[PING #2] [PATCH] enhance -Wrestrict to handle string built-ins (PR 78918)

[Martin Sebor <msebor at gmail dot com>]

Richard,

in discussing this work Jeff mentioned that your comments on
the tree-ssa-alias.c parts would be helpful.  When you have
a chance could you please give it a once over and let me know
if you have any suggestions or concerns?  There are no visible
changes to existing clients of the pass, just extensions that
are relied on only by the new diagnostics.

  https://gcc.gnu.org/ml/gcc-patches/2017-07/msg01264.html

I expect to revisit the sprintf %s patch mentioned below and
see how to simplify it by taking advantage of the changes
implemented here.

Thanks
Martin

On 07/24/2017 09:13 PM, Martin Sebor wrote:
> Ping:
>
>   https://gcc.gnu.org/ml/gcc-patches/2017-07/msg01264.html
>
> This change is related to
>
>   [PATCH] enhance -Wrestrict for sprintf %s arguments
>   https://gcc.gnu.org/ml/gcc-patches/2017-07/msg01176.html
>
> On 07/20/2017 02:45 PM, Martin Sebor wrote:
> > With more testing (building GDB, Glibc, Busybox, and the Linux
> > kernel) I found a few bugs and weaknesses in the initial patch.
> > Attached is version 2 that fixes the uncovered problems and makes
> > further enhancements to handle more cases.
> >
> > Martin
> >
> > On 07/16/2017 05:47 PM, Martin Sebor wrote:
> > > Being implemented in the front end, the -Wrestrict warning
> > > detects only trivial instances of violations.  The attached
> > > patch extends the implementation to the middle-end where
> > > data flow and alias analysis can be combined to detect even
> > > complex cases of overlap.  This work is independent of but
> > > follows on the patch below (waiting for review):
> > >
> > >   https://gcc.gnu.org/ml/gcc-patches/2017-07/msg00036.html
> > >
> > > The changes rely on extending the tree-ssa-{alias,structalias}
> > > machinery in a simple way to answer the "must-alias" kind of
> > > question in addition to the current "may-alias."
> > >
> > > The rest of the changes are in gimple-fold.c, tree-ssa-strlen.c,
> > > and builtins.c.
> > >
> > > Even though this change makes -Wrestrict a lot more useful, it's
> > > not a complete implementation.  Not all built-ins are handled yet
> > > (e.g., strncat), and support for user-defined functions is still
> > > subject to the limitations of the front end implementation.  To
> > > complete the support, handlers for the missing string built-ins
> > > will need to be added to tree-ssa-strlen.c, and the remaining
> > > bits should be moved from the front end to somewhere in
> > > the middle-end (e.g., calls.c).
> > >
> > > Martin