[PATCH v6 0/5] tracing/fprobe: Fix fprobe_ip_table related bugs

[Masami Hiramatsu (Google)]

Hi,

Here is the 6th version of fprobe bugfix series.
The previous version is here.

https://lore.kernel.org/all/177606956628.929411.17392736689322577701.stgit@devnote2/

In this version, I added yet another bugfixes for rejecting registered
fprobe correctly [1/5], and unregistering ftrace_ops if it is empty on
module unloading [5/5]. Also, add RCU synchronization for failure path
[2/5] and retry rhashtable walking outside of rhltable_walk_enter/exit
[3/5].

The [5/5] is to handle the module unloading problem I reported
previously [1]. This was handled solely within fprobe to avoid
impacting the live patch.

[1] https://lore.kernel.org/all/20260408110237.08d982fab08356bc8d48af45@xxxxxxxxxx/

Thank you!

Masami Hiramatsu (Google) (5):
      tracing/fprobe: Reject registration of a registered fprobe before init
      tracing/fprobe: Remove fprobe from hash in failure path
      tracing/fprobe: Avoid kcalloc() in rcu_read_lock section
      tracing/fprobe: Check the same type fprobe on table as the unregistered one
      tracing/fprobe: Fix to unregister ftrace_ops if it is empty on module unloading


 kernel/trace/fprobe.c |  462 +++++++++++++++++++++++++++++++++----------------
 1 file changed, 308 insertions(+), 154 deletions(-)


base-commit: 4346be6577aaa04586167402ae87bbdbe32484a4
--
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>