Date: Mon, 25 Dec 2017 13:22:01 -0700 From: Warner Losh <imp@bsdimp.com> To: Hans Petter Selasky <hps@selasky.org> Cc: Warner Losh <imp@freebsd.org>, src-committers <src-committers@freebsd.org>, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: Re: svn commit: r327177 - in head/usr.sbin: . devmatch Message-ID: <CANCZdfrky3aaYLVZAxOCGeuFUtE6ZpJ2TDKUqrhJ3VNuDGR_cQ@mail.gmail.com> In-Reply-To: <72cf7768-9577-24af-1635-085bca9383d1@selasky.org> References: <201712251856.vBPIu8av032814@repo.freebsd.org> <72cf7768-9577-24af-1635-085bca9383d1@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Dec 25, 2017 1:14 PM, "Hans Petter Selasky" <hps@selasky.org> wrote: Hi, Did you try sorting the hints, making them candidate for a binary search? I see the is a small bug hooking the utility into the parent Makefile. I guess this will be fixed soon. No. I've been contemplating what the best data structure to use would be since we have a number of and conditions coupled with don't care values as well as ranges. The bigger issue today is that we convert the pnp values for every single row... I wanted something working over a spped demon that wouldn't hit the tree for months... I'm hoping that this spurs the pnp table decoration in drivers... any optimization should also play well with my desire to do this for potential root devices in the boot loader... Warner
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfrky3aaYLVZAxOCGeuFUtE6ZpJ2TDKUqrhJ3VNuDGR_cQ>