There was never any investigation done to determine whether using
jemalloc was actually a net benefit for nvim. It has been a portability
limitation and adds another factor to consider when triaging issues.
Reverts previous experiment. PVS root is working correctly, one can
observe this in the PVS-studio.err file, for example:
/usr/local/clang-7.0.0/lib/clang/7.0.0/include/stddef.h:51:1: warning: V677 ...
/usr/local/clang-7.0.0/lib/clang/7.0.0/include/stddef.h:132:1: warning: V677 ...
./src/nvim/fileio.c:1382:1: warning: V1026 ...
./src/nvim/fileio.c:1388:1: warning: V1026 ...
The "./src/nvim/…" paths are correctly rooted, yet PVS somehow still
thinks it should analyze "/usr/local/clang-7.0.0/…".
See also: https://stackoverflow.com/q/44906903
Rework-of: ea7491586f
Helped-by: Björn Linse <bjorn.linse@gmail.com>
- The old (Vim) use of (char_u **)"" before ea7491586f is garbage,
which hints that this value was never used.
- The necessary condition is next to the NULL assigmnent, the pointer
would only be started to be accessed, if the length assignment next to
it is also changed.
The Vim version of Test_help_tagjump() tests for `:help sm?le` here. That
command got removed from Nvim, so the test was changed to check against `:help
sp?it` instead.
The new test already handled the case that on Win `:h split` would jump to the
entry for split() and on all other systems to the entry for :split.
Then this commit happened:
bb3aa824b lua/stdlib: vim.inspect, string functions
Since then `:h split` would jump to split() for macOS as well! I'm not sure why.
Anyway, instead of adding another check for has('mac'), we change the test once
more to be more akin to the original test. Instead of testing for :smile, which
is exclusive to Vim, we check against :checkhealth, which is exclusive to Nvim.
In vim, scrolling a window might mess up the cmdline. To keep it simple,
cmdline was always cleared for any window scroll. In nvim, where safe scrolling
is implemented in the TUI layer, this problem doesn't exist.
Clearing the message on scrolling, when we not do it e.g when switching tabs
is a bit weird, as the former is a much smaller context change.
A vim patch introduced the possibility to avoid the cmdlline clear for
redraws caused by async events. This case will now trivially be covered,
as the redraw is always avoided.
vim-patch:8.0.0592: if a job writes to a buffer screen is not updated
Implement lazy loading for vim.submodule, this would be over-engineering
for inspect only, but we expect to use this solution also for more and
larger modules.
Instead of eager-loading during plugin/* sourcing, define runtime
modules such as `vim.inspect` as lazy builtins. Otherwise non-builtin
Lua modules such as `vim.inspect` would not be available during startup
(init.vim, `-c`, `--cmd`, …).
ref #6580
ref #8677
- When setting a fixed/locked/readonly var, it is more relevant to
report on the key, not its container dict. If its container dict (v:)
is readonly, that does not mean the key itself is readonly.
- Allow modifying a "fixed" var. "fixed" only prevents deletion.
This maintenance release fixes some issues found in v0.3.3.
FIXES:
8a7b6200fb#9487 provider: improve error message if provider is missing
44ea903ca5#9468 checkhealth: detect broken pip
b4028056fa Windows: nvim-qt v0.2.12 (fix potential "blank screen" at startup)
There is various places where 'conceallevel' and 'concealcursor'
necessitates additional redraws. This tries to separate the different
cases and handle each accordingly:
- Share code with 'cursorline' for the common case: vertical move of
cursor within the same window (concealcursor not active)
- Improve the logic for managing 'concealcursor' and switching modes:
test for the case where the new mode behaves differently from the
last one.
- Clarify the special case for horizontal movement within a line when
'concealcursor' is active, now there is an if-statement only for this
and not hidden in larger check mostly for the first point.
- Keep the special case for moving between windows as is.
Move `has_eval_provider()` check to `eval_call_provider()` to make sure that
every code path calls it first.
Previously we would, when pynvim was missing, get a nice error message for
`:python3 1`, but not for `:py3file blah`.
Fixes https://github.com/neovim/neovim/issues/9485
