Opened 16 years ago
Closed 13 years ago
#3120 closed bug (fixed)
3com Combo nic doesn't work
Reported by: | apprentice | Owned by: | axeld |
---|---|---|---|
Priority: | normal | Milestone: | R1 |
Component: | Drivers/Network/3com | Version: | R1/Development |
Keywords: | Cc: | adek336@…, mattmadia@… | |
Blocked By: | Blocking: | ||
Platform: | x86 |
Description
Just read the attached syslog...
Attachments (7)
Change History (38)
by , 16 years ago
comment:1 by , 16 years ago
comment:2 by , 16 years ago
Seems so. I am closing the other one though, since this one has got the syslog and mentions only one issue.
comment:3 by , 16 years ago
Cc: | added |
---|
by , 16 years ago
Attachment: | 3com.diff.1.diff added |
---|
comment:4 by , 16 years ago
When you'll feel like testing, let me know what traced contains with this patch.
by , 16 years ago
Attachment: | 3com-diff.jpg added |
---|
comment:6 by , 16 years ago
Cc: | added |
---|
comment:7 by , 16 years ago
You've got 3c900 10Mbps Combo [Boomerang] (vendor/device 10b7/9001) and it doesn't attach
I've got 3c900B-Combo Etherlink XL Combo [Cyclone] (vendor/device 10b7/9005) and it attaches and sends/receives but I don't get an IP.
comment:9 by , 16 years ago
What does "ints" in KDL say, does removing usb devices or disabling *HCI help?
comment:10 by , 16 years ago
Could you also try the fix proposed http://unix.derkeiler.com/Mailing-Lists/FreeBSD/current/2004-04/1049.html
if (did == TC_DEVICEID_TORNADO_10_100BT_920B || did == TC_DEVICEID_TORNADO_10_100BT_920B_WNM) sc->xl_flags |= XL_FLAG_PHYOK; switch (did) { + case TC_DEVICEID_BOOMERANG_10BT_COMBO: /* added by rodrigc */ case TC_DEVICEID_BOOMERANG_10_100BT: /* 3c905-TX */ case TC_DEVICEID_HURRICANE_575A: case TC_DEVICEID_HURRICANE_575B: case TC_DEVICEID_HURRICANE_575C: sc->xl_flags |= XL_FLAG_NO_MMIO;
comment:11 by , 16 years ago
I disabled usb & acpi & apm and about everything else you can disable from the bios - and also acpi and apic - from haiku safe mode options and still be able to boot but with the 3com diff it still crashes. I added the ints output but can't see anything network related in it. I'll try the fix when I get enough time to test it.
comment:12 by , 16 years ago
Sorry, can't attach the image... trac complains something about not being able to give an unique name to the file.
comment:13 by , 16 years ago
Trac is playing up right now. Try upload it to a external site, for the time being that is.
comment:14 by , 16 years ago
Could you try that patch?
Your card doesn't support MMIO but Haiku uses MMIO for you card anyway, the patch above disables MMIO for your model; unless there is some other problem, your card will now work.
comment:15 by , 16 years ago
I apologize for the long delay but I haven't had enough time lately to devote to this bug.
Adek336, thanks for the patch! It worked and ifconfig now lists /dev/net/3com/0. Networking still doesn't work, but I suspect the reason is that the freebsd driver defaults to using BNC port instead of TP and currently it seems that there is now way of changing the port. Maybe the driver should divert from the freebsd driver just by defaulting to TP. This would be a reasonable and quick fix and honestly, who uses BNC anymore anyway.
comment:16 by , 16 years ago
You haven't had /dev/net/3com/0 before if I recall correctly, right? That's great news, as it means we've progressed a lot with this patch!
I'll take a look at what I can do with TP/BNC.
by , 16 years ago
Attachment: | haiku.syslog added |
---|
by , 16 years ago
Attachment: | 3com-forcebt-disablemmio.diff added |
---|
comment:19 by , 16 years ago
checked and tested, the latest patch solves the problem completely. I get an IP and ping works. Thanks!
comment:20 by , 16 years ago
Cool. Nice work, guys! Am I correct in assuming that this patch cannot be applied as is?
follow-up: 27 comment:21 by , 16 years ago
Great[[br]] Indeed, the patch should be reworked a little bit so it uses the workaround only for this device id.
by , 16 years ago
Attachment: | 3com.1.diff added |
---|
by , 16 years ago
comment:23 by , 16 years ago
checked it. I think the screenshot speaks for itself. Again, thanks a _lot_
comment:24 by , 15 years ago
Version: | R1/pre-alpha1 → R1/alpha2 |
---|
It seems I'm running into the same problem with this card in alpha 2. Is the fix proposed in this ticket not in the tree yet?
comment:25 by , 14 years ago
Can someone have a look at if the patch is in the tree or if it can be added as is?
comment:26 by , 14 years ago
patch: | 0 → 1 |
---|
follow-up: 28 comment:27 by , 14 years ago
Replying to Adek336:
Great[[br]] Indeed, the patch should be reworked a little bit so it uses the workaround only for this device id.
Is http://dev.haiku-os.org/attachment/ticket/3120/3com.1.diff the reworked patch, ready to be committed?
comment:28 by , 14 years ago
No, the second hunk is clearly out of place; I can't rework it as I don't have the hardware anymore.
comment:29 by , 14 years ago
Had a look now and the patch are good as is. Same lines still exist in the same places (file version hrev34822)
comment:30 by , 13 years ago
Component: | Drivers/Network → Drivers/Network/3com |
---|
comment:31 by , 13 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Version: | R1/alpha2 → R1/Development |
Applied in hrev42952. Thanks!
duplicate of #2954