4c9e42d0 | 05-Nov-2012 |
Bert Belder |
windows: un-break the build It was broken in fb64948. |
6620e614 | 05-Nov-2012 |
Bert Belder |
windows: un-break the build It was broken in 1d85815. |
0ddf9d6b | 04-Nov-2012 |
Ben Noordhuis |
Revert "unix: use select() for specific fds on OS X" This reverts commit 5da380a5ca7f9ef02667ef3021fc97428df0b478. Contains a bug that effectively makes the select() thread busy-loo
Revert "unix: use select() for specific fds on OS X" This reverts commit 5da380a5ca7f9ef02667ef3021fc97428df0b478. Contains a bug that effectively makes the select() thread busy-loop. The file descriptor is polled for both reading and writing, regardless of what events the main thread wants to receive. Fixing that requires proper synchronization between the two threads. See #614.
show more ...
|
20bb1bfd | 04-Nov-2012 |
Ben Noordhuis |
Revert "unix: reopen tty as /dev/tty" This reverts commit 31f9fbce6321e2a26ab6d6c2d404aba1a7f161a7. The reverted commit depends on commit 5da380a ("use select() for specific fds
Revert "unix: reopen tty as /dev/tty" This reverts commit 31f9fbce6321e2a26ab6d6c2d404aba1a7f161a7. The reverted commit depends on commit 5da380a ("use select() for specific fds on OS X") which polls /dev/tty file descriptors in a separate thread to work around deficiencies in the kqueue API on OS X. Unfortunately, 5da380a has a bug that effectively makes the select() thread busy-loop. Revert this commit for now.
show more ...
|
b5fc9444 | 03-Nov-2012 |
Ben Noordhuis |
build: export _LARGEFILE_SOURCE, _FILE_OFFSET_BITS Export compile-time defines that influence the size of struct stat. This should avoid sizeof(struct stat) mismatches between libuv
build: export _LARGEFILE_SOURCE, _FILE_OFFSET_BITS Export compile-time defines that influence the size of struct stat. This should avoid sizeof(struct stat) mismatches between libuv and dependent projects. On OS X, _DARWIN_USE_64_BIT_INODE=1 is also exported.
show more ...
|
914185d6 | 03-Nov-2012 |
Ben Noordhuis |
unix: make some internal methods static |
fb649487 | 01-Nov-2012 |
Charlie McConnell |
unix: do not set environ unless one is provided Currently, `uv_spawn` will set `environ` to the value of `options.env`, even if `options.env` is `NULL`. This results in child processes
unix: do not set environ unless one is provided Currently, `uv_spawn` will set `environ` to the value of `options.env`, even if `options.env` is `NULL`. This results in child processes for whom `environ == NULL`, which can cause a variety of unexpected issues. This is a back-port of commit 1d85815 from the master branch.
show more ...
|
1d858156 | 01-Nov-2012 |
Charlie McConnell |
unix: do not set environ unless one is provided Currently, `uv_spawn` will set `environ` to the value of `options.env`, even if `options.env` is `NULL`. This results in child processes
unix: do not set environ unless one is provided Currently, `uv_spawn` will set `environ` to the value of `options.env`, even if `options.env` is `NULL`. This results in child processes for whom `environ == NULL`, which can cause a variety of unexpected issues.
show more ...
|
dcce1eab | 02-Nov-2012 |
Ben Noordhuis |
test: fix -Wmissing-field-initializers warnings |
894a8523 | 02-Nov-2012 |
Ben Noordhuis |
test: fix signed/unsigned comparison warnings |
9c7ae2e7 | 02-Nov-2012 |
Ben Noordhuis |
darwin: don't use deprecated AbsoluteToNanoseconds |
be597ba6 | 02-Nov-2012 |
Ben Noordhuis |
unix: set req type to UV_FS |
b0bcbdfe | 01-Nov-2012 |
Ben Noordhuis |
unix: NULL pipe_fname in uv__pipe_close() Pro-actively avoid use-after-free errors, set the pipe_fname field to NULL. |
b8aa5b9b | 01-Nov-2012 |
Ben Noordhuis |
test: don't assert on UV_EPIPE in echo-server.c UV_EPIPE is not an error per se, it simply indicates that the other end of the connection - i.e. the test case - has gone away. P
test: don't assert on UV_EPIPE in echo-server.c UV_EPIPE is not an error per se, it simply indicates that the other end of the connection - i.e. the test case - has gone away. Pro-actively ignore UV_ECANCELED errors. They're not actually emitted right now because there's only ever one pending write but let's be forward compatible.
show more ...
|
225c6f17 | 01-Nov-2012 |
Ben Noordhuis |
unix, windows: fix several error messages |
0d5d6504 | 01-Nov-2012 |
Ben Noordhuis |
unix, windows: fix EAGAIN error message |
97c527ac | 25-Oct-2012 |
Leonard Hecker |
darwin: make it possible to compile for iOS Relocate the include of TargetConditionals.h and fixe the use of TARGET_OS_IPHONE. Furthermore, uv__fsevents_init() and uv__fsevents_close are
darwin: make it possible to compile for iOS Relocate the include of TargetConditionals.h and fixe the use of TARGET_OS_IPHONE. Furthermore, uv__fsevents_init() and uv__fsevents_close are now empty functions for iOS, since the FSEvents API is not available there.
show more ...
|
149b16f1 | 25-Oct-2012 |
Bert Belder |
windows: closing handles should always keep the loop alive This makes the tcp-ref2 and udp-ref2 tests pass again. Also adds another reference count test. |
6150feda | 25-Oct-2012 |
Ben Noordhuis |
unix: fix ‘fd’ undeclared build error Reapplies commit b5028c5b, failed to merge in 5cb4197. |
5cb4197d | 25-Oct-2012 |
Ben Noordhuis |
Merge branch 'v0.8'
|
f43ad85e | 25-Oct-2012 |
Ben Noordhuis |
include: fix ngx_queue_foreach() macro Guard against the possibility that the queue is emptied while we're iterating over it. Simple test case: #include "ngx-queue.h" #i
include: fix ngx_queue_foreach() macro Guard against the possibility that the queue is emptied while we're iterating over it. Simple test case: #include "ngx-queue.h" #include <assert.h> int main(void) { ngx_queue_t h; ngx_queue_t v[2]; ngx_queue_t* q; unsigned n = 0; ngx_queue_init(&h); ngx_queue_insert_tail(&h, v + 0); ngx_queue_insert_tail(&h, v + 1); ngx_queue_foreach(q, &h) { ngx_queue_remove(v + 0); ngx_queue_remove(v + 1); n++; } assert(n == 1); // *not* 2 return 0; } Fixes #605.
show more ...
|
0dbab845 | 24-Oct-2012 |
Bert Belder |
benchmark: async_pummel should not call uv_async_send on closed handle This fixes an assertion that triggered in debug builds on Windows. |
a54b9e29 | 24-Oct-2012 |
Bert Belder |
benchmark: timed_udp_pummel should not write to closed udp handle This fixes intermittent assertion failures when running the benchmark. |
31f9fbce | 24-Oct-2012 |
Ben Noordhuis |
unix: reopen tty as /dev/tty Reopen the file descriptor when it refers to a tty. This lets us put the tty in non-blocking mode without affecting other processes that share it with us.
unix: reopen tty as /dev/tty Reopen the file descriptor when it refers to a tty. This lets us put the tty in non-blocking mode without affecting other processes that share it with us. Fixes #601.
show more ...
|
b7f38b1e | 24-Oct-2012 |
Ben Noordhuis |
Revert "unix: avoid iterating over all async handles" This reverts commit 209abbab27abdc63200cbd707f72fed4245ab4d0. Fixes the following SIGSEGV: (gdb) f 1 #1 0x000
Revert "unix: avoid iterating over all async handles" This reverts commit 209abbab27abdc63200cbd707f72fed4245ab4d0. Fixes the following SIGSEGV: (gdb) f 1 #1 0x00007fc084683aec in uv__async_io (loop=0x7fc0848e0b40, handle=0x7fc0848e0c78, events=1) at src/unix/async.c:175 175 ASYNC_CB(h) (gdb) list 170 171 /* If we need to sweep all handles anyway - skip this loop */ 172 if (!loop->async_sweep_needed) { 173 for (i = 0; i < end; i += sizeof(h)) { 174 h = *((uv_async_t**) (buf + i)); 175 ASYNC_CB(h) 176 } 177 } 178 179 bytes -= end; (gdb) print *h $1 = {close_cb = 0x184e1b0, data = 0x18d9520, loop = 0x7fc0848e0b40, type = 49, handle_queue = {prev = 0x18dae10, next = 0x7860c0}, flags = 32, next_closing = 0x1863b40, pending = 0, async_cb = 0x31, queue = {prev = 0x18dae50, next = 0x7860c0}} (gdb) It looks like the async handle gets closed or otherwise becomes invalid before the sweep is executed. Fixes #603.
show more ...
|