T0b1/wasmtime
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
7,692 Commits 2 Branches 0 Tags
03466d04cee5552ecb92f0eaaa3541eab83afccd
Commit Graph

1277 Commits

Author SHA1 Message Date
Pat Hickey
03466d04ce temporarily use my branch of the wasi crate 2021-01-27 19:05:52 -08:00
Pat Hickey
be108c7a93 will this work for platform-specific errnos? WIP 2021-01-27 18:18:12 -08:00
Pat Hickey
2ae1dee642 convert all errno assertions to assert_errno! 2021-01-27 18:10:38 -08:00
Pat Hickey
4f655001c0 make a macro that automates errno assertions 2021-01-27 18:10:34 -08:00
Pat Hickey
4f2696fedf more docs 2021-01-27 16:16:10 -08:00
Pat Hickey
cb171712cc various windows test notes 2021-01-27 15:32:23 -08:00
Pat Hickey
d1160cb9b5 split path rename trailing slash test into files and dirs 
files fail on both platform, dirs pass on both! so, same bug!
 2021-01-27 14:55:39 -08:00
Pat Hickey
6c0f312a1b fix bug: i was ignoring the symlink follow flag in path_filestat_{get, 
set_times}
 2021-01-27 12:42:15 -08:00
Pat Hickey
b2e0597ae3 narrow the bug 2021-01-27 12:27:13 -08:00
Pat Hickey
6fcbd3f15d split path_filestat test into two, improve variable names 2021-01-27 12:18:27 -08:00
Pat Hickey
220509f3ed latest system-interface 2021-01-27 12:17:42 -08:00
Pat Hickey
8cd21eed23 elaborate 2021-01-26 12:27:07 -08:00
Pat Hickey
1eb8a8a7fe integrate GetSetFdFlags! 
change reopen_with_fdflags(&self, fdflags) -> Result<Box<dyn WasiFile>>
to set_fdflags(&mut self, fdflags) -> Result<()>.

this makes way more sense than my prior hare-brained schemes.
 2021-01-26 12:23:52 -08:00
Pat Hickey
c98d6f6201 fix warning 2021-01-26 12:18:39 -08:00
Pat Hickey
f4a6b95965 notes 2021-01-26 11:45:26 -08:00
Pat Hickey
17f43d4cc3 path_link test: we no longer support symlink following 
just assert that setting the symlink following lookupflag gives an
ERRNO_INVAL.
 2021-01-25 17:15:32 -08:00
Pat Hickey
fded424e68 Dir::hard_link: dont support symlink following 
if we're asked to follow symlinks, give ERRNO_INVAL instead.
 2021-01-25 17:08:02 -08:00
Pat Hickey
a993090e30 fix warning 2021-01-25 17:07:39 -08:00
Pat Hickey
42ccb11f15 cap-std is going to use winerror::ERROR_STOPPED_ON_SYMLINK 
to communicate eloop on windows
 2021-01-25 17:02:11 -08:00
Pat Hickey
8f4fecacdc interesting paths: loosen up errno requirements 2021-01-25 15:28:01 -08:00
Pat Hickey
b0b263f68e find a symlink error 2021-01-25 15:20:46 -08:00
Pat Hickey
996d49c81f nofollow test: permit ACCES as well as LOOP when opening with nofollow 2021-01-25 15:20:35 -08:00
Pat Hickey
144ca2f535 add a test of symlink happy path to isolate the windows failure 2021-01-25 15:14:26 -08:00
Pat Hickey
a8426125d2 kick the can 2021-01-25 14:56:58 -08:00
Pat Hickey
72b207de2e path_link: some improvements required by windows 
* need to close the handle to the subdirectory before its legal to
  delete it
* windows doesnt give us a way to distinguish between an ERRNO_PERM and
  an ERRNO_ACCES, so lets accept either one
 2021-01-25 14:35:02 -08:00
Pat Hickey
2c6bde5ee4 elaborate! 2021-01-25 12:54:13 -08:00
Pat Hickey
7fd3d58b23 enumerate windows test failures, move the file 2021-01-25 12:20:10 -08:00
Pat Hickey
17a82c5b88 table: fix up errnos. trap on internal failures. 2021-01-25 12:16:07 -08:00
Pat Hickey
957c434b67 ErrorExt: trap convenience method 2021-01-25 12:16:01 -08:00
Pat Hickey
634e911a4b tests: directory seeking is actually prohibited! but the test was wrong 
* path_open of a directory without OFLAGS_DIRECTORY worked on linux,
  but fortunately not on windows!
* the errno is BADF instead of NOTCAPABLE for fd_seek on a directory
* no way for a directory to have the FD_SEEK right.
 2021-01-25 12:13:02 -08:00
Pat Hickey
2b70ea8b91 windows scheduler passes tests 
unfortunately this does essentially nothing to convince me of its
correctness
 2021-01-22 18:26:45 -08:00
Pat Hickey
19254bfd72 error mapping for windows! 2021-01-22 15:44:32 -08:00
Pat Hickey
423973a8ea push the error types conversion all the way through 2021-01-22 15:29:09 -08:00
Pat Hickey
1d8070b734 wasi-c2: completely redo how errors work 2021-01-22 15:03:16 -08:00
Pat Hickey
a06c4fbe1a make a pipe filetype, reorder filetype enum 2021-01-22 11:37:39 -08:00
Pat Hickey
86d369c72a inline poll from yanix 2021-01-21 18:17:49 -08:00
Pat Hickey
9e0d295f8c use unsafe to mark reopen_with_fdflags as having special safety features 2021-01-21 18:08:15 -08:00
Pat Hickey
69b7352332 various stdio fixes 2021-01-21 17:57:12 -08:00
Pat Hickey
f1a5dce716 change set_fdflags to reopen_with_fdflags 2021-01-21 17:56:03 -08:00
Pat Hickey
a46c2ad0aa split poll tests up, only one requires "real" stdio 2021-01-21 16:54:40 -08:00
Pat Hickey
12056885eb fix preopen dir to work on windows 2021-01-21 16:08:49 -08:00
Pat Hickey
16b42a5707 fix 2021-01-21 14:14:05 -08:00
Pat Hickey
85c3b73b4e stdio compiles on windows! 2021-01-21 13:51:18 -08:00
Pat Hickey
fcd00f5de1 move more deps to cap-std-sync, define own SystemTimeSpec 2021-01-21 12:35:55 -08:00
Pat Hickey
61885b7071 stub in windows scheduler, get to some interesting errors 2021-01-20 21:03:33 -08:00
Pat Hickey
7ec03631da inline only what we need of cap-std/system-interface traits 2021-01-20 20:53:44 -08:00
Pat Hickey
f3e40e2fc4 restructure cap-std impls into their own crate 2021-01-20 19:09:15 -08:00
Pat Hickey
016ed8966a poll now handles monotonic clocks, not system clocks 
I initially had it backwards. It is not sensible to poll on a system
clock timeout because the system clock is not necessarily monotonic!
only a monotonic clock makes sense for a timeout.
 2021-01-19 15:11:54 -08:00
Pat Hickey
21713d3468 Revert "TEMPORARY: poll_oneoff test uses CLOCKID_REALTIME instead of MONOTONIC" 
This reverts commit f667263d9c.
 2021-01-19 14:57:10 -08:00
Pat Hickey
657024bd0d latest system-interface provides ReadReady::num_ready_bytes 2021-01-19 14:39:20 -08:00