On Mon, 10 Oct 2022 at 16:18, Prathamesh Kulkarni
<prathamesh.kulkarni@linaro.org> wrote:
>
> On Fri, 30 Sept 2022 at 21:38, Richard Sandiford
> <richard.sandiford@arm.com> wrote:
> >
> > Richard Sandiford via Gcc-patches <gcc-patches@gcc.gnu.org> writes:
> > > Prathamesh Kulkarni <prathamesh.kulkarni@linaro.org> writes:
> > >> Sorry to ask a silly question but in which case shall we select 2nd vector ?
> > >> For num_poly_int_coeffs == 2,
> > >> a1 /trunc n1 == (a1 + 0x) / (n1.coeffs[0] + n1.coeffs[1]*x)
> > >> If a1/trunc n1 succeeds,
> > >> 0 / n1.coeffs[1] == a1/n1.coeffs[0] == 0.
> > >> So, a1 has to be < n1.coeffs[0] ?
> > >
> > > Remember that a1 is itself a poly_int.  It's not necessarily a constant.
> > >
> > > E.g. the TRN1 .D instruction maps to a VEC_PERM_EXPR with the selector:
> > >
> > >   { 0, 2 + 2x, 1, 4 + 2x, 2, 6 + 2x, ... }
> >
> > Sorry, should have been:
> >
> >   { 0, 2 + 2x, 2, 4 + 2x, 4, 6 + 2x, ... }
> Hi Richard,
> Thanks for the clarifications, and sorry for late reply.
> I have attached POC patch that tries to implement the above approach.
> Passes bootstrap+test on x86_64-linux-gnu and aarch64-linux-gnu for VLS vectors.
>
> For VLA vectors, I have only done limited testing so far.
> It seems to pass couple of tests written in the patch for
> nelts_per_pattern == 3,
> and folds the following svld1rq test:
> int32x4_t v = {1, 2, 3, 4};
> return svld1rq_s32 (svptrue_b8 (), &v[0])
> into:
> return {1, 2, 3, 4, ...};
> I will try to bootstrap+test it on SVE machine to test further for VLA folding.
With the attached patch it seems to pass bootstrap+test with SVE enabled.
The only difference w.r.t previous patch is it adds check in
get_vector_for_pattern
if S is constant otherwise returns NULL_TREE.

I added this check because 930325-1.c ICE'd with previous patch
because it had following vec_perm_expr,
where S was non-constant:
vect__16.13_70 = VEC_PERM_EXPR <vect__16.12_69, vect__16.12_69, {
POLY_INT_CST [3, 4], POLY_INT_CST [6, 8], POLY_INT_CST [9, 12], ...
}>;
I am not sure how to proceed in this case, so chose to bail out.

Thanks,
Prathamesh

>
> I have a couple of questions:
> 1] When mask selects elements from same vector but from different patterns:
> For eg:
> arg0 = {1, 11, 2, 12, 3, 13, ...},
> arg1 = {21, 31, 22, 32, 23, 33, ...},
> mask = {0, 0, 0, 1, 0, 2, ... },
> All have npatterns = 2, nelts_per_pattern = 3.
>
> With above mask,
> Pattern {0, ...} selects arg0[0], ie {1, ...}
> Pattern {0, 1, 2, ...} selects arg0[0], arg0[1], arg0[2], ie {1, 11, 2, ...}
> While arg0[0] and arg0[2] belong to same pattern, arg0[1] belongs to different
> pattern in arg0.
> The result is:
> res = {1, 1, 1, 11, 1, 2, ...}
> In this case, res's 2nd pattern {1, 11, 2, ...} is encoded with:
> with a0 = 1, a1 = 11, S = -9.
> Is that expected tho ? It seems to create a new encoding which
> wasn't present in the input vector. For instance, the next elem in
> sequence would be -7,
> which is not present originally in arg0.
> I suppose it's fine since if the user defines mask to have pattern {0,
> 1, 2, ...}
> they intended result to have pattern with above encoding.
> Just wanted to confirm if this is correct ?
>
> 2] Could you please suggest a test-case for S < 0 ?
> I am not able to come up with one :/
>
> Thanks,
> Prathamesh
> >
> > > which is an interleaving of the two patterns:
> > >
> > >   { 0, 2, 4, ... }                  a0 = 0, a1 = 2, S = 2
> > >   { 2 + 2x, 4 + 2x, 6 + 2x }        a0 = 2 + 2x, a1 = 4 + 2x, S = 2