Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision? - Portable

This is a discussion on Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision? - Portable ; Hi everybody I have this fairly new Acer TravelMate 290 notebook running kernel 2.4. Most of the hardware work perfectly as described in this page http://perso.wanadoo.es/jtur/AcerTM2...elmate290.html except for two things: 1. WLAN ------- I'm not interested in the build in ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

  1. Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

    Hi everybody

    I have this fairly new Acer TravelMate 290 notebook running
    kernel 2.4.

    Most of the hardware work perfectly as described in this page
    http://perso.wanadoo.es/jtur/AcerTM2...elmate290.html
    except for two things:

    1. WLAN
    -------
    I'm not interested in the build in "Centrino" system, since
    two PCMCIA cards are available. They are Netgear MA401 and
    ZyAIR B-100, both are known to work with prism2_cs driver.

    This notebook won't load the driver 9 out of 10 insertions.
    /var/log/messages logs the following problem:
    Feb 26 08:26:44 matali kernel: prism2_init: dev_info is: prism2_cs
    Feb 26 08:26:44 matali kernel: prism2_cs: RequestIRQ: Resource in use
    Feb 26 08:26:44 matali kernel: prism2sta_config: NextTuple failure? It's
    probably a Vcc mismatch.
    Feb 26 08:26:44 matali kernel: prism2sta_event: prism2_cs:
    Initialization failed!

    I've disabled ACPI at boot time but the problem persists.

    Any clue? The kernel is 2.4.21-192-default from SuSE


    2. External Monitor
    -------------------
    I get a very weak and poor quality screen on the external monitor.
    Again suspected ACPI (see the note from Jürgen Quade in the web
    page mentioned above) but the problem remains even without ACPI.

    Interestingly, Tanguy Carrabin writes that it works fine "using
    Mandrake 9.2 & self-compiled vanilla 2.6.1"

    This is an important issue for me, since notebook is used for
    teaching.

    I would be much thankfull for any suggestions.

    Visvanath
    --
    visvanath at ratnaweera dot net


  2. Re: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

    In comp.os.linux.portable Visvanath Ratnaweera wrote:

    > This notebook won't load the driver 9 out of 10 insertions.
    > /var/log/messages logs the following problem:
    > Feb 26 08:26:44 matali kernel: prism2_init: dev_info is: prism2_cs
    > Feb 26 08:26:44 matali kernel: prism2_cs: RequestIRQ: Resource in use


    Back up and show the startup messages from the PCMCIA drivers. They
    should list what ISA interrupts are thought to be available for PCMCIA
    cards. Also look at /proc/interrupts. By default PCMCIA reserves irq
    4 and irq 7 for serial and parallel ports (in /etc/pcmcia/config.opts)
    so keep that in mind too.

    Do you really mean that you get this message 9 times out of 10, and
    randomly, 1 out of 10 does *not* give a RequestIRQ error?? I do not
    see how that's possible.

    -- Dave

  3. Re: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

    Hi Dave

    dhinds@sonic.net wrote:
    >
    > Back up and show the startup messages from the PCMCIA drivers. They
    > should list what ISA interrupts are thought to be available for PCMCIA


    I've added 20 lines at the bottom.

    > cards. Also look at /proc/interrupts. By default PCMCIA reserves irq
    > 4 and irq 7 for serial and parallel ports (in /etc/pcmcia/config.opts)
    > so keep that in mind too.


    0: 1472534 XT-PIC timer
    1: 9601 XT-PIC keyboard
    2: 0 XT-PIC cascade
    3: 181632 XT-PIC prism2_cs
    5: 0 XT-PIC Intel 82801DB-ICH4
    8: 2 XT-PIC rtc
    9: 88 XT-PIC acpi
    10: 0 XT-PIC ehci_hcd, usb-uhci, usb-uhci
    11: 18 XT-PIC eth0, ohci1394, usb-uhci, PCI device
    1524:1410 (ENE Technology Inc)
    12: 222143 XT-PIC PS/2 Mouse
    14: 39259 XT-PIC ide0
    15: 182 XT-PIC ide1


    > Do you really mean that you get this message 9 times out of 10, and
    > randomly, 1 out of 10 does *not* give a RequestIRQ error?? I do not
    > see how that's possible.


    I can not really reproduce or prove any connections to ACPI.
    The SuSE starting scripts add more to the confusion. For example
    though I disable ACPI giving that as a kernel boot-parameter,
    I have the feeling SuSE still activates ACPI behind the scenes

    Thanks for taking the trouble
    Visvanath.
    --

    /var/log/messages
    Feb 26 08:26:43 matali kernel: spurious 8259A interrupt: IRQ7.
    Feb 26 08:26:43 matali cardmgr[734]: socket 0: Bromax OEM 11Mbps 802.11b
    WLAN Card (Prism 3)
    Feb 26 08:26:43 matali kernel: cs: memory probe 0xa0000000-0xa0ffffff:
    clean.
    Feb 26 08:26:44 matali cardmgr[734]: module
    /lib/modules/2.4.21-192-default//pcmcia/p80211.o not available
    Feb 26 08:26:44 matali cardmgr[734]: executing: 'modprobe -v p80211'
    Feb 26 08:26:44 matali cardmgr[734]: + /sbin/insmod
    /lib/modules/2.4.21-192-default/kernel/drivers/net/wireless/p80211.o
    Feb 26 08:26:44 matali cardmgr[734]: + Using
    /lib/modules/2.4.21-192-default/kernel/drivers/net/wireless/p80211.o
    Feb 26 08:26:44 matali cardmgr[734]: + Symbol version prefix ''
    Feb 26 08:26:44 matali cardmgr[734]: executing: 'insmod -v
    /lib/modules/2.4.21-192-default//pcmcia/prism2_cs.o'
    Feb 26 08:26:44 matali /sbin/hotplug[1236]: no runnable
    /etc/hotplug/wlan.agent is installed
    Feb 26 08:26:44 matali cardmgr[734]: + Using
    /lib/modules/2.4.21-192-default//pcmcia/prism2_cs.o
    Feb 26 08:26:44 matali cardmgr[734]: + Symbol version prefix ''
    Feb 26 08:26:44 matali kernel: prism2_init: prism2_cs.o: 0.2.1-pre9 Loaded
    Feb 26 08:26:44 matali kernel: prism2_init: dev_info is: prism2_cs
    Feb 26 08:26:44 matali kernel: prism2_cs: RequestIRQ: Resource in use
    Feb 26 08:26:44 matali kernel: prism2sta_config: NextTuple failure? It's
    probably a Vcc mismatch.
    Feb 26 08:26:44 matali kernel: prism2sta_event: prism2_cs:
    Initialization failed!
    Feb 26 08:26:44 matali kernel: eth0: no IPv6 routers present
    Feb 26 08:26:45 matali cardmgr[734]: get dev info on socket 0 failed:
    Resource temporarily unavailable
    Feb 26 08:26:45 matali kernel: Unable to handle kernel NULL pointer
    dereference at virtual address 00000000


  4. Re: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

    Visvanath Ratnaweera wrote:
    > Hi Dave


    > 0: 1472534 XT-PIC timer
    > 1: 9601 XT-PIC keyboard
    > 2: 0 XT-PIC cascade
    > 3: 181632 XT-PIC prism2_cs
    > 5: 0 XT-PIC Intel 82801DB-ICH4
    > 8: 2 XT-PIC rtc
    > 9: 88 XT-PIC acpi
    > 10: 0 XT-PIC ehci_hcd, usb-uhci, usb-uhci
    > 11: 18 XT-PIC eth0, ohci1394, usb-uhci, PCI device
    > 1524:1410 (ENE Technology Inc)
    > 12: 222143 XT-PIC PS/2 Mouse
    > 14: 39259 XT-PIC ide0
    > 15: 182 XT-PIC ide1


    Ok, out of all possible irq's, this shows 4, 6, 7, and 13 free. As I
    said, 4 and 7 are excluded in the default /etc/pcmcia/config.opts
    file, but if you do not use one or the other of your serial or
    parallel ports, you can comment out those exclusions, and then those
    irq's will be available for PCMCIA. Irq 6 is usually reserved for a
    floppy drive and 13 is for the math coprocessor, so those are not
    really free. Now, here, irq 3 is allocated for PCMCIA. I suppose
    that the times you get the RequestIRQ error, some other device is
    using that irq.

    >> Do you really mean that you get this message 9 times out of 10, and
    >> randomly, 1 out of 10 does *not* give a RequestIRQ error?? I do not
    >> see how that's possible.


    > I can not really reproduce or prove any connections to ACPI.
    > The SuSE starting scripts add more to the confusion. For example
    > though I disable ACPI giving that as a kernel boot-parameter,
    > I have the feeling SuSE still activates ACPI behind the scenes


    My question didn't have anything to do with ACPI.

    > Feb 26 08:26:44 matali kernel: prism2_init: prism2_cs.o: 0.2.1-pre9 Loaded
    > Feb 26 08:26:44 matali kernel: prism2_init: dev_info is: prism2_cs
    > Feb 26 08:26:44 matali kernel: prism2_cs: RequestIRQ: Resource in use
    > Feb 26 08:26:44 matali kernel: prism2sta_config: NextTuple failure? It's
    > probably a Vcc mismatch.
    > Feb 26 08:26:44 matali kernel: prism2sta_event: prism2_cs:
    > Initialization failed!
    > Feb 26 08:26:44 matali kernel: eth0: no IPv6 routers present
    > Feb 26 08:26:45 matali cardmgr[734]: get dev info on socket 0 failed:
    > Resource temporarily unavailable
    > Feb 26 08:26:45 matali kernel: Unable to handle kernel NULL pointer
    > dereference at virtual address 00000000


    These were not the messages I was asking about, but, you did not
    mention the kernel fault before. That might be a bug in the prism2_cs
    driver.

    -- Dave

  5. Re: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision?

    On Thu, 26 Feb 2004 10:56:54 +0100, Visvanath Ratnaweera wrote:

    > Hi everybody
    >
    > I have this fairly new Acer TravelMate 290 notebook running
    > kernel 2.4.
    >
    > Most of the hardware work perfectly as described in this page
    > http://perso.wanadoo.es/jtur/AcerTM2...elmate290.html
    > except for two things:
    >
    > 1. WLAN
    > -------
    > I'm not interested in the build in "Centrino" system, since
    > two PCMCIA cards are available. They are Netgear MA401 and
    > ZyAIR B-100, both are known to work with prism2_cs driver.
    >

    Why not use the Centrino system. I believe it is supported in the open
    source ndiswrapper code. http://ndiswrapper.sourceforge.net

    --
    Entwisi


  6. Re: Acer TravelMate 290: ACPI, PCMCIA, Ext. Video collision? [50%solved!]

    Hi Dave, hi all

    dhinds@sonic.net wrote:
    > ...
    >
    > Ok, out of all possible irq's, this shows 4, 6, 7, and 13 free. As I
    > said, 4 and 7 are excluded in the default /etc/pcmcia/config.opts
    > file, but if you do not use one or the other of your serial or
    > parallel ports, you can comment out those exclusions, and then those
    > irq's will be available for PCMCIA. Irq 6 is usually reserved for a
    > ...


    Yes, I still had the default config.opts file where IRQ 4 and 7
    were excluded. I freed 4 (no serial ports) and the couple of times
    I inserted the card it worked 100%!

    >
    >>I can not really reproduce or prove any connections to ACPI.
    >>The SuSE starting scripts add more to the confusion. For example
    >>though I disable ACPI giving that as a kernel boot-parameter,
    >>I have the feeling SuSE still activates ACPI behind the scenes

    >
    > My question didn't have anything to do with ACPI.


    I had the fixed idea, that ACPI could hijack an IRQ and cause
    this lock-up. And didn't check the pcmcia/config.opts


    >>Feb 26 08:26:44 matali kernel: prism2_init: prism2_cs.o: 0.2.1-pre9 Loaded
    >>Feb 26 08:26:44 matali kernel: prism2_init: dev_info is: prism2_cs
    >>Feb 26 08:26:44 matali kernel: prism2_cs: RequestIRQ: Resource in use
    >>Feb 26 08:26:44 matali kernel: prism2sta_config: NextTuple failure? It's
    >>probably a Vcc mismatch.
    >>Feb 26 08:26:44 matali kernel: prism2sta_event: prism2_cs:
    >>Initialization failed!
    >>Feb 26 08:26:44 matali kernel: eth0: no IPv6 routers present
    >>Feb 26 08:26:45 matali cardmgr[734]: get dev info on socket 0 failed:
    >>Resource temporarily unavailable
    >>Feb 26 08:26:45 matali kernel: Unable to handle kernel NULL pointer
    >>dereference at virtual address 00000000

    >
    >
    > These were not the messages I was asking about, but, you did not
    > mention the kernel fault before. That might be a bug in the prism2_cs
    > driver.


    I too think "kernel Oops" is not a neat error message

    Thanks a lot for your help!


    I wonder whether anybody has a clue on the second problem:

    2. External Monitor
    -------------------
    I get a very weak and poor quality screen on the external monitor.
    Again suspected ACPI (see the note from Jürgen Quade in the web
    page mentioned above) but the problem remains even without ACPI.

    http://perso.wanadoo.es/jtur/AcerTM2...elmate290.html

    Interestingly, Tanguy Carrabin writes that it works fine "using
    Mandrake 9.2 & self-compiled vanilla 2.6.1"

    This is an important issue for me, since notebook is used for
    teaching.

    I would be much thankfull for any suggestions.

    Visvanath.


+ Reply to Thread