nforce 630i chipset USB controller behaves strangely under 8.0-RELEASE

ReverendRed · Dec 6, 2009

Hi there,

I have a problem with the USB controller of my board (GA-73PVM-S2H, nforce 630i chipset) under 8.0-RELEASE.

Whenever I want to create a new fs on an external hard drive (using newfs), the process hangs after a second or so and I cannot do anything except unplugging the drive. (Then I'll return to the prompt immediately, device lost...)

I plugged a different USB controller (VIA chipset) into an empty PCI slot on my board and tried again, using this one, and everything went well and without any problems.

As I encountered another strange issue when trying to install 8.0 from USB (the copy process was sloooow as hell after a while, it was unbearable...), I figure that 8.0-RELEASE seems to have some severe problems with nforce 630i chipsets.

Has anyone made a similar experience or maybe a hint for me how to find a workaround for this?

It feels so incomplete so be dependent on a secondary USB controller. x(

ktux · Jan 11, 2010

I can confirm the USB controller failure for my nForce 630i / GeForce 7100 based mainboard. FreeBSD also lacks SATA support for this board. My IDE cable is connected with internal hard drives, so I can't really tell if FreeBSD has any support whatsoever for this board.

ktux · Jan 11, 2010

It's been over a month since ReverendRed's post and this chip set has been unsupported since it's inception four years ago. I have a few questions before I give up on BSD altogether :

- does anybody read either this forum or the actual bug reports ?

- is more information needed - yes or no, or are desktop PCs even a priority ?

-if there is no intention of fixing things for this chip set before it becomes completely obsolete, why not say so ?

-why is bizarro hardware that can only be found at some geek flea market actively maintained, but mainstream hardware by major manufacturers ignored ?

-why is it that every Linux distro on the planet can handle this chip set, but FreeBSD has actually taken a step backwards to being completely non-functional ?

jb_fvwm2 · Jan 11, 2010

I am just curious whether you have *all* the
new v8 usb drivers loaded. (On my local
machine(s) I had to add one of them manually,
unknown to me whether it would be the case
for a GENERIC install or not). That final
kldload fixed an able-to-mount-thumbdrive
problem...

ktux · Jan 12, 2010

I don't know how to load additional USB drivers during the install process. I was assuming that if they were present, that they would be loaded. SATA support, I know for sure, is broken.

ReverendRed · Jan 12, 2010

Thanks, ktux, for picking this one up again... I'm still a little frustrated because of this issue as I consider the nforce 630i to be a fairly well-known chipset and do not want to plug a PCI USB controller into this machine just to get things done. It does not seem to be the sole fault of FreeBSD, I think...

Concerning SATA, I do not have any problems running my SATA hard drives under FreeBSD 8 using the same board/chipset. However, my SATA DVD writer just completely *beep**beep**beep**beep*s up... but this seems to happen with most SATA DVD hardware under FreeBSD, anyway, and nobody really seems to care (it is not a recommended desktop OS, after all, at least not primarily). A lot of these problems seem to result from incompatibilities with atapicam, but really, DVD drives are negligible from my point of view, so I don't care.

To get this USB thing right, I must say that I do not encounter any problems mounting drives etc... that works fine. But writing to and reading from USB disks causes severe trouble. Writing, e.g. performing a mkfs on a new USB harddisk, just hangs after writing for a second or something - and reading from USB memsticks (e.g. for installing FreeBSD8) goes slower and slower with every read block and finally stops working at all. Feels like something's stuck in the pipeline, if you ask me... Seems to be a chipset bug, though, as some Windows 7 drivers had similar problems which caused, e.g., iTunes to hang while syncing my iPhone using USB ports "powered by crapforce 630i".

Aww, now I'm enraged, I guess I'm gonna buy some Intel mainboard and throw this Gigabyte/nforce-stuff into the deep blue sea. :e

jb_fvwm2 · Jan 12, 2010

Code:

kldstat -v | grep ehci

if echi appears, it is loaded.
I am not sure but you may be able to run
that command before sysintall finishes...
...
re the writing-to-usb issue directly above,
I workaround a similar issue (maybe similar?)
using the "bwlimit" parameter to rsync ( you
can search the forum for that), where writing
to a pci controller-- writes to the sata disk hose the ufs
filesystem if it is done too quickly (controller maybe
not upto it, but works famously as a backup with that
parameter (sata receives at 1/10 the speed of the ata
disk sending, before the controller, *all day* in
the background, another tty !yay! )
...........
It may be useful to test the writes to usb
on that chipset with that parameter. The
cli I use results in the source disk sending
at 10 times the rate of rsync copying to the
target, works without fail so far...
.........
Apologies if it is irrelevant to the posts above.

ktux · Jan 13, 2010

ReverendRed said:
Concerning SATA, I do not have any problems running my SATA hard drives under FreeBSD 8 using the same board/chipset. However, my SATA DVD writer just completely *beep**beep**beep**beep*s up...
:e

I can't get past my SATA DVD writer or USB DVD writer to see if the HDs are actually detected. I'm getting the same old install medium not found on FreeBSD and only FreeBSD. Not being able to run a simple DVD writer off USB is just plain sad.

CrazyAlexNT · Jan 24, 2010

I`m sorry - my English very bad.

I`m have problem to usb flash working motherboard GA-M56S-S3.

Working da0 (flash disk) very slow (freeze)... I`m turn on debug:

Code:

[root@MAX /usr/home/max]# sysctl hw.usb.ehci.no_hs=1
hw.usb.ehci.no_hs: 0 -> 1

Code:

[root@MAX /usr/home/max]# tail -f /var/log/messages
Jan 24 10:39:40 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:40 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:44 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:44 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:48 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:48 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:52 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:52 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:55 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:55 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:59 MAX kernel: ehci_timeout:1553: xfer=0xc743d3c8
Jan 24 10:39:59 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:59 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:39:59 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:39:59 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:40:03 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:40:03 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:40:04 MAX kernel: ehci_timeout:1553: xfer=0xc743d6b8
Jan 24 10:40:04 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:40:04 MAX kernel: ehci_set_hw_power:3842: Async is active
Jan 24 10:40:06 MAX kernel: ehci_root_intr:2040: port 8 changed
Jan 24 10:40:06 MAX kernel: ehci_set_hw_power:3830:
Jan 24 10:40:06 MAX kernel: ehci_set_hw_power:3842: Async is active

Working not good...

I`m change hw.usb.ehci.no_hs=1

Code:

ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:
ehci_set_hw_power:3830:

Work NORMAL (usb 1.0?)...

Motherboard chipset nForce MCP65 (nForce 560)...