Re: master: x86-64: TRUNCATE using the Lemire, Kaser, Kurz transform
Douglas Katzman via Sbcl-commits <[email protected]> Sat, 18 Apr 2026 22:51:54 -0400
| Newsgroups | gmane.lisp.steel-bank.cvs,gmane.lisp.steel-bank.devel |
|---|---|
| Message-ID | <CAOrNaszKPoFDfKaobEaR9Ld+8VX8yLf=EfoH2B5TmJYeH39F4Q@mail.gmail.com> |
--===============5201170701124308666== Content-Type: multipart/alternative; boundary="0000000000007505ce064fc74417" --0000000000007505ce064fc74417 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Sat, Apr 18, 2026 at 10:40=E2=80=AFPM Stas Boukarev <[email protected]>= wrote: > The VOP gets defeated by one of the 23 transforms, when (speed 3): > https://github.com/sbcl/sbcl/blob/master/src/compiler/srctran.lisp#L5135 > > Good catch. Can we make that transform give-up if it knows that the quotient isn't used, and that the backend supports fastrem and that the argument types would be acceptable ? Or should we try to just add a 24th transform that (miraculously) is squeezed in between two other transforms? --0000000000007505ce064fc74417 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><br></div><br><div class= =3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Sat, Apr 18, 2026= at 10:40=E2=80=AFPM Stas Boukarev <<a href=3D"mailto:[email protected]= " target=3D"_blank">[email protected]</a>> wrote:<br></div><blockquote = class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px sol= id rgb(204,204,204);padding-left:1ex">The VOP gets defeated by one of the 2= 3 transforms, when (speed 3):<br> <a href=3D"https://github.com/sbcl/sbcl/blob/master/src/compiler/srctran.li= sp#L5135" rel=3D"noreferrer" target=3D"_blank">https://github.com/sbcl/sbcl= /blob/master/src/compiler/srctran.lisp#L5135</a><br> <br></blockquote><div>Good catch.</div><div>Can we make that transform give= -up if it knows that the quotient isn't used, and that the backend supp= orts fastrem and that the argument types would be acceptable ?=C2=A0 Or sho= uld we try to just add a 24th transform that (miraculously) is squeezed in = between two other transforms?</div></div></div> </div> --0000000000007505ce064fc74417-- --===============5201170701124308666== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline --===============5201170701124308666== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Sbcl-commits mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/sbcl-commits --===============5201170701124308666==--