unix: Support atomic compare & swap xlC on AIX

On AIX with the native xlC compiler, the compiler supports sync
compare swap and compare in two flavors:
1 __compare_and_swap
2 _

unix: Support atomic compare & swap xlC on AIX

On AIX with the native xlC compiler, the compiler supports sync
compare swap and compare in two flavors:
1 __compare_and_swap
2 __compare_and_swaplp

This differs from the gcc version of this function in name. The
second version of this function supports 8-byte boundary for
a doubleword. Therefore, the macro logic checks:
if (AIX OS AND using xlC compiler)
if (64bit)
__compare_and_swaplp /* 64 bit version */
else
__compare_and_swap
endif
endif

Signed-off-by: Neil Mushell <nmushell@bloomberg.net>

PR-URL: https://github.com/libuv/libuv/pull/521
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

Working on the next version

Add SHA to ChangeLog

2015.09.12, Version 1.7.4 (Stable)

Changes since version 1.7.3:

* doc: uv_read_start and uv_read_cb clarifications (Ben Trask)

* freebsd: obtain true uptime through clock_g

2015.09.12, Version 1.7.4 (Stable)

Changes since version 1.7.3:

* doc: uv_read_start and uv_read_cb clarifications (Ben Trask)

* freebsd: obtain true uptime through clock_gettime() (Jianghua Yang)

* win, tty: do not convert \r to \r\n (Colin Snover)

* build,gyp: add DragonFly to the list of OSes (Michael Neumann)

* fs: fix bug in sendfile for DragonFly (Michael Neumann)

* doc: add uv_dlsym() return type (Brian White)

* tests: fix fs tests run w/o full getdents support (Jeremy Whitlock)

* doc: fix typo (Devchandra Meetei Leishangthem)

* doc: fix uv-unix.h location (Sakthipriyan Vairamani)

* unix: fix error check when closing process pipe fd (Ben Noordhuis)

* test,freebsd: fix ipc_listen_xx_write tests (Santiago Gimeno)

* win: fix unsavory rwlock fallback implementation (Bert Belder)

* doc: clarify repeat timer behavior (Eli Skeggs)

show more ...

doc: clarify repeat timer behavior

Signed-off-by: Eli Skeggs <skeggse@gmail.com>

Fixes: https://github.com/libuv/libuv/issues/181
PR-URL: https://github.com/libuv/libuv/pull/520

doc: clarify repeat timer behavior

Signed-off-by: Eli Skeggs <skeggse@gmail.com>

Fixes: https://github.com/libuv/libuv/issues/181
PR-URL: https://github.com/libuv/libuv/pull/520
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

win: fix unsavory rwlock fallback implementation

Before this patch an uv_mutex_t (backed by a critical section) could be
released by a tread different from the thread that acquired it, w

win: fix unsavory rwlock fallback implementation

Before this patch an uv_mutex_t (backed by a critical section) could be
released by a tread different from the thread that acquired it, which is
not allowed. This is fixed by using a semaphore instead.

Note that the affected code paths were used on Windows XP and Windows
Server 2003 only.

Fixes: https://github.com/libuv/libuv/issues/515
PR-URL: https://github.com/libuv/libuv/pull/516
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

test,freebsd: fix ipc_listen_xx_write tests

In FreeBSD if connect is called and the queue of pending connections(backlog)
is full, the call is automatically rejected with ECONNRESET. In

test,freebsd: fix ipc_listen_xx_write tests

In FreeBSD if connect is called and the queue of pending connections(backlog)
is full, the call is automatically rejected with ECONNRESET. In the tests, 100
connections were tried and the backlog of both servers is 12, so the error.

To fix ipc_listen_before_write and ipc_listen_after_write, the backlog of the
server has been increased to 128.

PR-URL: https://github.com/libuv/libuv/pull/504
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

unix: fix error check when closing process pipe fd

uv__close() does not set errno, it returns the error. It also never
returns EINTR, it always maps it to EINPROGRESS.

PR-URL:

unix: fix error check when closing process pipe fd

uv__close() does not set errno, it returns the error. It also never
returns EINTR, it always maps it to EINPROGRESS.

PR-URL: https://github.com/libuv/libuv/pull/512
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

doc: fix uv-unix.h location

uv-unix.h is in include directory, not in src.

PR-URL: https://github.com/libuv/libuv/pull/511
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

doc: fix typo

PR-URL: https://github.com/libuv/libuv/pull/509
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

tests: fix fs tests run w/o full getdents support

fs_access, fs_async_dir and fs_symlink_dir could fail when the
filesystem where the tests are run does not fully support getting the

tests: fix fs tests run w/o full getdents support

fs_access, fs_async_dir and fs_symlink_dir could fail when the
filesystem where the tests are run does not fully support getting the
file type via getdents. Having getdents support does not imply that the
underlying filesystem fully supports what is necessary for these tests
to have passed prior to this commit. So the failing tests were updated
to check for UV_DIRENT_FILE as it did previously but it will also check for
UV_DIRENT_UNKNOWN as well to handle the filesystems that do not fully support
getting the file type via getdents. For OS X and Windows, which are
known to fully support getting the file type, the tests work as they did
before and will not check for UV_DIRENT_UNKNOWN. We could/should update the
preprocessor directive that chooses "rigorous" or "lax" checks
accordingly when we learn of new environments that should always do the
original "rigorous" checks.

Fixes: https://github.com/libuv/libuv/issues/501
PR-URL: https://github.com/libuv/libuv/pull/508
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

doc: add uv_dlsym() return type

PR-URL: https://github.com/libuv/libuv/pull/503
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

fs: fix bug in sendfile for DragonFly

The code made an implicit assumption that the 'len` variable passed
to the sendfile(2) syscall is not modified by the operating system
in case o

fs: fix bug in sendfile for DragonFly

The code made an implicit assumption that the 'len` variable passed
to the sendfile(2) syscall is not modified by the operating system
in case of an error other than EAGAIN or EINTR.

The man page leaves this unspecified on FreeBSD, DragonFly and
Darwin, so better check the error code which returns a valid
value in `len` explicitly (only EAGAIN and EINTR).

This fixes the test case for sendfile on DragonFly.

PR-URL: https://github.com/libuv/libuv/pull/466
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

build,gyp: add DragonFly to the list of OSes

PR-URL: https://github.com/libuv/libuv/pull/466
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Saúl Ibarra Corretgé <saghul

build,gyp: add DragonFly to the list of OSes

PR-URL: https://github.com/libuv/libuv/pull/466
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

win, tty: do not convert \r to \r\n

\r is a single carriage return without line feed on all platforms,
including Windows.

PR-URL: https://github.com/libuv/libuv/pull/472
Rev

win, tty: do not convert \r to \r\n

\r is a single carriage return without line feed on all platforms,
including Windows.

PR-URL: https://github.com/libuv/libuv/pull/472
Reviewed-By: Bert Belder <bertbelder@gmail.com>

show more ...

freebsd: obtain true uptime through clock_gettime()

Obtain true uptime through clock_gettime() instead of subtracting
'bootime' from 'now'.

PR-URL: https://github.com/libuv/libu

freebsd: obtain true uptime through clock_gettime()

Obtain true uptime through clock_gettime() instead of subtracting
'bootime' from 'now'.

PR-URL: https://github.com/libuv/libuv/pull/497
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

doc: uv_read_start and uv_read_cb clarifications

PR-URL: https://github.com/libuv/libuv/pull/500
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

Add SHA to ChangeLog

2015.08.28, Version 1.7.3 (Stable)

Changes since version 1.7.2:

* threadpool: fix thread starvation bug (Ben Noordhuis)

threadpool: fix thread starvation bug

Commit 0f1bdb6 ("threadpool: send signal only when queue is empty")
introduces a regression where work is not evenly distributed across
the thre

threadpool: fix thread starvation bug

Commit 0f1bdb6 ("threadpool: send signal only when queue is empty")
introduces a regression where work is not evenly distributed across
the thread pool because the work queue's condition variable is only
signalled when the queue is empty, even when there are waiting workers.

It doesn't turn into outright deadlock because there is always
at least one thread making forward progress but it does degrade
throughput, sometimes massively so.

Signalling whenever there are waiting workers fixes the throughput
issue while still keeping the number of uv_cond_signal() calls low,
which was the motivation for commit 0f1bdb6.

Fixes: https://github.com/libuv/libuv/pull/490
Fixes: https://github.com/libuv/libuv/issues/492
PR-URL: https://github.com/libuv/libuv/pull/493
Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>

show more ...

Add SHA to ChangeLog

2015.08.25, Version 1.7.2 (Stable)

Changes since version 1.7.1:

* unix, win: make uv_loop_init return on error (Willem Thiart)

* win: reset pipe handle for pipe servers (Sa

2015.08.25, Version 1.7.2 (Stable)

Changes since version 1.7.1:

* unix, win: make uv_loop_init return on error (Willem Thiart)

* win: reset pipe handle for pipe servers (Saúl Ibarra Corretgé)

* win: fix replacing pipe handle for pipe servers (Saúl Ibarra Corretgé)

* win: fix setting pipe pending instances after bind (Saúl Ibarra
Corretgé)

show more ...

win: fix setting pipe pending instances after bind

It's an error to do so, so just ignore it. The test would cause an
invalid memory access if the fix is undone.

PR-URL: https:/

win: fix setting pipe pending instances after bind

It's an error to do so, so just ignore it. The test would cause an
invalid memory access if the fix is undone.

PR-URL: https://github.com/libuv/libuv/pull/488
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...

win: fix replacing pipe handle for pipe servers

On Windows we create multiple pipe handles (system handles) which are
attached to pending accept requests. Each of these will take turns i

win: fix replacing pipe handle for pipe servers

On Windows we create multiple pipe handles (system handles) which are
attached to pending accept requests. Each of these will take turns in
replacing the reference in handle->handle, so make sure we allow for
that **only** for pipe servers.

PR-URL: https://github.com/libuv/libuv/pull/488
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...

win: reset pipe handle for pipe servers

On pipe servers handle->handle takes turns between the different server
handles, so after closing all of them just reset the dandling reference.

win: reset pipe handle for pipe servers

On pipe servers handle->handle takes turns between the different server
handles, so after closing all of them just reset the dandling reference.

PR-URL: https://github.com/libuv/libuv/pull/488
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>

show more ...