Gitea 1.23.7 frequently crashes with PANIC

[amix307]

Description

Hi! After upgrading from Gitea 1.23.5 to 1.23.7, the instance frequently crashes with graceful-shutdown.

In gitea.log:

2025/05/15 15:17:58 .../graceful/manager.go:86:func1() [E] PANIC during RunWithShutdownContext: runtime error: invalid memory address or nil pointer dereference
Stacktrace: /usr/local/go/src/runtime/panic.go:262 (0x479a98)
/usr/local/go/src/runtime/signal_unix.go:917 (0x479a68)
/source/models/issues/comment.go:347 (0x20ee00a)
/source/services/issue/assignee.go:256 (0x20ee016)
/source/services/issue/assignee.go:246 (0x20ede58)
/source/services/pull/pull.go:453 (0x211c414)
/source/modules/graceful/manager.go:92 (0xc8b97a)
/source/services/pull/pull.go:366 (0x211b58f)
/usr/local/go/src/runtime/asm_amd64.s:1700 (0x480020)
2025/05/15 15:17:58 ...queue/workergroup.go:195:func1() [D] Queue "pr_auto_merge" stops idle worker
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "notification-service" is shutting down
2025/05/15 15:17:58 ...s/graceful/server.go:156:Serve() [D] Waiting for connections to finish... (PID: 3805066)
2025/05/15 15:17:58 ...queue/workergroup.go:195:func1() [D] Queue "notification-service" stops idle worker
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "repo_license_updater" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "task" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:350:doRun() [D] Queue "notification-service" stops running
2025/05/15 15:17:58 ...queue/workergroup.go:195:func1() [D] Queue "repo_license_updater" stops idle worker
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "repo-archive" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:350:doRun() [D] Queue "task" stops running
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "tag_sync" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "repo_stats_update" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "pr_patch_checker" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:350:doRun() [D] Queue "repo-archive" stops running
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "actions_ready_job" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:195:func1() [D] Queue "pr_patch_checker" stops idle worker
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "mail" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:349:doRun() [D] Queue "push_update" is shutting down
2025/05/15 15:17:58 ...queue/workergroup.go:195:func1() [D] Queue "repo_stats_update" stops idle worker

I'm unable to provide exact reproduction steps because my instance is heavily loaded with:

• More than 500 organizations
• Hundreds of repositories
• Many active users (>1000)
• Lot of teams (5-15 per Org)
• Frequent pull request and issue activity

The crashes appear to occur during normal high-load operations, but I cannot isolate a specific trigger due to the complexity of my instance

Gitea Version

1.23.7

Can you reproduce the bug on the Gitea demo site?

No

Log Gist

No response

Screenshots

No response

Git Version

No response

Operating System

CentOS9-Stream

How are you running Gitea?

systemd

Database

PostgreSQL

[amix307]

#34254