skge driver, 2.6.12-9-k7, Asus A7V600

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

skge driver, 2.6.12-9-k7, Asus A7V600

Anders Karlsson-2
Hi there,

In Breezy, the skge driver is overriding the sk98lin driver. I believe
that may be a mistake. On my A7V600 motherboard I have experienced
hard (keyboard unresponsive, SysRq not working) lockups. Upgrading the
BIOS of the motherboard did nothing to aleviate the situation.

If I can get my hands on a null modem cable, I can potentially debug
these crashes.

In the meantime, you may want to consider giving the sk98lin driver
priority over the skge driver.

Regards,

--
Anders Karlsson <[hidden email]>

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Sebastian Dröge-3
On So, 2005-11-13 at 17:45 +0000, Anders Karlsson wrote:

> Hi there,
>
> In Breezy, the skge driver is overriding the sk98lin driver. I believe
> that may be a mistake. On my A7V600 motherboard I have experienced
> hard (keyboard unresponsive, SysRq not working) lockups. Upgrading the
> BIOS of the motherboard did nothing to aleviate the situation.
>
> If I can get my hands on a null modem cable, I can potentially debug
> these crashes.
>
> In the meantime, you may want to consider giving the sk98lin driver
> priority over the skge driver.
Well, the skge driver works fine here with the same board and nic... are
you sure it's caused by that driver? did you already try switching back
to sk98lin on the same kernel?

Bye

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel

signature.asc (196 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Anders Karlsson-2
On 11/14/05, Sebastian Dröge <[hidden email]> wrote:

> On So, 2005-11-13 at 17:45 +0000, Anders Karlsson wrote:
> > Hi there,
> >
> > In Breezy, the skge driver is overriding the sk98lin driver. I believe
> > that may be a mistake. On my A7V600 motherboard I have experienced
> > hard (keyboard unresponsive, SysRq not working) lockups. Upgrading the
> > BIOS of the motherboard did nothing to aleviate the situation.
> >
> > If I can get my hands on a null modem cable, I can potentially debug
> > these crashes.
> >
> > In the meantime, you may want to consider giving the sk98lin driver
> > priority over the skge driver.
>
> Well, the skge driver works fine here with the same board and nic... are
> you sure it's caused by that driver? did you already try switching back
> to sk98lin on the same kernel?
Interesting. I have a Silicon Image PCI0680 IDE card, a LSI/Symbios
53C896/897 card and a nVidia NV38 card added to the system. The system
had worked perfectly on Hoary.

When I started copying 3+ GB of ISO's across that NIC the box would
hang after approx a gig of traffic. On the 2.6.14.2 kernel I compiled
with sk98lin, I have copied the same 3+ GB of ISO's, 4+ GB of ogg's
and mp3's as well as a bit over a gig of images from my digital
camera. No lockup so far.

On 2.6.12-9-k7, it did not work putting "skge" in
/etc/hotplug/blacklist to prevent it loading, and even after unloading
skge, "modprobe sk98lin" would not load the sk98lin driver. insmod
with a full path would do it however.

As the box in question is running multiple services for me I'd prefer
it to be stable, and I appear to have achieved that with 2.6.14.2.
Once I have located a null modem cable, I will boot 2.6.12-9-k7 again,
attempt to crash it and submit a bug.

Regards,

--
Anders Karlsson <[hidden email]>

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Stephen Hemminger
In reply to this post by Anders Karlsson-2
On Sun, 13 Nov 2005 17:45:17 +0000
Anders Karlsson <[hidden email]> wrote:

> Hi there,
>
> In Breezy, the skge driver is overriding the sk98lin driver. I believe
> that may be a mistake. On my A7V600 motherboard I have experienced
> hard (keyboard unresponsive, SysRq not working) lockups. Upgrading the
> BIOS of the motherboard did nothing to aleviate the situation.

It just happens that way since both drivers support the same hardware.

How soon does the crash happen, right away or only under load.

> If I can get my hands on a null modem cable, I can potentially debug
> these crashes.

Please report driver problems to me (the maintainer) and [hidden email].

Also include lspci and console boot output. It really helps to
know the chip version and revision.

> In the meantime, you may want to consider giving the sk98lin driver
> priority over the skge driver.

There is no easy way to give priority, it is an either or situation.
 
Also the long term plan for 2.6 mainline kernel is to deprecate the sk98lin driver.

--
Stephen Hemminger <[hidden email]>
OSDL http://developer.osdl.org/~shemminger

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Anders Karlsson-2
On 11/15/05, Stephen Hemminger <[hidden email]> wrote:

[skge vs sk98lin, lockup, A7V600]

> It just happens that way since both drivers support the same hardware.

I do realise that. :-)

> How soon does the crash happen, right away or only under load.

It can be provoked fairly fast under load, an hour or so, but just
leaving the interface configured and the box alone causes lockups as
well. It just takes longer, couple of days, maybe a week.

> > If I can get my hands on a null modem cable, I can potentially debug
> > these crashes.
>
> Please report driver problems to me (the maintainer) and [hidden email].

Will do.

> Also include lspci and console boot output. It really helps to
> know the chip version and revision.

The lspci output is below, once I can capture any Oops, I will attach
boot logs as well.

0000:00:09.0 Ethernet controller: 3Com Corporation 3c940
10/100/1000Base-T [Marvell] (rev 12)
        Subsystem: Asustek Computer, Inc. P4P800 Mainboard
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop-
ParErr- Stepping- SERR- FastB2B-
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium
>TAbort- <TAbort- <MAbort- >SERR- <PERR-
        Latency: 32 (5750ns min, 7750ns max), Cache Line Size: 0x08 (32 bytes)
        Interrupt: pin A routed to IRQ 20
        Region 0: Memory at cd800000 (32-bit, non-prefetchable) [size=16K]
        Region 1: I/O ports at d800 [size=256]
        Capabilities: [48] Power Management version 2
                Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA
PME(D0+,D1+,D2+,D3hot+,D3cold+)
                Status: D0 PME-Enable- DSel=0 DScale=1 PME-
        Capabilities: [50] Vital Product Data


> > In the meantime, you may want to consider giving the sk98lin driver
> > priority over the skge driver.
>
> There is no easy way to give priority, it is an either or situation.

Understood, I was just floating an idea that worked for me. Once I
compiled 2.6.14.2, I could compile in just the driver I wanted, and
leave out what I did not need/want.

> Also the long term plan for 2.6 mainline kernel is to deprecate the sk98lin driver.

What type of timescale are we talking about? Weeks, months or year+?

--
Anders Karlsson <[hidden email]>

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Stephen Hemminger
Since the skge driver uses NAPI and the sk98lin does not,
make sure that the BIOS has correctly the interrupt to level triggered.

You can check by looking at /proc/interrupts


--
Stephen Hemminger <[hidden email]>
OSDL http://developer.osdl.org/~shemminger

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Anders Karlsson-2
On 11/15/05, Stephen Hemminger <[hidden email]> wrote:
> Since the skge driver uses NAPI and the sk98lin does not,
> make sure that the BIOS has correctly the interrupt to level triggered.
>
> You can check by looking at /proc/interrupts

Well, with sk98lin loaded, /proc/interrupts looks like this:

anders@lenin:~$ cat /proc/interrupts
           CPU0
  0:   55022946    IO-APIC-edge  timer
  1:      29087    IO-APIC-edge  i8042
  4:     150418    IO-APIC-edge  serial
  8:          4    IO-APIC-edge  rtc
  9:          0   IO-APIC-level  acpi
 12:      11109    IO-APIC-edge  i8042
 14:    1975722    IO-APIC-edge  ide2
 16:   21033710   IO-APIC-level  ide0, ide1
 17:          0   IO-APIC-level  libata
 18:         31   IO-APIC-level  sym53c8xx
 19:   16807057   IO-APIC-level  sym53c8xx, nvidia
 20:   15734609   IO-APIC-level  SysKonnect SK-98xx
 21:   44503739   IO-APIC-level  uhci_hcd:usb1, uhci_hcd:usb2,
uhci_hcd:usb3, uhci_hcd:usb4, ehci_hcd:usb5
 22:     229705   IO-APIC-level  VIA8237
NMI:          0
LOC:   55010327
ERR:          0
MIS:          0

If skge was loaded instead, I would assume it would still be level, right?


--
Anders Karlsson <[hidden email]>

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Stephen Hemminger
On Wed, 16 Nov 2005 00:28:15 +0000
Anders Karlsson <[hidden email]> wrote:

> On 11/15/05, Stephen Hemminger <[hidden email]> wrote:
> > Since the skge driver uses NAPI and the sk98lin does not,
> > make sure that the BIOS has correctly the interrupt to level triggered.
> >
> > You can check by looking at /proc/interrupts
>
> Well, with sk98lin loaded, /proc/interrupts looks like this:
>
> anders@lenin:~$ cat /proc/interrupts
>            CPU0
>   0:   55022946    IO-APIC-edge  timer
>   1:      29087    IO-APIC-edge  i8042
>   4:     150418    IO-APIC-edge  serial
>   8:          4    IO-APIC-edge  rtc
>   9:          0   IO-APIC-level  acpi
>  12:      11109    IO-APIC-edge  i8042
>  14:    1975722    IO-APIC-edge  ide2
>  16:   21033710   IO-APIC-level  ide0, ide1
>  17:          0   IO-APIC-level  libata
>  18:         31   IO-APIC-level  sym53c8xx
>  19:   16807057   IO-APIC-level  sym53c8xx, nvidia
>  20:   15734609   IO-APIC-level  SysKonnect SK-98xx
>  21:   44503739   IO-APIC-level  uhci_hcd:usb1, uhci_hcd:usb2,
> uhci_hcd:usb3, uhci_hcd:usb4, ehci_hcd:usb5
>  22:     229705   IO-APIC-level  VIA8237
> NMI:          0
> LOC:   55010327
> ERR:          0
> MIS:          0
>
> If skge was loaded instead, I would assume it would still be level, right?

Yes it is controlled by BIOS/ACPI, not driver

--
Stephen Hemminger <[hidden email]>
OSDL http://developer.osdl.org/~shemminger

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Anders Karlsson-2
Good morning,

[Re: skge, A7V600, lockups]

On 11/16/05, Stephen Hemminger <[hidden email]> wrote:
> Yes it is controlled by BIOS/ACPI, not driver

After several days of crashes, lots of swearing and troubleshooting
blind, friday morning resulted in me tweaking yet another thing to see
if it helped. Switched from the NVIDIA proprietary driver to the open
'nv' driver. System has been stable since.

Many apologies for pointing the finger at skge when I should have
pointed it at NVIDIA. I have this morning in the post received the
null modem cable, so I can now troubleshoot things with more detail.

Have a nice weekend all,

--
Anders Karlsson <[hidden email]>

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: skge driver, 2.6.12-9-k7, Asus A7V600

Stephen Hemminger
On Sat, 19 Nov 2005 08:04:58 +0000
Anders Karlsson <[hidden email]> wrote:

> Good morning,
>
> [Re: skge, A7V600, lockups]
>
> On 11/16/05, Stephen Hemminger <[hidden email]> wrote:
> > Yes it is controlled by BIOS/ACPI, not driver
>
> After several days of crashes, lots of swearing and troubleshooting
> blind, friday morning resulted in me tweaking yet another thing to see
> if it helped. Switched from the NVIDIA proprietary driver to the open
> 'nv' driver. System has been stable since.
>
> Many apologies for pointing the finger at skge when I should have
> pointed it at NVIDIA. I have this morning in the post received the
> null modem cable, so I can now troubleshoot things with more detail.
>

Thanks, for following up, it makes one less thing for me to worry about.

--
Stephen Hemminger <[hidden email]>
OSDL http://developer.osdl.org/~shemminger

--
ubuntu-devel mailing list
[hidden email]
http://lists.ubuntu.com/mailman/listinfo/ubuntu-devel