Vim-fork focused on extensibility and usability
Go to file
Justin M. Keyes 0e75438a38 tui: update_size(): Fix race condition.
bridge.width and bridge.height reach ui.c:ui_refresh() when it iterates
through all UIs, so they do not need to be set directly by
tui.c:update_size().

Race found by helgrind:

==18532== Helgrind, a thread error detector
==18532== Copyright (C) 2007-2015, and GNU GPL'd, by OpenWorks LLP et al.
==18532== Using Valgrind-3.12.0 and LibVEX; rerun with -h for copyright info
==18532== Command: ./build/bin/nvim -u NONE --cmd set\ rtp+=~/.vim/bundle/vimfiler.vim,~/.vim/bundle/unite.vim --cmd runtime\ plugin/vimfiler.vim --cmd runtime\ plugin/unite.vim
==18532== Parent PID: 6477
==18532==
==18532== ---Thread-Announcement------------------------------------------
==18532==
==18532== Thread #2 was created
==18532==    at 0x68FA98E: clone (clone.S:73)
==18532==    by 0x5270179: create_thread (createthread.c:102)
==18532==    by 0x5271BE2: pthread_create@@GLIBC_2.2.5 (pthread_create.c:679)
==18532==    by 0x4C32B07: pthread_create_WRK (hg_intercepts.c:427)
==18532==    by 0x4E53A3F: uv_thread_create (in /usr/lib/x86_64-linux-gnu/libuv.so.1.0.0)
==18532==    by 0x6A7154: ui_bridge_attach (ui_bridge.c:89)
==18532==    by 0x6A164C: tui_start (tui.c:116)
==18532==    by 0x6A4CFC: ui_builtin_start (ui.c:89)
==18532==    by 0x55A825: main (main.c:433)
==18532==
==18532== ---Thread-Announcement------------------------------------------
==18532==
==18532== Thread #1 is the program's root thread
==18532==
==18532== ----------------------------------------------------------------
==18532==
==18532== Possible data race during write of size 4 at 0x770E7B4 by thread #2
==18532== Locks held: none
==18532==    at 0x6A3071: update_size (tui.c:759)
==18532==    by 0x6A30DB: sigwinch_cb (tui.c:269)
==18532==    by 0x4D0A54: signal_event (signal.c:44)
==18532==    by 0x4CDDB6: multiqueue_process_events (multiqueue.c:146)
==18532==    by 0x4CD135: loop_poll_events (loop.c:56)
==18532==    by 0x6A2451: tui_main (tui.c:239)
==18532==    by 0x6A857A: ui_thread_run (ui_bridge.c:112)
==18532==    by 0x4E539F6: ??? (in /usr/lib/x86_64-linux-gnu/libuv.so.1.0.0)
==18532==    by 0x4C32D06: mythread_wrapper (hg_intercepts.c:389)
==18532==    by 0x5271423: start_thread (pthread_create.c:333)
==18532==    by 0x68FA9BE: clone (clone.S:105)
==18532==
==18532== This conflicts with a previous read of size 4 by thread #1
==18532== Locks held: none
==18532==    at 0x6A542A: ui_refresh (ui.c:169)
==18532==    by 0x6A5870: ui_refresh_event (ui.c:181)
==18532==    by 0x4CDDB6: multiqueue_process_events (multiqueue.c:146)
==18532==    by 0x4CD135: loop_poll_events (loop.c:56)
==18532==    by 0x5DEDB4: os_breakcheck (input.c:150)
==18532==    by 0x59263D: line_breakcheck (misc1.c:2667)
==18532==    by 0x621AE5: nfa_regmatch (regexp_nfa.c:6171)
==18532==    by 0x61DCF7: nfa_regtry (regexp_nfa.c:6240)
==18532==  Address 0x770e7b4 is 4 bytes inside a block of size 352 alloc'd
==18532==    at 0x4C2EFE5: calloc (vg_replace_malloc.c:711)
==18532==    by 0x57C962: xcalloc (memory.c:119)
==18532==    by 0x6A6E29: ui_bridge_attach (ui_bridge.c:53)
==18532==    by 0x6A164C: tui_start (tui.c:116)
==18532==    by 0x6A4CFC: ui_builtin_start (ui.c:89)
==18532==    by 0x55A825: main (main.c:433)
==18532==  Block was alloc'd by thread #1
==18532==
==18532== ----------------------------------------------------------------
==18532==
==18532== Possible data race during write of size 4 at 0x770E7B8 by thread #2
==18532== Locks held: none
==18532==    at 0x6A3085: update_size (tui.c:760)
==18532==    by 0x6A30DB: sigwinch_cb (tui.c:269)
==18532==    by 0x4D0A54: signal_event (signal.c:44)
==18532==    by 0x4CDDB6: multiqueue_process_events (multiqueue.c:146)
==18532==    by 0x4CD135: loop_poll_events (loop.c:56)
==18532==    by 0x6A2451: tui_main (tui.c:239)
==18532==    by 0x6A857A: ui_thread_run (ui_bridge.c:112)
==18532==    by 0x4E539F6: ??? (in /usr/lib/x86_64-linux-gnu/libuv.so.1.0.0)
==18532==    by 0x4C32D06: mythread_wrapper (hg_intercepts.c:389)
==18532==    by 0x5271423: start_thread (pthread_create.c:333)
==18532==    by 0x68FA9BE: clone (clone.S:105)
==18532==
==18532== This conflicts with a previous read of size 4 by thread #1
==18532== Locks held: none
==18532==    at 0x6A5455: ui_refresh (ui.c:170)
==18532==    by 0x6A5870: ui_refresh_event (ui.c:181)
==18532==    by 0x4CDDB6: multiqueue_process_events (multiqueue.c:146)
==18532==    by 0x4CD135: loop_poll_events (loop.c:56)
==18532==    by 0x5DEDB4: os_breakcheck (input.c:150)
==18532==    by 0x59263D: line_breakcheck (misc1.c:2667)
==18532==    by 0x621AE5: nfa_regmatch (regexp_nfa.c:6171)
==18532==    by 0x61DCF7: nfa_regtry (regexp_nfa.c:6240)
==18532==  Address 0x770e7b8 is 8 bytes inside a block of size 352 alloc'd
==18532==    at 0x4C2EFE5: calloc (vg_replace_malloc.c:711)
==18532==    by 0x57C962: xcalloc (memory.c:119)
==18532==    by 0x6A6E29: ui_bridge_attach (ui_bridge.c:53)
==18532==    by 0x6A164C: tui_start (tui.c:116)
==18532==    by 0x6A4CFC: ui_builtin_start (ui.c:89)
==18532==    by 0x55A825: main (main.c:433)
==18532==  Block was alloc'd by thread #1
2017-02-03 14:04:38 +01:00
.ci win: ci/AppVeyor: Rearrange PATH setup for python. (#5946) 2017-01-15 14:45:14 +01:00
cmake build: prevent in-tree builds for the time being, as it's not supported 2016-10-11 20:20:31 -04:00
config build: Do not check for obsolete <termio.h> 2016-12-24 17:41:06 +01:00
contrib Revert "Makefile: add PREFIX variable" 2016-10-16 12:17:00 +02:00
man tui: Enable mode-sensitive cursor by default. 2017-01-20 15:48:41 +01:00
runtime vim-patch:063b9d1 2017-02-01 18:35:12 -05:00
scripts vim-patch:8.0.0121 2017-01-13 23:40:59 +01:00
src tui: update_size(): Fix race condition. 2017-02-03 14:04:38 +01:00
test Merge #5975 from jamessan/execute-with-attrs 2017-02-03 12:36:17 +01:00
third-party third-party: Add helper to fetch binary dependencies 2017-01-20 18:29:17 +00:00
unicode Update unicode files 2016-09-24 14:03:22 -04:00
.editorconfig add .editorconfig 2016-02-23 18:03:27 +09:00
.gitignore ignore .res files generated by the tests in testdir 2016-08-11 02:14:02 -04:00
.travis.yml Merge pull request #5827 from justinmk/travisupdate 2016-12-27 23:25:24 -05:00
appveyor.yml ci: Update appveyor artifact path. 2016-12-09 17:09:17 +01:00
BACKERS.md Update backer URL 2015-11-11 19:50:33 -08:00
CMakeLists.txt doc: QuickBuild 2017-01-31 08:58:43 +01:00
CONTRIBUTING.md doc: QuickBuild 2017-01-31 08:58:43 +01:00
ISSUE_TEMPLATE.md doc: precision, concision, elision 2016-06-03 17:11:10 -04:00
LICENSE LICENSE: add LuaJIT notice. #899 2014-06-30 13:59:56 -04:00
Makefile build: Make clean targets more thorough 2016-12-02 20:38:55 -05:00
README.md doc: README.md: update appveyor URL 2016-12-11 01:32:28 +01:00

Neovim

Wiki | Documentation | Twitter | Community | Gitter Chat

Travis Build Status AppVeyor Build status Pull requests waiting for review Coverage Status Coverity Scan Build Clang Scan Build Debian

Neovim is a project that seeks to aggressively refactor Vim in order to:

For more details, see the wiki!

Throughput Graph

What's been done so far

See :help nvim-features for a comprehensive list.

What's being worked on now

See the Roadmap.

How do I get it?

There is a Debian package, homebrew formula, PKGBUILD for Arch Linux, RPM, and more. See the wiki!

License

Neovim is licensed under the terms of the Apache 2.0 license, except for parts that were contributed under the Vim license.

  • Contributions committed before b17d96 remain under the Vim license.

  • Contributions committed after b17d96 are licensed under Apache 2.0 unless those contributions were copied from Vim (identified in the commit logs by the vim-patch token).

See LICENSE for details.

Vim is Charityware.  You can use and copy it as much as you like, but you are
encouraged to make a donation for needy children in Uganda.  Please see the
kcc section of the vim docs or visit the ICCF web site, available at these URLs:

        http://iccf-holland.org/
        http://www.vim.org/iccf/
        http://www.iccf.nl/

You can also sponsor the development of Vim.  Vim sponsors can vote for
features.  The money goes to Uganda anyway.