36 Commits

Author SHA1 Message Date
bjorn3
9d25e851ce Remove const_bitcast from ConstMethods 2024-06-21 19:26:07 +00:00
Ralf Jung
d565d3d23d ScalarInt: size mismatches are a bug, do not delay the panic 2024-06-10 13:43:16 +02:00
Scott McMurray
1a05106f94 Add+Use mir::BinOp::Cmp 2024-03-23 23:23:41 -07:00
Guillaume Gomez
faebf73983 Merge commit 'b385428e3ddf330805241e7758e773f933357c4b' into subtree-update_cg_gcc_2024-03-05 2024-03-05 19:58:36 +01:00
r0cky
92de9d4210 Remove dead codes 2023-12-11 23:24:32 +08:00
Ralf Jung
cb32ffd8da ctfe interpreter: extend provenance so that it can track whether a pointer is immutable 2023-12-07 17:46:36 +01:00
Antoni Boucher
9f4f90b19a Merge commit 'e4fe941b11a55c5005630696e9b6d81c65f7bd04' into subtree-update_cg_gcc_2023-10-25 2023-10-26 17:42:02 -04:00
Erik Desjardins
bc4a65d47f cg_ssa: remove pointee types and pointercast/bitcast-of-ptr 2023-07-29 13:18:20 -04:00
DonoughLiu
9176f51a9b Support 128-bit enum variant in debuginfo codegen 2023-06-10 03:39:24 +08:00
Oli Scherer
80efecf18c Stop creating intermediate places just to immediate convert them to operands 2023-05-26 15:01:29 +00:00
Nikita Popov
0c115bf8b8 Use poison instead of undef
In cases where it is legal, we should prefer poison values over
undef values.

This replaces undef with poison for aggregate construction and
for uninhabited types. There are more places where we can likely
use poison, but I wanted to stay conservative to start with.

In particular the aggregate case is important for newer LLVM
versions, which are not able to handle an undef base value during
early optimization due to poison-propagation concerns.
2023-03-16 15:07:04 +01:00
Antoni Boucher
d725cfb6ab Merge commit '08a6d6e16b5efe217123e780398969946266268f' into sync-cg_gcc-2023-03-04 2023-03-05 12:03:19 -05:00
Erik Desjardins
559b57e50a abi: add AddressSpace field to Primitive::Pointer
...and remove it from `PointeeInfo`, which isn't meant for this.

There are still various places (marked with FIXMEs) that assume all pointers
have the same size and alignment. Fixing this requires parsing non-default
address spaces in the data layout string, which will be done in a followup.
2023-01-22 23:41:39 -05:00
Jeremy Stucki
2d09c9bc94 Add missing anonymous lifetime 2022-12-20 22:34:42 +01:00
Jeremy Stucki
b4e9ba5cf3 rustc: Remove needless lifetimes 2022-12-20 22:10:40 +01:00
Oli Scherer
c3dce60ac7 Remove dead broken code from const zst handling in backends 2022-09-06 14:09:49 +00:00
Ralf Jung
59d223d1ed slightly cleaner, if more verbose, vtable handling in codegen backends 2022-07-20 17:12:07 -04:00
Ralf Jung
3f24eb870d consistently use VTable over Vtable (matching stable stdlib API RawWakerVTable) 2022-07-20 17:12:07 -04:00
Ralf Jung
8f290184d2 add a Vtable kind of symbolic allocations 2022-07-20 16:57:31 -04:00
Ralf Jung
33e8d4c2d4 fix cranelift and gcc backends 2022-07-09 07:27:29 -04:00
bjorn3
c0ec18bd0d Avoid unnecessary string interning for const_str 2022-06-28 18:38:36 +00:00
Antoni Boucher
fac57d9a06 Merge commit 'e8dca3e87d164d2806098c462c6ce41301341f68' into sync_from_cg_gcc 2022-06-06 22:04:37 -04:00
Oli Scherer
d71f633d38 Mark scalar layout unions so that backends that do not support partially initialized scalars can special case them. 2022-04-05 13:18:21 +00:00
David Morrison
88c058b614 make memcmp return a value of c_int_width instead of i32 2022-04-02 17:21:08 -07:00
bjorn3
3888aafe3a Merge commit '39683d8eb7a32a74bea96ecbf1e87675d3338506' into sync_cg_gcc-2022-03-26 2022-03-26 18:29:37 +01:00
bors
0e36868bc7 Auto merge of #94638 - erikdesjardins:noextranull, r=nagisa
cleanup: remove unused ability to have LLVM null-terminate const strings

(and the copied function in rustc_codegen_gcc)

Noticed this while writing https://github.com/rust-lang/rust/pull/94450#issuecomment-1059687348.

r? `@nagisa`
2022-03-07 02:07:36 +00:00
Nicholas Nethercote
92d1850f17 Introduce ConstAllocation.
Currently some `Allocation`s are interned, some are not, and it's very
hard to tell at a use point which is which.

This commit introduces `ConstAllocation` for the known-interned ones,
which makes the division much clearer. `ConstAllocation::inner()` is
used to get the underlying `Allocation`.

In some places it's natural to use an `Allocation`, in some it's natural
to use a `ConstAllocation`, and in some places there's no clear choice.
I've tried to make things look as nice as possible, while generally
favouring `ConstAllocation`, which is the type that embodies more
information. This does require quite a few calls to `inner()`.

The commit also tweaks how `PartialOrd` works for `Interned`. The
previous code was too clever by half, building on `T: Ord` to make the
code shorter. That caused problems with deriving `PartialOrd` and `Ord`
for `ConstAllocation`, so I changed it to build on `T: PartialOrd`,
which is slightly more verbose but much more standard and avoided the
problems.
2022-03-07 08:25:50 +11:00
Erik Desjardins
a19138f1f9 cleanup: remove unused ability to have LLVM null-terminate const strings 2022-03-06 12:28:46 -05:00
bjorn3
54d2ec1a82 Merge commit '1411a98352ba6bee8ba3b0131c9243e5db1e6a2e' into sync_cg_clif-2021-12-31 2021-12-31 16:26:32 +01:00
Nicholas Nethercote
bba8b97b83 Remove unnecessary sigils around Symbol::as_str() calls. 2021-12-15 17:32:14 +11:00
antoyo
11c2023ef5
Fix/count trailing zeroes (#95)
* Fix count trailing zeroes
* Fix pop count
* Fix bit reverse
2021-09-27 20:35:45 -04:00
antoyo
ab4ff2dfe0
Cleanup fix for global initialization (#93)
* Cleanup fix for global initialization
* Remove linker script hack
* Use v0 symbol mangling
* Fix warnings
2021-09-27 09:34:06 -04:00
antoyo
64c561dc22
Fix global initialization (#91)
* Make define_global() return a RValue directly
* Return LValue in functions declaring a global variable
* Remove useless cast
* Fix bytes_in_context to use an array rvalue
* Remove global_names which is unused
* Make const_struct create a constant struct
* Correctly initialize global in static_addr_of_mut
* Fix global variable initialization
* Remove workaround for ARGV
2021-09-26 12:20:02 -04:00
Commeownist
48d60ab7c5
Update to nightly-2021-09-11 (#79)
* Implement `black_box` as intrinsic

Responsibility of implementing the black box is now lies on backend

* Remove some TODOs

* Update to nightly-2021-09-17

* CI: don't fail on warnings
2021-09-17 17:19:25 -04:00
antoyo
e228f0c16e
Cleanup (#67) 2021-08-15 08:28:46 -04:00
Antoni Boucher
afae271d5d Initial commit 2021-08-12 21:46:50 -04:00