> -----Original Message-----
> From: Robin Dapp <[email protected]>
> Sent: 24 September 2026 15:55
> To: Tamar Christina <[email protected]>; Robin Dapp
> <[email protected]>; [email protected]
> Cc: nd <[email protected]>; [email protected]; [email protected]
> Subject: Re: [patch 9/14][isel]: form conditional vector compare IFNs
>
> >> At some point here we should ensure that the else value is zero
> >> before
> >> performing the replacement. I realize it's not necessary for SVE but if
> >> we have the (fixed-zero) else value like I mentioned in 7/14, we should
> >> make sure here.
> >
> > But there's no else value to emit. The OPTAB and ifn do not have an else.
> > since any non-zero else value don't make sense unless as I mentioned in
> > the earlier patch you can make sure all consumer are LEN masked which
> > today there is no guaranteed.
>
> Heh, that was of coursed based on the assumption that we'd add an else
> to them, if we don't then it's moot. Maybe let's continue under the
> assumption that COND_VEC_CMP is non-RVV and can enforce zeroing
> without
> target cost.
>
> > I'm happy to drop it, but I don't think you want to replace m & cmp with a
> > opaque conditional operation. Support for VLA explicitly didn't do this as
> > you lose all optimizations. i.e. {0,0, ... } & cmp is obviously {0,0, ..}
> > and such
> > optimizations are important when we unroll, without this we can't remove
> > dead edges.
> >
> > So again I'm happy to drop this, but you'd be fighting missed optimizations
> > for a long time by doing this for no obvious gain to me. I really think you
> > should pick a defined else value in GIMPLE for these CFG operations.
>
> In the RVV path today, we don't have the m & cmp but emit cmp +
> VCOND_MASK_LEN. So there's no mask to be elided or optimized.
>
> Somewhat related but not to the RVV case: Why can't
> try_conditional_simplification and friends optimize {0,0, ... } & cmp
> (with cmp being an IFN)? Shouldn't it, conceptually, even if actual
> support is missing?
I'm not sure how that would work when you have a non zero else.
a = {1,1,1,1, ...}
b = {2,2,2,2, ...}
mask = {1,1,0,0,..}
c = VEC_COND_LEN_CMP (GT_EXPR, a, b, mask, len, else_value)
the vector Boolean result here is undefined. The value of else_value affects
the result of c.
So a normal GIMPLE expression c != 0 which checks, for every value in c, is any
value not zero, can't reduce because it doesn't know for the inactive elements
what the result will be.
For normal VLA code, this would have been folded to false.
Thanks,
Tamar
>
> --
> Regards
> Robin
