USB problems on 2.6 kernels.
Graeme Nichols
gnichols at tpg.com.au
Sun Aug 8 05:02:01 UTC 2004
Hello Rick, I thought I would update you on my progress with my USB
problems.
The work-around I outlined in my last post seemed reliable enough during
my testing but ultimately I couldn't get the HD to be recognised at all
while the Zip was recognised every time. Very disappointing!
After spending a lot of time googling around the archives I came across
lots of posts about the problem and only one answer that has worked
reliably for several days now, remove the ehci_hcd module before turning
on/plugging in the USB device.
What I am now doing is:-
1. Leave the Zip drive plugged in, no disk loaded, at boot.
2. Leave the HD plugged in, no power on at boot.
3. Leave the scanner plugged in, no power on at boot.
4. /sbin/rmmod ehci_hcd before turning on the HD.
5. This results in the Zip drive being recognised first (sda4) and the
HD being recognised second (sdb1). The important thing is it is
consistent so I know what device I am mounting.
The output of /sbin/lsmod after removing the ehci_hcd module is as
follows:-
[graeme at barney graeme]$ /sbin/lsmod
Module Size Used by
nls_utf8 1536 0
nls_cp437 5376 0
vfat 9984 0
fat 33600 1 vfat
usb_storage 53576 0
sd_mod 16384 0
snd_mixer_oss 13824 1
snd_ens1371 17380 1
snd_rawmidi 17444 1 snd_ens1371
snd_seq_device 6152 1 snd_rawmidi
snd_pcm 69256 1 snd_ens1371
snd_page_alloc 8072 1 snd_pcm
snd_timer 17284 1 snd_pcm
snd_ac97_codec 54788 1 snd_ens1371
gameport 3328 1 snd_ens1371
snd 39396 7
snd_mixer_oss,snd_ens1371,snd_rawmidi,snd_seq_device,snd_pcm,snd_timer,snd_ac97_codec
soundcore 6112 2 snd
ipv6 184672 10
parport_pc 19392 1
lp 8236 0
parport 29640 2 parport_pc,lp
autofs4 12932 0
sunrpc 109924 1
8139too 17920 0
mii 3584 1 8139too
ipt_REJECT 4736 1
ipt_state 1536 2
ip_conntrack 24968 1 ipt_state
iptable_filter 2176 1
ip_tables 13568 3 ipt_REJECT,ipt_state,iptable_filter
floppy 47440 0
sg 27680 0
scsi_mod 91984 3 usb_storage,sd_mod,sg
dm_mod 32800 0
uhci_hcd 24472 0
ext3 103656 2
jbd 40728 1 ext3
[graeme at barney graeme]$
The output of cat /etc/modprobe.conf is as follows:-
[graeme at barney graeme]$ cat /etc/modprobe.conf
# Note: for use under 2.4, changes must also be made to modules.conf!
alias parport_lowlevel parport_pc
alias snd-card-0 snd-ens1371
alias usb-controller1 uhci-hcd
alias usb-controller2 ehci-hcd
install sound-slot-0 /sbin/modprobe --first-time --ignore-install
sound-slot-0 && { /bin/aumix-minimal -f /etc/.aumixrc -L >/dev/null 2>&1
|| :; }
remove sound-slot-0 { /bin/aumix-minimal -f /etc/.aumixrc -S >/dev/null
2>&1 || :; } ; /sbin/modprobe -r --first-time --ignore-remove
sound-slot-0
alias eth0 8139too
[graeme at barney graeme]$
The output of modules.conf is as follows:-
[graeme at barney graeme]$ cat /etc/modules.conf
alias parport_lowlevel parport_pc
alias usb-controller usb-ohci
alias sound-slot-0 es1371
post-install sound-slot-0 /bin/aumix-minimal -f /etc/.aumixrc -L
>/dev/null 2>&1 || :
pre-remove sound-slot-0 /bin/aumix-minimal -f /etc/.aumixrc -S
>/dev/null 2>&1 || :
alias usb-controller1 usb-uhci
alias usb-controller2 ehci-hcd
alias eth0 8139too
options 8139too irq=5
# Note: for use under 2.6, changes must also be made to modprobe.conf!
[graeme at barney graeme]$
Now the question is, can I simply comment out the line in modprobe.conf
to prevent the ehci_hcd module from loading at boot time? or is there
some other way it needs to be done?
As a matter of interest, after removing the ehci_hcd module the output
of /sbin/lsusb has changed as follows:-
[graeme at barney graeme]$ /sbin/lsusb
Bus 002 Device 004: ID 0dc4:0013 Macpower Peripherals, Ltd
Bus 002 Device 003: ID 059b:0033 Iomega Corp.
Bus 002 Device 001: ID 0000:0000
Bus 001 Device 002: ID 04b8:010c Seiko Epson Corp. Perfection 640
Bus 001 Device 001: ID 0000:0000
[graeme at barney graeme]$
I only have 2 busses now instead of 3. There has never been a problem
with the scanner, not being a block device I guess.
If any others are having problems with the 2.6 kernels and USB perhaps
the above is the answer for them as well.
BTW I posted a query regarding 2.6 kernels and USB problems on the
linux-usb-devel list and got the following reply:-
> Is this problem fixed yet in the 2.6 kernels?
None of us is likely to know that answer. Either ask RedHat,
or just try 2.6.8-rc3 yourself.
Will have to query RedHat I guess. In the meantime the above procedure
works well for me at the moment and is painless enough.
--
----------------------------------------------------------------------
Kind regards, Graeme Nichols.
---------------------------------------------------------------------- A
priest asked: What is Fate, Master? And the Master answered: It is that
which gives a beast of burden its reason for existence. It is that which
men in former times had to bear upon their backs. It is that which has
caused nations to build byways from City to City upon which carts and
coaches pass, and alongside which inns have come to be built to stave
off Hunger, Thirst and Weariness. And that is Fate? said the priest.
Fate... I thought you said Freight, responded the Master. That's all
right, said the priest. I wanted to know what Freight was too. -- Kehlog
Albran, "The Profit"
----------------------------------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://listman.redhat.com/archives/redhat-install-list/attachments/20040808/aaacc5ca/attachment.sig>
More information about the Redhat-install-list
mailing list