Intermittent unability to start PCMCIA at boot (2nd posting)
Hello world, Unfortunately nobody reacted to my previous posting, so I'm trying a second time : During boot of my SuSE 8.2 on a laptop it is sometimes not possible to start my PCMCIA wireless card. In /var/log/messages I see : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' kernel: pcmf502r: RequestIRQ: Resource in use ---------------- I wonder "in use" by what ? When I subsequently do a reboot then sometimes the same thing happens again, and some other time all works well the second boot. If it works, I see in in /var/log/messages : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' cardmgr[1124]: executing: './network start eth1' /etc/hotplug/net.agent[1135]: Interface eth1 is handled by service PCMCIA ... exiting ----------------- If it didn't work the second time, then the third reboot ALWAYS goes well ! So apparently after two or three times the IRQ blocking resource has evaporated. Please note, that many times the first boot is OK right away. Any ideas what is happening here and how to prevent this ? TIA and greetings, -- Jan Elders Nuenen the Netherlands
On Thursday 07 August 2003 04:30 pm, Jan Elders wrote:
During boot of my SuSE 8.2 on a laptop it is sometimes not possible to start my PCMCIA wireless card. In /var/log/messages I see : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' kernel: pcmf502r: RequestIRQ: Resource in use <snip>
I'm no expert, but I can offer some paths to follow. For one, when the system is up look at the /proc/interrupts file. That will tell you the IRQ number that the PCMCIA card is using, as well as all the other IRQ's in use. Then what you could try is to edit the /etc/sysconfig/pcmcia file to force it to use a certain IRQ (preferably one that isn't already being used!) I'm not positive this will work, but I know that I have to force my PCMCIA to use IRQ 11 on my laptop otherwise it won't work at all. Good Luck, Mike Sacco
On Friday 08 August 2003 00:42, Michael Sacco wrote:
On Thursday 07 August 2003 04:30 pm, Jan Elders wrote:
During boot of my SuSE 8.2 on a laptop it is sometimes not possible to start my PCMCIA wireless card. In /var/log/messages I see : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' kernel: pcmf502r: RequestIRQ: Resource in use
<snip>
I'm no expert, but I can offer some paths to follow. For one, when the system is up look at the /proc/interrupts file. /proc/interrupts ????????????? That file is empty on my system ! In fact most of the files in /proc are empty. There is only some contents in : config.gz , kcore and mtrr. Would this be a major flaw in my installation ? It is in general running pretty smooth, though.
That will tell you the IRQ number that the PCMCIA card is using, as well as all the other IRQ's in use. Then what you could try is to edit the /etc/sysconfig/pcmcia file to force it to use a certain IRQ (preferably one that isn't already being used!) I'm not positive this will work, but I know that I have to force my PCMCIA to use IRQ 11 on my laptop otherwise it won't work at all.
Good Luck, Mike Sacco
-- Jan Elders Nuenen the Netherlands
Question: are you sure ACPI is running on your laptop? (assuming of course that your laptop provides ACPI support.) What model laptop is is, and how old? On Thursday 07 August 2003 12:30, Jan Elders wrote:
Hello world, Unfortunately nobody reacted to my previous posting, so I'm trying a second time :
During boot of my SuSE 8.2 on a laptop it is sometimes not possible to start my PCMCIA wireless card. In /var/log/messages I see : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' kernel: pcmf502r: RequestIRQ: Resource in use ---------------- I wonder "in use" by what ?
When I subsequently do a reboot then sometimes the same thing happens again, and some other time all works well the second boot. If it works, I see in in /var/log/messages : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' cardmgr[1124]: executing: './network start eth1' /etc/hotplug/net.agent[1135]: Interface eth1 is handled by service PCMCIA ... exiting -----------------
If it didn't work the second time, then the third reboot ALWAYS goes well ! So apparently after two or three times the IRQ blocking resource has evaporated. Please note, that many times the first boot is OK right away.
Any ideas what is happening here and how to prevent this ? TIA and greetings, -- Jan Elders Nuenen the Netherlands
-- _____________________________________ John Andersen
On Friday 08 August 2003 07:46, John Andersen wrote:
Question: are you sure ACPI is running on your laptop? (assuming of course that your laptop provides ACPI support.) The answer is yes. But why do you ask ? Has ACPI something to do with it ?
Now you mention this, I notice that right behind the pcmf502r error-message there are some ACPI related messages : kernel : ACPI: AC Adapter [ACAD] (on-line) kernel: ACPI: Fan [FN1] (off) kernel: ACPI: Fan [FN2] (off) kernel: ACPI: Processor [CPU0] (supports C1 C2, 2 throttling states) insmod: /lib/modules/2.4.20-4GB/kernel/drivers/acpi/toshiba_acpi.o: init_module: No such device Could ACPI be causing the conflict - randomly ?
What model laptop is it, and how old?
It is a MEDION (German brand) with an Intel Pentium 4 2.6GHz It is just a few months old.
//snip
During boot of my SuSE 8.2 on a laptop it is sometimes not possible to start my PCMCIA wireless card. In /var/log/messages I see : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' kernel: pcmf502r: RequestIRQ: Resource in use ---------------- I wonder "in use" by what ?
When I subsequently do a reboot then sometimes the same thing happens again, and some other time all works well the second boot. If it works, I see in in /var/log/messages : ----------------- cardmgr[1124]: socket 0: SITECOM WL-011 Wireless Network PC Card cardmgr[1124]: executing: 'insmod -v /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o' cardmgr[1124]: + Using /lib/modules/2.4.20-4GB//pcmcia/pcmf502r.o cardmgr[1124]: + Symbol version prefix '' cardmgr[1124]: executing: './network start eth1' /etc/hotplug/net.agent[1135]: Interface eth1 is handled by service PCMCIA ... exiting ----------------- //snip
-- Jan Elders Nuenen the Netherlands
On Friday 08 August 2003 05:15, Jan Elders wrote:
Question: are you sure ACPI is running on your laptop? (assuming of course that your laptop provides ACPI support.)
The answer is yes. But why do you ask ? Has ACPI something to do with it ?
Now you mention this, I notice that right behind the pcmf502r error-message there are some ACPI related messages : kernel : ACPI: AC Adapter [ACAD] (on-line) kernel: ACPI: Fan [FN1] (off) kernel: ACPI: Fan [FN2] (off) kernel: ACPI: Processor [CPU0] (supports C1 C2, 2 throttling states) insmod: /lib/modules/2.4.20-4GB/kernel/drivers/acpi/toshiba_acpi.o: init_module: No such device Could ACPI be causing the conflict - randomly ?
What model laptop is it, and how old?
It is a MEDION (German brand) with an Intel Pentium 4 2.6GHz It is just a few months old.
The reason I ask is that pcmcia requires an interrupt and acpi is the mechanism by which these are assigned in modern machines, especially laptops. SuSE 8.2 usually detects acpi capable machines correctly, and it looks like yours was properly detectes and pcmcia started. That part's fine. You could boot with acpi turned off, but I suggest you not mess with the acpi, lots of hardware might stop working without it. But there is another item to check that just came to mind. In one of the SuSE releases (i think it was 8.2) where I discovered that SuSE installs with external pcmcia drivers, but then switches to kernel drivers at first boot. (Why? - No Clue). I had a nic that did not like the kernel drivers but I new it worked because it was there during install. Go to /etc/sysconfig/pcmcia and see which driver type you are useing (first paragraph). Swap to the othe and see if that helps. You can also set which interrupts the pcmcia code will try to use by adding these to the list... PCMCIA_PCIC_OPTS="pci_int=1 pci_csc=1 irq_list=9,10" German Laptop, German Distro... It oughta work!!! -- _____________________________________ John Andersen
On Saturday 09 August 2003 09:21, John Andersen wrote:
On Friday 08 August 2003 05:15, Jan Elders wrote:
Question: are you sure ACPI is running on your laptop? (assuming of course that your laptop provides ACPI support.)
The answer is yes. But why do you ask ? Has ACPI something to do with it ?
Now you mention this, I notice that right behind the pcmf502r error-message there are some ACPI related messages : kernel : ACPI: AC Adapter [ACAD] (on-line) kernel: ACPI: Fan [FN1] (off) kernel: ACPI: Fan [FN2] (off) kernel: ACPI: Processor [CPU0] (supports C1 C2, 2 throttling states) insmod: /lib/modules/2.4.20-4GB/kernel/drivers/acpi/toshiba_acpi.o: init_module: No such device Could ACPI be causing the conflict - randomly ?
//snip The reason I ask is that pcmcia requires an interrupt and acpi is the mechanism by which these are assigned in modern machines, especially laptops.
SuSE 8.2 usually detects acpi capable machines correctly, and it looks like yours was properly detectes and pcmcia started. That part's fine.
You could boot with acpi turned off, but I suggest you not mess with the acpi, lots of hardware might stop working without it.
Are you sure you mean acpi - the power management ? Or do you mean apic ? I have apic turned off in my bootloader ( noapic ). Greetings, -- Jan Elders Nuenen the Netherlands
On Sunday 10 August 2003 07:40, Jan Elders wrote:
Are you sure you mean acpi - the power management ? Or do you mean apic ?
No, I mean acpi. Take a look at /proc/acpi some time, there's lost of stuff in there to play with...., Cpu speets, battery charge states, Chip Temperatures, etc. Modern laptops are pretty dependent on acpi. -- _____________________________________ John Andersen
On Monday 11 August 2003 10:32, John Andersen wrote:
On Sunday 10 August 2003 07:40, Jan Elders wrote:
Are you sure you mean acpi - the power management ? Or do you mean apic ?
No, I mean acpi. Take a look at /proc/acpi some time, there's lost of stuff in there to play with...., Cpu speets, battery charge states, Chip Temperatures, etc. Modern laptops are pretty dependent on acpi.
John, /proc/acpi is empty on my system ! In fact most of the files in /proc are empty. There is only some contents in : config.gz , kcore and mtrr. Would this be a major flaw in my installation ? It is in general running pretty smooth, though. (This Linux system keeps puzzling me) Greetings, -- Jan Elders Nuenen the Netherlands
On Saturday 09 August 2003 09:21, John Andersen wrote: <snip>
But there is another item to check that just came to mind. In one of the SuSE releases (i think it was 8.2) where I discovered that SuSE installs with external pcmcia drivers, but then switches to kernel drivers at first boot. (Why? - No Clue). I had a nic that did not like the kernel drivers but I new it worked because it was there during install.
Go to /etc/sysconfig/pcmcia and see which driver type you are useing (first paragraph). Swap to the other and see if that helps.
You can also set which interrupts the pcmcia code will try to use by adding these to the list... PCMCIA_PCIC_OPTS="pci_int=1 pci_csc=1 irq_list=9,10"
Well, I just changed /etc/sysconfig/pcmcia : PCMCIA_SYSTEM="external" (was "kernel") PCMCIA_PCIC_OPTS="pci_int=1 pci_csc=1 irq_list=9" But no, after the first boot, I got again the errormessage that the resource was in use. cat /proc/interrupts tells me : CPU0 0: 12660 XT-PIC timer 1: 83 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 691 XT-PIC ehci-hcd, usb-ohci, usb-ohci, usb-ohci, SiS SI7012 8: 2 XT-PIC rtc 10: 1 XT-PIC ohci1394, PCI device 1524:1410 (ENE Technology Inc) 11: 0 XT-PIC acpi 12: 1995 XT-PIC PS/2 Mouse 14: 8281 XT-PIC ide0 15: 369 XT-PIC ide1 NMI: 0 LOC: 0 ERR: 0 MIS: 0 After the second or third reboot, it suddenly does work. cat /proc/interrupts now tells me : CPU0 0: 11748 XT-PIC timer 1: 97 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 668 XT-PIC ehci-hcd, usb-ohci, usb-ohci, usb-ohci, SiS SI7012 8: 2 XT-PIC rtc 9: 1364 XT-PIC pcmf502r 10: 1 XT-PIC ohci1394, PCI device 1524:1410 (ENE Technology Inc) 11: 0 XT-PIC acpi 12: 441 XT-PIC PS/2 Mouse 14: 8719 XT-PIC ide0 15: 329 XT-PIC ide1 NMI: 0 LOC: 0 ERR: 1 MIS: 0 So,apparently, IRQ 9 is being used, but why did it fail at the first attempt ? Any ideas/suggestions ? Greetings, -- Jan Elders Nuenen the Netherlands
Hello world, Anybody out there who has a clue what might be going on with the variations in my PCMCIA startup events ? On Sunday 17 August 2003 15:34, Jan Elders wrote:
On Saturday 09 August 2003 09:21, John Andersen wrote: <snip>
But there is another item to check that just came to mind. In one of the SuSE releases (i think it was 8.2) where I discovered that SuSE installs with external pcmcia drivers, but then switches to kernel drivers at first boot. (Why? - No Clue). I had a nic that did not like the kernel drivers but I new it worked because it was there during install.
Go to /etc/sysconfig/pcmcia and see which driver type you are useing (first paragraph). Swap to the other and see if that helps.
You can also set which interrupts the pcmcia code will try to use by adding these to the list... PCMCIA_PCIC_OPTS="pci_int=1 pci_csc=1 irq_list=9,10"
Well, I just changed /etc/sysconfig/pcmcia : PCMCIA_SYSTEM="external" (was "kernel") PCMCIA_PCIC_OPTS="pci_int=1 pci_csc=1 irq_list=9"
But no, after the first boot, I got again the errormessage that the resource was in use. cat /proc/interrupts tells me : CPU0 0: 12660 XT-PIC timer 1: 83 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 691 XT-PIC ehci-hcd, usb-ohci, usb-ohci, usb-ohci, SiS SI7012 8: 2 XT-PIC rtc 10: 1 XT-PIC ohci1394, PCI device 1524:1410 (ENE Technology Inc) 11: 0 XT-PIC acpi 12: 1995 XT-PIC PS/2 Mouse 14: 8281 XT-PIC ide0 15: 369 XT-PIC ide1 NMI: 0 LOC: 0 ERR: 0 MIS: 0
After the second or third reboot, it suddenly does work. cat /proc/interrupts now tells me : CPU0 0: 11748 XT-PIC timer 1: 97 XT-PIC keyboard 2: 0 XT-PIC cascade 5: 668 XT-PIC ehci-hcd, usb-ohci, usb-ohci, usb-ohci, SiS SI7012 8: 2 XT-PIC rtc 9: 1364 XT-PIC pcmf502r 10: 1 XT-PIC ohci1394, PCI device 1524:1410 (ENE Technology Inc) 11: 0 XT-PIC acpi 12: 441 XT-PIC PS/2 Mouse 14: 8719 XT-PIC ide0 15: 329 XT-PIC ide1 NMI: 0 LOC: 0 ERR: 1 MIS: 0
So,apparently, IRQ 9 is being used, but why did it fail at the first attempt ? It seems to be free there. Any ideas/suggestions ?
TIA -- Jan Elders Nuenen the Netherlands
On Friday 08 August 2003 05:15, Jan Elders wrote:
insmod: /lib/modules/2.4.20-4GB/kernel/drivers/acpi/toshiba_acpi.o: init_module: No such device
Don't worry about the above bit, Its not a toshiba. My Sony throws that error too... -- _____________________________________ John Andersen
participants (3)
-
Jan Elders -
John Andersen -
Michael Sacco