Libuv is not supported on Windows XP, as stated in the
SUPPORTED_PLATFORMS document. It also can't possibly work because
`ReOpenFile()` in src/win/fs.c is not weakly linked, so any executable
that links libuv would simply fail to load on XP.
This reverts commit 17eaa956bd.
Refs: https://github.com/libuv/libuv/pull/1810
PR-URL: https://github.com/libuv/libuv/pull/1854
Reviewed-By: Bartosz Sosnowski <bartosz@janeasystems.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Libuv does not support multiple handles watching the same file
descriptor. That condition is caught by an assert but it's detached
from the call site and therefore not always trivial to track down.
This commit turns cases where we can easily detect duplicates into
runtime `UV_EEXIST` errors. More work is needed to catch _all_ cases.
Partially addresses https://github.com/libuv/libuv/issues/1172.
PR-URL: https://github.com/libuv/libuv/pull/1851
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
* Don't silently ignore ERROR_OPERATION_ABORTED.
Code to silently ignore this error was added in c42a4ca, under the
false premise that this error is somehow equivalent to EINTR on posix
platforms. This isn't true; ERROR_OPERATION_ABORTED doesn't happen
unless the application explicitly aborts an I/O operation.
Silently ignoring this error elsewhere could potentially hide bugs,
hence libuv shouldn't do it. Instead, explicitly deal with it where
it is expected.
* Don't mark aborted reads as successful.
The worker thread used to call ReadFile() on synchronous pipes would
incorrectly mark cancelled read requests as successful, leading to
issues later on.
* Rely on main thread to restart aborted reads.
After a blocking ReadFile() call was cancelled, the worker thread
would previously attempt to restart it immediately, making
synchronization logic needlessly complex. Instead, we simply cancel
the operation, and leave it to the main loop to restart it if so
desired.
Since we now realy on the main thread to restart interrupted
ReadFile() calls, we can now have a single function interrupts a
synchronous read until the event loop restarts it again.
* Clean up uv__pipe_read_stop().
A single function to interrupt blocking reads also allows us to
remove weird logic in uv__pipe_read_stop() that quickly pauses and
unpauses a read operation in the hope of cancelling it.
* Assume CancelIo() and CancelSynchronousIo() are always available.
Since libuv doesn't support windows XP and Server 2003 any more, we
can assume that these APIs are always available, and do away with
branching around them.
PR-URL: https://github.com/libuv/libuv/pull/1843
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Bartosz Sosnowski <bartosz@janeasystems.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Back in the day I wrote comments in a really unusual way. Nowadays it
makes my eyes bleed, and clang-format doesn't know how to deal with it.
PR-URL: https://github.com/libuv/libuv/pull/1853
Reviewed-By: Bartosz Sosnowski <bartosz@janeasystems.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
PR-URL: https://github.com/libuv/libuv/pull/1840
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
uv__nonblock_fcntl() and uv__nonblock_ioctl() do not commute when
O_NDELAY is not equal to O_NONBLOCK. Case in point: linux/sparc32
and linux/sparc64, where O_NDELAY is O_NONBLOCK + another bit.
Libuv uses uv__nonblock_fcntl() directly sometimes so ensure that it
commutes with uv__nonblock().
Fixes: https://github.com/libuv/libuv/issues/1830
PR-URL: https://github.com/libuv/libuv/pull/1832
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
If not initialized to zero, revents could carry dirty bits in the
output from poll.
PR-URL: https://github.com/libuv/libuv/pull/1820
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
This commit attempts to dynamically load
ConvertInterfaceIndexToLuid() and ConvertInterfaceLuidToNameW()
from iphlpapi.dll before using them. If they are not available,
UV_ENOSYS is returned.
PR-URL: https://github.com/libuv/libuv/pull/1810
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
The Reg* APIs on Windows don't use GetLastError() to report
failures. The errors are returned directly from the call.
For systems which don't have one of the values GetLastError() can
end up returning 0 to the caller, indicating success. The caller
then assumes that the data is valid and can attempt to execute on
garbage data. This change fixes the flow to correctly return the
error to the caller.
PR-URL: https://github.com/libuv/libuv/pull/1811
Reviewed-By: Bartosz Sosnowski <bartosz@janeasystems.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
The System V semaphores on z/OS require explicit ending of the worker
threads and cleanup at process exit. The user will have to manually
cleanup these resources that are left behind. Instead use the custom
semaphore implementation which uses posix mutexes and condition
variables which are cleaned up automatically on process exit.
PR-URL: https://github.com/libuv/libuv/pull/1805
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
OpenBSD does not support setting IPV6_ONLY to anything but 1; OpenBSD
does not support IPv4-mapped IPv6 addresses.
Upstreamed from OpenBSD Ports Collection.
PR-URL: https://github.com/libuv/libuv/pull/1797
Reviewed-By: Aaron Bieber <deftly@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Hack around https://sourceware.org/bugzilla/show_bug.cgi?id=12674
by providing a custom implementation for glibc < 2.21 in terms of other
concurrency primitives.
The glibc implementation on these versions is inherently unsafe.
So, while libuv and Node.js support those versions, it seems to make
sense for libuv in its functionality as a platform abstraction library
to provide a working version.
Fixes: https://github.com/nodejs/node/issues/19903
PR-URL: https://github.com/libuv/libuv/pull/1795
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
According to its man page, there is a bug in fstat()
on macOS related to pipes. This commit replaces a
fstat() call in uv_pipe_chmod() with a stat() call in
order to get the correct permissions.
PR-URL: https://github.com/libuv/libuv/pull/1635
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
uv_fs_unlink would fail for read-only files with Archive attribute
cleared. This fixes this issue.
PR-URL: https://github.com/libuv/libuv/pull/1774
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
UV_FS_COPYFILE_FICLONE attemps to use copy-on-write
semantics in uv_fs_copyfile(). If CoW is not available,
it falls back to a normal copy operation.
Refs: https://github.com/libuv/libuv/pull/1465
PR-URL: https://github.com/libuv/libuv/pull/1491
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
uv_process_tty_accept_req() and uv_process_tty_connect_req()
are marked with "TODO: remove me" comments. This commit
updates the comments to give a better explanation of why
they exist.
PR-URL: https://github.com/libuv/libuv/pull/1335
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
This commit preserves errno properly in the following cases:
- new_socket() was previously discarding errno in some cases,
and replacing it with the result of getsockname().
- uv__close() was not preserving errno when __MVS__ is defined.
Fixes: https://github.com/libuv/libuv/issues/1756
PR-URL: https://github.com/libuv/libuv/pull/1763
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Fix bug in `uv__pollfds_del()` where only one invalidated fd is purged.
Calls to `uv__platform_invalidate_fd()` will invalidate descriptors
(rather than removing them) when `poll()` is iterating. Under stressful
conditions, this may happen more frequently than `uv__pollfds_del()` is
called. So, when uv__pollfds_del() is called with a value of -1, it
should iterate the entire array to remove all invalidated descriptors.
PR-URL: https://github.com/libuv/libuv/pull/1738
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
This commit refactors uv_os_homedir() to to use uv_os_getenv()
to retrieve the USERPROFILE environment variable. This removes
a fair amount of duplicated code, while introducing the following
changes:
1. The function no longer returns UV_EIO when USERPROFILE is
longer than MAX_PATH. This can be reinstated by adding a
check for r == 0 && *size >= MAX_PATH.
2. The USERPROFILE string is now a char*, meaning that it must
be converted from utf8 to utf16 by uv_os_getenv().
Refs: https://github.com/libuv/libuv/pull/1760
PR-URL: https://github.com/libuv/libuv/pull/1761
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
- Remove the `active_reqs` queue, which is never used. There are more
efficient per-stream queues that `libuv` uses whenever it needs this
information, so duplicating it and managing it here seems
like unnecessary extra space and work.
- Unix `uv_loop_init` didn't explicitly initialize.
`loop->active_handles` (although it did memset the whole struct
to 0, so it wasn't wrong previously, just inconsistent).
- Consolidate repeated code for `uv__has_active_reqs`.
- Change `uv__loop_alive` to use the helper functions (mirroring the
unix copy of the same function).
- Initialize some more uv_stream_t fields in init, rather than waiting
for the connection callback. This helps surface bugs in libuv or the
caller better, since it ensures libuv doesn't see uninitialized
memory if asked to look at these fields before it thinks the socket
is connected.
- Fixes what appears to be a double-counting of `handle->reqs_pending`,
in the highly-unlikely event that the code wants to emulate IOCP,
but `RegisterWaitForSingleObject` somehow managed to fail.
PR-URL: https://github.com/libuv/libuv/pull/1746
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
`UV_READABLE_PIPE` and `UV_WRITABLE_PIPE` flags should be honored on
unices.
Windows is better about already doing this, so this will make the
behavior of these flags more consistent across platforms.
It also is just better to set these flags to reflect the actual
mode of the stream, rather than guessing at it based on typical usage.
Refs: https://github.com/libuv/libuv/pull/1655
Refs: https://github.com/nodejs/node/pull/18701
PR-URL: https://github.com/libuv/libuv/pull/1741
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
On OSX when sending handles via `sendmsg()` it can return `EMSGSIZE` if
there isn't room in the socket output buffer to store the whole message.
If that's the case, return control to the loop and try again in the next
iteration.
Fixes: https://github.com/nodejs/node/issues/14828
PR-URL: https://github.com/libuv/libuv/pull/1739
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
The pthread_cond_timedwait requires a timeout relative to the Epoch.
So don't use uv__hrtime to set the timeout because it is relative to an
arbitrary time in the past. Use gettimeofday instead.
PR-URL: https://github.com/libuv/libuv/pull/1711
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Jamie Davis <davisjam@vt.edu>
This commit addresses POD variable initialization issues with older
Visual Studio compilers.
PR-URL: https://github.com/libuv/libuv/pull/1740
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Using -errno, -E**, and -pthread_function() can be
error prone, and breaks compatibility with some operating
systems that already negate errno's (e.g. Haiku).
This commit adds a UV__ERR() macro that ensures libuv
errors are negative.
Fixes: https://github.com/libuv/help/issues/39
PR-URL: https://github.com/libuv/libuv/pull/1687
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reverted for breaking `test/parallel/test-child-process-cwd.js` from the
Node.js test suite. Instead of ENOENT when trying to remove a directory
that does not exist, it started failing with ENOTDIR.
This reverts commit 15f29dc08f.
PR-URL: https://github.com/libuv/libuv/pull/1717
Refs: https://github.com/nodejs/node/issues/18014
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
