Date: Mon, 21 Jul 1997 12:32:37 -0600 From: Warner Losh <imp@rover.village.org> To: "Jordan K. Hubbard" <jkh@time.cdrom.com> Cc: Sean Eric Fagan <sef@Kithrup.COM>, current@freebsd.org Subject: Re: I am contemplating the following change... Message-ID: <E0wqNG9-0001rg-00@rover.village.org> In-Reply-To: Your message of "Mon, 21 Jul 1997 11:19:37 PDT." <29037.869509177@time.cdrom.com> References: <29037.869509177@time.cdrom.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <29037.869509177@time.cdrom.com> "Jordan K. Hubbard" writes: : I agree, but after 3 years I think I can be forgiven for assuming that : this particular Godot is not going to show up. :-) :-). What is delaying the arrival of Gadot? :-) Heck, I'd even settle for something that I can run in configure mode that would tell me what I needed to know. Even if it did fubar other cards in the process, it is a lot easier to reboot from the config mode than from a full system. :-) Barring that, I'd even settle for something that says "I told the card to interrupt, but the IRQ you told me to use didn't come within a second (or 10 seconds), so your ethernet card is likely fubar'd." The biggest problem that I've had is forgetting that just because the kernel prints it, doesn't mean that it is the proper settings. Most of these cards come with something that tests the interrupt. If it can't make the card interrupt, then it will tell you that that test failed. Is this possible during the probing of the card? If not, it is possible just before it sends its first packet? I'd be happy if the ed driver did this... So my preferences are: 1) Have the kernel tell me where the card is config'd for. 2) Have the kernel tell me that the interrupt doesn't seem to be right at probe time. 3) Have the kernel tell me as soon as it can that the card isn't interrupting. 4) The status quo. Are my desires even close to being able to be realized (well, #4 obviously is :-). Warner P.S. Regardless, I think that jordan's changes are good.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E0wqNG9-0001rg-00>