2.5.25 uhci-hcd very bad

public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed

* 2.5.25  uhci-hcd  very bad
@ 2002-07-13 20:14 Pierre Rousselet
  2002-07-17 21:33 ` Greg KH
  0 siblings, 1 reply; 9+ messages in thread
From: Pierre Rousselet @ 2002-07-13 20:14 UTC (permalink / raw)
  To: lkml; +Cc: Greg KH

usb inboard Abit BE6, ADSL modem Speedtouch usb.

I used to use satisfactorily usb-uhci-hcd before 2.5.25, switching to 
uhci-hcd halts the adsl modem after a few minutes with this message:

kernel: uhci-hcd.c: c000: host controller process error. something bad 
happened
kernel: uhci-hcd.c: c000: host controller halted. very bad

2.5.24 is not happy too with uhci-hcd.

What should I look at ?

Pierre
-- 
------------------------------------------------
  Pierre Rousselet <pierre.rousselet@wanadoo.fr>
------------------------------------------------

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-13 20:14 2.5.25 uhci-hcd very bad Pierre Rousselet
@ 2002-07-17 21:33 ` Greg KH
  2002-07-18  6:35   ` Pierre ROUSSELET
  0 siblings, 1 reply; 9+ messages in thread
From: Greg KH @ 2002-07-17 21:33 UTC (permalink / raw)
  To: Pierre Rousselet; +Cc: lkml

On Sat, Jul 13, 2002 at 10:14:40PM +0200, Pierre Rousselet wrote:
> usb inboard Abit BE6, ADSL modem Speedtouch usb.
> 
> I used to use satisfactorily usb-uhci-hcd before 2.5.25, switching to 
> uhci-hcd halts the adsl modem after a few minutes with this message:
> 
> kernel: uhci-hcd.c: c000: host controller process error. something bad 
> happened
> kernel: uhci-hcd.c: c000: host controller halted. very bad
> 
> 2.5.24 is not happy too with uhci-hcd.
> 
> What should I look at ?

Is this the usbfs userspace ASDL driver?  Or the kernel driver?

thanks,

greg k-h

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-17 21:33 ` Greg KH
@ 2002-07-18  6:35   ` Pierre ROUSSELET
  2002-07-18  6:05     ` Greg KH
  2002-07-18  6:10     ` Duncan Sands
  0 siblings, 2 replies; 9+ messages in thread
From: Pierre ROUSSELET @ 2002-07-18  6:35 UTC (permalink / raw)
  To: greg; +Cc: linux-kernel

The driver is made of a kernel module speedtch.o (built outside of the tree) and of userspace modem firmware loader and management daemon speedmgt.

>Messsage du 17/07/2002 23:33
>De : Greg KH <greg@kroah.com>
>A : Pierre Rousselet <pierre.rousselet@wanadoo.fr>
>Copie à : lkml <linux-kernel@vger.kernel.org>
>Objet : Re: 2.5.25  uhci-hcd  very bad  
>
> On Sat, Jul 13, 2002 at 10:14:40PM +0200, Pierre Rousselet wrote:
> > usb inboard Abit BE6, ADSL modem Speedtouch usb.
> > 
> > I used to use satisfactorily usb-uhci-hcd before 2.5.25, switching to 
> > uhci-hcd halts the adsl modem after a few minutes with this message:
> > 
> > kernel: uhci-hcd.c: c000: host controller process error. something bad 
> > happened
> > kernel: uhci-hcd.c: c000: host controller halted. very bad
> > 
> > 2.5.24 is not happy too with uhci-hcd.
> > 
> > What should I look at ?
> 
> Is this the usbfs userspace ASDL driver?  Or the kernel driver?
> 
> thanks,
> 
> greg k-h
>

Pierre


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  6:35   ` Pierre ROUSSELET
@ 2002-07-18  6:05     ` Greg KH
  2002-07-18  7:36       ` Pierre ROUSSELET
  2002-07-18 11:18       ` Peter Osterlund
  2002-07-18  6:10     ` Duncan Sands
  1 sibling, 2 replies; 9+ messages in thread
From: Greg KH @ 2002-07-18  6:05 UTC (permalink / raw)
  To: Pierre ROUSSELET; +Cc: linux-kernel

On Thu, Jul 18, 2002 at 08:35:37AM +0200, Pierre ROUSSELET wrote:
> The driver is made of a kernel module speedtch.o (built outside of the
> tree) and of userspace modem firmware loader and management daemon
> speedmgt.

I'd suggest asking the authors of the driver about this.

Good luck,

greg k-h

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  6:05     ` Greg KH
@ 2002-07-18  7:36       ` Pierre ROUSSELET
  2002-07-18 11:18       ` Peter Osterlund
  1 sibling, 0 replies; 9+ messages in thread
From: Pierre ROUSSELET @ 2002-07-18  7:36 UTC (permalink / raw)
  To: greg; +Cc: linux-kernel

The driver is OK with usb-uhci-hcd up to 2.5.24, you might as well suggest to un-plug it...


>Messsage du 18/07/2002 08:05
>De : Greg KH <greg@kroah.com>
>A : Pierre ROUSSELET <pierre.rousselet@wanadoo.fr>
>Copie à :  <linux-kernel@vger.kernel.org>
>Objet : Re: 2.5.25  uhci-hcd  very bad  
>
> On Thu, Jul 18, 2002 at 08:35:37AM +0200, Pierre ROUSSELET wrote:
> > The driver is made of a kernel module speedtch.o (built outside of the
> > tree) and of userspace modem firmware loader and management daemon
> > speedmgt.
> 
> I'd suggest asking the authors of the driver about this.
> 
> Good luck,
> 
> greg k-h
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  6:05     ` Greg KH
  2002-07-18  7:36       ` Pierre ROUSSELET
@ 2002-07-18 11:18       ` Peter Osterlund
  1 sibling, 0 replies; 9+ messages in thread
From: Peter Osterlund @ 2002-07-18 11:18 UTC (permalink / raw)
  To: Greg KH; +Cc: Pierre ROUSSELET, Johannes Erdfelt, linux-kernel

Greg KH <greg@kroah.com> writes:

> On Thu, Jul 18, 2002 at 08:35:37AM +0200, Pierre ROUSSELET wrote:
> > The driver is made of a kernel module speedtch.o (built outside of the
> > tree) and of userspace modem firmware loader and management daemon
> > speedmgt.
> 
> I'd suggest asking the authors of the driver about this.

I also get the "very bad" message when booting 2.5.25 and 2.5.26, even
if no usb devices are connected. From dmesg:

        Freeing unused kernel memory: 200k freed
        Adding 104824k swap on /dev/hda3.  Priority:-1 extents:1
        usb.c: registered new driver usbfs
        usb.c: registered new driver hub
        uhci-hcd.c: USB Universal Host Controller Interface driver v2.0
        PCI: Assigned IRQ 10 for device 00:07.2
        hcd-pci.c: uhci-hcd @ 00:07.2, Intel Corp. 82371AB PIIX4 USB
        hcd-pci.c: irq 10, io base 0000f300
        hcd.c: new USB bus registered, assigned bus number 1
        uhci-hcd.c: detected 2 ports
        hcd.c: 00:07.2 root hub device address 1
        usb.c: new device strings: Mfr=3, Product=2, SerialNumber=1
        Product: Intel Corp. 82371AB PIIX4 USB
        Manufacturer: Linux 2.5.26-packet uhci-hcd
        SerialNumber: 00:07.2
        hub.c: USB hub found at /
        hub.c: 2 ports detected
        hub.c: standalone hub
        hub.c: ganged power switching
        hub.c: global over-current protection
        hub.c: Port indicators are not supported
        hub.c: power on to power good time: 2ms
        hub.c: hub controller current requirement: 0mA
        hub.c: port removable status: RR
        hub.c: local power source is good
        hub.c: no over-current condition exists
        hub.c: enabling power on all ports
        usb.c: hub driver claimed interface c10ce460
        usb.c: kusbd: /sbin/hotplug add 1
        uhci-hcd.c: f300: suspend_hc
        uhci-hcd.c: f300: wakeup_hc
        uhci-hcd.c: f300: suspend_hc
        uhci-hcd.c: f300: wakeup_hc
        ...                             (usually repeated about 50 times)
        uhci-hcd.c: f300: suspend_hc
        uhci-hcd.c: f300: host controller halted. very bad
        uhci-hcd.c: f300: wakeup_hc

This happens both with preemtible and non-preemtible kernels. I can
still use a Freecom usb cdrw drive though, so maybe this is not a
problem, even though the message indicates that it is.

I also have a performance problem with the uhci-hcd driver. When
writing data to a CDRW in packet writing mode, usb_stor_freecom_reset
is called quite often, see logs below. This happens in both 2.5.25 and
2.5.26. I tested with the usb-uhci-hcd driver in 2.5.25 (edited the
Makefile) just to rule out kernel changes and other usb changes, but
that driver didn't have this problem. I saw one or two freecom reset
messages though, so it's possible that the new driver just makes an
old problem much more frequent.

pengo:/home/petero# tail -f -n 10000 /var/log/debug | egrep 'Command |reset'
        ...
        Jul 18 12:57:04 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:05 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:05 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:05 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:05 pengo kernel: freecom reset called
        Jul 18 12:57:05 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:06 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:06 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:06 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:07 pengo kernel: freecom reset called
        Jul 18 12:57:07 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:08 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:08 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:08 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:08 pengo kernel: freecom reset called
        Jul 18 12:57:08 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:10 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:10 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:10 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:10 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:11 pengo kernel: freecom reset called
        Jul 18 12:57:11 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:13 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:13 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:13 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:13 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:15 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:15 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:16 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:16 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:16 pengo kernel: freecom reset called
        Jul 18 12:57:16 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:17 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:20 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:20 pengo kernel: freecom reset called
        Jul 18 12:57:20 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:20 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:21 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
        Jul 18 12:57:21 pengo kernel: freecom reset called
        ...

>From /var/log/debug:

Jul 18 12:18:11 pengo kernel: usb-storage: queuecommand() called
Jul 18 12:18:11 pengo kernel: usb-storage: *** thread sleeping.
Jul 18 12:18:11 pengo kernel: usb-storage: *** thread awakened.
Jul 18 12:18:11 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
Jul 18 12:18:11 pengo kernel: usb-storage: 2a 00 00 00 1b e0 00 00 20 00 00 00
Jul 18 12:18:11 pengo kernel: usb-storage: Freecom TRANSPORT STARTED
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 2a 00 00 00 1b e0 00 00 - 20 00 00 00             - *....... ...
Jul 18 12:18:11 pengo kernel: usb-storage: foo Status result 0 4
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 4a 00 00 f8                                       - J...
Jul 18 12:18:11 pengo kernel: usb-storage: Device indicates that it has 63488 bytes available
Jul 18 12:18:11 pengo kernel: usb-storage: SCSI requested 65536
Jul 18 12:18:11 pengo kernel: usb-storage: Write data Freecom! (c=65536)
Jul 18 12:18:11 pengo kernel: usb-storage: Done issuing write request: 0 64
Jul 18 12:18:11 pengo kernel: usb-storage: Start of write
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 0
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): xfer 4096 bytes
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_bulk_msg() returned 0 xferred 4096/4096
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): transfer complete
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 4096
... [cut]
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 61440
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): xfer 4096 bytes
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_bulk_msg() returned 0 xferred 4096/4096
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): transfer complete
Jul 18 12:18:11 pengo kernel: usb-storage: freecom_writedata done!
Jul 18 12:18:11 pengo kernel: usb-storage: FCM: Waiting for status
Jul 18 12:18:11 pengo kernel: usb-storage: operation failed
Jul 18 12:18:11 pengo kernel: usb-storage: -- transport indicates command failure
Jul 18 12:18:11 pengo kernel: usb-storage: Issuing auto-REQUEST_SENSE
Jul 18 12:18:11 pengo kernel: usb-storage: Freecom TRANSPORT STARTED
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 03 00 00 00 12 00 00 00 - 20 00 00 00             - ........ ...
Jul 18 12:18:11 pengo kernel: usb-storage: foo Status result 0 4
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 52 03 ff ff                                       - R...
Jul 18 12:18:11 pengo kernel: usb-storage: Device indicates that it has 65535 bytes available
Jul 18 12:18:11 pengo kernel: usb-storage: SCSI requested 18
Jul 18 12:18:11 pengo kernel: usb-storage: Truncating request to match buffer length: 18
Jul 18 12:18:11 pengo kernel: usb-storage: SCSI wants data, drive doesn't have any
Jul 18 12:18:11 pengo kernel: usb-storage: -- auto-sense failure
Jul 18 12:18:11 pengo kernel: freecom reset called
Jul 18 12:18:11 pengo kernel: usb-storage: scsi cmd done, result=0x70000
Jul 18 12:18:11 pengo kernel: usb-storage: queuecommand() called
Jul 18 12:18:11 pengo kernel: usb-storage: *** thread sleeping.
Jul 18 12:18:11 pengo kernel: usb-storage: *** thread awakened.
Jul 18 12:18:11 pengo kernel: usb-storage: Command WRITE_10 (10 bytes)
Jul 18 12:18:11 pengo kernel: usb-storage: 2a 00 00 00 1b e0 00 00 20 00 00 00
Jul 18 12:18:11 pengo kernel: usb-storage: Freecom TRANSPORT STARTED
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 2a 00 00 00 1b e0 00 00 - 20 00 00 00             - *....... ...
Jul 18 12:18:11 pengo kernel: usb-storage: foo Status result 0 4
Jul 18 12:18:11 pengo kernel: usb-storage: 00000000: 4a 00 00 f8                                       - J...
Jul 18 12:18:11 pengo kernel: usb-storage: Device indicates that it has 63488 bytes available
Jul 18 12:18:11 pengo kernel: usb-storage: SCSI requested 65536
Jul 18 12:18:11 pengo kernel: usb-storage: Write data Freecom! (c=65536)
Jul 18 12:18:11 pengo kernel: usb-storage: Done issuing write request: 0 64
Jul 18 12:18:11 pengo kernel: usb-storage: Start of write
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 0
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): xfer 4096 bytes
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_bulk_msg() returned 0 xferred 4096/4096
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): transfer complete
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 4096
... [cut]
Jul 18 12:18:11 pengo kernel: usb-storage: transfer_amount: 65536 and total_transferred: 61440
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): xfer 4096 bytes
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_bulk_msg() returned 0 xferred 4096/4096
Jul 18 12:18:11 pengo kernel: usb-storage: usb_stor_transfer_partial(): transfer complete
Jul 18 12:18:11 pengo kernel: usb-storage: freecom_writedata done!
Jul 18 12:18:11 pengo kernel: usb-storage: FCM: Waiting for status
Jul 18 12:18:12 pengo kernel: usb-storage: Transfer happy
Jul 18 12:18:12 pengo kernel: usb-storage: scsi cmd done, result=0x0
Jul 18 12:18:12 pengo kernel: usb-storage: *** thread sleeping.

-- 
Peter Osterlund - petero2@telia.com
http://w1.894.telia.com/~u89404340

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  6:35   ` Pierre ROUSSELET
  2002-07-18  6:05     ` Greg KH
@ 2002-07-18  6:10     ` Duncan Sands
  2002-07-18  7:37       ` Pierre ROUSSELET
  1 sibling, 1 reply; 9+ messages in thread
From: Duncan Sands @ 2002-07-18  6:10 UTC (permalink / raw)
  To: Pierre ROUSSELET, greg; +Cc: linux-kernel

On Thursday 18 July 2002 08:35, Pierre ROUSSELET wrote:
> The driver is made of a kernel module speedtch.o (built outside of the
> tree) and of userspace modem firmware loader and management daemon
> speedmgt.

Is your kernel preemptive?

Duncan.

^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  6:10     ` Duncan Sands
@ 2002-07-18  7:37       ` Pierre ROUSSELET
  2002-07-18  7:02         ` Duncan Sands
  0 siblings, 1 reply; 9+ messages in thread
From: Pierre ROUSSELET @ 2002-07-18  7:37 UTC (permalink / raw)
  To: duncan.sands, greg; +Cc: linux-kernel

Yes it is.



>Messsage du 18/07/2002 08:10
>De : Duncan Sands <duncan.sands@math.u-psud.fr>
>A : Pierre ROUSSELET <pierre.rousselet@wanadoo.fr>,  <greg@kroah.com>
>Copie à :  <linux-kernel@vger.kernel.org>
>Objet : Re: 2.5.25  uhci-hcd  very bad  
>
> On Thursday 18 July 2002 08:35, Pierre ROUSSELET wrote:
> > The driver is made of a kernel module speedtch.o (built outside of the
> > tree) and of userspace modem firmware loader and management daemon
> > speedmgt.
> 
> Is your kernel preemptive?
> 
> Duncan.
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 


^ permalink raw reply	[flat|nested] 9+ messages in thread

* Re: 2.5.25  uhci-hcd  very bad
  2002-07-18  7:37       ` Pierre ROUSSELET
@ 2002-07-18  7:02         ` Duncan Sands
  0 siblings, 0 replies; 9+ messages in thread
From: Duncan Sands @ 2002-07-18  7:02 UTC (permalink / raw)
  To: Pierre ROUSSELET, greg; +Cc: linux-kernel

In answer to the question, "is your kernel preemptive":

On Thursday 18 July 2002 09:37, Pierre ROUSSELET wrote:
> Yes it is.

Well, the original module code contains several races.  I spent
some time cleaning these up (because it would sometimes oops
on a preemptive system), until I cracked and went over to the user
space driver...  But maybe you are not using the original code?

All the best,

Duncan.

^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2002-07-18 11:19 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-07-13 20:14 2.5.25 uhci-hcd very bad Pierre Rousselet
2002-07-17 21:33 ` Greg KH
2002-07-18  6:35   ` Pierre ROUSSELET
2002-07-18  6:05     ` Greg KH
2002-07-18  7:36       ` Pierre ROUSSELET
2002-07-18 11:18       ` Peter Osterlund
2002-07-18  6:10     ` Duncan Sands
2002-07-18  7:37       ` Pierre ROUSSELET
2002-07-18  7:02         ` Duncan Sands

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox