On Sat, Dec 16, 2023 at 04:53:23PM +1100, NeilBrown wrote:
> On Sat, 16 Dec 2023, Ahelenia Ziemiańska wrote:
> > On Sat, Dec 16, 2023 at 03:27:16PM +1100, NeilBrown wrote:
> > > On Sat, 16 Dec 2023, Ahelenia Ziemiańska wrote:
> > > > To make it self-documenting, the referenced commit added the space
> > > > for the null terminator as sizeof('\0'). The message elaborates on
> > > > why only one byte is needed, so this is clearly a mistake.
> > > > Spell it as 1 /* NUL */ instead.
> > > >
> > > > Fixes: commit 1e360a60b24a ("SUNRPC: Address buffer overrun in
> > > > rpc_uaddr2sockaddr()")
> > > It isn't clear to me that "Fixes" is appropriate as that patch isn't
> > > harmful, just confused and sub-optimal.
> > I definitely agree, I don't like Fixes here at all,
> > but I don't really see another trailer in the documentation
> > or in the log that could be used for this.
> >
>
> Make up a new Trailer?
>
> I would probably just write
>
> To make it self-documenting,
> commit 1e360a60b24a ("SUNRPC: Address buffer overrun in rpc_uaddr2sockaddr()")
> added the space for the null terminator as sizeof('\0') which is 4. The commit
> elaborates on why only one byte is needed, so this is clearly a mistake.
> Spell it as 1 /* NUL */ instead.
Agreed; if Fixes: is overkill, simply spell out the commit that
introduced the issue in the patch description as Neil does here.
--
Chuck Lever