Date: Mon, 16 Jun 2003 22:34:56 -0700 From: John-Mark Gurney <gurney_j@efn.org> To: "M. Warner Losh" <imp@bsdimp.com> Cc: arch@freebsd.org Subject: Re: make /dev/pci really readable Message-ID: <20030617053456.GG73854@funkthat.com> In-Reply-To: <20030616.231258.116352275.imp@bsdimp.com> References: <20030616074122.GF73854@funkthat.com> <20030616.225216.115910026.imp@bsdimp.com> <20030617050905.GE73854@funkthat.com> <20030616.231258.116352275.imp@bsdimp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
M. Warner Losh wrote this message on Mon, Jun 16, 2003 at 23:12 -0600:
> In message: <20030617050905.GE73854@funkthat.com>
> John-Mark Gurney <gurney_j@efn.org> writes:
> : M. Warner Losh wrote this message on Mon, Jun 16, 2003 at 22:52 -0600:
> : > In message: <20030616074122.GF73854@funkthat.com>
> : > John-Mark Gurney <gurney_j@efn.org> writes:
> : > : Does anyone have an objection to making /dev/pci really honor the
> : > : permissions, and giving normal users (or just group wheel) premission
> : > : to run pciconf -l. Right now the code requires the write bit set for
> : > : any operation.
> : >
> : > Yes. That's too dangerous and will panic a machine.
> :
> : I'm not talking about -r... and anyways, I have fixed part of the
> : problem with unaligned reads/writes which has been posted to this
> : thread already....
>
> Saw that. Still not sure it is a good idea, but only if the code is
> reviewed heavily...
You can take a look at it if you want (I'll send you my patchse
shortly.). Shouldn't take but about 5-10 minutes of your time. It's
a VERY simple interface/code.
> : > You can get the same information that pciconf -l does by going through
> : > the alternative interface of devinfo.
> :
> : So, I hear you proposing to remove pciconf and the /dev/pci interface?
>
> Not really. I'm saying that we should beef up devinfo interface so
> that it can get PNP information from more databases that just the PCI
> one. pciconf is the wrong place for -v to be placed. But it was the
> only place to place it when it was written. Now that other busses
> have begun to implement pnp info to userland, we should look at a good
> way to deal.
I agree... devinfo is much beter than pciconf for device information,
but this whole bit was just started to get one small piece of code
committed, at least it'll be better while it lasts.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030617053456.GG73854>