Bug 13665
| Summary: | xl driver does not work with 3C900 10-T cards. It only works with 3c90x 10/100 cards | ||
|---|---|---|---|
| Product: | Base System | Reporter: | dave <dave> |
| Component: | misc | Assignee: | freebsd-bugs (Nobody) <bugs> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | Unspecified | ||
| Hardware: | Any | ||
| OS: | Any | ||
|
Description
dave
1999-09-09 18:00:00 UTC
Of all the gin joints in all the towns in all the world, dave@dugard.org had to walk into mine and say: > >Number: 13665 > >Category: misc > >Synopsis: xl driver does not work with 3C900 10-T cards. It only works with 3c90x 10/100 cards You know, I scream and yell and harrass people in an attempt to get them to provide accurate and detailed information when they file so-called bug reports, and by now I would have though everyone would have noticed the screaming and yelling and gotten the point. But nooooo.... Well that's just fine. I can scream and yell all day. Just remember: you asked for it. > >Description: > Machine started rebooting What does it say when it reboots! Does it print a panic message?! If so, what!? What does it say when you do a verbose boot?! When exactly does it reboot! Show us what it says on the screen!! Write it down! Set up a serial console! Paint it on a cave wall! Chisel it into a stone tablet! Do *anything* besides just sit there like a bump on a log and complain that "it don't work!" You can't just say "the machine reboots." You have to provide a *DETAILED* description of what you observe. This means write down everything that you see: all the panic messages, all the error messages, all of the things that indicate something has gone wrong. Why? Because I can't see your computer from here, so if you don't explain exactly what happend I can't help you! > right after it looked at the NIC It's not "the NIC." It's a particular model. There are several 3Com 10Mbps-only PCI cards; it's important that I know which one you have. The kernel tells you what model it is when it probes it, then you share this information with us so we know exactly what you're talking about! > after make world and then the kernel config. > After reading /sys/i386/conf/LINT about the MII bus device I guessed > that it might be a problem with 3c900. Well you guessed wrong. I just compiled the -current xl driver and loaded it on my test box with a 3c900-COMBO: xl0: <3Com 3c900-COMBO Etherlink XL> irq 9 at device 13.0 on pci0 xl0: Ethernet address: 00:60:08:08:f5:4a xl0: selecting 10baseT transceiver, half duplex Seems to work fine for me. This means that whatever your problem is, it's more complicated than you make it out to be, which means we need more information, which you didn't provide! Look: the one thing I can't stand is having to go several rounds of e-mail with people in order to painstakingly extract the details about their particular problem. People seem to think it's okay to just post a little blurb first and ask "does anybody have any ideas?", and then, after several hours of interrogation, then finally disclose more details. This is *WRONG*. You want to provide all of the details *UP* *FRONT* so people don't have to go tracking you down trying to get more information out of you! It wastes time! Now I want you to provide more information. I want you to provide it a timely fashion (i.e. within the next 24 hours, not next week or next month, or whenever you feel like it), I want it chock full of details, I want it verbatim without any attempts on your part to interpret, paraphrase or otherwise mangle it, I want it in complete sentences and I want it in 80 columns or less. I don't want you to say things like "I can't remember" or "I forgot to write it down" or "I have to go to Fiji for a week, I'll try it when I get back." Take a memory improvement class, get a pencil and paper, cancel the reservations at the Fiji Hilton, put the card *back* in the machine, recreate the problem and document the crap out of it! -Bill -- ============================================================================= -Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu Work: wpaul@ctr.columbia.edu | Center for Telecommunications Research Home: wpaul@skynet.ctr.columbia.edu | Columbia University, New York City ============================================================================= "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness" ============================================================================= State Changed From-To: open->closed I still haven't heard anything from you yet, but I managed to figure out your problem anyway. Let me explain to you the things you *didn't* tell me which would have saved me a lot of time: - You only said the system reboots. This is wrong. In fact, it panics, and the panic comes from ifmedia_set. It tells you this quite plainly, and it would have taken you all of one minute to grab a piece of paper and write this information down. You have no excuse for not doing this. When you're standing before St. Peter at the pearly gates, this will be counted against you. - You didn't check the card with the 3C90XCFG.EXE utility and inspect the transceiver setting. If you had, you would have discovered that the setting is for "auto" which is part of the problem. I'm not going to flame you too hard for this since it's not exactly obvious. The problem is that the driver doesn't support autoselection for non-NWAY transceivers. For the 10/100 cards, autoselection is done with NWAY. For the TPC/TPO/COMBO/FL cards, the driver just selects the default transceiver selection from the EEPROM, and you can override it with ifconfig if you need to. However if the EEPROM says to use "auto," the driver doesn't handle it correctly. (The one exception is the 3c905B-COMBO, for which there is already a little bit of special handling.) I patched the driver to detect the case where "auto" has been selected for the case where there is no NWAY media and select a reasonable default based on the card type. I just noticed this bug, which can probably be closed: I currently have a 3c900 10baseT card which is working just fine with FreeBSD 4.4-STABLE. Incidentally, the card does not work with Windows 2000, because of an obscure error. Hopefully, a real system can make it work :-) Samuel Tardieu wrote:
> I just noticed this bug, which can probably be closed: I currently have
> a 3c900 10baseT card which is working just fine with FreeBSD 4.4-STABLE.
>
> Incidentally, the card does not work with Windows 2000, because of an
> obscure error. Hopefully, a real system can make it work :-)
>
Forget about this notice: I hadn't seen that the bug had been closed a
long time ago.
|
