Felix Miata wrote:

On 2008/04/05 21:14 (GMT+0100) Sid Boyce apparently typed:

...

I saw the problem reported in LKML some weeks ago, but no related kernel patches so far. I also have the same problem on a number of boxes. I messed with kbdrate, but that didn't alleviate the problem.

Any recollection how many weeks? I've not had any luck so far searching for discussion in the lkml archive. Too many addresses and trap screens full of 00000000 strings.

Very occasionally happens on my laptops and quite often on the desktop boxes. 24-31 March, 15 posts, no resoltion or further posts. =========== http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/index.html http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/2567.html http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/2582.html http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/2588.html http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/2588.html Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Felix Miata wrote:

On 2008/04/06 20:38 (GMT+0100) Sid Boyce apparently typed:

...

Felix Miata wrote:

...

...

On 2008/04/05 21:14 (GMT+0100) Sid Boyce apparently typed:

...

...

...

I saw the problem reported in LKML some weeks ago, but no related kernel patches so far. I also have the same problem on a number of boxes. I messed with kbdrate, but that didn't alleviate the problem.

...

...

Any recollection how many weeks? I've not had any luck so far searching for discussion in the lkml archive. Too many addresses and trap screens full of 00000000 strings.

...

Very occasionally happens on my laptops and quite often on the desktop boxes. 24-31 March, 15 posts, no resoltion or further posts. =========== http://www.ussg.iu.edu/hypermail/linux/kernel/0803.3/index.html

OK, thanks. I found that thread on http://lkml.org/lkml/2008/3/29/96 where it's rather longer, but no more help. It doesn't indicate if particular keys are involved, or if it's only zeros, which is all I see. Do any other keys happen for you? What CPUs, CPU speeds, & chipsets do you see it with, and what kernel were you using when if first happened?

As it has been only a minor nuisance, I haven't paid close attention to it. It's never happened with zeros, it's always has been a letter and not the same one either. I can only remember "lllll" from the last time it happened. Laptop CPU 64x2 Mobile Tl-58 1.9Ghz (ATI chipset), laptop Athlon XP3000+ Mobile (VIA VT82xxxx chipset) and 2 boxes with 64x2 6400 - NVidia MCP55 chipsets. I think it was with 2.6.25-rc?, not exactly sure which -rc or -rc-git. Almost certainly happens during software compiling. When next it happens, I shall dig a bit deeper. I have just been trying powersave to see if I could induce a failure at full speed. Getting a hal error, don't know how long it's been around. I shall check bugs to see if it's known and open one if it isn't. Currently the laptops are running at 800Mhz and the desktops at 1000 Mhz. # powersave -A liblazy (liblazy_dbus_send_method_call:97): Received error reply: The name org.freedesktop.Hal was not provided by any .service files liblazy (liblazy_dbus_send_method_call:97): Received error reply: The name org.freedesktop.Hal was not provided by any .service files Could not set CPUFreq policy to dynamic. barrabas:/usr/src/linux-2.6.25-rc8-git4 # rpm -q hal hal-0.5.10_git20080319-4 # rpm -q hal hal-0.5.10_git20080319-7 tindog:/usr/src/linux-2.6.25-rc8-git4 # powersave -A liblazy (liblazy_dbus_send_method_call:97): Received error reply: Method "SetCPUFreqGovernor" with signature "s" on interface "org.freedesktop.Hal.Device.CPUFreq" doesn't exist liblazy (liblazy_dbus_send_method_call:97): Received error reply: Method "SetCPUFreqGovernor" with signature "s" on interface "org.freedesktop.Hal.Device.CPUFreq" doesn't exist Could not set CPUFreq policy to dynamic. sepulot:~ # rpm -q hal hal-0.5.10_git20080319-6 sepulot:~ # powersave -A liblazy (liblazy_dbus_send_method_call:97): Received error reply: Method "SetCPUFreqGovernor" with signature "s" on interface "org.freedesktop.Hal.Device.CPUFreq" doesn't exist liblazy (liblazy_dbus_send_method_call:97): Received error reply: Method "SetCPUFreqGovernor" with signature "s" on interface "org.freedesktop.Hal.Device.CPUFreq" doesn't exist Could not set CPUFreq policy to dynamic. bumble:~ # rpm -q hal hal-0.5.10_git20080319-6 bumble:~ # powersave -f liblazy (liblazy_dbus_send_method_call:97): Received error reply: The name org.freedesktop.Hal was not provided by any .service files Could not set CPUFreq policy to performance. hal-0.5.10_git20080319-6 Boycie:~ # powersave -f liblazy (liblazy_dbus_send_method_call:97): Received error reply: The name org.freedesktop.Hal was not provided by any .service files Could not set CPUFreq policy to performance. Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

...

...

On 4/7/2008 at 02:11, Felix Miata wrote: On 2008/04/04 23:35 (GMT-0400) Felix Miata apparently typed:

...

That's what I get at all prompts on ttys, and in Konsole, and in KDM, and in MC in Konsole. The rate between chars varies, sometimes allowing some short command to be typed and entered in between.

...

It also happens in Mandriva Cooker on same machine, also using 2.6.24.x kernel, but not in DOS, OS/2, 10.2, or Mandriva 2007 on same machine. I tried 2 PS/2 & 1 USB keyboard without change. Machine has NForce2 chipset, mga agp, 1 PATA HD, e100 NIC.

...

I'm still using 2.6.24.1-6-default, but otherwise I've been keeping everything updated at two week or shorter intervals. I currently have another update happening by having written a mini-script while booted to 10.2.

...

Anyone have any idea what could be wrong, or how to fix?

This seems solved by including maxcpus=1 on cmdline.

I would never dare to call maxcpus=1 being a 'solution' to any problem. Everybody going out to buy some C2d, 3core, 4 core cpus and then use maxcpus=1? I hope not! Dominique --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Puh...

...

...

On 4/7/2008 at 14:05, Felix Miata wrote: On 2008/04/07 12:11 (GMT+0200) Dominique Leuenberger apparently typed:

...

...

...

...

On 4/7/2008 at 02:11, Felix Miata wrote:

...

...

This seems solved by including maxcpus=1 on cmdline.

...

I would never dare to call maxcpus=1 being a 'solution' to any problem. Everybody going out to buy some C2d, 3core, 4 core cpus and then use maxcpus=1? I hope not!

Do you know the meaning of "seems"? Did you see anyone using multiple CPUs or multicore CPU reply in thread that he sees the same problem? Did you see anyone reply in thread that he sees the same problem? Did you see anyone but me still using 2.6.24 kernels? Don't you suppose anyone who might see this problem on a multiple core system would rather have a suboptimal system than a non-functional system?

I believe "seems solved" is a clue that this could be a problem only for those with single core/single CPUs, and maybe only for those whose hardware closely matches mine Also it's better than the many threads where problems are described but never have any confirmation or solution or workaround of any kind noted for Googlers who encounter the same or similar problem.

Frankly, this reads quiet heavy and has a very aggressive tone... but let's skip this and get back to normal, shall we? I agree it's a workaround and I agree with you that is has to be stated. But the worst I can EVER see on whatever site / list / forum is talking about 'something being solved' / 'something seeming to be solved' ( which actually just means that 'it looks like solved ro me' and not stating any side effects or telling honestly that he's talking about a work around. Please try to distinguish between solution and workaround.. they're clearly NOT the same. Dominique --------------------------------------------------------------------- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org