Commit message (Collapse) | Author | Age | |
---|---|---|---|
* | Add a byte counter to chttp2_stream and use that for timestamps | Yash Tibrewal | 2018-11-29 |
| | |||
* | Do not add the TCP buffer length. | Yash Tibrewal | 2018-11-29 |
| | |||
* | Add the length of the buffer that is traced | Yash Tibrewal | 2018-11-28 |
| | |||
* | Merge pull request #17244 from guantaol/ev_epollbg | Guantao Liu | 2018-11-28 |
|\ | | | | | Preparation for the new background poller 'epollbg' | ||
* | | Use grpc_event_engine_can_track_errors | Yash Tibrewal | 2018-11-27 |
| | | |||
* | | Always set fd to be readable/writable on receiving EPOLLERR | Yash Tibrewal | 2018-11-27 |
| | | |||
| * | Revert "Use a static local flag to memorize whether the grpc event engine ↵ | Guantao Liu | 2018-11-27 |
| | | | | | | | | | | | | runs in background or not" This reverts commit 9bbda894cbd845cea48a76c536d9731436c6313f. | ||
| * | Revert "Resolve naming conflicts" | Guantao Liu | 2018-11-27 |
| | | | | | | | | This reverts commit c1af11fbd622c640b15a7c8e5977e0c40a546969. | ||
* | | Enable errqueue support for linux kernel versions 4.0.0 and above | Yash Tibrewal | 2018-11-27 |
| | | |||
| * | Resolve naming conflicts | Guantao Liu | 2018-11-26 |
| | | |||
| * | Use a static local flag to memorize whether the grpc event engine runs in ↵ | Guantao Liu | 2018-11-26 |
| | | | | | | | | background or not | ||
* | | Add can_track_err methods to other platforms too | Yash Tibrewal | 2018-11-21 |
| | | |||
* | | Reviewer comments | Yash Tibrewal | 2018-11-20 |
| | | |||
| * | Also extend iomgr_windows.cc | Guantao Liu | 2018-11-19 |
| | | |||
| * | Extend ev_posix.* to prepare for the new background poller 'epollbg', | Guantao Liu | 2018-11-19 |
| | | | | | | | | | | | | | | | | | | and get rid of the dependency loop on the grpc shutdown path. Make sure all background closures are complete before shutting down the other grpc modules. Avoid using the backup poller in TCP endpoints if using the background poller. | ||
* | | Cleanup | Yash Tibrewal | 2018-11-16 |
| | | |||
* | | Merge master | Yash Tibrewal | 2018-11-16 |
|\| | |||
| * | Enable SO_REUSEADDR in client sockets | Guantao Liu | 2018-11-09 |
| | | |||
| * | Allow setsockopt and getsockopt to softly fail for TCP_USER_TIMEOUT | Yash Tibrewal | 2018-11-06 |
| | | |||
| * | Remeve memset(0) from arena allocated memory. | Soheil Hassas Yeganeh | 2018-11-05 |
| | | | | | | | | | | | | | | Callers are updated to properly initialize the memory. This behavior can be overridden using GRPC_ARENA_INIT_STRATEGY environment variable. | ||
| * | Merge pull request #17083 from yashykt/tcpusertimeout | Yash Tibrewal | 2018-11-02 |
| |\ | | | | | | | Do not make TCP_USER_TIMEOUT failure an error | ||
| | * | Do not make TCP_USER_TIMEOUT failure an error | Yash Tibrewal | 2018-11-02 |
| | | | |||
| * | | Merge pull request #16823 from AspirinSJL/rq | Juanli Shen | 2018-10-25 |
| |\ \ | | | | | | | | | Account the memory usage of channel and call by resource quota | ||
| | * | | Fix OOM by expanding resource quota coverage | Juanli Shen | 2018-10-25 |
| | | | | |||
| * | | | Don't unnecessarily create exec_ctx | Yash Tibrewal | 2018-10-25 |
| |/ / | |||
| * | | Use union to make alignement robust. | Soheil Hassas Yeganeh | 2018-10-19 |
| | | | | | | | | | | | | Suggested-by: vjpai@google.com | ||
| * | | Do not waste cache lines with unnecessary paddings. | Soheil Hassas Yeganeh | 2018-10-19 |
| | | | | | | | | | | | | | | | | | | Make sure paddings are GPR_CACHELINE_SIZE minus the size of fields in that cache line. Otherwise, we will waste an additional cache line for no good reason. | ||
| * | | Add comment for grpc_error_get_int(). | Soheil Hassas Yeganeh | 2018-10-18 |
| | | | | | | | | | | | | Explain that the input parameter cannot be nullptr. | ||
| * | | Optimize error handling for special cases. | Soheil Hassas Yeganeh | 2018-10-18 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit contains a few improvements: 1. Using a consequetive range of [0..4], will allow us to merge all branches of error_is_special into one comparison. 2. With (1), we can remove the for loops to find entries in error_status_map with a single O(1) lookup. 3. grpc_error_is_special() code paths should be inlined for ref and unref to avoid callq for the majority of cases where speical error is used. 4. grpc_error_get_int() should never accept a nullptr argument to avoid an expensive branch in the hot path. Callers should all allocate a dummy int on the stack when calling. | ||
| * | | Merge pull request #16885 from ncteisen/tcp-user-timeout | Noah Eisen | 2018-10-16 |
| |\ \ | | | | | | | | | Protect Spammy Log with Tracer | ||
| * \ \ | Merge pull request #10859 from haorenfsa/master | hcaseyal | 2018-10-16 |
| |\ \ \ | | | | | | | | | | | fix: while using eventfd, a random fd will be closed when error occurs during creating eventfd | ||
| | | * | | protect spammy log with tracer | ncteisen | 2018-10-15 |
| | |/ / | |/| | | |||
| * | | | Merge pull request #16838 from soheilhy/worktree-tcp-read | Soheil Hassas Yeganeh | 2018-10-12 |
| |\ \ \ | | | | | | | | | | | Simple optimizations in TCP read | ||
| | * | | | Do not wait for allocation if buffer is less than half the target. | Soheil Hassas Yeganeh | 2018-10-11 |
| | | | | | | | | | | | | | | | | | | | | | | | | | We overallocate by 2x for target. Unless buffer is more than half full, we should not delay read for more allocation. | ||
| | * | | | Update TCP read estimates as soon as we read the whole buffer. | Soheil Hassas Yeganeh | 2018-10-11 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If we have a continous stream of bytes on the socket, we will never grow the buffer, because we will never get EAGAIN, and call finish. This is a serious performance issue, which can be misued. As soon as we have a full buffer, update the estimate. | ||
| * | | | | Merge pull request #16753 from jtattermusch/unify_fallthrough | Jan Tattermusch | 2018-10-11 |
| |\ \ \ \ | | |/ / / | |/| | | | Unify fallthrough comment in switch statements | ||
| * | | | | Remove grpc_use_signal which is no longer relevant | Vijay Pai | 2018-10-10 |
| | | | | | |||
| | * | | | unify fallthrough comment in switch statements | Jan Tattermusch | 2018-10-10 |
| |/ / / | |||
| * | | | Merge pull request #16750 from AspirinSJL/timer | Juanli Shen | 2018-10-08 |
| |\ \ \ | | | | | | | | | | | Clean some timer code | ||
| * \ \ \ | Merge pull request #16795 from yashykt/disable_error_tracking | Yash Tibrewal | 2018-10-08 |
| |\ \ \ \ | | | | | | | | | | | | | Disable error tracking until Fathom changes are commited | ||
| * \ \ \ \ | Merge pull request #16714 from muxi/add-timer-manager-debug | Muxi Yan | 2018-10-05 |
| |\ \ \ \ \ | | | | | | | | | | | | | | | Core infrastructure for timer manager debug | ||
| | | * | | | | Disable error tracking till Fathom changes are commited | Yash Tibrewal | 2018-10-05 |
| | | | | | | | |||
| | * | | | | | Clean-up unused variable | Muxi Yan | 2018-10-03 |
| | | | | | | | |||
| | * | | | | | Polish macro again | Muxi Yan | 2018-10-02 |
| | | | | | | | |||
| * | | | | | | Shard channelz stats by CPU | ncteisen | 2018-10-02 |
| | |/ / / / | |/| | | | | |||
| | | * | | | Clean some timer code | Juanli Shen | 2018-10-01 |
| | | | | | | |||
| | * | | | | polish macro | Muxi Yan | 2018-10-01 |
| | | | | | | |||
| | * | | | | Add alias for macro for internal app | Muxi Yan | 2018-10-01 |
| | | | | | | |||
| * | | | | | Address feedback comment | Sree Kuchibhotla | 2018-09-28 |
| | | | | | | |||
| | | | * | | read_fd should always have a certain value when create error occurs [fix: ↵ | haorenfsa | 2018-09-29 |
| | | |/ / | | | | | | | | | | | | | | | | while using eventfd, when error occurs during creating eventfd, a random fd will be closed] |