Detached tasks spawned on every config generation set and restored the
placeholder DNS without ordering, so a set could back up the placeholder
itself and a later restore reported success while leaving it in place.
Reconcile from the applied config under the lifecycle lock, and restore
whenever the core stops. Keep one backup per network service, never back
up the placeholder, and restore only while the service still holds it.
Scripts verify every read and write and are killed after 5 s.
Refs #8070, #8168
Same-name nodes from different proxy-providers were marked ambiguous and
greyed out. Resolve each group member against the group's own providers
(use / include-all, exclude-type) and bind the n-th occurrence of a name to
the n-th candidate, matching Mihomo.
Closes#8021
When the service's start type was changed from automatic, startup waited
about two minutes for it before falling back to Sidecar. Detect this
immediately, skip the TUN readiness wait, and offer a reinstall that
restores automatic start.
Refs #8129
Dispatch blocking window restoration outside HTTP workers to prevent notification timeouts and repeated window activation in #8118. Remove the misleading failure log, update the changelog, and remove server tests as requested.
When the startup probe finds the service refusing the approved core, the
app falls back to Sidecar with only the generic "could not start the
core" notice. The reason and repair guide appeared only after a
reinstall, which cannot fix a writable volume root (#8045).
Carry the fallback reason to the window and show the same permission
notice as a rejected installation when the core was refused.
Refs #8045
A confirmed override on a profile with provider DNS was turned off at
every startup; it now stays on until that profile's provider DNS changes.
Refs #7993
Pick up tauri-plugin-mihomo@ba8434c: provider updates now wait out
mihomo's 20s fetch instead of failing at 5s with "error sending request",
so slow subscriptions update and real failures show mihomo's own error.
Refs #8046
A registered but stopped service made every version probe back off for
about two minutes, blocking startup before the window and core (#8059).
Check the SCM first; a service that is starting, or not yet started this
boot, still goes through the IPC retries.
Replace the local path pin from fbc4b516 with the release tag, which
ships the installer group lookup fix (#8043) along with the residual
helper (#8031) and app data owner (#8026) service changes.
When the service watchdog gave up or the core vanished, the owner
monitor marked the core stopped without telling the user, and on macOS
left the system proxy on the dead mixed port (#8029).
- On SameOwnerFailure, clear the proxy through the helper while the
session is still ours; displaced or unreachable owners still leave the
machine-wide setting alone.
- Retain the stop as a core failure with its own message, recorded
under the lifecycle lock and shown now or on next focus; a running
core resets the notice dedup.
- Log each watchdog restart the service reports, since launchd discards
the service's own log.
When the service refuses the app data root because another uid owns it
(code 1012), show a sticky notice with the chown command for this
account. Repairing the service cannot fix this.
Refs #8026
A macOS helper left running without IPC made Sidecar refuse to start
while the service counted as not installed, so no repair was offered
(#8031, #8043). Treat such a helper as install evidence and record the
Sidecar refusal as unavailable so the repair dialog appears.
Show unresolved startup failures when the window becomes visible and clear them when either core mode starts successfully. Cover recovery followed by exit and WebView recreation, use neutral service permission wording, and document the Windows startup fix.
Restore non-empty DNS page fields and hosts after Merge/Script processing while preserving extension values for empty fields. Keep provider protection tied to the original profile and retain per-profile confirmations.
Delegate service and core installation to Service IPC so replacement uses
one elevation. Share fallback occupancy checks and retain the execution
guard until the Sidecar exits.
Track missing approved cores and start refusals as unavailable, and show
when no core is running. Build matching service tools from the local IPC
dependency and pass both core executables through the Dev installer.
Save the DNS override switch and session confirmation per profile so automatic protection only disables the affected profile. Preserve the legacy switch as the initial preference and show the profile scope beside the settings gear and in the DNS editor.
Refs #7993
Replace DNS map fields and hosts instead of retaining old entries, while preserving unspecified DNS fields. Keep DNS override priority and explicit empty values intact.
Refs #7993
Address #7983 with a recovery dialog for Windows service error 1003, including startup refusals received before the UI subscribes. Reinstall the service and require service mode before reporting recovery.
Pass available cores and their SHA-256 digests explicitly during service installation and repair, instead of relying on automatic discovery that can skip untrusted source locations. Preserve the arguments across root, pkexec, and sudo execution paths.
Fixes#7995
Upgrade clash_verge_service_ipc and bundled service downloads to v2.7.2,
which labels runtime-installed executables before publication and repairs
existing core labels instead of timing out waiting for service IPC.
Register persistent bin_t rules for production and development service
binaries and cores in the RPM post-install script. Support older semanage
versions, apply rules to existing directories, and warn when optional
SELinux tools are unavailable.
Refs: #7965
Preserve TUN settings and restore proxy operation after fallback.
Deliver fallback notices after UI subscription and stop retrying
deterministic bundle refusals. Publish fallback state atomically
and avoid conflicting service startup notifications.
The service installer now admits the staged core copies through Windows
Firewall, so TUN system/mixed stacks pass traffic in service mode again
without users re-allowing the new path by hand.
Refs #7915
Limit GUI authority to saved TUN fields while preserving custom overrides.
Sync provider caches after service startup, reloads and updates.
Publish copies atomically with session, runtime and cache metadata checks.
The switch read enable_tun_mode while the Core was given
enable_tun_mode && !session_suppression, so falling back to Sidecar left it
showing a TUN that was not running, with nothing to write the flag back.
Drop the session flag: the stored preference is the only truth, re-checked
at startup and written off when the conditions for it are not met, and the
switch renders it against current capability.
Windows publishes the core by moving the running binary aside, and only the
success path could delete it afterwards. A failed restart left it on disk,
and the uninstaller did not know about it either.
mihomo's updater truncates its own running executable in place; macOS kills
that under Hardened Runtime and leaves a 0-byte core. Verge now downloads the
release, verifies a staged copy next to the target, and renames it into place,
which never touches the inode the running core executes from.
Closes#6834
The startup fallback persisted its replacement, so a port blocked once was never
asked for again. It now overrides the Runtime Config for the session only.
Most of those conflicts were not real: SO_REUSEADDR cannot reuse a TIME_WAIT
socket owned by another user, and a service-managed core runs as root while the
app does not. A refused bind is now confirmed by asking whether anything answers.
Also restores the borrowed port when saving proxy ports fails.
Each auto-update reloads the core, which drops every connection running through a
`proxy-providers` entry. Existing profiles are corrected once at startup; the form
warns about shorter intervals but still saves them.
The queue got a full interval from now while the UI counted from `updated`, so a
restart pushed every deadline back and the countdown could read "update failed".
Retiring a task disabled mid-update also keeps its state, so re-enabling it no
longer starts a second run alongside the first.
wonfen