Date: Fri, 19 Oct 2007 18:10:23 -0400 From: Gary Palmer <gpalmer@freebsd.org> To: Scott Long <scottl@samsco.org> Cc: FreeBSD Current <freebsd-current@freebsd.org>, d@delphij.net Subject: Re: twa driver needs updating Message-ID: <20071019221023.GC54662@in-addr.com> In-Reply-To: <47192969.20809@samsco.org> References: <85AFE2F9-33F6-40E8-B760-990970B7A791@emeraldcityeg.com> <4718EC36.7090401@delphij.net> <4E478C75-35D0-47EA-A4A3-05D06A167A51@emeraldcityeg.com> <47192969.20809@samsco.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Oct 19, 2007 at 04:02:17PM -0600, Scott Long wrote: > ECEG / Daniel Duerr wrote: > >Thank you for redirecting my message. I've been keeping these machines > >updated on RELENG_6_2 because I have a couple pre-compiled hardware > >drivers I must use which I do not have kernel source for. I just did a > >csup/buildkernel last week on the latest p8 source tree and tested the > >included twa driver and it was not up to date. Would it be possible to > >get the twa driver pushed back into RELENG_6_2? > > > > There are plans to get the new device support pushed back to RELENG_6 in > time for the 6.3 release. I know that's not the best answer for you at > this moment. What is really, truly needed here is a "driver update > disk" setup for the installer that allows you to load vendor-supplied > drivers at install time and have those drivers integrated into the > installed system. Sysinstall has a clunky feature that kinda sorta > satisfies the first requirement, but completely falls down on the second > requirement, making it completely useless. If anyone is serious about > sitting down and making such a feature happen, let me know. It would be > an incredibly positive step towards better vendor support. I'd like to > be involved because I've been in the vendor role in the past with > supporting linux, and I'd like to make sure that the linux mistakes > aren't repeated. Since we still build a bunch of drivers into the kernel, how do you propose replacing a driver that is part of GENERIC? I think implementing the above feature will mean fundamentally changing the way we ship kernels.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20071019221023.GC54662>