Problem:
Cross-compilation issues encountered when building Neovim for WASM.
When cross-compiling, three issues occur:
1. `-Wl,--no-undefined` — not supported by `wasm-ld`
2. `-lutil` — not available in the Emscripten sysroot
3. Doc generation fails because CMake tries to execute `$<TARGET_FILE:nvim_bin>` on the host machine, which fails because the binary is not native to the host. It fails with `/bin/sh: nvim.js: Permission denied`
Solution:
The fix includes skipping `-Wl,--no-undefined` and `-lutil` with `NOT CMAKE_CROSSCOMPILING` and adding `NVIM_HOST_PRG` variable to `runtime/CMakeLists.txt` so when cross-compiling, it uses a host native nvim binary for doc generation instead of using the cross-compiled target.
Problem:
In aec3d7915c Vim changed prompt-buffers
to respect 'modified' so the termdebug plugin can "control closing the
window". But for most use-cases (REPL, shell, AI "chat", …),
prompt-buffers are in practice always "modified", and no way to "save"
them, so *implicitly* setting 'modified' is noisy and annoying.
Solution:
Don't implicitly set 'modified' when a prompt-buffer is updated.
Plugins/users can still explicitly set 'modified', which will then
trigger the "E37: No write since last change" warning.
Problem: When reconfiguring a float reallocates the grid before the old
area is cleared, artifacts are left on the screen.
Solution: Use the last known compositor dimensions of a grid when
clearing the area covered by the old position.
Co-authored-by: glepnir <glephunter@gmail.com>
Problem: Since the "[Process exited]" msg is no longer part of buffer
contents, `jobstart`'s reuse of unmodified finished terminal buffers
does not clear the msg.
Solution: Delete the extmark if `term` is already closed.
Problem: update_window_hl dereferenced the pmap_get result without a
NULL check. Also `tmp` had the wrong type and attr_id==0 was rejected
twice.
Solution: check the pointer before dereferencing, change `tmp` to
bool, and update the attr_id check to `< 0`.
Problem: Exception error message only prints the first chunk of a
multi-chunk nvim_echo() message.
Solution: Concatenate consecutive message chunks in the exception
message list.
Problem:
When a terminal process exits, "[Process Exited]" text is added
to the buffer contents.
Solution:
- Return `exitcode` field from `nvim_get_chan_info`.
- Show it in the default 'statusline'.
- Show exitcode as virtual text in the terminal buffer.
Problem:
Cursor style is not set to default when guicursor is disabled. See
return early when guicursor is disabled (rather than resetting the
cursor style to terminal default).
Solution:
Send the appropriate DECSCUSR sequence to reset the cursor style when
guicursor is disabled.
Note that this does not fix#23122. The CSI sequence sent out depends on
the terminal in question. Ideally, it would would send `\x1b[0 q` to
reset the cursor style; the behavior of this sequence depends on the
terminal and configuration in question. See [Ghostty
docs](https://ghostty.org/docs/vt/csi/decscusr) for more details.
In practice, it sends out `\x1b[2 q` (steady block) for Ghostty, which
seems to be coming from the Unibilium database. I'm not sure what it may
send out for other terminals, but it doesn't exactly reset to the
default style.
Problem: Potential buffer overflow in regdump()
Solution: Add the size to the compiled regular expression and ensure we
don't read over the limit.
Note: this is not a security issue, because regdump() is typically not
compiled in any version of Vim, so should not affect anybody.
supported by AI claude.
9360647715
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem: :tab accepts a tab address range but doesn't specify this in
the command specification.
Solution: Add EX_RANGE and EX_ZEROR to the command specification and use
ADDR_TABS (Doug Kearns).
As command modifers are handled separately before these flags are tested
in the ex-command parser they have no effect. However, it's better to
use an accurate description and the command specification table has uses
in other areas like runtime file generation for the Vim filetype.
closes: vim/vim#1910049b8d9903b
Co-authored-by: Doug Kearns <dougkearns@gmail.com>
Problem: String handling can be improved
Solution: Pass string length where it is known to avoid strlen() calls,
do a few minor refactors (John Marriott).
This commit changes some calls to function `set_vim_var_string()` to pass
the string length where it is known or can be easily calculated.
In addition:
In `evalvars.c`:
* In function `set_reg_var()` turn variable `regname` into a C string
because that is how it used.
* Small cosmetics.
In `option.c`:
* Slightly refactor function `apply_optionset_autocmd()` to move some
variables closer to where they are used.
In `getchar.c`:
* Slightly refactor function `do_key_input_pre()`:
-> change call to `dict_add_string()` to `dict_add_string_len()` and
pass it the length of `buf`.
-> only call `get_vim_var_string()` once.
In `message.c`:
* Use a `string_T` to store local variable `p`.
In `normal.c`:
* Move some variables closer to where they are used.
closes: vim/vim#19618727f6e2686
Co-authored-by: John Marriott <basilisk@internode.on.net>
Problem: vim.ui_attach() is unable to display streamed shell output,
and will display it as individual messages.
Unwanted newlines in "shell_ret" message.
Solution: Treat the "shell_*" kinds as non-fast and set msg_show->append
for the streamed stdout/err messages.
Remove leading newline from (translated) message with
ext_messages, remove trailing newline altogether.
Problem:
There is an inconsistency between extmarks/highlights regarding the
`end_col` param.
Solution:
Allow end_col=-1 to mean "end of line" (if strict=false).
Co-authored-by: Justin M. Keyes <justinkz@gmail.com>
Problem:
After #37722 splice optimization for :packadd, nvim crashes with SIGSEGV
on startup while running `runtime! lua/xxx/*` and sourced file executes
multiple `:packadd`.
Solution:
While `do_in_cached_path` is executing, it doesn't expect reference to
runtime_search_path changes. But when callback is called, and
add_pack_dir_to_rtp does 'splice' it may trigger realloc, and change
address. Check runtime_search_path_ref to prevent ref held by
do_in_cached_path changes.
vim-patch:9.2.0062: Using the wrong field with DAP channel mode
vim-patch:9.2.0064: popup: opacity feature causes flickering
vim-patch:9.2.0071: Vim9: lambda function deleted on re-sourcing
vim-patch:9.2.0082: Patch v9.2.0052 was wrong
vim-patch:9.2.0084: Vim9: isn_get_calltype() can be improved
vim-patch:a6550d571 translation: regenerate po/vim.pot after v9.2.0083
vim-patch:9.2.0086: Coverity complains that ScreenLines can be NULL
vim-patch:9.2.0087: popup: redrawing can be improved when moving popups
vim-patch:5a4291d34 runtime(osc52): Omit paste from the osc52 provider when g:osc52_disable_paste is enabled
vim-patch:93cb5e544 translation: regenerate po/vim.pot after v9.2.0088
vim-patch:9166f1c09 translation: regenerate po/vim.pot after v9.2.0093
vim-patch:f7a568ebe runtime(syntax-tests): regenerate dump files (after v9.2.0093)
vim-patch:22c62e269 runtime(syntax-tests): regenerate dump files (after v9.2.0110)
vim-patch:bd77d327d runtime(doc): remove thoughts of generics for a function
vim-patch:9.2.0114: MS-Windows: terminal output may go to wrong terminal
vim-patch:49b81793f runtime(docs): Update CONTRIBUTING.md and mention #vim on libera
vim-patch:9.2.0123: GTK: using deprecated gdk_pixbuf_new_from_xpm_data()
Co-authored-by: marvim <marvim@users.noreply.github.com>
Problem: CompleteDone fires with reason="discard" even when the candidate
text was inserted and left in the buffer, because reason was determined
solely by the terminating keycode (Ctrl-Y).
Solution: Check compl_used_match to detect whether inserted
text remains in the buffer, and set reason="accept" accordingly.
Problem: nvim_win_set_config may raise unnecessary errors when not moving a
split.
Solution: skip checks related to moving when only maybe resizing a split.
Problem: No explicit redraw after setting style=minimal in nvim_open_win or
nvim_win_set_config, which may cause it to appear like it's not set.
Solution: call changed_window_setting after applying the style, which should be
enough for these options.
vim-patch:9.2.0124: auto-format may swallow white space
Problem: With auto paragraph formatting enabled, when a user makes an
attempt to add a new word before the end of a line and with
the following space bringing the line width over 'textwidth',
the space ends up just getting swallowed by the editor.
Solution: Detect such a constellation and do not auto-format in that
case (Daniel Müller).
closes: vim/vim#1959324fd698045
Problem: nvim_win_set_config may merge configs despite failing to configure a
split, and without applying necessary side-effects (like setting style=minimal
options). Plus, autocommands may apply a different config after the merge,
causing side-effects to apply for an outdated config.
Solution: merge configs last, only on success. Include fields only relevant to
splits. Properly set _cmdline_offset for splits.
Maybe better to disallow _cmdline_offset for splits instead, as the pum is
relative to cmdline_row anyway? (I didn't want to change behaviour too much)
Also use expect_unchanged in an unrelated test to quash a warning.
Problem: nvim_win_set_config with style="minimal" does not take immediate effect
when a split is not moved.
Solution: don't skip nvim_win_set_config's epilogue when only a resize may be
needed. De-duplicate resize logic and remove unnecessary redraw. (win_set*_win
already handles that)
Problem:
helptag.html is used to search for the online help documentation of a
specified tag. The previous URL was `/doc/user/helptag.html` but that
switched to `/doc/user/helptag/` in
https://github.com/neovim/neovim.github.io/pull/437. The alias of the
.html page was added to all other doc pages but forgotten for the
helptag.html page
Solution:
Add the alias to the helptag.html page too
Problem: Calling nvim_set_hl() with url= crashes because it tries to
free arena-owned string memory.
Solution: Remove the bad free and return a validation error instead.
This reverts commit ab8371a26c.
Need to think of a different solution, which may require adding new
flags to nvim_ui_attach() (e.g. passing stdout or stderr fd).
using the GNU compiler we just get a bunch of const warnings we can fix.
clang, however, gets really upset that the standard library suddenly
starts using a lot of c11 features, despite us being in -std=gnu99 mode.
Basically, _GNU_SOURCE which we set is taken as a _carte blanche_ by the
glibc headers to do whatever they please, and thus we must inform clang
that everything is still OK.
Problem:
On FreeBSD, output written to TTY may be lost on exit.
Example test failure:
FAILED
test/functional/terminal/tui_spec.lua @
2521:
TUI no assert failure on deadly signal #21896
test/functional/terminal/tui_spec.lua:2523: Row 1 did not match.
Expected:
|*Nvim: Caught deadly signal 'SIGTERM' |
|* |
|*[Process exited 1]^ |
|* |
|* |
| |
|{5:-- TERMINAL --} |
Actual:
|* |
|*[Process exited 1]{100:^ }|
|*{100:~ }|
|*{100:~ }|
|*{3:[No Name] }|
| |
|{5:-- TERMINAL --} |
To print the expect() call that would assert the current screen state, use
screen:snapshot_util(). In case of non-deterministic failures, use
screen:redraw_debug() to show all intermediate screen states.
Snapshot:
screen:expect([[
|
[Process exited 1]{100:^ }|
{100:~ }|*2
{3:[No Name] }|
|
{5:-- TERMINAL --} |
]])
stack traceback:
test/functional/ui/screen.lua:909: in function '_wait'
test/functional/ui/screen.lua:537: in function 'expect'
test/functional/terminal/tui_spec.lua:2523: in function <test/functional/terminal/tui_spec.lua:2521>
Solution:
Call tcdrain() on stdout and stderr on exit.
This problem is only observed on FreeBSD, but it probably doesn't hurt
to do this on all platforms with termios.h. In fact using tcdrain() on
PTY slave is no-op on Linux according to Linux kernel source code.
Problem: byteidx_common() and f_utf16idx() are calling ptr2len() twice
per iteration, instead of reusing the already computed clen.
Solution: Reuse clen for pointer advancement in both functions
(Yasuhiro Matsumoto).
closes: vim/vim#19573499e93d09a
N/A patches:
vim-patch:9.2.0109: VIM_BACKTICK is always defined except for tiny builds
Co-authored-by: Yasuhiro Matsumoto <mattn.jp@gmail.com>
Problem: memory leak in expand_findfunc() (after v9.1.0811)
Solution: Free list variable l on early return (Huihui Huang).
closes: vim/vim#19564648240fe9a
N/A patches:
vim-patch:9.2.0105: memory leak in heredoc_get() in src/evalvars.c
Co-authored-by: Huihui Huang <625173@qq.com>
Problem: Memory leak in qf_push_dir() (after v9.2.0091)
Problem: free dirname, if it is not a directory.
closes: vim/vim#19552e352bb632a
Co-authored-by: Christian Brabandt <cb@256bit.org>
Problem: 'listchars' "leadtab" not used in :list (after 9.2.0088).
Solution: Also check for "leadtab" when using :list. Fix memory leak on
E1572 if "multispace" or "leadmultispace" is set (zeertzjq).
closes: vim/vim#195575845741d69
Problem: Unable to tell whether a msg_show event is emitted as a result
a command typed on the cmdline (UI may want to represent these
differently from other messages).
Solution: Add trigger parameter that is set to "typed_cmd" for
a message emitted due to an interactively typed command.
Possible extensions are mapping/timer/event but it's hard to
imagine a UI distinguishing those so not added here.
