Re: [PATCH] tracing/kprobe: Fix to support kretprobe events on unloaded modules

From: Masami Hiramatsu
Date: Wed Jan 27 2021 - 19:11:46 EST


On Wed, 27 Jan 2021 17:29:50 -0500
Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:

> On Thu, 28 Jan 2021 00:37:51 +0900
> Masami Hiramatsu <mhiramat@xxxxxxxxxx> wrote:
>
> > Fix kprobe_on_func_entry() returns error code instead of false so that
> > register_kretprobe() can return an appropriate error code.
> >
> > append_trace_kprobe() expects the kprobe registration returns -ENOENT
> > when the target symbol is not found, and it checks whether the target
> > module is unloaded or not. If the target module doesn't exist, it
> > defers to probe the target symbol until the module is loaded.
> >
> > However, since register_kretprobe() returns -EINVAL instead of -ENOENT
> > in that case, it always fail on putting the kretprobe event on unloaded
> > modules. e.g.
> >
> > Kprobe event:
> > /sys/kernel/debug/tracing # echo p xfs:xfs_end_io >> kprobe_events
> > [ 16.515574] trace_kprobe: This probe might be able to register after target module is loaded. Continue.
> >
> > Kretprobe event: (p -> r)
> > /sys/kernel/debug/tracing # echo r xfs:xfs_end_io >> kprobe_events
> > sh: write error: Invalid argument
> > /sys/kernel/debug/tracing # cat error_log
> > [ 41.122514] trace_kprobe: error: Failed to register probe event
> > Command: r xfs:xfs_end_io
> > ^
> >
> > To fix this bug, change kprobe_on_func_entry() to detect symbol lookup
> > failure and return -ENOENT in that case. Otherwise it returns -EINVAL
> > or 0 (succeeded, given address is on the entry).
> >
> > Reported-by: Jianlin Lv <Jianlin.Lv@xxxxxxx>
> > Signed-off-by: Masami Hiramatsu <mhiramat@xxxxxxxxxx>
>
> Is this something that should go to stable? And if so, can you supply a
> Fixes tag?

Yes, I thought that had not supported previously, but that's wrong.
I found below commit introduced -ENOENT check in trace_kprobe side.

Fixes: 59158ec4aef7 ("tracing/kprobes: Check the probe on unloaded module correctly")

Thank you,


--
Masami Hiramatsu <mhiramat@xxxxxxxxxx>