linux: fix sscanf() overflows reading from /proc

Use `"%" PRIu64` and uint64_t for sscanf-ing large values, not `"%lu"`
and unsigned long. The latter can overflow when libuv is a 32 bits

linux: fix sscanf() overflows reading from /proc

Use `"%" PRIu64` and uint64_t for sscanf-ing large values, not `"%lu"`
and unsigned long. The latter can overflow when libuv is a 32 bits
application running on a 64 bits machine.

Fixes: https://github.com/libuv/libuv/issues/2297
PR-URL: https://github.com/libuv/libuv/pull/2305
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

test: increase test timeout

The test threadpool_multiple_event_loops has been timing
out consistently on FreeBSD in the CI. This commit attempts
to mitigate the problem by extending

test: increase test timeout

The test threadpool_multiple_event_loops has been timing
out consistently on FreeBSD in the CI. This commit attempts
to mitigate the problem by extending the test timeout.

PR-URL: https://github.com/libuv/libuv/pull/2304
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>

show more ...

unix: simplify uv/posix.h include logic

This commit includes uv/posix.h in one #elif instead of four.

PR-URL: https://github.com/libuv/libuv/pull/2302
Reviewed-By: Refael Ackerm

unix: simplify uv/posix.h include logic

This commit includes uv/posix.h in one #elif instead of four.

PR-URL: https://github.com/libuv/libuv/pull/2302
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>

show more ...

Now working on 1.29.1

Fixes: https://github.com/libuv/libuv/issues/2293

Add SHA to ChangeLog

2019.05.16, Version 1.29.0 (Stable)

Changes since version 1.28.0:

* ibmi: read memory and CPU usage info (Xu Meng)

* doc: update the cmake testing instruction (zlargon)

2019.05.16, Version 1.29.0 (Stable)

Changes since version 1.28.0:

* ibmi: read memory and CPU usage info (Xu Meng)

* doc: update the cmake testing instruction (zlargon)

* unix: fix race condition in uv_async_send() (Ben Noordhuis)

* linux: use O_CLOEXEC instead of EPOLL_CLOEXEC (Ben Noordhuis)

* doc: mark uv_async_send() as async-signal-safe (Ben Noordhuis)

* linux: init st_flags and st_gen when using statx (Oscar Waddell)

* linux: read free/total memory from /proc/meminfo (Ben Noordhuis)

* test: test zero-sized uv_fs_sendfile() writes (Ben Noordhuis)

* unix: don't assert on UV_PROCESS_WINDOWS_* flags (Ben Noordhuis)

* linux: set correct mac address for IP-aliases (Santiago Gimeno)

* win,util: fix null pointer dereferencing (Tobias Nießen)

* unix,win: fix `uv_fs_poll_stop()` when active (Anna Henningsen)

* doc: add missing uv_fs_type entries (Michele Caini)

* doc: fix build with sphinx 2.x (FX Coudert)

* unix: don't make statx system call on Android (George Zhao)

* unix: fix clang scan-build warning (Kyle Edwards)

* unix: fall back to kqueue on older macOS systems
(ken-cunningham-webuse)

* unix,win: add uv_get_constrained_memory() (Kelvin Jin)

* darwin: fix thread cancellation fd leak (Ben Noordhuis)

* linux: fix thread cancellation fd leak (Ben Noordhuis)

show more ...

linux: fix thread cancellation fd leak

Same as the previous commit but this time for Linux.

Specifically, the glibc close() wrapper is a cancellation point;
the thread is unwoun

linux: fix thread cancellation fd leak

Same as the previous commit but this time for Linux.

Specifically, the glibc close() wrapper is a cancellation point;
the thread is unwound when close() is called when the thread is
in the "cancel" state.

That's according to spec and therefore arguably less severe than
MacOS's "EINTR without actually closing" behavior but we can avoid
the file descriptor leak altogether by sidestepping glibc and
making the system call directly.

Musl libc is unaffected, its close() wrapper doesn't check
the cancel state.

PR-URL: https://github.com/libuv/libuv/pull/2291
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

darwin: fix thread cancellation fd leak

The close() system call tests for thread cancellation first. That has
the unfortunate side effect of aborting the sytem call with EINTR
withou

darwin: fix thread cancellation fd leak

The close() system call tests for thread cancellation first. That has
the unfortunate side effect of aborting the sytem call with EINTR
without actually closing the file descriptor when the thread is in
the "cancel" state. Work around that by calling close$NOCANCEL().

This might well qualify as an academic bug because approximately no one
uses thread cancellation but let's aim for correctness anyway.

PR-URL: https://github.com/libuv/libuv/pull/2291
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

unix,win: add uv_get_constrained_memory()

Fixes: https://github.com/libuv/libuv/issues/2286
PR-URL: https://github.com/libuv/libuv/pull/2289
Reviewed-By: Ben Noordhuis <info@bnoordhu

unix,win: add uv_get_constrained_memory()

Fixes: https://github.com/libuv/libuv/issues/2286
PR-URL: https://github.com/libuv/libuv/pull/2289
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

unix: fall back to kqueue on older macOS systems

Newer macOS systems (10.7+) use fsevents() for the fs_event backend,
but older macOS systems don't have full functionality for this in

unix: fall back to kqueue on older macOS systems

Newer macOS systems (10.7+) use fsevents() for the fs_event backend,
but older macOS systems don't have full functionality for this in
the macOS fsevents() API. Fall back to kqueue for these systems.

PR-URL: https://github.com/libuv/libuv/pull/2290
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

unix: fix clang scan-build warning

When building libuv with clang scan-build, an "Assigned value is
garbage or undefined" warning is emitted unless the return value
of open() is veri

unix: fix clang scan-build warning

When building libuv with clang scan-build, an "Assigned value is
garbage or undefined" warning is emitted unless the return value
of open() is verified to be >= 0. This commit fixes the warning.

PR-URL: https://github.com/libuv/libuv/pull/2140
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...

unix: don't make statx system call on Android

Fixes: https://github.com/libuv/libuv/issues/2282
PR-URL: https://github.com/libuv/libuv/pull/2284
Reviewed-By: Ben Noordhuis <info@bnoo

unix: don't make statx system call on Android

Fixes: https://github.com/libuv/libuv/issues/2282
PR-URL: https://github.com/libuv/libuv/pull/2284
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

doc: fix build with sphinx 2.x

app.info() was deprecated on Jan 4, 2017 (sphinx-doc/sphinx#3267),
and removed as of Sphinx 2.0.0. This commit removes the usage of
app.info().

doc: fix build with sphinx 2.x

app.info() was deprecated on Jan 4, 2017 (sphinx-doc/sphinx#3267),
and removed as of Sphinx 2.0.0. This commit removes the usage of
app.info().

PR-URL: https://github.com/libuv/libuv/pull/2265
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

doc: add missing uv_fs_type entries

PR-URL: https://github.com/libuv/libuv/pull/2280
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

unix,win: fix `uv_fs_poll_stop()` when active

Fix `uv_fs_poll_stop()` for active handles by not attempting to
mark the `uv_fs_poll_t` handle as closing when `uv_close()`
hasn’t been

unix,win: fix `uv_fs_poll_stop()` when active

Fix `uv_fs_poll_stop()` for active handles by not attempting to
mark the `uv_fs_poll_t` handle as closing when `uv_close()`
hasn’t been called on it.

Fixes: https://github.com/libuv/libuv/issues/2287
PR-URL: https://github.com/libuv/libuv/pull/2288
Refs: https://github.com/libuv/libuv/pull/1875
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>

show more ...

win,util: fix null pointer dereferencing

If uv__calloc returns NULL, the function jumps to the error label and
attempts to access array elements of cpu_infos (which is NULL).

PR

win,util: fix null pointer dereferencing

If uv__calloc returns NULL, the function jumps to the error label and
attempts to access array elements of cpu_infos (which is NULL).

PR-URL: https://github.com/libuv/libuv/pull/2264
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...

linux: set correct mac address for IP-aliases

IP-alias format is the interface name followed by a colon and a string
(usually a number). Set the physical address to the one in the 'base'

linux: set correct mac address for IP-aliases

IP-alias format is the interface name followed by a colon and a string
(usually a number). Set the physical address to the one in the 'base'
interface.

PR-URL: https://github.com/libuv/libuv/pull/2248
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...

unix: don't assert on UV_PROCESS_WINDOWS_* flags

UV_PROCESS_WINDOWS_HIDE_CONSOLE and UV_PROCESS_WINDOWS_HIDE_GUI were
whitelisted on Windows but not Unices. Now they are.

Bug in

unix: don't assert on UV_PROCESS_WINDOWS_* flags

UV_PROCESS_WINDOWS_HIDE_CONSOLE and UV_PROCESS_WINDOWS_HIDE_GUI were
whitelisted on Windows but not Unices. Now they are.

Bug introduced in commit 4c2dcca27 ("win: support more fine-grained
windows hiding") which I reviewed but where I failed to spot it. Mea
culpa.

Fixes: https://github.com/libuv/libuv/issues/2266
PR-URL: https://github.com/libuv/libuv/pull/2278
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Kyle Edwards <kyle.edwards@kitware.com>
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

test: test zero-sized uv_fs_sendfile() writes

This was reported as a bug in November 2018 but it appears to be working
now. Add a regression test to ensure it stays that way.

Fi

test: test zero-sized uv_fs_sendfile() writes

This was reported as a bug in November 2018 but it appears to be working
now. Add a regression test to ensure it stays that way.

Fixes: https://github.com/libuv/libuv/issues/2076
PR-URL: https://github.com/libuv/libuv/pull/2279
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Jameson Nash <vtjnash@gmail.com>
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>

show more ...

linux: read free/total memory from /proc/meminfo

It was reported that uv_get_free_memory() and uv_get_total_memory()
report the wrong values inside an lxc container.

Libuv calls

linux: read free/total memory from /proc/meminfo

It was reported that uv_get_free_memory() and uv_get_total_memory()
report the wrong values inside an lxc container.

Libuv calls sysinfo(2) but that isn't intercepted by lxc. /proc/meminfo
however is because /proc is a FUSE fs inside the container.

This commit makes libuv try /proc/meminfo first and fall back to
sysinfo(2) in case /proc isn't mounted.

Fixes: https://github.com/libuv/libuv/issues/2249
PR-URL: https://github.com/libuv/libuv/pull/2258
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

linux: init st_flags and st_gen when using statx

Explicitly initialize uv_stat_t fields st_flags and st_gen when using
statx as uv__to_stat does when statx is not available. This makes

linux: init st_flags and st_gen when using statx

Explicitly initialize uv_stat_t fields st_flags and st_gen when using
statx as uv__to_stat does when statx is not available. This makes
valgrind happier.

PR-URL: https://github.com/libuv/libuv/pull/2263
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Jameson Nash <vtjnash@gmail.com>

show more ...

doc: mark uv_async_send() as async-signal-safe

Refs: https://github.com/libuv/libuv/issues/2173
PR-URL: https://github.com/libuv/libuv/pull/2273
Reviewed-By: Anna Henningsen <anna@ad

doc: mark uv_async_send() as async-signal-safe

Refs: https://github.com/libuv/libuv/issues/2173
PR-URL: https://github.com/libuv/libuv/pull/2273
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

linux: use O_CLOEXEC instead of EPOLL_CLOEXEC

It was reported that EPOLL_CLOEXEC is not defined on Android API < 21,
a.k.a. Lollipop. Since EPOLL_CLOEXEC is an alias for O_CLOEXEC on all

linux: use O_CLOEXEC instead of EPOLL_CLOEXEC

It was reported that EPOLL_CLOEXEC is not defined on Android API < 21,
a.k.a. Lollipop. Since EPOLL_CLOEXEC is an alias for O_CLOEXEC on all
architectures, we just use that instead.

Fixes: https://github.com/libuv/libuv/issues/2167
PR-URL: https://github.com/libuv/libuv/pull/2272
Refs: https://github.com/libuv/libuv/pull/2216
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

show more ...

unix: fix race condition in uv_async_send()

After inspection of the code, it seems like there is a race window
between the cmpxchgi() and uv__async_send() calls.

If the event lo

unix: fix race condition in uv_async_send()

After inspection of the code, it seems like there is a race window
between the cmpxchgi() and uv__async_send() calls.

If the event loop thread is already busy looping over the async handles,
it can invoke the callback - which in turn can close the handle - before
the other thread reaches the uv__async_send() call. That's bad because
it accesses the handle that at that point might not be valid anymore.

Fix that by introducing an ad hoc spinlock that blocks the event loop
thread until the sending thread is done. It's not pretty or elegant
but it fixes the immediate bug and appears to have no measurable
impact on the async handle benchmarks.

Fixes: https://github.com/libuv/libuv/issues/2226
PR-URL: https://github.com/libuv/libuv/pull/2231
Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

doc: update the cmake testing instruction

PR-URL: https://github.com/libuv/libuv/pull/2253
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>