aboutsummaryrefslogtreecommitdiff
path: root/src/os/unix
Commit message (Collapse)AuthorAge
...
* Core: fix typo in error message.Piotr Sikora2015-12-02
| | | | Signed-off-by: Piotr Sikora <piotrsikora@google.com>
* Used the pwritev() syscall for writing files where possible.Valentin Bartenev2015-11-17
| | | | It is more effective, because it doesn't require a separate lseek().
* Moved file writev() handling code to a separate function.Valentin Bartenev2015-11-17
| | | | No functional changes.
* Handled EINTR from write() and pwrite() syscalls.Valentin Bartenev2015-11-17
| | | | This is in addition to 6fce16b1fc10.
* Adjusted file->sys_offset after the write() syscall.Valentin Bartenev2015-11-17
| | | | | | | | | This fixes suboptimal behavior caused by surplus lseek() for sequential writes on systems without pwrite(). A consecutive read after write might result in an error on systems without pread() and pwrite(). Fortunately, at the moment there are no widely used systems without these syscalls.
* Fixed building --with-debug, broken by 6fce16b1fc10.Valentin Bartenev2015-09-02
|
* Writing to some file systems can be interrupted.Valentin Bartenev2015-09-02
| | | | | At least such behavior was observed with CephFS, see: http://mailman.nginx.org/pipermail/nginx/2015-July/048188.html.
* Core: idle connections now closed only once on exiting.Valentin Bartenev2015-08-11
| | | | | | | | | | | | Iterating through all connections takes a lot of CPU time, especially with large number of worker connections configured. As a result nginx processes used to consume CPU time during graceful shutdown. To mitigate this we now only do a full scan for idle connections when shutdown signal is received. Transitions of connections to idle ones are now expected to be avoided if the ngx_exiting flag is set. The upstream keepalive module was modified to follow this.
* Core: added support for writing to stdout.Vladimir Homutov2015-06-16
|
* Configure: GNU Hurd properly recognized.Maxim Dounin2015-05-25
| | | | | | | | | | With this change it's no longer needed to pass -D_GNU_SOURCE manually, and -D_FILE_OFFSET_BITS=64 is set to use 64-bit off_t. Note that nginx currently fails to work properly with master process enabled on GNU Hurd, as fcntl(F_SETOWN) returns EOPNOTSUPP for sockets as of GNU Hurd 0.6. Additionally, our strerror() preloading doesn't work well with GNU Hurd, as it uses large numbers for most errors.
* Introduced worker number, ngx_worker.Maxim Dounin2015-05-20
|
* Removed the "worker_rlimit_sigpending" directive.Ruslan Ermilov2015-04-23
| | | | It was only needed by the just removed rtsig module.
* Removed the obsolete rtsig module.Ruslan Ermilov2015-04-23
|
* Removed the obsolete aio module.Ruslan Ermilov2015-04-22
|
* Used the correct type for the AIO preload handler return value.Valentin Bartenev2015-04-06
|
* Removed busy locks.Ruslan Ermilov2015-03-20
|
* Removed unix ngx_threaded and related ngx_process_changes.Ruslan Ermilov2015-03-20
|
* Removed old pthread implementation.Ruslan Ermilov2015-03-20
|
* Removed old FreeBSD rfork() thread implementation.Ruslan Ermilov2015-03-20
|
* Added support for offloading Linux sendfile() in thread pools.Valentin Bartenev2015-03-14
|
* Added support for offloading read() in thread pools.Valentin Bartenev2015-03-14
|
* Thread pools implementation.Valentin Bartenev2015-03-14
|
* Renamed NGX_THREADS to NGX_OLD_THREADS because of deprecation.Ruslan Ermilov2015-03-04
| | | | It's mostly dead code and the original idea of worker threads has been rejected.
* Refactored ngx_linux_sendfile_chain() even more.Valentin Bartenev2015-02-27
| | | | | | The code that calls sendfile() was cut into a separate function. This simplifies EINTR processing, yet is needed for the following changes that add threads support.
* Refactored sendfile() AIO preload.Valentin Bartenev2015-02-11
| | | | | | | | This reduces layering violation and simplifies the logic of AIO preread, since it's now triggered by the send chain function itself without falling back to the copy filter. The context of AIO operation is now stored per file buffer, which makes it possible to properly handle cases when multiple buffers come from different locations, each with its own configuration.
* Fixed sendfile() trailers on OS X (8e903522c17a, 1.7.8).Maxim Dounin2015-01-13
| | | | | | | The trailer.count variable was not initialized if there was a header, resulting in "sendfile() failed (22: Invalid argument)" alerts on OS X if the "sendfile" directive was used. The bug was introduced in 8e903522c17a (1.7.8).
* Fixed building with musl libc (ticket #685).Maxim Dounin2014-12-24
|
* Style.Valentin Bartenev2014-11-19
|
* Fixed type of sendfile() return value on Linux.Valentin Bartenev2014-11-19
| | | | | There was no real problem since the amount of bytes can be sent is limited by NGX_SENDFILE_MAXSIZE to less than 2G. But that can be changed in the future
* Moved writev() handling code to a separate function.Valentin Bartenev2014-08-13
| | | | | This reduces code duplication and unifies debug logging of the writev() syscall among various send chain functions.
* Merged conditions in the ngx_*_sendfile_chain() functions.Valentin Bartenev2014-08-13
| | | | No functional changes.
* Moved the code for coalescing file buffers to a separate function.Valentin Bartenev2014-08-13
|
* Refactored ngx_solaris_sendfilev_chain().Valentin Bartenev2014-11-19
| | | | | | | Though ngx_solaris_sendfilev_chain() shouldn't suffer from the problem mentioned in d1bde5c3c5d2 since currently IOV_MAX on Solaris is 16, but this follows the change from 3d5717550371 in order to make the code look similar to other systems and potentially eliminates the problem in the future.
* Introduced the ngx_output_chain_to_iovec() function.Valentin Bartenev2014-10-07
| | | | | It deduplicates code of the send chain functions and uses only preallocated memory, which completely solves the problem mentioned in d1bde5c3c5d2.
* Renamed ngx_handle_sent_chain() to ngx_chain_update_sent().Valentin Bartenev2014-11-19
| | | | No functional changes.
* Events: introduced cancelable timers.Valentin Bartenev2014-08-13
|
* Events: removed broken thread support from event timers.Valentin Bartenev2014-08-25
| | | | | It's mostly dead code. And the idea of thread support for this task has been deprecated.
* Core: added limit to recv_chain().Roman Arutyunyan2014-10-28
|
* Fixed counting of sent bytes in the send chain functions on EINTR.Valentin Bartenev2014-08-27
| | | | | | Previously, a value of the "send" variable wasn't properly adjusted in a rare case when syscall was interrupted by a signal. As a result, these functions could send less data than the limit allows.
* Removed duplicate initialization of the "rev" variable.Valentin Bartenev2014-09-22
|
* Generalized definitions of the number of preallocated iovec's.Valentin Bartenev2014-08-13
| | | | No functional changes.
* Reduced difference between the send chain functions.Valentin Bartenev2014-08-13
| | | | No functional changes. This follows the change from ad137a80919f.
* Merged implementations of ngx_readv_chain().Valentin Bartenev2014-08-13
| | | | | There's no real need in two separate implementations, with and without kqueue support.
* Removed the "complete" variable from various send chain functions.Valentin Bartenev2014-08-13
| | | | | It was made redundant by the previous change, since the "sent" variable is no longer modified.
* Moved the code for adjusting sent buffers in a separate function.Valentin Bartenev2014-08-13
|
* Fixed writev() debug log message in ngx_darwin_sendfile_chain().Valentin Bartenev2014-08-13
|
* Added ngx_init_setproctitle() return code check.Maxim Dounin2014-09-08
| | | | | | | | The ngx_init_setproctitle() function, as used on systems without setproctitle(3), may fail due to memory allocation errors, and therefore its return code needs to be checked. Reported by Markus Linnala.
* Events: removed broken thread support from posted events.Valentin Bartenev2014-09-01
| | | | | It's mostly dead code. And the idea of thread support for this task has been deprecated.
* Core: exit on ngx_pnalloc() failure.Piotr Sikora2014-08-01
| | | | Signed-off-by: Piotr Sikora <piotr@cloudflare.com>
* Bigger iovec buffer in ngx_readv_chain().Maxim Dounin2014-07-28
| | | | | | This helps to reduce likelyhood of memory allocations in ngx_readv_chain(), which are known to lead to noticeable effects in some cases, see http://mailman.nginx.org/pipermail/nginx/2014-July/044512.html.