Commit graph

230 commits

Author SHA1 Message Date
Jacob Young
ac68d72d24 x86_64: implement aggregate init of a packed struct 2023-04-02 04:49:53 -04:00
Jacob Young
1e080e5056 x86_64: implement atomic loops 2023-03-25 16:23:55 -04:00
Jacob Young
8f385e77ca x86_64: implement struct_field_val for packed containers 2023-03-25 16:23:55 -04:00
Jacob Young
0cfc0d0d13 x86_64: implement struct_field_ptr for packed containers 2023-03-25 16:23:55 -04:00
Jacob Young
f95faac5ae x86_64: (re)implement optional ops
Note that this commit also changes the layout of optional for all
other backends using `src/codegen.zig` without updating them!
2023-03-21 08:49:54 +01:00
Jacob Young
c51930b060 behavior: enable passing behavior tests on stage2_x86_64 2023-03-15 01:04:21 -04:00
Jacob Young
434c6f42ca behavior: enable passing CBE tests 2023-02-21 09:43:23 -05:00
Andrew Kelley
ffdce5f98c add test coverage for fixed bug. closes #5497 2023-02-19 21:18:27 -07:00
Veikka Tuominen
5259d11e3b Sema: resolve fields before checking tuple len
Closes #14400
2023-01-22 00:12:37 +02:00
Veikka Tuominen
342bae02d8 Sema: automatically optimize order of struct fields
This is a simple starting version of the optimization described in #168
where the fields are just sorted by order of descending alignment.
2023-01-16 19:46:41 +02:00
Veikka Tuominen
58c1d98c14 add tests for fixed stage1 bugs
Closes #4144
Closes #4255
Closes #4372
Closes #4375
Closes #4380
Closes #4417
Closes #4423
Closes #4476
Closes #4528
Closes #4562
Closes #4572
Closes #4597
Closes #4639
Closes #4672
Closes #4782
Closes #4955
Closes #4984
Closes #4997
Closes #5010
Closes #5114
Closes #5166
Closes #5173
Closes #5276
2022-12-31 20:49:02 -05:00
Veikka Tuominen
9a0c593a54 add tests for fixed stage1 bugs
Closes #1957
Closes #1994
Closes #2140
Closes #2746
Closes #2802
Closes #2855
Closes #2895
Closes #2981
Closes #3054
Closes #3158
Closes #3234
Closes #3259
Closes #3371
Closes #3376
Closes #3387
Closes #3529
Closes #3653
Closes #3750
Closes #3778
Closes #3882
Closes #3915
Closes #3929
Closes #3961
Closes #3988
Closes #4123
Closes #7448
2022-12-29 12:42:44 +02:00
Jakub Konka
aea3460cf5
Merge pull request #14084 from joachimschmidt557/stage2-aarch64
stage2 AArch64: small improvements
2022-12-27 18:59:59 +01:00
joachimschmidt557
d6e6162081
stage2 AArch64: unify callee-preserved regs on all targets
also enables many passing behavior tests
2022-12-27 21:17:52 +08:00
Jacob Young
81318e8704 llvm: add asserts and behavior tests for #14063
Closes #14063
2022-12-27 00:12:56 -05:00
Jacob Young
64865679cf Sema: add missing resolveLazyValue cases
Closes #14032
2022-12-26 04:39:32 -05:00
Jacob Young
f5b6019646 Sema: fix missing struct layout for llvm backend
Closes #14063
2022-12-25 20:18:15 -05:00
Jacob Young
0e3feebb04 codegen: fix taking the address of a zero-bit field in a zero-bit struct
Normally when we want a pointer to the end of a struct we just add 1 to
the struct pointer.  However, when it is a zero-bit struct, the pointer
type being used during lowering is often a dummy pointer type that
actually points to a non-zero-bit type, so we actually want to add 0
instead, since a zero-bit struct begins and ends at the same address.
2022-12-18 22:11:26 -05:00
Jacob Young
e96f65db77 llvm: fix lowering pointer to final zero-width field of a comptime value
* Handle a `null` return from `llvmFieldIndex`.
 * Add a behavior test to test this code path.
 * Reword this test name, which incorrectly described how pointers to
   zero-bit fields behave, and instead describe the actual test.
2022-12-18 22:11:26 -05:00
Andrew Kelley
6ed0910d6d Revert "llvm: fix lowering pointer to final zero-width field of a comptime value"
This reverts commit e0bc5f65b9.

Caused an assertion failure when running the behavior tests:

```
zig: llvm/lib/IR/Type.cpp:729: static llvm::PointerType* llvm::PointerType::get(llvm::Type*, unsigned int): Assertion `isValidElementType(EltTy) && "Invalid type for pointer element!"' failed.
Aborted (core dumped)
```
2022-12-18 13:51:15 -07:00
Jacob Young
e0bc5f65b9 llvm: fix lowering pointer to final zero-width field of a comptime value
* Handle a `null` return from `llvmFieldIndex`.
 * Add a behavior test to test this code path.
 * Reword this test name, which incorrectly described how pointers to
   zero-bit fields behave, and instead describe the actual test.
2022-12-18 02:17:11 -05:00
Koakuma
f9e9ba784f stage2: sparc64: Skip unimplemented tests 2022-12-10 21:51:46 +07:00
Andrew Kelley
c8aba15c22 remove references to stage1 in behavior tests
Good riddance.
2022-12-06 19:06:48 -07:00
Jacob Young
ec3116f573 cbe: fix zero-bit struct field pointer 2022-12-03 08:40:08 -05:00
Andrew Kelley
610d604f5b
Merge pull request #13748 from jacobly0/c-unaligned
cbe: use memcpy for underaligned loads and stores
2022-12-03 02:18:07 -05:00
Jacob Young
fdedd62365 cbe: use memcpy for underaligned loads and stores 2022-12-02 22:21:24 -05:00
Veikka Tuominen
7f9e841f74 Sema: do not forcibly canonicalize unresolved pointer element type
Closes #13308
2022-12-03 00:09:23 +02:00
Luuk de Gram
090deae41d
wasm: enable behavior tests for packed structs 2022-11-30 21:01:09 +01:00
Veikka Tuominen
d2cc55109a llvm: correct calculation of index of zero-bit field
If the field comes before any non-zero-bit field then the index of
the next field should be returned.

Closes #13363
2022-11-11 17:59:53 +02:00
Ali Chraghi
f5f1f8c666 all: rename i386 to x86 2022-11-04 00:09:27 +03:30
Jacob Young
ab468d57e3 cbe: implement packed structs
Sometimes you have to break a test to make progress :)
2022-10-25 05:11:29 -04:00
Jacob Young
4765294ca4 cbe: get enough things working to support basic programs
* Enable advanced start support.
 * Enable advanced test_runner support.
 * Zig Language Reference's Hello World now works.
2022-10-25 05:11:29 -04:00
Jacob Young
912b84bbad cbe: fix atomics 2022-10-25 05:11:29 -04:00
Jacob Young
87d432328e cbe: implement aggregate_init of struct 2022-10-25 05:11:29 -04:00
Jacob Young
5eafc10bf4 cbe: fix global access fix 2022-10-25 05:11:29 -04:00
Jacob Young
a12535f501 cbe: fix global access 2022-10-25 05:11:28 -04:00
Jacob Young
6a4266d62a cbe: fix infinite recursion on recursive types 2022-10-25 05:11:28 -04:00
Jacob Young
525dcaecba behavior: enable stage2_c tests that are currently passing
Also fix C warnings triggered by these tests.
2022-10-25 05:11:28 -04:00
r00ster91
51d9db8569 fix(text): hyphenate "comptime" adjectives 2022-10-05 21:19:30 +02:00
Veikka Tuominen
541b3e3a31 Sema: check pointer qualifiers before implicit cast
Closes #12881
2022-09-20 00:50:13 +03:00
Koakuma
0910cb6fcb stage2: sparc64: Skip compile-failing tests for now 2022-09-14 16:58:12 -07:00
joachimschmidt557
3794f2c493
stage2 ARM: implement struct_field_val for registers 2022-09-09 19:17:18 +02:00
joachimschmidt557
261fec8036
stage2 ARM: amend implementation of various AIR instructions
- unwrap_errunion_err for registers
- unwrap_errunion_payload for registers
- ptr_slice_len_ptr for all MCValues
- ptr_slice_ptr_ptr for all MCValues
2022-09-09 19:17:17 +02:00
Veikka Tuominen
b55a5007fa Sema: fix parameter of type 'T' must be comptime error
Closes #12519
Closes #12505
2022-08-22 11:16:36 +03:00
Meghan Denny
0879cbeed2 remove 'builtin.stage2_arch', Sema is smart enough now 2022-08-07 16:07:08 -07:00
Veikka Tuominen
e47706f344 Sema: validate packed struct field types 2022-08-01 23:59:40 +03:00
Veikka Tuominen
ff7ec4efb5 Sema: bad union field access safety 2022-07-23 15:40:11 +03:00
Andrew Kelley
f550c29c4e LLVM: fix lowering of structs with underaligned fields
When lowering a struct type to an LLVM struct type, keep track of
whether there are any underaligned fields. If so, then make it a packed
llvm struct. This works because we already insert manual padding bytes
regardless.

We could unconditionally use an LLVM packed struct; the reason we bother
checking for underaligned fields is that it is a conservative choice, in
case LLVM handles packed structs less optimally. A future improvement
could simplify this code by unconditionally using packed LLVM structs
and then make sure measure perf is unaffected.

closes #12190
2022-07-21 22:51:17 -07:00
Andrew Kelley
4a28c1d5c3 stage2: lower each struct field type, align, init separately
Previously, struct types, alignment values, and initialization
expressions were all lowered into the same ZIR body, which caused false
positive "depends on itself" errors when the initialization expression
depended on the size of the struct.

This also uses ResultLoc.coerced_ty for struct field alignment and
initialization values. The resulting ZIR encoding ends up being roughly
the same, neither smaller nor larger than previously.

Closes #12029
2022-07-13 02:38:13 -04:00
Veikka Tuominen
cb5d2b691a Sema: validate equality on store to comptime field 2022-06-06 13:11:50 -07:00