Commit Graph

1827 Commits

Author SHA1 Message Date
Veikka Tuominen
078e4afdaf
Merge pull request #5500 from Vexu/fix
Fix std.zig rejecting literal tabs in comments
2020-06-01 17:24:30 +03:00
Veikka Tuominen
eb687810cf
Merge pull request #4976 from xackus/std-lib-docs
std lib docs: fix null and int values
2020-06-01 14:50:05 +03:00
Veikka Tuominen
8c5c860b00
Merge pull request #5443 from data-man/mem_len_vectors
Support vectors in mem.len
2020-06-01 14:44:38 +03:00
Vexu
a47257d9b0
fix std.zig rejecting literal tabs in comments 2020-06-01 14:37:36 +03:00
Veikka Tuominen
899d79f8f0
Merge pull request #5478 from vrischmann/fix-iouring
linux: fix IOSQE_BIT decl and io_uring_sqe flags
2020-05-31 16:59:04 +03:00
Veikka Tuominen
30b2bc86b9
Merge pull request #5477 from alexnask/patch-1
Fix WSARecvFrom signature
2020-05-31 16:49:08 +03:00
Veikka Tuominen
21544267ee
Merge pull request #5481 from kubkon/typo-packedintarray
Fix typo in PackedIntArray tests
2020-05-31 16:48:19 +03:00
Jakub Konka
13c405127d Fix typo in PackedIntArray tests
Both `PackedIntArray` and `PackedIntSlice` tests defined `max_bits`
but didn't use it in the predicate in the inline while loop.
2020-05-30 23:17:55 +02:00
Vexu
010c58e303
fix zig fmt out of bounds on empty file 2020-05-30 23:07:51 +03:00
Vincent Rischmann
636d3ba780
linux: fix IOSQE_BIT decl and io_uring_sqe flags 2020-05-30 16:05:46 +02:00
Alexandros Naskos
dbdacb0938
Fix WSARecvFrom signature
The lpFromLen should be a pointer.
2020-05-30 16:46:32 +03:00
Andrew Kelley
89a97a7a27 cleanups 2020-05-29 18:30:09 -04:00
Jonathan S
c209da1589 Document the failure to improve selfExe{,Dir}PathAlloc 2020-05-29 18:27:39 -04:00
Jonathan S
0674b51453 In getCwdAlloc, geometrically allocate larger buffers to find an appropriate size. 2020-05-29 18:27:39 -04:00
Jonathan S
ab3931fa95 Prefer Files to paths in std.debug. Additionally [breaking] add a flags parameter to openSelfExe and stop exporting openElfDebugInfo.
This should save a call to readlink in openSelfDebugInfo and support executables in overlong paths on Linux.
2020-05-29 18:27:39 -04:00
Jonathan S
631633b252 Document and reduce usage of MAX_PATH_BYTES, lifting arbitrary buffer size requirements 2020-05-29 18:23:06 -04:00
Andrew Kelley
65c3833ec2
Merge pull request #5388 from kubkon/wasi-args-iter
Add ArgIteratorWasi and integrate it with ArgIterator
2020-05-29 15:55:14 -04:00
Noam Preil
e9f344dcd4 Add include dirs to translate-c (close #5098) 2020-05-29 13:39:16 -04:00
Jakub Konka
6e347e6180 Fix args iterator test 2020-05-29 17:12:19 +02:00
Jakub Konka
6f48842ddb Make ArgIterator.init() a compile error in WASI
Given that the previous design would require the use of a default
allocator to have `ArgIterator.init()` work in WASI, and since in
Zig we're trying to avoid default allocators, I've changed the design
slightly in that now `init()` is a compile error in WASI, and instead
in its message it points to `initWithAllocator(*mem.Allocator)`.
The latter by virtue of requiring an allocator as an argument can
safely be used in WASI as well as on other OSes (where the allocator
argument is simply unused). When using `initWithAllocator` it is then
natural to remember to call `deinit()` after being done with the
iterator. Also, to make use of this, I've also added `argsWithAllocator`
function which is equivalent to `args` minus the requirement of supplying
an allocator and being fallible.

Finally, I've also modified the WASI only test `process.ArgWasiIterator`
to test all OSes.
2020-05-29 10:48:03 +02:00
Jakub Konka
f1a4e1a70f Add ArgIteratorWasi and integrate it with ArgIterator
This commit pulls WASI specific implementation of args extraction
from the runtime from `process.argsAlloc` and `process.argsFree`
into a new iterator struct `process.ArgIteratorWasi`. It also
integrates the struct with platform-independent `process.ArgIterator`.
2020-05-29 10:48:03 +02:00
data-man
dc4fea983d Use tuples in multiTrait 2020-05-28 23:10:44 -04:00
Dmitry Atamanov
0328537ca6
Support stringify for vectors (#5441)
* use array's pointer
2020-05-28 23:10:16 -04:00
Walter Mays
8630ae7523 Remove reliance on hashing algorithm for iterator reset test. 2020-05-28 23:09:21 -04:00
Andrew Kelley
1e0de896b8
Merge pull request #5452 from squeek502/comptime-string-map
Add std.ComptimeStringMap based on the tokenizer optimization in #5442
2020-05-28 23:06:40 -04:00
Jakub Konka
4f24ab9cfc Fix skipping condition (skip when wasm32) 2020-05-29 00:06:26 +02:00
Jakub Konka
3dbe376c77 Link issue to the failing test case targeting wasm 2020-05-28 23:32:17 +02:00
Ryan Liptak
3cac0a5614 Not sure how a tab snuck in there 2020-05-26 23:26:19 -07:00
Ryan Liptak
b683498ae8 Use ComptimeStringMap in std.meta.stringToEnum when feasible 2020-05-26 23:10:13 -07:00
Ryan Liptak
dfafafac7b std.ComptimeStringMap: Add support for void value type (i.e. a set) 2020-05-26 23:10:12 -07:00
Ryan Liptak
62cfc68d2f Use std.ComptimeStringMap in the C tokenizer 2020-05-26 23:10:08 -07:00
Ryan Liptak
a9dd79d293 Use std.ComptimeStringMap in zig's tokenizer
Should have no performance change, the implementation was just made generic/re-usable
2020-05-26 21:36:58 -07:00
Ryan Liptak
0865e5d360 Add std.ComptimeStringMap 2020-05-26 21:34:55 -07:00
Jakub Konka
e61e8c94be Reenable zig parser tests disabled targeting Wasm
I'm not sure why I disabled them when landing extended Wasm/WASI
support, but they pass the parser tests just fine now, so I'm gonna
go ahead and re-enable them.
2020-05-26 21:01:54 -04:00
Andrew Kelley
ba41a9d5d7 different strategy for tokenizing keywords
throughput: 279 MiB/s => 347 MiB/s
2020-05-26 18:59:49 -04:00
Walter Mays
e1186c88ea Remove unimplemented init call from ArrayListUnmanaged. 2020-05-26 15:59:56 -04:00
data-man
49dd2cbd9a Support vectors in mem.len 2020-05-27 00:00:19 +05:00
Dmitry Atamanov
dd62f63c04
fmt padding correction (#5403)
* Make .Left as default
2020-05-26 20:53:51 +03:00
Walter Mays
19a04d8ebd
Add writeToSlice method to SegmentedList. (#5405) 2020-05-26 13:04:25 -04:00
Andrew Kelley
62fefe8648 std.fmt: non-exhaustive enums: avoid the loop when it's not necessary 2020-05-26 12:15:08 -04:00
daurnimator
5cdeac3b0e std: in fmt, use tag names of non-exhaustive enums when available 2020-05-26 12:00:33 -04:00
daurnimator
292d515e91 std: allow specifiying enums to be formatted as numbers 2020-05-26 12:00:33 -04:00
data-man
d10e407977 More vector support in std.meta 2020-05-26 10:56:29 -04:00
Veikka Tuominen
4b8077ea8e
Merge pull request #5383 from Vexu/fix
Fix missing compile error on while/for missing block
2020-05-26 11:13:09 +03:00
Andrew Kelley
7c8d0cc678 fix pwrite on 32-bit linux 2020-05-25 19:59:39 -04:00
Andrew Kelley
a83aab5209 fix std lib tests for WASI 2020-05-25 19:46:28 -04:00
Andrew Kelley
cda102be02 improvements to self-hosted cache hash system
* change miscellaneous things to more idiomatic zig style
 * change the digest length to 24 bytes instead of 48. This is
   still 70  more bits than UUIDs. For an analysis of probability of
   collisions, see:
   https://en.wikipedia.org/wiki/Universally_unique_identifier#Collisions
 * fix the API having the possibility of mismatched allocators
 * fix some error paths to behave properly
 * modify the guarantees about when file contents are loaded for input files
 * pwrite instead of seek + write
 * implement isProblematicTimestamp
 * fix tests with regards to a working isProblematicTimestamp function.
   this requires sleeping until the current timestamp becomes
   unproblematic.
 * introduce std.fs.File.INode, a cross platform type abstraction
   so that cache hash implementation does not need to reach into std.os.
2020-05-25 19:29:03 -04:00
Vexu
e07b467c7c
fix missing compile error on while/for missing block 2020-05-25 23:25:06 +03:00
Andrew Kelley
69ff89fd12 stage2 parser: heuristics to pre-allocate token arrays
throughput: 72.2 MiB/s => 75.3 MiB/s

I also tried the idea from the deleted comment in this commit and it
made the throughput worse.
2020-05-25 15:12:23 -04:00
Andrew Kelley
6d5ec184ab stage2 parser: heuristics to pre-allocate token arrays
throughput: 72.2 MiB/s => 75.3 MiB/s
2020-05-25 15:02:02 -04:00