
bjorn3
47c311a705 Use Symbol for target features in asm handling
This saves a couple of Symbol::intern calls
2022-01-17 18:06:27 +01:00
Tomasz Miąsko
68cbb46913 Remove deprecated LLVM-style inline assembly 2022-01-12 18:51:31 +01:00
Aaron Hill
09afd64969 Store a Symbol instead of an Ident in VariantDef/FieldDef
The field is also renamed from `ident` to `name. In most cases,
we don't actually need the `Span`. A new `ident` method is added
to `VariantDef` and `FieldDef`, which constructs the full `Ident`
using `tcx.def_ident_span()`. This method is used in the cases
where we actually need an `Ident`.

This makes incremental compilation properly track changes
to the `Span`, without all of the invalidations caused by storing
a `Span` directly via an `Ident`.
2022-01-11 10:16:22 -05:00
Josh Triplett
37069f1a08 Update references to -Z symbol-mangling-version to use -C
Replace `-Z symbol-mangling-version=v0` with `-C symbol-mangling-version=v0`.

Replace `-Z symbol-mangling-version=legacy` with
`-Z unstable-options -C symbol-mangling-version=legacy`.
2022-01-01 15:53:11 -08:00
bors
da4a39264b Auto merge of #92419 - erikdesjardins:coldland, r=nagisa
Mark drop calls in landing pads `cold` instead of `noinline`

Now that deferred inlining has been disabled in LLVM (#92110), this shouldn't cause catastrophic size blowup.

I confirmed that the test cases from https://github.com/rust-lang/rust/issues/41696#issuecomment-298696944 still compile quickly (<1s) after this change. ~Although note that I wasn't able to reproduce the original issue using a recent rustc/llvm with deferred inlining enabled, so those tests may no longer be representative. I was also unable to create a modified test case that reproduced the original issue.~ (edit: I reproduced it on CI by accident--the first commit timed out on the LLVM 12 builder, because I forgot to make it conditional on LLVM version)

r? `@nagisa`
cc `@arielb1` (this effectively reverts #42771 "mark calls in the unwind path as !noinline")
cc `@RalfJung` (fixes #46515)

edit: also fixes #87055
2022-01-01 13:28:13 +00:00
bjorn3
e6e46055df Remove unused extern crate rustc_symbol_mangling 2021-12-31 16:57:33 +01:00
bjorn3
54d2ec1a82 Merge commit '1411a98352ba6bee8ba3b0131c9243e5db1e6a2e' into sync_cg_clif-2021-12-31 2021-12-31 16:26:32 +01:00
Erik Desjardins
f3e1f33e3c keep noinline for system llvm < 14 2021-12-30 00:15:51 -05:00
Erik Desjardins
2fec53053e Mark drop calls in landing pads cold instead of noinline
Now that deferred inlining has been disabled in LLVM,
this shouldn't cause catastrophic size blowup.
2021-12-29 15:47:49 -05:00
bors
c5d0023afb Auto merge of #91957 - nnethercote:rm-SymbolStr, r=oli-obk
Remove `SymbolStr`

This was originally proposed in https://github.com/rust-lang/rust/pull/74554#discussion_r466203544. As well as removing the icky `SymbolStr` type, it allows the removal of a lot of `&` and `*` occurrences.

Best reviewed one commit at a time.

r? `@oli-obk`
2021-12-19 09:31:37 +00:00
Nicholas Nethercote
bba8b97b83 Remove unnecessary sigils around Symbol::as_str() calls. 2021-12-15 17:32:14 +11:00
Matthias Krüger
f87a8f88e3 Rollup merge of #91868 - tmiasko:llvm-time-trace-out, r=oli-obk
Use `OutputFilenames` to generate output file for `-Zllvm-time-trace`

The resulting profile will include the crate name and will be stored in
the `--out-dir` directory.

This implementation makes it convenient to use LLVM time trace together
with cargo, in the contrast to the previous implementation which would
overwrite profiles or store them in `.cargo/registry/..`.
2021-12-15 01:28:06 +01:00
Tomasz Miąsko
379fb40277 Use OutputFilenames to generate output file for -Zllvm-time-trace
The resulting profile will include the crate name and will be stored in
the `--out-dir` directory.

This implementation makes it convenient to use LLVM time trace together
with cargo, in the contrast to the previous implementation which would
overwrite profiles or store them in `.cargo/registry/..`.
2021-12-13 00:00:00 +00:00
Amanieu d'Antras
eec5f919e7 Stabilize asm! and global_asm!
They are also removed from the prelude as per the decision in
https://github.com/rust-lang/rust/issues/87228.

stdarch and compiler-builtins are updated to work with the new, stable
asm! and global_asm! macros.
2021-12-12 11:20:03 +00:00
Matthias Krüger
c6009ca30f Rollup merge of #91625 - est31:remove_indexes, r=oli-obk
Remove redundant [..]s
2021-12-10 22:40:36 +01:00
Matthias Krüger
10a9b1202b Rollup merge of #90796 - Amanieu:remove_reg_thumb, r=joshtriplett
Remove the reg_thumb register class for asm! on ARM

Also restricts r8-r14 from being used on Thumb1 targets as per #90736.

cc ``@Lokathor``

r? ``@joshtriplett``
2021-12-09 05:08:32 +01:00
est31
b1d329ba55 Remove redundant [..]s 2021-12-09 00:01:29 +01:00
bors
2e7bc96321 Auto merge of #91604 - nikic:section-flags, r=nagisa
Use object crate for .rustc metadata generation

We already use the object crate for generating uncompressed .rmeta
metadata object files. This switches the generation of compressed
.rustc object files to use the object crate as well. These have
slightly different requirements in that .rmeta should be completely
excluded from any final compilation artifacts, while .rustc should
be part of shared objects, but not loaded into memory.

The primary motivation for this change is #90326: In LLVM 14, the
current way of setting section flags (and in particular, preventing
the setting of SHF_ALLOC) will no longer work. There are other ways
we could work around this, but switching to the object crate seems
like the most elegant, as we already use it for .rmeta, and as it
makes this independent of the codegen backend. In particular, we
don't need separate handling in codegen_llvm and codegen_gcc.
codegen_cranelift should be able to reuse the implementation as
well, though I have omitted that here, as it is not based on
codegen_ssa.

This change mostly extracts the existing code for .rmeta handling
to allow using it for .rustc as well, and adjusts the codegen
infrastructure to handle the metadata object file separately: We
no longer create a backend-specific module for it, and directly
produce the compiled module instead.

This does not `fix` #90326 by itself yet, as .llvmbc will need to be
handled separately.

r? `@nagisa`
2021-12-08 14:58:48 +00:00
Amanieu d'Antras
0d936e19c9 Remove the reg_thumb register class for asm! on ARM
Also restricts r8-r14 from being used on Thumb1 targets as per #90736.
2021-12-07 23:54:09 +00:00
Nikita Popov
98afc30b95 Use object crate for .rustc metadata generation
We already use the object crate for generating uncompressed .rmeta
metadata object files. This switches the generation of compressed
.rustc object files to use the object crate as well. These have
slightly different requirements in that .rmeta should be completely
excluded from any final compilation artifacts, while .rustc should
be part of shared objects, but not loaded into memory.

The primary motivation for this change is #90326: In LLVM 14, the
current way of setting section flags (and in particular, preventing
the setting of SHF_ALLOC) will no longer work. There are other ways
we could work around this, but switching to the object crate seems
like the most elegant, as we already use it for .rmeta, and as it
makes this independent of the codegen backend. In particular, we
don't need separate handling in codegen_llvm and codegen_gcc.
codegen_cranelift should be able to reuse the implementation as
well, though I have omitted that here, as it is not based on
codegen_ssa.

This change mostly extracts the existing code for .rmeta handling
to allow using it for .rustc as well, and adjust the codegen
infrastructure to handle the metadata object file separately: We
no longer create a backend-specific module for it, and directly
produce the compiled module instead.

This does not fix #90326 by itself yet, as .llvmbc will need to be
handled separately.
2021-12-07 09:39:05 +01:00
Andrew Dona-Couch
4e68093d8c Implement inline asm! for AVR platform 2021-12-06 01:02:49 -05:00
cynecx
4fa64fd1c3 rustc_codegen_gcc: proper check for may_unwind 2021-12-03 23:51:49 +01:00
cynecx
b7cb08d4e7 rustc_codegen_gcc: error on unwinding inline asm 2021-12-03 23:51:49 +01:00
bors
48fc7d9351 Auto merge of #91003 - psumbera:sparc64-abi, r=nagisa
fix sparc64 ABI for aggregates with floating point members

Fixes #86163
2021-12-02 02:59:44 +00:00
Petr Sumbera
dcd48250d7 fix sparc64 ABI for aggregates with floating point members 2021-12-01 10:03:45 +01:00
rdambrosio
0d61135280 Feat: make cg_ssa get_param borrow the builder mutable 2021-11-23 22:30:20 -05:00
Andreas Jonson
a912fedd0f Remove workaround for the forward progress handling in LLVM 2021-11-14 16:35:09 +01:00
bors
d5a6aafb88 Auto merge of #89652 - rcvalle:rust-cfi, r=nagisa
Add LLVM CFI support to the Rust compiler

This PR adds LLVM Control Flow Integrity (CFI) support to the Rust compiler. It initially provides forward-edge control flow protection for Rust-compiled code only by aggregating function pointers in groups identified by their number of arguments.

Forward-edge control flow protection for C or C++ and Rust -compiled code "mixed binaries" (i.e., for when C or C++ and Rust -compiled code share the same virtual address space) will be provided in later work as part of this project by defining and using compatible type identifiers (see Type metadata in the design document in the tracking issue #89653).

LLVM CFI can be enabled with -Zsanitizer=cfi and requires LTO (i.e., -Clto).

Thank you, `@eddyb` and `@pcc,` for all the help!
2021-10-27 09:19:42 +00:00
Yuki Okushi
8094e2e5ae Properly check target_features not to trigger an assertion 2021-10-26 11:02:51 +09:00