Fixes an autoscroll bug that sometimes obscures the current row behind a
sticky header.
The bug was caused by `Editor::autoscroll_vertically` using the
`sticky_header_line_count` value that was cached during the previous
render, which isn't necessarily the number of sticky headers that the
scroll target needs, e.g.
- when jumping to a definition
- when pressing the up arrow key when the selection is in the topmost
visible row with `"vertical_scroll_margin": 0`
This fixes that by not caching `sticky_header_line_count` and instead
querying Tree-sitter on the fly to get the number of sticky headers that
the scroll target needs. Performance-wise this seem okay, I measured it
to take less than 200µs on my machine using a very large Rust file (and
there are still some possible ways to optimize this if necessary). In
particular, the pathological huge single-line file case as discussed in
#48450 shouldn't be an issue here because unlike the main sticky header
Tree-sitter query, here we query the outline items that contain a
particular point rather than those that intersect an entire row.
The `ScrollCursorTop` handler is also simplified to use the same shared
function for counting the number of sticky headers, since that action
doesn't rely on autoscroll.
Before:
https://github.com/user-attachments/assets/efb12776-82d9-4b94-baa5-347ec769fb98
After:
https://github.com/user-attachments/assets/236deb9f-fe06-43bd-b167-7bd3ab719e4c
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#50803
Release Notes:
- Fixed sticky headers sometimes obscuring the current row.
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the UI/UX checklist
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#52352
Release Notes:
- Fixed system clipboard being overwritten when pasting over a visual
selection with Cmd-V/Ctrl-V in vim mode
Special thanks to @lingyaochu for finding out the root cause of the
issue.
Took the test from: https://github.com/zed-industries/zed/pull/54233
The git graph was using `buffer_ui_size` when calculating the canvas row
height and the table row height. This is wrong because elements such as
`Labels` in the graph table were rendered using `ui_font_size`. This PR
normalizes the row height calculation by always using `ui_font_size` for
the canvas and table row height calculation, and taking into account the
scaling factor.
This PR also fixes a bug where the bottom of the canvas could flicker on
its first redraw because the uniform list hadn't cached the viewport
size yet, and we underestimated the visible size of the canvas and
underdrew it. We now fallback to the window height as the viewport size.
This means we'll overdraw for a single frame whenever the uniform list
hasn't cached the last item size, but it avoids the flicker!
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#53492Closes#53469
Release Notes:
- git_graph: Fix misalignment issues between the graph canvas and the
graph table
---------
Co-authored-by: lingyaochu <zx0@mail.ustc.edu.cn>
## Summary
When `debugger::Start` runs a build task (via the `"build"` field in
`debug.json`), it was bypassing the task's `"save"` property and calling
`project.create_terminal_task()` directly. This meant unsaved files were
never saved before the build ran, even when `"save": "all"` was set
unlike `task: spawn` which correctly saved first.
**Changes:**
- Extracted `Workspace::save_for_task(workspace, strategy, cx)` as a
shared async helper in `workspace/src/tasks.rs`
- Refactored `schedule_resolved_task` to call this helper (no behavior
change, removes inline duplication)
- Called the same helper in `RunningState::resolve_scenario` before
`create_terminal_task`, so the debugger build path now respects the
task's `save` strategy
- Added `test_save_for_task_all`, `test_save_for_task_current`, and
`test_save_for_task_none` tests covering the extracted helper directly
## Self-Review Checklist
- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#53284
## Video after fix
[Screencast from 2026-04-08
08-04-18.webm](https://github.com/user-attachments/assets/b0c20164-3670-440a-b43a-8457fb57bfbd)
## Release Notes:
- Fixed debugger not saving files before running a build task when
`"save": "all"` is set in the task definition
Fixes#51456
## Problem
When opening a recent project with an unsaved scratch buffer (untitled
file), the project switch silently fails. The log shows:
```
failed to save contents of buffer
Caused by: FOREIGN KEY constraint failed
```
## Root Cause
When a user creates a new file from the welcome screen and edits it
without saving, the buffer is dirty but has no file path. When opening a
recent project, `prepare_to_close` calls
`save_all_internal(SaveIntent::Close)`, which tries to serialize dirty
items to the database for hot-exit functionality.
The serialization fails with a FOREIGN KEY constraint error because
workspace serialization is throttled - the workspace row may not exist
in the database yet when the editor tries to INSERT into the editors
table referencing that workspace_id.
The previous code used `try_join_all` on all serialize tasks, so a
single serialization failure would abort the entire close flow,
preventing the project switch.
## Fix
Replace `try_join_all` with individual task awaiting. If a serialize
task fails, the item is moved back to the `remaining_dirty_items` list
so the user gets a proper save/discard prompt instead of the action
silently failing.
This is a minimal change (7 insertions, 2 deletions) that:
- Handles the FOREIGN KEY error gracefully
- Preserves the save/discard prompt UX for items that fail serialization
- Logs the serialization error for debugging via `.log_err()`
- Does not change behavior for items that serialize successfully
## Test Plan
- [ ] Open Zed with no active workspace
- [ ] Create a new file from the welcome screen
- [ ] Edit the buffer to make it dirty (do not save)
- [ ] Open a recent project via `projects: open recent`
- [ ] Verify the project opens (previously it silently failed)
- [ ] Verify a save/discard prompt appears for the dirty scratch buffer
Release Notes:
- Fixed opening a recent project silently failing when an unsaved
scratch buffer is present (#51456).
---
This PR was written with the assistance of AI (Claude).
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
This reverts #54329 and the part of #52499 that was an earlier attempt
at the same thing, which caused us to incorrectly miss git state updates
on Windows. cc @Veykril it seems like we need to find a different way to
fix the problem of `.git` scanning cycles.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Fixed a bug causing stale git state on Windows.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
Fixes a user-reported bug where a thread could be missing from the
sidebar even though it was still present in the metadata store and still
visible in Thread History. The thread reappears in the sidebar only
after the user sends a message.
### Scenario
A single multi-root workspace whose roots are `[/cloud,
/worktrees/zed/wt_a/zed]`, where:
- `/cloud` is a standalone git repo (main == folder).
- `/worktrees/zed/wt_a/zed` is a linked worktree of a separate `/zed`
repo.
The project group normalizes to `main_worktree_paths = [/cloud, /zed]`.
A thread created in this workspace is written with `main=[/cloud, /zed],
folder=[/cloud, /worktrees/zed/wt_a/zed]` and the sidebar finds it via
`entries_for_main_worktree_path`.
If the thread's stored `main_worktree_paths` ever drifts from the group
key — e.g. a stale row loaded from the store on startup, a legacy write,
or a row persisted with `main == folder` — all three existing lookups in
`Sidebar::rebuild_contents` miss:
1. `entries_for_main_worktree_path([/cloud, /zed])` — the thread's
stored main doesn't equal the group key.
2. `entries_for_path([/cloud, /zed])` — the thread's folder paths don't
equal the group key either.
3. The linked-worktree fallback iterates the group workspaces'
`linked_worktrees()` snapshots. Those yield *sibling* linked worktrees
of the repo, not the workspace's own roots, so `/worktrees/zed/wt_a/zed`
doesn't match.
The row falls out of the sidebar entirely even though the metadata is
intact and the thread's folder paths exactly equal the open workspace's
roots. The store heals the stored row on the next `RootThreadUpdated`
event, which is why sending a message makes the row reappear — but until
then the sidebar misrepresents the state.
### Fix
Add a fourth lookup to `Sidebar::rebuild_contents`: for each open
workspace in the group, query the store by the workspace's own root
paths. Any thread whose `folder_paths` matches an open workspace's roots
belongs under that group, regardless of what its `main_worktree_paths`
say.
This covers the gap between stale-row load and store self-heal, matches
the principle that the sidebar should reflect state that exists in a
reasonable way, and is symmetric with the existing lookups (same store
API, one more iterator).
### Commits
1. `sidebar: Add failing repro for thread disappearing from sidebar` —
adds `test_sidebar_keeps_multi_root_thread_with_stale_main_paths` which
reproduces the bug. Sets up the multi-root + linked-worktree layout,
persists a thread in the stale shape, and asserts the row is still
visible in the sidebar.
2. `sidebar: Show threads whose folder paths match an open workspace` —
the fourth-lookup fix. 31 lines in `crates/sidebar/src/sidebar.rs`, no
deletions.
### Verification
- `cargo test -p sidebar`: 103 passed, 0 failed (the new test was
failing before commit 2 and is passing after).
- `./script/clippy -p sidebar`: clean.
### Follow-ups
The three existing lookups in `rebuild_contents` are each covering a
different historical shape the store can be in. A real cleanup — do a
one-shot migration on reload that fills in `main_worktree_paths` for any
row missing it, then retire the two legacy-shape lookups — is worth
doing as a separate PR, but out of scope here.
Release Notes:
- Fixed an issue where agent threads could go missing from the sidebar.
The recent-projects picker, sidebar, welcome screen, and agent thread
store all called `recent_workspaces_on_disk`, which combined listing
with deleting stale rows. Its retention predicate rejected workspaces
with no on-disk directory, including empty workspaces holding unsaved
scratch buffers, and the resulting `delete_workspace_by_id` call
cascaded into `items`, `pane_groups`, and the per-editor tables. For
clarity, the method has been renamed to `recent_workspaces_for_ui`.
Meanwhile, `last_session_workspace_locations` used a slightly different
form of the same predicate. The two disagreeing on what counts as a
valid workspace caused #48799, `Workspace WorkspaceId(N) not found`
errors on repeated launches (#50409), the
`test_window_edit_state_restoring_enabled` flake (#50871), and the
foreign key constraint fail on `projects: open recent` with a dirty
scratch buffer (#51456).
Note that for the last issue mentioned (#51456) there is no save prompt
for scratch buffers. This seems out of scope for this PR so I'll fix
that after this is addressed.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [ ] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#48799Closes#50409Closes#50871Closes#51456
Release Notes:
- Fixed unsaved scratch buffers being lost across restarts and an
occasional error when opening a recent project.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
We were just creating a window without showing it. However, wayland does
not respect this setting, so it seemed to work.
Now we actually activate the window :)
Rather than changing the single call-site, we always do this when
calling `Workspace::new_local`, since there were no code paths in the
repo that pass `NewWindow` without immediately activating it
Release Notes:
- N/A or Added/Fixed/Improved ...
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
Use stored thread metadata during reset when no root thread view exists,
so retries keep the original session id and title instead of starting a
new session.
Add a regression test for reconnecting after an initial custom agent
load failure.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes #ISSUE
Release Notes:
- N/A or Added/Fixed/Improved ...
---------
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#54391
Release Notes:
- Fixed onboarding UI scrollbar placement so the vertical scrollbar now
appears at the right edge of the onboarding pane, while keeping
onboarding content centered.
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
This PR fixes an issue where we'd show incorrect worktree labels when
opening a linked worktree directly. The linked worktree name would be
displayed in the project button but the dedicated worktree button would
be displaying "main", which is incorrect.
Now that we have a dedicated worktree button in the title bar, we can
make the project button always show the root repository name,
effectively matching the threads sidebar.
Release Notes:
- Fixed the title bar worktree button showing "main" when opening a
linked git worktree directly.
This will be fully replaced by
https://github.com/zed-industries/zed/pull/54342 - however, this will
not help for the next week on the stable branch, since we have plenty of
non-signed commits on that branch currently.
Thus, adding this temporary check here to fix this and which is a
cherry-pickable change. This **will** be fully replaced by the linked
PR, however, I cannot cherry pick that since we would otherwise need to
force-push the branches to remove the bad commits.
Release Notes:
- N/A
This helps with two things: Testing changes locally against real commits
as well as laying the groundwork for making things less convoluted for
the GitHub worker.
Also removes some useless wrapping left from an earlier direction.
Release Notes:
- N/A
When images are resized to meet provider size constraints (Anthropic's
1568px limit or the 5MB encoded-PNG cap), the stored ImageSize was still
recording the original width/height rather than the final post-downscale
dimensions. This caused incorrect token estimation via estimate_tokens()
since it uses width * height / 750.
Use processed_image.dimensions() after all downscale passes so that
ImageSize reflects the actual image sent to the provider.
Release Notes:
- Fixed an issue where token estimation would be incorrect in case where
the thread contained downscaled images.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes #ISSUE
Release Notes:
- N/A
Closes ZED-6PK
The issue here was that we could hit cases where the amount of segments
< amount of CPUs, e.g. for 5 candidates and 4 CPUs, we would have 2
candidates per matcher, so for the fourth matcher, we would start
slicing at 3 * 2 = 6 > 5, which is out of bounds.
Instead, make it so that we distribute the candidates across all
matchers so that all matchers have either n or n + 1 candidates.
No release notes since this is only on Nightly.
Release Notes:
- N/A
Fixes#52453Fixes#53246
Both issues were introduced by #51208 ("Handle Linux FS Rescan Events"),
which added `PathEventKind::Rescan` handling.
## Rules files not loading (#52453)
`load_worktree_info_for_system_prompt` called `load_worktree_rules_file`
synchronously, which uses `entry_for_path()` on the current worktree
snapshot. If the background scanner hasn't finished its initial scan,
the entry doesn't exist yet and the lookup returns `None` — the code
concludes no rules file exists. This was always a latent race condition,
but became more visible after Rescan events were introduced, since they
can trigger additional `WorktreeUpdatedEntries` churn that interacts
with the refresh mechanism.
The fix awaits `scan_complete()` on local worktrees before performing
the rules file lookup, ensuring the full directory tree is indexed
first.
## Config file rescan clearing OAuth tokens (#53246)
The `Rescan` handlers in `watch_config_dir` used
`fs.load(file_path).await.unwrap_or_default()`, which turns any
file-read error into an empty string. This empty string flows to
consumers like `CopilotChat`, where `extract_oauth_token("")` returns
`None`, causing the OAuth token to be unconditionally overwritten with
`None` — triggering re-authentication.
The fix changes both Rescan handlers to skip files that fail to load
(using `if let Ok(contents) = ...`), matching the pattern already used
by the `Created`/`Changed` handler.
Release Notes:
- Fixed rules files (AGENTS.md, CLAUDE.md, .rules, etc.) sometimes not
being applied in agent threads.
- Fixed GitHub Copilot re-prompting for authentication after filesystem
rescan events.
## Summary
Fix completion ordering when two items are otherwise tied and only
differ by letter case.
In cases like `abc` vs `ABC`, if the user types `a`, Zed should prefer
`abc`. If the user types `A`, Zed should prefer `ABC`. This matches the
expectation described in #37081 and #27993, where `subscription` should
rank above `Subscription` for query `s`.
## What changed
In `CompletionsMenu::sort_string_matches`, I added a tie-breaker that
prefers completions with more exact case-sensitive matches at the
fuzzy-match positions.
This only applies after the existing higher-priority sort keys, so it
does not replace fuzzy score, match positions, snippet ordering, or LSP
`sortText`. It only resolves ambiguous ties more intuitively.
## Tests
Added regression coverage in `code_completion_tests` for abstract
case-only examples:
- `a` prefers `abc` over `ABC`
- `A` prefers `ABC` over `abc`
- `ab` prefers `abc` over `ABC`
- `AB` prefers `ABC` over `abc`
- mixed-case multi-letter queries like `Ab` and `aB`
## Verification
Ran:
```bash
cargo test -p editor code_completion_tests
```
PS: all code and description is generated by Codex
Release Notes:
- Fixed completions not tie braking by case
Closes#48729, closes#27263, closes#45954
This PR aims to make zed responsive on symlinked directory events
outside the editor.
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
new-linked-folder is inside /zed-test/zed-project
output of ls -ld new-linked-folder
`lrwxr-xr-x 1 prayanshchhablani staff 42 28 Mar 23:20 new-linked-folder
-> /Users/prayanshchhablani/new-target-folder`
this shows new-linked-folder is a symlink folder whose target is
new-target-folder which is outside the root dir of the project opened in
zed.
https://github.com/user-attachments/assets/ffebafc3-2fc4-4293-bdbf-3a894a45e276
Release Notes:
- Fixed file watching of symlinks that point outside of the
project/watched directory. Zed should now properly respond to changes in
files in symlinked directories
## Summary
Paths inside parentheses without a preceding space (e.g.
`Update(.claude/skills/sandbox/SKILL.md)` or `Write(/test/cool.rs)` from
Claude Code output) were not clickable in the terminal. The `(`
character was allowed as a middle character in the default path
hyperlink regex, causing the entire `Update(.path/here` to be matched as
a single invalid path.
**Changes:**
- Remove `(` from the middle-chars alternation (`[:(]` → `:`) in the
default path hyperlink regex, so `(` always acts as a path boundary —
consistent with it already being excluded from first and last character
sets. Preserves upstream's space exclusion after `:`.
- Iterate all regex matches in the line instead of only the first, so
the correct path (which may be the second match after a prefix like
`Update`) is found. This also simplifies the code by removing the
separate hovered-word search logic.
**Known tradeoff:**
Filenames with parentheses in the middle (e.g.
`docker-compose.prod(copy).yml`) are no longer matched as a single path.
This is uncommon in terminal output contexts (compiler errors, stack
traces, tool output) and is documented with a `should_panic` test.
**What doesn't break:**
- `(/path/file.js:321:13)` — `(` at word start is excluded by the
first-char rule (unchanged)
- Node.js stack traces like `at fn (/path/file.js:10:5)` — space before
`(` makes it a separate word
- All 38 terminal hyperlink tests pass
Related: #18556, #23774
Release Notes:
- Fixed terminal path detection for paths inside parentheses without
preceding space (e.g. `Update(path)` or `Write(path)` patterns from CLI
tool output)
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Closes#46551
Zed was using the `terminal.shell` setting for some actions unrelated to
the terminal GUI, like environment capture, ACP agent startup, context
server startup, and vim `:!`, and so if a user set it to `tmux` or
`nushell`, those paths would fail.
This PR ensures paths unrelated to Zed's terminal use the system path
instead.
Manual testing:
Set `terminal.shell` to `tmux` or any another non-bash executable.
Starting an external agent thread shouldn't break.
Release Notes:
- Fixed ACP agent and other breakage when `terminal.shell` was set to a
non-shell program like `tmux`
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Partially addresses #51455
Release Notes:
- Fixed terminal child processes surviving after closing Zed by sending
SIGTERM synchronously on terminal drop
Closes#45271
Zed hardcodes `ControlMaster=yes` with a `ControlPath` in a random temp
directory, so it can never find a ControlMaster session the user already
has open. This means you get prompted for credentials again even if
you're already authenticated.
This patch checks for an existing master before spawning a new one. It
runs `ssh -G` to resolve the user's configured `controlpath` (with `%h`,
`%p`, etc. already expanded), then `ssh -O check` to verify it's alive.
If it is, Zed skips askpass and uses the existing socket. If not, the
current behavior is unchanged. Both commands are local and don't hit the
network.
Also adds a `killed: AtomicBool` to `SshRemoteConnection` because
`has_been_killed()` was using `master_process.is_none()` as a proxy for
"connection is dead." When reusing an external master, `master_process`
starts as `None`, so the connection pool would discard it immediately.
Tested against a university SSH setup with `ControlMaster auto`
configured. The reuse path connects without prompting, and the fallback
path works normally when no master exists. Windows is unaffected.
Release Notes:
- Zed now reuses existing SSH ControlMaster sessions instead of
prompting for credentials again (#45271).
When opening a remote session, sometimes you get a message:
/bin/bash: line 1: cd: /Users/pep: No such file or directory
which suggests that that the local home dir (/Users/pep) is used for the
remote terminal session, where it should be something like /home/ubuntu.
Release Notes:
- Fixed remote terminals incorrectly trying to change to a local home
directory.
### Context:
- Having a unified way of searching would allow for better debugging as
we move forward here. Right now we have remote/headless specific
searching and it's getting messy. This should allow for a more intuitive
function graph in the head for debugging environment related issues in
remote repl. The implementation mirrors python.rs approach.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#50892
Release Notes:
- N/A
Self-Review Checklist:**
- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [ ] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [X] Performance impact has been considered and is acceptable
## What
When terminal cursor shape is set via escape sequences (e.g. `\e[6 q`
for bar, `\e[4 q` for underline), the cursor looks identical whether the
terminal is focused or unfocused. Block cursors already convert to a
hollow outline when unfocused, but bar and underline were missing this
treatment.
## How
Added unfocused guards for `Beam` and `Underline` in the cursor shape
match in `terminal_element.rs`, converting them to `Hollow` when the
terminal loses focus. This follows the same pattern already used for
`Block`.
## Why this approach
A hollow block outline is the clearest unfocused signal and stays
consistent with the existing Block behavior. Alternative approaches like
dimming opacity or drawing outlined versions of bar/underline were
considered, but a 2px-wide outlined bar would be nearly invisible at
normal font sizes.
## Demo:
Before:
https://github.com/user-attachments/assets/81d49899-0837-42fe-a68f-4eb745892af6
After:
https://github.com/user-attachments/assets/a4444e28-5835-4c9f-872f-e9ce8c4805a0Closes#52716
Release Notes:
- Fixed terminal bar and underline cursors set via escape sequences not
visually distinguishing between focused and unfocused states
