Do we really need this for 1.9?

I describe about rC6f8b1d4cb798: ecc: Consistently handle parameters as unsigned value..

rC6f8b1d4cb798 is important (for me), as it is the removal of the last place which (somehow explicitly) allows use of negative value.

After this change, the ECC computation can assume that integers are all non-negative.

My intention of this task directly relates to:

• curve specific optimization
• introduce of redundant representation of integer (in future version of libgcrypt), for internal computation of modern ECC

which uses fixed number of limbs, assuming non-negative.

For printing SEXP, it would be good to have this change:

diff --git a/src/sexp.c b/src/sexp.c
index de28aedc..864916be 100644
--- a/src/sexp.c
+++ b/src/sexp.c
@@ -1483,6 +1483,12 @@ do_vsexp_sscan (gcry_sexp_t *retsexp, size_t *erroff,
                 }
               else
                 {
+                  if (mpifmt == GCRYMPI_FMT_USG && mpi_cmp_ui (m, 0) < 0)
+                    {
+                      err = GPG_ERR_INV_ARG;
+                      goto leave;
+                    }
+
                   err = _gcry_mpi_print (mpifmt, NULL, 0, &nm, m);
                   if (err)
                     goto leave;

This change makes sure not to write wrong value, but to raise an error.

I agree to the sexp change - but it should not be backported to 1.8