On Thu, 10 Nov 2022 at 21:47, Palmer Dabbelt wrote:
>
> On Thu, 10 Nov 2022 07:09:35 PST (-0800), philipp.toms...@vrull.eu wrote:
> > On Thu, 10 Nov 2022 at 02:46, Palmer Dabbelt wrote:
> >>
> >> On Tue, 08 Nov 2022 11:54:34 PST (-0800), philipp.toms...@vrull.eu wrote:
> >> > The strength-reducti
On Thu, 10 Nov 2022 07:09:35 PST (-0800), philipp.toms...@vrull.eu wrote:
On Thu, 10 Nov 2022 at 02:46, Palmer Dabbelt wrote:
On Tue, 08 Nov 2022 11:54:34 PST (-0800), philipp.toms...@vrull.eu wrote:
> The strength-reduction implementation in expmed.c will assess the
> profitability of using s
On Thu, 10 Nov 2022 at 02:46, Palmer Dabbelt wrote:
>
> On Tue, 08 Nov 2022 11:54:34 PST (-0800), philipp.toms...@vrull.eu wrote:
> > The strength-reduction implementation in expmed.c will assess the
> > profitability of using shift-and-add using a RTL expression that wraps
> > a MULT (with a powe
On Tue, 08 Nov 2022 11:54:34 PST (-0800), philipp.toms...@vrull.eu wrote:
The strength-reduction implementation in expmed.c will assess the
profitability of using shift-and-add using a RTL expression that wraps
a MULT (with a power-of-2) in a PLUS. Unless the RISC-V rtx_costs
function recognizes
The strength-reduction implementation in expmed.c will assess the
profitability of using shift-and-add using a RTL expression that wraps
a MULT (with a power-of-2) in a PLUS. Unless the RISC-V rtx_costs
function recognizes this as expressing a sh[123]add instruction, we
will return an inflated cos
