Re: [PATCH 1/8] mm: introduce for_each_process_rcu and for_each_thread_rcu

[Steven Rostedt]

On Tue, 8 Sep 2026 10:38:17 +0200
Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:

> Right, so GCC will accept and miscompile jumps into a scope. Clang will
> warn/error about this, so this should never happen.
> 
> Aside from the ASM GOTO issue, there is also an issue with computed
> goto, those don't work either.

So should we have Sashiko stop complaining about it?

   [Severity: Low]
   Does this mix scope-based cleanup with manual goto-based error handling?

   By replacing the manual RCU lock with for_each_process_thread_rculock()
   (which uses scoped_guard(rcu)), the function now relies on scope-based
   cleanup for the RCU lock, but still jumps to a manual "goto free" label
   for the memory cleanup below.

   The cleanup guidelines state that for a given routine, we should convert all
   resources that need a "goto" cleanup to scope-based cleanup, or convert none
   of them, avoiding mixed usage in the same function.

https://lore.kernel.org/linux-trace-kernel/20260911081013.E1ACF1F000FF@xxxxxxxxxxxxxxx/

-- Steve