bfredl

kbtree: add kb_itr_prev and more precise kb_itr_getp

6

It doesn't look to me like `kb_itr_get` is working: for instance the iterator will always store index 0 for the root node no matter what is the correct index in...

current device does not stay active between cells / command lines

6

I'm using the kernel with a frontend that does not support graphics. Therefore I want to use the X11 display. However `options(device="X11")` does not seem to work: when I run...

sema: allow compare .lazy_size with a signed rhs value. fixes #12498

7

not sure if this is the right fix. but resolving the layout unconditionally causes another test to fail (`@sizeOf(T) == 0 doesn't force resolving struct size`) and based on @andrewrk...

stdlib: make linux.PERF.TYPE non-exhaustive

1

perf_event_attr.type needs to take a runtime defined value to enable dynamic PMU:s, such as kprobe and uprobe. This value can exceed predefined values defined in the linux headers. reference: perf_event_open(2)...

[WIP] anticonceal (virtual text occupying horizontal space)

19

More of TODO than WIP to be honest, but at least "conceptual" preview of something that can become a generalization of virttext, where virtual text also can be placed in...

feat(ui): inline virtual text

1

Note: This is just the first patch. there are many vim patches to go through which fixes various edge cases and bugs.

test(ui): add proper support for measuring redraw performance

2

currently our screen test suite is very good at catching final redraw state errors. It cares very little about redraw performance tho (i e don't invalidate a larger region of...

perf(redraw): remove redraw panic in insert-mode completion

1

- order of showing the pum and updating curwin is irrelevant in nvim - no need to ins_redraw() in the middle of calculating longest match