Re: [Linux-gpib-general] Petalinux Kernel Compatability

[dave p. <dpe...@gm...>]

OK it is hitting EFAULT at the IBONL ioctl gpib_config. To be sure that
everything is compatible, you can download, build and install the standard
current combined  4.3.6 user and kernel packages from sourceforge
<https://sourceforge.net/projects/linux-gpib/files/latest/download> and see
if the problem persists. This package supports 6.1.xx kernels. BTW the
/dev/gpib* files are set up by the gpib_common module and the access
permissions in the  98-gpib-generic.rules udev script, so there should be
no problem there.
cheers,
-Dave

On Mon, 11 Nov 2024 at 21:43, Michael Jaggers <mgj...@gm...> wrote:

> Yes, there is quite a bit in the console log. I have the debug turned on
> (and added some statements in myself), but this is pretty much what happens
> when I plug in the device:
> [ 9748.976545] usb 1-1.4: new high-speed USB device number 5 using xhci-hcd
> [ 9749.084858] usb 1-1.4: config 1 interface 0 altsetting 0 endpoint 0x81
> has an invalid bInterval 0, changing to 7
> [ 9749.095353] usb 1-1.4: New USB device found, idVendor=3923,
> idProduct=709b, bcdDevice= 1.01
> [ 9749.103769] usb 1-1.4: New USB device strings: Mfr=1, Product=2,
> SerialNumber=3
> [ 9749.111094] usb 1-1.4: Product: GPIB-USB-HS
> [ 9749.115289] usb 1-1.4: Manufacturer: National Instruments
> [ 9749.120690] usb 1-1.4: SerialNumber: 01D9F4A4
> [ 9749.210590] Linux-GPIB 4.3.6 Driver
> [ 9749.386255] ni_usb_gpib driver loading
> [ 9749.402265] ni_usb_gpib: probe succeeded for path:
> usb-xhci-hcd.1.auto-1.4
> [ 9749.414241] usbcore: registered new interface driver ni_usb_gpib
> [ 9749.423413] gpib: registered ni_usb_b interface
> [ 9749.927552] gpib debug: pid 0, gpib: opening minor 0
> [ 9749.935625] gpib debug: pid 0, gpib: request module returned 256
> [ 9749.942207] gpib debug: pid 0, minor 0, ioctl 39, interface=, use=0,
> onl=0, arg=f523e908
> [ 9749.970918] gpib debug: pid 0, minor 0, ioctl 39, interface=, use=0,
> onl=0
> [ 9749.978472] gpib debug: pid 0, gpib: closing minor 0
> [ 9808.492473] gpib debug: pid 0, gpib: opening minor 0
> [ 9808.500324] gpib debug: pid 0, gpib: request module returned 256
> [ 9808.506433] gpib debug: pid 0, minor 0, ioctl 39, interface=, use=0,
> onl=0, arg=fb3f8278
>
> When I try to run "gpib_config", it only shows this line:
> [11325.680770] gpib debug: pid 0, gpib: opening minor 0
> [11325.688909] gpib debug: pid 0, gpib: request module returned 256
> [11325.695036] gpib debug: pid 0, minor 0, ioctl 39, interface=, use=0,
> onl=0, arg=ef9596d8
> [11325.723526] gpib debug: pid 0, minor 0, ioctl 39, interface=, use=0,
> onl=0
> [11325.730940] gpib debug: pid 0, gpib: closing minor 0
>
> Best Regards,
> Michael
>
> On Mon, Nov 11, 2024 at 1:28 PM dave penkler <dpe...@gm...> wrote:
>
>> Hi Michael,
>> The gpib.conf looks perfect. Somewhere during the configure we are
>> getting an EFAULT - Bad address. This would point to some incompatibility
>> between user space gpib_config and the gpib_common kernel module. Is there
>> anything in the console log (dmesg) ?
>> cheers,
>> -dave
>>
>>
>>
>> On Tue, 12 Nov 2024, 05:28 Michael Jaggers, <mgj...@gm...> wrote:
>>
>>> Hey Dave,
>>>
>>> I went back and double checked on the configure argument you asked for,
>>> and I did provide the path to it. I also checked it by putting a syntax
>>> error into the gpib.conf file and that output an error and the path it was
>>> pointing to.
>>>
>>> Here is the conf file that I'm using:
>>>
>>> /* This section configures the configurable driver characteristics
>>>  * for an interface board, such as board address, and interrupt level.
>>>  * minor = 0 configures /dev/gpib0, minor = 1 configures /dev/gpib1,
>>> etc.
>>>  */
>>>
>>> interface {
>>>     minor = 0
>>>     board_type = "ni_usb_gpib"
>>>     name = "gpib0"
>>>     pad = 2
>>>     timeout = T1s
>>>     eos = 0x0A
>>>     set-reos = yes
>>>     set-bin = no
>>>     master = yes
>>>
>>> }
>>>
>>> Thanks!
>>> Michael
>>>
>>> On Sat, Nov 9, 2024 at 2:24 AM dave penkler <dpe...@gm...> wrote:
>>>
>>>> Hi Michael,
>>>> It looks like a gpib.conf problem. Did you provide a sysconfdir
>>>> argument to .configure ?
>>>> Please send me your gpib.conf file.
>>>> cheers,
>>>> -dave
>>>>
>>>> On Sat, 9 Nov 2024, 04:38 Michael Jaggers, <mgj...@gm...> wrote:
>>>>
>>>>> Dave,
>>>>>
>>>>> I've run the gpib_config, but no luck on that end. Here is what
>>>>> happens to me:
>>>>> 1. First, I restart the device without plugging in the gpib (drivers
>>>>> aren't loaded).
>>>>> root@tester0:~/git/linux-gpib-git# gpib_config
>>>>> failed to open device file '/dev/gpib0'
>>>>> main: No such file or directory
>>>>>
>>>>> 2. Now, I plug the gpib dongle in.
>>>>> root@tester0:~/git/linux-gpib-git# gpib_config
>>>>> failed to bring board offline
>>>>> failed to configure board
>>>>> main: Bad address
>>>>>
>>>>> 3. Finally, I unplug the gpib dongle.
>>>>> root@tester0:~/git/linux-gpib-git# gpib_config
>>>>> failed to bring board offline
>>>>> failed to configure board
>>>>> main: Bad address
>>>>>
>>>>> To me, this looks like the dongle isn't attaching to /dev/gpib0. I
>>>>> know the driver recognizes the device, it correctly starts the ni_usb_gpib
>>>>> driver, and probes the correct location on the device tree, and the driver
>>>>> in the device tree points to the correct place the dongle is attached to.
>>>>> What's got me scratching my head is why it's not attached to gpib0 (or even
>>>>> how to circumvent the need to attach to gpib0).
>>>>>
>>>>> Could it be related to my udev? In the rules, in file
>>>>> "/etc/udev/rules.d/98-gpib-generic.rules" I added:
>>>>> ACTION=="add|change",
>>>>> DEVPATH=="/devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4/1-1.4:1.0",
>>>>> ENV{GPIB_CONFIG_OPTIONS}="--minor 0"
>>>>>
>>>>> That's the path that udev recognizes. When I run udevadm monitor, I
>>>>> get this after plugging in the dongle:
>>>>> KERNEL[1122.600699] add
>>>>> /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4
>>>>> (usb)
>>>>> KERNEL[1122.650504] add
>>>>> /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4/1-1.4:1.0
>>>>> (usb)
>>>>> KERNEL[1122.671551] bind
>>>>>  /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4/1-1.4:1.0
>>>>> (usb)
>>>>> KERNEL[1122.671684] bind
>>>>>  /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4
>>>>> (usb)
>>>>> UDEV  [1122.676333] add
>>>>> /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4
>>>>> (usb)
>>>>> UDEV  [1122.769461] add
>>>>> /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4/1-1.4:1.0
>>>>> (usb)
>>>>> UDEV  [1122.772939] bind
>>>>>  /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4/1-1.4:1.0
>>>>> (usb)
>>>>> UDEV  [1122.777553] bind
>>>>>  /devices/platform/axi/ff9d0000.usb/fe200000.usb/xhci-hcd.1.auto/usb1/1-1/1-1.4
>>>>> (usb)
>>>>>
>>>>> Could the version of udev be the cause?
>>>>> root@tester0:~/git/linux-gpib-git# udevd --version
>>>>> 251.8+
>>>>> root@tester0:~/git/linux-gpib-git# udevadm --version
>>>>> 251
>>>>>
>>>>> Thanks for your responses!
>>>>> Michael
>>>>>
>>>>> On Thu, Nov 7, 2024 at 9:04 PM dave penkler <dpe...@gm...>
>>>>> wrote:
>>>>>
>>>>>> It looks like the board is not attached. Did you run # gpib_config ?
>>>>>> Normally the supplied udev scripts do this for you.
>>>>>> cheers,
>>>>>> -dave
>>>>>>
>>>>>> On Fri, 8 Nov 2024, 05:24 Charles Lane, <la...@dc...> wrote:
>>>>>>
>>>>>>> Setting up the /dev/gpib* devices is something that
>>>>>>> you might want to look at udev for...but as I recall
>>>>>>> there's some problems. That isn't necessarily in the
>>>>>>> linux-gpib git project, maybe in linux-gpib-packaging
>>>>>>> project....where the idea is to wrap the linux-gpib
>>>>>>> code in an rpm package (so that installation deals
>>>>>>> with udev scripts, protection issues, etc)
>>>>>>>
>>>>>>> On Thu, 7 Nov 2024 09:51:32 -0700
>>>>>>> Michael Jaggers <mgj...@gm...> wrote:
>>>>>>>
>>>>>>> > Hello,
>>>>>>> >
>>>>>>> > I'm working on installing the gpib kernel drivers to the Petalinux
>>>>>>> > platform. So far, I've managed to get the headers and get the gpib
>>>>>>> > drivers compiled + installed. The modules show up with modprobe,
>>>>>>> and
>>>>>>> > when I plug my gpib dongle in the driver loads properly and it
>>>>>>> seems
>>>>>>> > to recognize it. Below is the message I get when plugging in the
>>>>>>> > dongle: [19163.768240] usb 1-1.4: new high-speed USB device number
>>>>>>> 11
>>>>>>> > using xhci-hcd [19164.020393] usb 1-1.4: config 1 interface 0
>>>>>>> > altsetting 0 endpoint 0x81 has an invalid bInterval 0, changing to
>>>>>>> 7
>>>>>>> > [19164.030876] usb 1-1.4: New USB device found, idVendor=3923,
>>>>>>> > idProduct=709b, bcdDevice= 1.01
>>>>>>> > [19164.039247] usb 1-1.4: New USB device strings: Mfr=1, Product=2,
>>>>>>> > SerialNumber=3
>>>>>>> > [19164.046560] usb 1-1.4: Product: GPIB-USB-HS
>>>>>>> > [19164.050739] usb 1-1.4: Manufacturer: National Instruments
>>>>>>> > [19164.056138] usb 1-1.4: SerialNumber: 01D9F4A4
>>>>>>> >
>>>>>>> > Here is where the issue comes up. When ibopen goes to set things up
>>>>>>> > with the dongle, the /dev/gpib0 doesn't seem to be properly
>>>>>>> > configured. I get this message when I run the ibtest utility:
>>>>>>> > [84391.688217] gpib debug: pid 0, gpib: opening minor 0
>>>>>>> > [84391.696029] gpib debug: pid 0, gpib: request module returned 256
>>>>>>> > [84391.702071] gpib debug: pid 0, minor 0, ioctl 3, interface=,
>>>>>>> use=0,
>>>>>>> > onl=0, arg=d8a171a0
>>>>>>> > [84391.710090] gpib: no gpib board configured on /dev/gpib0
>>>>>>> > [84391.715398] gpib debug: pid 0, minor 0, ioctl 3, interface=,
>>>>>>> > use=0, onl=0 [84391.722212] gpib debug: pid 0, minor 0, ioctl 5,
>>>>>>> > interface=, use=0, onl=0, arg=d8a17120
>>>>>>> > [84391.730218] gpib: no gpib board configured on /dev/gpib0
>>>>>>> > [84391.735527] gpib debug: pid 0, minor 0, ioctl 5, interface=,
>>>>>>> > use=0, onl=0 [84391.742357] gpib debug: pid 0, minor 0, ioctl 5,
>>>>>>> > interface=, use=0, onl=0, arg=d8a17180
>>>>>>> > [84391.750363] gpib: no gpib board configured on /dev/gpib0
>>>>>>> > [84391.755673] gpib debug: pid 0, minor 0, ioctl 5, interface=,
>>>>>>> > use=0, onl=0 [84391.762675] audit: type=1701
>>>>>>> > audit(1730996390.861:20): auid=0 uid=0 gid=0 ses=2 pid=10840
>>>>>>> > comm="ibtest" exe="/usr/local/bin/ibtest" sig=6 res=1
>>>>>>> [84391.764574]
>>>>>>> > gpib debug: pid 0, gpib: closing minor 0
>>>>>>> >
>>>>>>> > I've been debugging this for a week or so now. I'm fairly confident
>>>>>>> > that the probing of the dongle works. In fact, when I find the
>>>>>>> dongle
>>>>>>> > in the sys path, I see that the module driver is in that path of
>>>>>>> the
>>>>>>> > device itself: root@tester0:~# ls -ll
>>>>>>> > /sys/bus/usb/drivers/ni_usb_gpib/1-1.4\:1.0/ total 0
>>>>>>> > -rw-r--r-- 1 root root 4096 Nov  7 16:25 authorized
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bAlternateSetting
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bInterfaceClass
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bInterfaceNumber
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bInterfaceProtocol
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bInterfaceSubClass
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 bNumEndpoints
>>>>>>> > lrwxrwxrwx 1 root root    0 Nov  7 16:25 driver ->
>>>>>>> > ../../../../../../../../../../bus/usb/drivers/ni_usb_gpib
>>>>>>> > drwxr-xr-x 3 root root    0 Nov  7 16:25 ep_02
>>>>>>> > drwxr-xr-x 3 root root    0 Nov  7 16:25 ep_06
>>>>>>> > drwxr-xr-x 3 root root    0 Nov  7 16:25 ep_81
>>>>>>> > drwxr-xr-x 3 root root    0 Nov  7 16:25 ep_84
>>>>>>> > drwxr-xr-x 3 root root    0 Nov  7 16:25 ep_88
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 modalias
>>>>>>> > drwxr-xr-x 2 root root    0 Nov  7 16:25 power
>>>>>>> > lrwxrwxrwx 1 root root    0 Nov  7 16:18 subsystem ->
>>>>>>> > ../../../../../../../../../../bus/usb
>>>>>>> > -r--r--r-- 1 root root 4096 Nov  7 16:25 supports_autosuspend
>>>>>>> > -rw-r--r-- 1 root root 4096 Nov  7 16:18 uevent
>>>>>>> >
>>>>>>> > root@tester0:~# grep ^
>>>>>>> > /sys/bus/usb/drivers/ni_usb_gpib/1-1.4\:1.0/modalias
>>>>>>> > usb:*v3923p709B*d0101dc00dsc00dp00icFFisc00ip00in00
>>>>>>> >
>>>>>>> > The driver itself tells me that the interface registers properly
>>>>>>> too
>>>>>>> > (my apologies for the extra debug messages):
>>>>>>> > [85375.091233] ni_usb_gpib driver loading
>>>>>>> > [85375.098301] ni_usb_driver_probe
>>>>>>> > [85375.101436] set bus interface 0 to address 0x00000000bca009b3
>>>>>>> > [85375.107200] ni_usb_gpib: probe succeeded for path:
>>>>>>> > usb-xhci-hcd.1.auto-1.4
>>>>>>> > [85375.119145] usbcore: registered new interface driver ni_usb_gpib
>>>>>>> > [85375.128308] gpib: registered ni_usb_b interface
>>>>>>> >
>>>>>>> > However, everywhere I look in the code doesn't indicate to me where
>>>>>>> > the disconnect occurs. I'm fairly certain the issue is that the
>>>>>>> > /dev/gpib* section isn't pointing to the dongle area and that's my
>>>>>>> > problem. I've been trying to create a workaround to this that shows
>>>>>>> > the device working at all, but I'm not having any luck there.
>>>>>>> > So I'm at a loss for where to go next. Any suggestions? RIght now,
>>>>>>> I'm
>>>>>>> > trying to figure out why I'm getting "interface=". I feel like
>>>>>>> that's
>>>>>>> > the underlying problem, but I can't confirm it.
>>>>>>> > Just in case, here is some extra information about he system I'm
>>>>>>> on:
>>>>>>> > Operating System: PetaLinux 2023.2+update-61_04172258- (langdale)
>>>>>>> >           Kernel: Linux 6.1.30-xilinx-v2023.2
>>>>>>> >     Architecture: arm64
>>>>>>> >
>>>>>>> > Thanks,
>>>>>>> > Michael
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Drexel University  |/             /    ·   ν ·   Chuck Lane   |D
>>>>>>> ======]－－－>－－－C－－－－－π+－－<                Disque 911  |U
>>>>>>> Particle Physics    \             \    ~~ e+~~   215-895-1545 |N
>>>>>>> la...@dc...                    －​μ+－－+νν                 |E
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Linux-gpib-general mailing list
>>>>>>> Lin...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/linux-gpib-general
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Linux-gpib-general mailing list
>>>>>> Lin...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/linux-gpib-general
>>>>>>
>>>>>