Commit Graph
378 Commits
Author SHA1 Message Date
Marshall Bowers 9735912965 Enable clippy::clone_on_copy (#8728)
This PR enables the
[`clippy::clone_on_copy`](https://rust-lang.github.io/rust-clippy/master/index.html#/clone_on_copy)
rule and fixes the outstanding violations.

Release Notes:

- N/A
2024-03-02 17:37:48 -05:00
Liam Murphy 0903062933 Fix unsafe precondition violation when building with nightly rustc (#8691)
Fixes #8658

Release Notes:

- N/A
2024-03-02 08:01:44 +01:00
Marshall Bowers ca2cda8d2a Remove unneeded 'static lifetimes on &strs in constants (#8698)
This PR removes unneeded `'static` lifetimes on `&str`s stored in
`const` declarations.

This addresses some Clippy lints about
[`redundant_static_lifetimes`](https://rust-lang.github.io/rust-clippy/master/index.html#/redundant_static_lifetimes).

In item-level `const` declarations we can rely on lifetime elision and
use the default `'static` lifetime.

Note that associated constants still require an explicit `'static`
lifetime, as explained in
https://github.com/rust-lang/rust/issues/115010.

Release Notes:

- N/A
2024-03-02 00:40:49 -05:00
Conrad Irwin 9bd5ebb74b Revert "Introduce a new ToggleGraphicsProfiler command (#7607)" (#8567)
This reverts commit 0cebf68306.

Although this thing is very cool, it is a top source of crashes.

Example crash:
```
Segmentation fault: 11 on thread 26
  objc_retain +16
  invocation function for block in Overlay::onCommandBufferCommit(id<MTLCommandBuffer>) +60
  MTLDispatchListApply +52
```

Release Notes:

- Removed "Toggle Graphics Profiler" as it crashes too much.
2024-02-28 16:39:51 -07:00
Antonio ScandurraandThorsten ca251babcd Drop Box<dyn PlatformWindow> when the OS closes the native window (#8016)
Closes #7973 

This fixes a leak in GPUI when the user didn't override
`on_should_close_window`.

Release Notes:

- N/A

---------

Co-authored-by: Thorsten <thorsten@zed.dev>
2024-02-19 15:38:50 +01:00
Dzmitry MalyshauandMikayla 9ad1862f2f Enable Blade on MacOS via "macos-blade" feature (#7669)
Depends on https://github.com/zed-industries/font-kit/pull/2 and
https://github.com/kvark/blade/pull/77

This change enables Blade to be also used on MacOS. It will also make it
easier to use it on Windows.

What works: most of the things. Zed loads as fast and appears equally
responsive to the current renderer.
<img width="306" alt="Screenshot 2024-02-11 at 12 09 15 AM"
src="https://github.com/zed-industries/zed/assets/107301/66d82f45-5ea2-4e2b-86c6-5b3ed333c827">

Things missing:
- [x] video streaming. ~~Requires a bit of plumbing on both Blade and
Zed sides, but all fairly straightforward.~~
  -  verified with a local setup
- [x] resize. ~~Not sure where exactly to hook up the reaction on the
window size change. Once we know where, the fix is one line.~~
- [ ] fine-tune CA Layer
- this isn't a blocker for merging the PR, but it would be a blocker if
we wanted to switch to the new path by default
- [ ] rebase on latest, get the dependency merged (need review/merge of
https://github.com/zed-industries/font-kit/pull/2!)

Update: I implemented resize support as well as "surface" rendering on
the Blade path (which will be useful on Linux/Windows later on). I
haven't tested the latter though - not sure how to get something
streaming. Would appreciate some help! I don't think this should be a
blocker to this PR, anyway.

The only little piece that's missing for the Blade on MacOS path to be
full-featured is fine-tuning the CALayer configuration. Zed does a lot
of careful logic in configuring the layer, such as switching the
"present with transaction" on/off intermittently, which Blade path
doesn't have yet.

Release Notes:
- N/A

---------

Co-authored-by: Mikayla <mikayla@zed.dev>
2024-02-16 13:39:40 -08:00
Conrad Irwin b800fe96d2 Fix dealloc of MacWindow (#7708)
We see some panics in the Drop handler for MacWindow

Looking into this, I noticed that our drop implementation was not
correctly
cleaning up the window state.

Release Notes:

- N/A
2024-02-12 20:01:09 -07:00
Greg Morenz 48a4cbad37 Don't move traffic lights while fullscreen
Fixes #4712, or at least works around it. As discussed in the issue,
setting the buttons frames while fullscreen can cause them to render
in the wrong location. This fixes that by simply not moving them when
fullscreen.
2024-02-12 14:17:10 -05:00
RomanandMikayla 266988adea gpui: Decouple X11 logic from LinuxPlatform (#7598)
Release Notes:

- Separated Linux platform and X11-specific code, so that we can add
Wayland support now.

---------

Co-authored-by: Mikayla <mikayla@zed.dev>
2024-02-12 10:44:34 -08:00
Antonio Scandurra 0cebf68306 Introduce a new ToggleGraphicsProfiler command (#7607)
On macOS, this will enable or disable the Metal HUD at runtime. Note
that this only works when Zed is bundled because it requires to set the
`MetalHudEnabled` key in the Info.plist.

Release Notes:

- Added a new `ToggleGraphicsProfiler` command that can be used as an
action (or via the `Help -> Toggle Graphics Profiler` menu) to
investigate graphics performance.
2024-02-09 13:29:40 +01:00
04e1641a29 Properly handle backspace when in dead key state (#7494)
Previously we wouldn't handle Backspace in dead key state correctly:
instead of removing what was typed, we'd insert the text that was in the
dead key state.

Example: on a US English layout, press `opt-u` to end up in a dead key
state with `¨` waiting for the next character to be typed. Type
`backspace`. The `¨` should be removed, but it's not.

With this change, the `backspace` is interpreted instead of being
ignored.

Release Notes:

- Fixed backspace not working for dead keys (i.e. when typing accents or
umlauts)

---------

Co-authored-by: Antonio <antonio@zed.dev>
Co-authored-by: Conrad <conrad@zed.dev>
2024-02-09 12:22:12 +01:00
67b96b2b40 Replace CADisplayLink with CVDisplayLink (#7583)
Release Notes:

- Fixed a bug that caused Zed to render at 60fps even on ProMotion
displays.
- Fixed a bug that could saturate the main thread event loop in certain
circumstances.

---------

Co-authored-by: Thorsten <thorsten@zed.dev>
Co-authored-by: Nathan <nathan@zed.dev>
Co-authored-by: Max <max@zed.dev>
2024-02-08 13:47:12 -05:00
Mikayla Maki d4be15b2b2 Suppress related warnings, fix nanoid, and get the build green (#7579)
This is in preparation for adding a Linux build step to our CI.

Release Notes:

- N/A
2024-02-08 09:32:53 -08:00
Mikayla 67555ee5b4 Merge branch 'main' into kvark-linux 2024-02-07 11:52:44 -08:00
Antonio ScandurraandConrad 07fce7aae1 Stop display link when window is occluded (#7511)
Release Notes:

- Fixed a bug that caused the window to become unresponsive after
foregrounding.

---------

Co-authored-by: Conrad <conrad@zed.dev>
2024-02-07 11:45:30 -07:00
Marshall Bowers c322179bb9 Initialize the SystemAppearance using the app's global window appearance (#7508)
This PR changes our approach to initializing the `SystemAppearance` so
that we can do it earlier in the startup process.

Previously we were using the appearance from the window, meaning that we
couldn't initialize the value until we first opened the window.

Now we read the `window_appearance` from the `AppContext`. On macOS this
is backed by the
[`effectiveAppearance`](https://developer.apple.com/documentation/appkit/nsapplication/2967171-effectiveappearance)
on the `NSApplication`.

We currently still watch for changes to the appearance at the window
level, as the only hook I could find in the documentation is
[`viewDidChangeEffectiveAppearance`](https://developer.apple.com/documentation/appkit/nsview/2977088-viewdidchangeeffectiveappearance),
which is at the `NSView` level.

In my testing this makes it so Zed appropriately chooses the correct
light/dark theme on startup.

Release Notes:

- N/A
2024-02-07 11:51:18 -05:00
b59e110c59 Attempt to fix random lag (#7506)
Co-Authored-By: Antonio <antonio@zed.dev>
Co-Authored-By: Thorsten <thorsten@zed.dev>
Co-Authored-By: Mikayla <mikayla@zed.dev>

Release Notes:

- N/A

**or**

- N/A

Co-authored-by: Antonio <antonio@zed.dev>
Co-authored-by: Thorsten <thorsten@zed.dev>
Co-authored-by: Mikayla <mikayla@zed.dev>
2024-02-07 09:45:49 -07:00
Antonio Scandurra 55129d4d6c Revert "Use Fx* variants of HashMap and HashSet everywhere in Zed" (#7492)
Reverts zed-industries/zed#7481

This would regress performance because we'd be using the standard
library's hash maps everywhere, so reverting for now.
2024-02-07 13:16:22 +01:00
Kirill Bulatov eb236302c2 Use Fx* variants of HashMap and HashSet everywhere in Zed (#7481)
Release Notes:

- N/A
2024-02-07 09:45:37 +02:00
Piotr Osiewicz 70e7ea365c chore: Fix up warnings from cargo +beta check. (#7453)
With upcoming release of 1.76 I did a check of current +beta (which
seems to already be at 1.77). These would cause CI pipeline failures
once 1.77 is out.

Release Notes:

- N/A
2024-02-06 20:36:02 +01:00
c7c4166724 Disable extra frames for ProMotion when screen is not active. (#7410)
This was causing an issue where windows about 1/10 of the way across the
display would hang for a fully second after being deactivated.

Release Notes:

- N/A

Co-authored-by: max <max@zed.dev>
Co-authored-by: nathan <nathan@zed.dev>
Co-authored-by: antonio <antonio@zed.dev>
2024-02-05 14:48:54 -08:00
Antonio ScandurraandNathan d742b3bfac Mark the window as dirty when first opening it (#7384)
Otherwise we won't display anything if the window never notifies.

Release Notes:

- N/A

Co-authored-by: Nathan <nathan@zed.dev>
2024-02-05 09:09:49 -07:00
3bf412feff Use window's screen rather than window itself to start display link
Co-Authored-By: Antonio Scandurra <antonio@zed.dev>
Co-Authored-By: Nathan Sobo <nathan@zed.dev>
2024-02-05 11:03:23 -05:00
Dzmitry Malyshau cf71fe8bf1 fix MacOS build, switch external RWH to 0.6
leaving blade-internal RWH as 0.5 until this is fixed:
https://github.com/ash-rs/ash/issues/864
2024-02-03 23:50:44 -08:00
Antonio ScandurraandNathan ae2c23bd8e Reintroduce ProMotion support (#7347)
This re-introduces the changes of #7305 but this time we create a
display link using the `NSScreen` associated with the window. We're
hoping we'll get these frame requests more reliably, and this seems
supported by the fact that awakening my laptop restores the frame
requests.

Release Notes:

- See #7305.

Co-authored-by: Nathan <nathan@zed.dev>
2024-02-03 16:33:08 -07:00
Kirill Bulatov 1ab0af2fa3 Revert the commit that broke Zed display capabilities (#7326) 2024-02-03 18:32:56 +02:00
Antonio ScandurraandNathan Sobo c906fd232d Reduce GPU memory usage (#7319)
This pull request decreases the size of each instance buffer and shares
instance buffers across windows.

Release Notes:

- Improved GPU memory usage.

---------

Co-authored-by: Nathan Sobo <nathan@zed.dev>
2024-02-03 05:27:08 -07:00
Antonio ScandurraandNathan Sobo 15edc46827 Maintain smooth frame rates when ProMotion and direct mode are enabled (#7305)
This is achieved by starting a `CADisplayLink` that will invoke the
`on_request_frame` callback at the refresh interval of the display.

We only actually draw frames when the window was dirty, or for 2 extra
seconds after the last input event to ensure ProMotion doesn't downclock
the refresh rate when the user is actively interacting with the window.

Release Notes:

- Improved performance when using a ProMotion display with fast key
repeat rates.

---------

Co-authored-by: Nathan Sobo <nathan@zed.dev>
2024-02-02 16:42:46 -07:00
Antonio Scandurra 659423a4a1 Use Mutex instead of a RefCell to acquire/release instance buffers (#7291)
This fixes a panic happening when releasing an instance buffer.
Releasing the buffer happens on a different thread but the borrow
checker was not catching it because the metal buffer completion handler
API doesn't have a `Send` marker on it.

Release Notes:

- N/A
2024-02-02 09:53:07 -07:00
3d76ed96f5 Use command_buffer.wait_until_scheduled in metal renderer (#7283)
This commit goes back to using `wait_until_scheduled` as opposed to
`wait_until_completed`. What this means, however, is that another draw
could take place before the previous one finished. When that happens we
don't want to reuse the same instance buffer because the GPU is actively
reading from it, so we use a pool instead.

Release Notes:

- Fixed a bug that caused inconsistent frame rate when scrolling on
certain hardware.

---------

Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Antonio <antonio@zed.dev>
2024-02-02 09:09:58 -07:00
Thorsten Ball 2940a0ebd8 Revert "Avoid excessive blocking of main thread when rendering in direct mode (#7253)" (#7272)
This reverts commit 020c38a891 because it
leads to glitches when selecting text.



https://github.com/zed-industries/zed/assets/1185253/78c2c184-bc15-4b04-8c80-a23ca5c96afa



Release Notes:

- N/A
2024-02-02 10:33:08 +01:00
020c38a891 Avoid excessive blocking of main thread when rendering in direct mode (#7253)
Release Notes:

- Fixed a bug that caused inconsistent frame rate when scrolling on
certain hardware.

Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Nathan Sobo <nathan@zed.dev>
2024-02-01 15:39:41 -08:00
Antonio Scandurra 5424c8bfd5 Introduce a fast path for drawing quads with no borders / corner radii (#7231)
This will introduce an extra conditional but saves us from doing a bunch
of math in the simple case of drawing simple rectangles that aren't
rounded or don't have borders.


![Figure_1](https://github.com/zed-industries/zed/assets/482957/cba95ce2-2d9a-46ab-a142-35368334eb75)

Release Notes:

- Improved rendering performance.
2024-02-01 09:49:27 -08:00
Piotr OsiewiczandNiklas 5941102aac gpui: Add runtime-shaders feature so that Xcode.app is no longer necessary for Nix-based workflows (#7148)
Release Notes:

- N/A

Co-authored-by: Niklas <niklas@niklaskorz.de>
2024-01-31 17:37:16 +01:00
Mikayla Maki a54eaaecee Add raw window handle implementations to GPUI (#7101)
This is in preparation for experiments with wgpu. This should have no
external effect.

Release Notes:

- N/A
2024-01-30 11:42:28 -08:00
Conrad Irwin c07355265f Add logging for the font_descriptor panic (#7097)
Release Notes:

- Fixed a panic caused by an inconsistency in font metrics.
2024-01-30 11:16:43 -07:00
Hans 843916d585 Fix two typos (#7056) 2024-01-30 10:17:06 +02:00
Piotr Osiewicz 8fbc88b708 Load embedded fonts directly from .rdata instead of cloning (#6932)
The fonts we embed in Zed binary (Zed Sans & Zed Mono) weigh about 30Mb in total and we are cloning them several times during startup and loading of embedded assets (once explicitly in Zed and then under the hood in font-kit). Moreover, after loading we have at least 2 copies of each font in our program; one in .rdata and the other on the heap for use by font-kit.
This commit does away with that distinction (we're no longer allocating the font data) and slightly relaxes the interface of `TextSystem::add_fonts` by expecting one to pass `Cow<[u8]>` instead of `Arc<Vec<u8>>`. Additionally, `AssetSource::get` now returns `Cow<'static, [u8]>` instead of `Cow<'self, [u8]>`; all existing implementations conform with that change.

Note that this optimization takes effect only in Release builds, as the library we use for asset embedding - rust-embed - embeds the assets only in Release mode and in Dev builds it simply loads data from disk. Thus it returns `Cow<[u8]>` in it's interface. Therefore, we still copy that memory around in Dev builds, but that's not really an issue. 
This patch makes no assumptions about the build profile we're running under, that's just an intrinsic property of rust-embed.

Tl;dr: this should shave off about 30Mb of memory usage and a fair chunk (~30ms) of startup time.

Release Notes:
- Improved startup time and memory usage.
2024-01-29 10:06:57 +01:00
Conrad Irwin 01424a62ea Allow prompts to have detail, and use for good
Make channel panel errors louder
2024-01-24 23:15:37 -07:00
Conrad Irwin 249a6da54a Fix circular locking in prompts
Sometimes Cocoa calls app delegate methods (notably the display link)
while we're calling Cocoa methods. This causes a deadlock unless we
are careful to run cocao methods while we're not holding our internal
locks
2024-01-24 19:40:01 -07:00
Joseph T. Lyons 40dbe15b2a Update all links point to community 2024-01-24 15:11:17 -05:00
Antonio Scandurra 25a7eb27d2 Move interaction with keychain off the main thread 2024-01-24 14:53:05 +01:00
Mikayla a241a13d60 fix test 2024-01-22 19:42:25 -08:00
Mikayla 938b84c045 Finish documenting GPUI 2024-01-22 19:33:45 -08:00
Mikayla a99d5b87e8 WIP: text_system 2024-01-22 19:30:40 -08:00
Kirill BulatovandPiotr a75fa35a08 Fix the fonts panic
co-authored-by: Piotr <piotr@zed.dev>
2024-01-23 00:23:48 +02:00
Conrad Irwin a1c4716eeb Potential fix for #2422 (#4197)
- it seems like layer.drawable_size() is 0,0 in metal_renderer.rs:189 
- we set this in two places:
    - in response to a CALayerDelegate setFrameSize: event
- in response to a CALayerDelegate viewDidChangeBackingProperties:
event.
- it looks like if we don't set it in either of these cases we get a
different failure mode: the view is zoomed just wrong.
- That said, I can reproduce the screenshot if .scale_factor() returns
0.
- This might happen if [nativeWindow screen] is nil, which happens when
the window is off screen.
    - possible that zed started with offscreen window?
- I would expect that viewDidChangeBackingProperties would fire when
that changed.
    - potential fix: default to 2.0

Release Notes:

- Fixed Zed occasionally rendering blank on laungh
([#2422](https://github.com/zed-industries/community/issues/2422)).
2024-01-22 11:25:02 -07:00
Conrad Irwin 41a99c686b Potential fix for #2422 2024-01-22 11:10:15 -07:00
Conrad Irwin b06e2eb6af Update handling of 'pending' keys
Before this change if you had a matching binding and a pending key,
the matching binding happened unconditionally.

Now we will wait a second before triggering that binding to give you
time to complete the action.
2024-01-21 20:12:01 -07:00
Mikayla 9da6b8c7f6 Lock down mac os platform type visibility in the rest of GPUI
Add documentation to all platform types
2024-01-20 14:38:03 -08:00