Problem:
Coverity reports use after free:
*** CID 352784: Memory - illegal accesses (USE_AFTER_FREE)
/src/nvim/buffer.c: 1508 in set_curbuf()
1502 if (old_tw != curbuf->b_p_tw) {
1503 check_colorcolumn(curwin);
1504 }
1505 }
1506
1507 if (bufref_valid(&prevbufref) && prevbuf->terminal != NULL) {
>>> CID 352784: Memory - illegal accesses (USE_AFTER_FREE)
>>> Calling "terminal_check_size" dereferences freed pointer "prevbuf->terminal".
1508 terminal_check_size(prevbuf->terminal);
1509 }
1510 }
1511
1512 /// Enter a new current buffer.
1513 /// Old curbuf must have been abandoned already! This also means "curbuf" may
Solution:
Change terminal_destroy and terminal_close to set caller storage to NULL,
similar to XFREE_CLEAR. This aligns with the pattern found already in:
terminal_destroy e897ccad3e
term_delayed_free 3e59c1e20d
The size of long varies depending on architecture, in contrast to the
MAXLNUM constant which sets the maximum allowable number of lines to
2^32-1. This discrepancy may lead to hard to detect bugs, for example
https://github.com/neovim/neovim/issues/18454. Setting linenr_T to a
fix maximum size of 2^32-1 will prevent this type of errors in the
future.
Also change the variables `amount` and `amount_after` to be linenr_T
since they're referring to "the line number difference" between two
texts.
* on_scrollback_option_changed renamed to adjust_scrollback. The
function name did not correspond to what it was doing. It is
called unconditionally in every refresh of the terminal
unrelated if the scrollback option was changed.
* new on_scrollback_option_changed function, which calls
refresh_terminal, which then calls adjust_scrollback
* terminal_check_size is not the appropriate function to call when the
option is changed since it only conditionally adjusts the scrollback.
Use the new on_scrollback_option_changed
fixes#15477fixes#11811
At the moment of comparison, the pointer save_curwin can be invalid (as
suggested by the comment) because it has been free'd. Worst, the new
curwin could have been re-allocated to that same pointer, altering the
execution flow unpredictably.
While there are many other potential similar cases to fix in the
codebase, the presented scenario is not hypothetical and does happen in
practice (while spawning new windows from fzf for instance).
There are numerous other instances of curwin comparisons in the
codebase, and they may need further investigation.
closes#16941
Most code in keymap.h is for keycode definitions, while most code in
keymap.c is for the parsing and conversion of keycodes.
The name "keymap" may also make people think these two files are for
mappings, while in fact keycodes are used even when no mappings are
involved, so "keycodes" should be a better file name than "keymap".
Problem: The mode #defines are not clearly named.
Solution: Prepend MODE_. Renumber them to put the mapped modes first.
249591057b
A hunk from the patch depends on patch 8.2.4861, which hasn't been
ported yet, but that should be easy to notice.
PROBLEM
------------------------------------------------------------------------
$NVIM_LISTEN_ADDRESS has conflicting purposes as both a parameter ("the
current process should listen on this address") and a descriptor ("the
current process is a child of this address").
This contradiction means the presence of NVIM_LISTEN_ADDRESS is
ambiguous, so child Nvim always tries to listen on its _parent's_
socket. This is the cause of lots of "Failed to start server" spam in
our test/CI logs:
WARN 2022-04-30… server_start:154: Failed to start server: address already in use: \\.\pipe\nvim-4480-0
WARN 2022-04-30… server_start:154: Failed to start server: address already in use: \\.\pipe\nvim-2168-0
SOLUTION
------------------------------------------------------------------------
1. Set $NVIM to the parent v:servername, *only* in child processes.
- Now the correct way to detect a "parent" Nvim is to check for $NVIM.
2. Do NOT set $NVIM_LISTEN_ADDRESS in child processes.
3. On startup if $NVIM_LISTEN_ADDRESS exists, unset it immediately after
server init.
4. Open a channel to parent automatically, expose it as v:parent.
Fixes#3118Fixes#6764Fixes#9336
Ref https://github.com/neovim/neovim/pull/8247#issuecomment-380275696
Ref #8696
Add space around arithmetic operators '+' and '-'.
Remove space between back-to-back parentheses, i.e. ')(' vs. ') ('.
Remove space between '((' or '))' of control statements.
Add space between ')' and '{' of control statements.
Remove space between function name and '(' on function declaration.
Collapse empty blocks between '{' and '}'.
Remove newline at the end of the file.
Remove newline between 'enum' and '{'.
Remove newline between '}' and ')' in a function invocation.
Remove newline between '}' and 'while' of 'do' statement.
Problem: Dropping modifier when putting a character back in typeahead.
Solution: Add modifier to ins_char_typebuf(). (closesvim/vim#6158)
b42c0d5427
Vim's test doesn't seem to work properly as the hit-enter prompt seems
to be delayed. Add a Lua screen test.
Problem: The ModeChanged autocmd event is inefficient.
Solution: Avoid allocating memory. (closesvim/vim#10134) Rename
trigger_modechanged() to may_trigger_modechanged().
2bf52dd065
Make v:event readonly for ModeChanged.
Problem: Syntax coloring and highlighting is in one big file.
Solution: Move the highlighting to a separate file. (Yegappan Lakshmanan,
closesvim/vim#4674)
f9cc9f209e
Name the new file highlight_group.c instead.
Co-authored-by: zeertzjq <zeertzjq@outlook.com>
- Fix the problem that chanclose() does not work for channel created by
nvim_open_term().
- Fix the problem that the loopback channel is not released.
- Fix the error message when sending raw data to the loopback channel.
Behavioral changes:
1. Added support for lua function in keymaps in
--------------------------------------------
- nvim_set_keymap
Can set lua function as keymap rhs like following:
```lua
vim.api.nvim_{buf_}set_keymap('n', '<leader>lr', '', {callback = vim.lsp.buf.references})
```
Note: lua function can only be set from lua . If api function being
called from viml or over rpc this option isn't available.
- nvim_{buf_}get_keymap
When called from lua, lua function is returned is `callback` key .
But in other cases callback contains number of the function ref.
- :umap, nvim_del_keymap & nvim_buf_del_keymap clears lua keymaps correctly.
- :map commands for displaing rhs .
For lua keymaps rhs is displayed as <Lua function ref_no>
Note: lua keymap cannot be set through viml command / functions.
- mapargs()
When dict is false it returns string in `<Lua function ref_no>`
format (same format as :map commands).
When dict is true it returns ref_no number in `callback` key.
- mapcheck()
returns string in `<Lua function ref_no>` format (same format as :map commands).
2. Added support for keymap description
---------------------------------------
- nvim_{buf_}set_keymap: added `desc` option in opts table .
```lua
vim.api.nvim_set_keymap('n', '<leader>w', '<cmd>w<cr>', {desc='Save current file'})
```
- nvim_{buf_}get_keymap: contains `desc` in returned list.
- commands like `:nmap <leader>w` will show description in a new line below rhs.
- `maparg()` return dict contains `desc`.
refresh_scrollback assumes pending scrollback rows exist only if the
terminal window height decreased (or the screen was full).
However, after accumulating scrollback, it's possible in some cases for
the terminal height to increase before refresh_scrollback is called via
invalidation (especially when the terminal buffer isn't initially
displayed in a window before nvim_open_term), which may crash.
As we'll have enough room for some scrollback rows, just append them to
the top of the buffer until it fills the window, then continue with the
previous logic for any remaining scrollback rows if necessary.
Problem: Internal error when ModeChanged is triggered when v:event is
already in use.
Solution: Save and restore v:event if needed.
3075a45592
In the vim codebase there is no occurrence of get_vim_var_dict(VV_EVENT)
after the above patch, so in order to hold the same invariant in the
neovim codebase we needed to replace more occurrences than the related
vim patch.
Use the (currently unused) 'destroy' field of the terminal struct as a
flag to indicate that the terminal's destruction is imminent (and
therefore it's close callback should not be called again).
Co-authored-by: Gregory Anders <greg@gpanders.com>
When entering terminal mode, cursorlineopt is no longer entirely
disabled. Instead, it's set to `number`. Doing so ensures that users
using `set cursorline` combined with `set cursorlineopt=number` have
consistent highlighting of the line numbers, instead of this being
disabled when entering terminal mode.
Co-authored-by: Gregory Anders <greg@gpanders.com>
Co-authored-by: Sean Dewar <seandewar@users.noreply.github.com>
Note: the reason for removing them is not that there after this refactor
is no use of them, but rather that having them available is an
anti-pattern: they manange an _extra_ heap allocation which has
nothing to do with the functionality of the map itself (khash
manages the real buffers internally). In case there happens to
be a reason to allocate the map structure itself later, this
should be made explicit using xcalloc/xfree calls.
Problem: Cannot use 'balloonexpr' in a terminal.
Solution: Add 'balloonevalterm' and add code to handle mouse movements in a
terminal. Initial implementation for Unix with GUI.
51b0f3701e
before, calling vim.schedule() from inside an event would execute
the scheduled callback immediately after this event without
checking for user input in between. Break event processing
whenever user input or an interrupt is available.
