Commit Graph

867 Commits

Author SHA1 Message Date
Andrew Kelley
52bb71867d implement vector negation
also fix vector behavior tests, they weren't actually testing
runtime vectors, but now they are.

See #903
2019-02-22 13:28:57 -05:00
Andrew Kelley
0c5f897904 fix @bitCast when src/dest types have mismatched handle_is_ptr
* separate BitCast and BitCastGen instructions
 * closes #991
 * closes #1934
 * unrelated: fix typo in docs (thanks gamester for pointing it out)
2019-02-22 08:49:27 -05:00
Andrew Kelley
cbce61a209 better field access of types which have one possible value
* When you do field access of a type which only has one possible
   value, the result is comptime-known.
 * StorePtr instructions which operate on pointers to types which
   only have one possible value, the result is a comptime no-op.

closes #1554
2019-02-21 16:09:14 -05:00
Andrew Kelley
1066004b79 better handling of arrays in packed structs
* Separate LoadPtr IR instructions into pass1 and pass2 variants.
 * Define `type_size_bits` for extern structs to be the same as
   their `@sizeOf(T) * 8` and allow them in packed structs.
 * More helpful error messages when trying to use types in
   packed structs that are not allowed.
 * Support arrays in packed structs even when they are not
   byte-aligned.
 * Add compile error for using arrays in packed structs when the
   padding bits would be problematic. This is necessary since
   we do not have packed arrays.

closes #677
2019-02-21 14:44:14 -05:00
Andrew Kelley
2bb795dc45 @sliceToBytes works at comptime
closes #262
2019-02-21 10:07:11 -05:00
Andrew Kelley
db31c2524d extern structs support comptime bitcasting 2019-02-21 08:46:43 -05:00
Andrew Kelley
3ee9d06cbd packed structs support comptime bitcasting
* `type_size_store` is no longer a thing. loading and storing a pointer
   to a value may dereference up to `@sizeOf(T)` bytes, even for
   integers such as `u24`.
 * fix `types_have_same_zig_comptime_repr` to not think that the
   same `ZigTypeId` means the `ConstExprValue` neccesarily has the
   same representation.
 * implement `buf_write_value_bytes` and `buf_read_value_bytes` for
   `ContainerLayoutPacked`

closes #1120
2019-02-20 22:40:41 -05:00
Andrew Kelley
079728752e deduplicate compile errors for undeclared identifiers
closes #111
2019-02-20 08:04:46 -05:00
Andrew Kelley
c8ce351ec9 pull request fixups 2019-02-19 15:34:44 -05:00
Andrew Kelley
1034af40f9 Merge branch 'slice-deref-failure' of https://github.com/matthew-mcallister/zig into matthew-mcallister-slice-deref-failure 2019-02-19 15:27:10 -05:00
Andrew Kelley
28bf768883 export _mh_execute_header with weak linkage
* also fix extern variables with initialiaztion values to generate runtime code
 * remove the workaround in example/shared_library/mathtest.zig
 * introduce the ability for global variables to have Weak and LinkOnce
   linkage
 * fix `@export` to work for non-functions. this code needs to be
   audited though.
 * fix comptime ptrcast not keeping bigger alignment
 * fix linker warnings when targeting darwin

closes #1903
2019-02-18 16:47:30 -05:00
Andrew Kelley
7a84fe79b9 pull request fixups 2019-02-18 13:05:26 -05:00
emekoi
9b3013d2f6 make @enumToInt work on union(enum)
closes #1711
2019-02-18 11:43:45 -05:00
Matthew McAllister
51783510b9 Deduplicate compile log statement warnings 2019-02-17 14:02:37 -08:00
Matthew McAllister
91989e70ba Fix lvalue dereference type checking
Previously, if a dereference instruction was an lvalue, it would fail to
typecheck that the value being dereferenced was indeed a pointer.
Although a little clunky, this change obviates the need for redundant
type checks scattered about the analysis.
2019-02-16 17:37:47 -08:00
Andrew Kelley
a05e224150 typecheck the panic function
this adds the prototype of panic to @import("builtin")
and then uses it to do an implicit cast of the panic
function to this prototype, rather than redoing all the
implicit cast logic.

closes #1894
closes #1895
2019-02-15 19:19:28 -05:00
Andrew Kelley
7293e012d7 breaking: fix @sizeOf to be alloc size rather than store size
* Fixes breaches of the guarantee that `@sizeOf(T) >= @alignOf(T)`
 * Fixes std.mem.secureZero for integers where this guarantee previously
   was breached
 * Fixes std.mem.Allocator for integers where this guarantee previously
   was breached

Closes #1851
Closes #1864
2019-02-15 18:05:50 -05:00
Andrew Kelley
567c9b688e Merge pull request #1965 from ziglang/c-pointer-type
implement C pointers
2019-02-15 02:20:42 -05:00
Andrew Kelley
d6e0d82c32 translate-c: back to *c_void for opaque types
See #1059
2019-02-14 23:09:31 -05:00
Andrew Kelley
d5bbd74871 allow C pointers to have alignment
clang/gcc support pointer alignment attribute:
https://clang.llvm.org/docs/AttributeReference.html#align-value
2019-02-14 20:04:13 -05:00
Andrew Kelley
cc7060d0d9 compile error for C pointer with align attribute
See #1059
2019-02-14 20:02:29 -05:00
Andrew Kelley
6769183a9d fix implicit cast error unions with non-optional to optional pointer
and update self hosted compiler for C pointers

See #1059
2019-02-14 15:48:28 -05:00
Andrew Kelley
e03c770145 compile error tests for implicit C pointer casting
See #1059
2019-02-14 12:28:50 -05:00
Andrew Kelley
c58b802034 remove the "top of the comptime stack" compile error
It's still best practice to put `@setEvalBranchQuota` at the top of
the comptime stack, but as Jimmi notes in #1949, when a function
can be called at comptime and also can be the top of the comptime stack,
this compile error is fundamentally unsound.

So now it's gone.

closes #1949
2019-02-14 10:51:59 -05:00
Andrew Kelley
59de24817e runtime safety check for casting null to pointer
see #1059
2019-02-14 01:09:33 -05:00
Andrew Kelley
d4d2718bca comptime detection of casting null to pointer
See #1059
2019-02-14 00:40:39 -05:00
Andrew Kelley
5699ab5e77 C pointers: errors for nested pointer casting regarding null
See #1059
2019-02-12 18:20:00 -05:00
Andrew Kelley
270933b1e9 compile error test for casting integer to c pointer
when the int has more bits than pointers

See #1059
2019-02-12 10:25:21 -05:00
Andrew Kelley
6f05e8d1be implicit casting between C pointer and optional non-C pointer
See #1059
2019-02-12 01:38:11 -05:00
Andrew Kelley
285e2f62ba disallow C pointers to non-C-ABI-compatible element types
See #1059
2019-02-12 00:51:06 -05:00
Andrew Kelley
0abe6d668e C pointers: delete dead code in ir_num_lit_fits_in_other_type 2019-02-12 00:39:08 -05:00
Andrew Kelley
069fc1a269 peer type resolution with C pointers
See #1059
2019-02-11 19:21:59 -05:00
Andrew Kelley
57a7ab0d33 comptime support for pointer arithmetic with hard coded addresses 2019-02-11 19:12:01 -05:00
Andrew Kelley
90b8cd4a45 add C pointer type to @typeInfo
See #1059
2019-02-11 16:07:40 -05:00
Andrew Kelley
342bca7f46 C pointer comparison and arithmetic
See #1059
2019-02-11 15:31:09 -05:00
Andrew Kelley
73e8e46257 casting between C pointers and normal pointers
See #1059
2019-02-10 01:11:40 -05:00
Andrew Kelley
b8cbe3872e added C pointer type and implicit int-to-ptr for this type
See #1059
2019-02-10 00:14:30 -05:00
Andrew Kelley
caf672c495 @truncate: comptime 0 when target type is 0 bits
also if the dest type is a comptime_int, then treat it
as an implicit cast.

also compile error for attempting to truncate undefined

closes #1568
2019-02-09 21:38:34 -05:00
Andrew Kelley
46ddd5f5f4 fix compiler assertion failure when returning value from test
closes #1935
2019-02-08 19:23:46 -05:00
Andrew Kelley
c2db077574 std.debug.assert: remove special case for test builds
Previously, std.debug.assert would `@panic` in test builds,
if the assertion failed. Now, it's always `unreachable`.

This makes release mode test builds more accurately test
the actual code that will be run.

However this requires tests to call `std.testing.expect`
rather than `std.debug.assert` to make sure output is correct.

Here is the explanation of when to use either one, copied from
the assert doc comments:

Inside a test block, it is best to use the `std.testing` module
rather than assert, because assert may not detect a test failure
in ReleaseFast and ReleaseSafe mode. Outside of a test block, assert
is the correct function to use.

closes #1304
2019-02-08 18:23:38 -05:00
Andrew Kelley
f330eebe4b fix using the result of @intCast to u0
closes #1817
2019-02-07 16:02:45 -05:00
Andrew Kelley
b1775ca168 thread local storage working for linux x86_64 2019-02-06 13:48:04 -05:00
Andrew Kelley
8c6fa982cd SIMD: array to vector, vector to array, wrapping int add
also vectors and arrays now use the same ConstExprVal representation

See #903
2019-02-04 20:30:00 -05:00
Matthew McAllister
c90c256868 Fix slice concatenation
This was causing an underflow error
2019-02-02 22:22:00 -05:00
Matthew McAllister
8bedb10939 Fix runtime assignment to comptime aggregate field
This was causing a segfault
2019-02-01 17:14:12 -05:00
Andrew Kelley
ae1ebe09b7 add compile errror for @bitCast when bit counts mismatch
fixes invalid LLVM IR from previous commit
2019-02-01 14:06:51 -05:00
Andrew Kelley
545064c1d9 introduce vector type for SIMD
See #903

 * create with `@Vector(len, ElemType)`
 * only wrapping addition is implemented

This feature is far from complete; this is only the beginning.
2019-01-30 23:39:25 -05:00
Andrew Kelley
581edd643f backport copy elision changes
This commit contains everything from the copy-elision-2
branch that does not have to do with copy elision directly,
but is generally useful for master branch.

 * All const values know their parents, when applicable, not
   just structs and unions.
 * Null pointers in const values are represented explicitly,
   rather than as a HardCodedAddr value of 0.
 * Rename "maybe" to "optional" in various code locations.
 * Separate DeclVarSrc and DeclVarGen
 * Separate PtrCastSrc and PtrCastGen
 * Separate CmpxchgSrc and CmpxchgGen
 * Represent optional error set as an integer, using the 0 value.
   In a const value, it uses nullptr.
 * Introduce type_has_one_possible_value and use it where applicable.
 * Fix debug builds not setting memory to 0xaa when storing
   undefined.
 * Separate the type of a variable from the const value of a variable.
 * Use copy_const_val where appropriate.
 * Rearrange structs to pack data more efficiently.
 * Move test/cases/* to test/behavior/*
 * Use `std.debug.assertOrPanic` in behavior tests instead of
   `std.debug.assert`.
 * Fix outdated slice syntax in docs.
2019-01-29 22:30:30 -05:00
Andrew Kelley
9c328b4291 simpler implementation of && and || hints
This accomplishes the same goal, but with less changes, so that
I can backport copy elision stuff easier.
2019-01-29 22:29:56 -05:00
kristopher tate
5bf9ffdc5b Hint at use of and/or when &&/|| is improperly used (#1886) 2019-01-25 21:10:40 +01:00