Re: [PATCH 4.19.y] tracing/kprobes: Return EADDRNOTAVAIL when func matches several symbols

[Greg KH <gregkh@linuxfoundation.org>]

On Mon, Nov 27, 2023 at 06:36:32PM +0100, Francis Laniel wrote:
> Hi!
> 
> 
> Le lundi 27 novembre 2023, 16:44:22 CET Greg KH a écrit :
> > On Mon, Nov 27, 2023 at 03:41:31PM +0100, Francis Laniel wrote:
> > > Hi!
> > > 
> > > Le vendredi 24 novembre 2023, 17:17:04 CET Greg KH a écrit :
> > > > On Fri, Nov 24, 2023 at 01:24:13PM +0100, Francis Laniel wrote:
> > > > > When a kprobe is attached to a function that's name is not unique (is
> > > > > static and shares the name with other functions in the kernel), the
> > > > > kprobe is attached to the first function it finds. This is a bug as
> > > > > the
> > > > > function that it is attaching to is not necessarily the one that the
> > > > > user wants to attach to.
> > > > > 
> > > > > Instead of blindly picking a function to attach to what is ambiguous,
> > > > > error with EADDRNOTAVAIL to let the user know that this function is
> > > > > not
> > > > > unique, and that the user must use another unique function with an
> > > > > address offset to get to the function they want to attach to.
> > > > > 
> > > > > Link:
> > > > > https://lore.kernel.org/all/20231020104250.9537-2-flaniel@linux.micros
> > > > > oft
> > > > > .com/
> > > > > 
> > > > > Cc: stable@vger.kernel.org
> > > > > Fixes: 413d37d1eb69 ("tracing: Add kprobe-based event tracer")
> > > > > Suggested-by: Masami Hiramatsu <mhiramat@kernel.org>
> > > > > Signed-off-by: Francis Laniel <flaniel@linux.microsoft.com>
> > > > > Link:
> > > > > https://lore.kernel.org/lkml/20230819101105.b0c104ae4494a7d1f2eea742@k
> > > > > ern
> > > > > el.org/ Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> > > > > Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
> > > > > (cherry picked from commit b022f0c7e404887a7c5229788fc99eff9f9a80d5)
> > > > > ---
> > > > > 
> > > > >  kernel/trace/trace_kprobe.c | 48
> > > > >  +++++++++++++++++++++++++++++++++++++
> > > > >  1 file changed, 48 insertions(+)
> > > > 
> > > > Again, we need a version for 5.4.y as well before we can take this
> > > > version.
> > > 
> > > I sent the 5.4.y patch some times ago, you can find it here:
> > > https://lore.kernel.org/stable/20231023113623.36423-2-flaniel@linux.micros
> > > oft.com/
> > > 
> > > With the recent batch I sent, I should have cover all the stable kernels.
> > > In case I miss one, please indicate it to me so I can fix this problem and
> > > ensure all stable kernels have a corresponding patch.
> > 
> > I only see the following in my stable mbox right now:
> > 
> >    1   C Nov 27 Francis Laniel  (4.4K) ┬─>[PATCH 5.10.y] tracing/kprobes:
> > Return EADDRNOTAVAIL when func matches several symbols 2 r C Nov 24 Francis
> > Laniel  (4.4K) └─>[PATCH 5.10.y] tracing/kprobes: Return EADDRNOTAVAIL when
> > func matches several symbols 3   F Nov 24 To Francis Lani (1.5K)   └─>
> >    4 r T Nov 27 Francis Laniel  (1.9K)     └─>
> >    5   F Nov 27 To Francis Lani (2.0K)       └─>
> >   23 r C Nov 24 Francis Laniel  (2.7K) [PATCH 4.19.y] tracing/kprobes:
> > Return EADDRNOTAVAIL when func matches several symbols 24 r + Nov 27
> > Francis Laniel  (2.0K) └─>
> > 
> > So could you resend them all just to be sure I have all of the latest
> > versions that you wish to have applied?
> 
> I normally sent again the patch for version 4.14 to 5.15 (it was already 
> present in 6.1 and 6.6).
> I tested all of them by building the corresponding kernel with the patch 
> applied before sending, so they should not break compilation or testing.
> 
> Can you please confirm me you received them?

Got them now, thanks, I'll dig through them after this latest round of
stable kernels gets released.

thanks,

greg k-h