Parport didn't survive upgrade to 2.6.5
04 Jan 2015 09:55 - 04 Jan 2015 09:57 #54535
by rip1980
Parport didn't survive upgrade to 2.6.5 was created by rip1980
So, been having some nagging issues with my setup and figured it was time to tear down the machine for maint and also rebuild my controller.
I had been running older code and decided to put 2.6.5 on the controller (a dual core 2.7GHz machine) with a NetMos 9900 Parallel Port card. This had worked well in the past.
I should say, this was a clean install, and not a stepped upgrade. I never had to do anything special to it before on old builds.
I've spent a couple days on and off poking at it, and looks like this chip is not getting a driver to stick to it anymore:
lspci:
02:00.0 Parallel controller: NetMos Technology Device 9900 (prog-if 03 [IEEE1284])
Subsystem: Device a000:2000
Flags: fast devsel, IRQ 7
I/O ports at d010 [disabled]
I/O ports at d000 [disabled]
[virtual] Memory at fea01000 (32-bit, non-prefetchable) [size=4K]
[virtual] Memory at fea00000 (32-bit, non-prefetchable) [size=4K]
Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit+
Capabilities: [78] Power Management version 3
Capabilities: [80] Express Legacy Endpoint, MSI 00
Capabilities: [100] Virtual Channel
Capabilities: [800] Advanced Error Reporting
dmesg:
[ 0.212612] pci 0000:02:00.0: [9710:9900] type 00 class 0x070103
[ 0.212640] pci 0000:02:00.0: reg 10: [io 0xd010-0xd017]
[ 0.212662] pci 0000:02:00.0: reg 14: [io 0xd000-0xd007]
[ 0.212683] pci 0000:02:00.0: reg 18: [mem 0xfea01000-0xfea01fff]
[ 0.212731] pci 0000:02:00.0: reg 24: [mem 0xfea00000-0xfea00fff]
[ 0.212781] pci 0000:02:00.0: PME# supported from D3hot D3cold
The last line in LSPCI should be Kernel driver in use: (Probably parport_pc).
There is a driver for it here:
www.asix.com.tw/FrootAttach/driver/AX991...3.0.0_Source.tar.bz2
But have resisted that because I don't know enough about the rtai kernel to blindly start messing with it. Seems to simply patch 8250_pci.c but in confusing language hints that this is already done in newer kernels anyway. (can someone point me at the 8250_pci used by the 2.6.5 live distro build?)
Been reading a lot and spending a lot of time in shell messing with it, so admitting some amount of defeat and posting because I doubt I'm the only one who has run into this. Tried just about all the hints in the wiki (netmos, parallel port, explicitly passing port and irq to parport_pc) and other distros. If I just made an dork out of myself for a trivial fix, it's not for a lack of trying before I post this.
Thanks!
I had been running older code and decided to put 2.6.5 on the controller (a dual core 2.7GHz machine) with a NetMos 9900 Parallel Port card. This had worked well in the past.
I should say, this was a clean install, and not a stepped upgrade. I never had to do anything special to it before on old builds.
I've spent a couple days on and off poking at it, and looks like this chip is not getting a driver to stick to it anymore:
lspci:
02:00.0 Parallel controller: NetMos Technology Device 9900 (prog-if 03 [IEEE1284])
Subsystem: Device a000:2000
Flags: fast devsel, IRQ 7
I/O ports at d010 [disabled]
I/O ports at d000 [disabled]
[virtual] Memory at fea01000 (32-bit, non-prefetchable) [size=4K]
[virtual] Memory at fea00000 (32-bit, non-prefetchable) [size=4K]
Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit+
Capabilities: [78] Power Management version 3
Capabilities: [80] Express Legacy Endpoint, MSI 00
Capabilities: [100] Virtual Channel
Capabilities: [800] Advanced Error Reporting
dmesg:
[ 0.212612] pci 0000:02:00.0: [9710:9900] type 00 class 0x070103
[ 0.212640] pci 0000:02:00.0: reg 10: [io 0xd010-0xd017]
[ 0.212662] pci 0000:02:00.0: reg 14: [io 0xd000-0xd007]
[ 0.212683] pci 0000:02:00.0: reg 18: [mem 0xfea01000-0xfea01fff]
[ 0.212731] pci 0000:02:00.0: reg 24: [mem 0xfea00000-0xfea00fff]
[ 0.212781] pci 0000:02:00.0: PME# supported from D3hot D3cold
The last line in LSPCI should be Kernel driver in use: (Probably parport_pc).
There is a driver for it here:
www.asix.com.tw/FrootAttach/driver/AX991...3.0.0_Source.tar.bz2
But have resisted that because I don't know enough about the rtai kernel to blindly start messing with it. Seems to simply patch 8250_pci.c but in confusing language hints that this is already done in newer kernels anyway. (can someone point me at the 8250_pci used by the 2.6.5 live distro build?)
Been reading a lot and spending a lot of time in shell messing with it, so admitting some amount of defeat and posting because I doubt I'm the only one who has run into this. Tried just about all the hints in the wiki (netmos, parallel port, explicitly passing port and irq to parport_pc) and other distros. If I just made an dork out of myself for a trivial fix, it's not for a lack of trying before I post this.
Thanks!
Last edit: 04 Jan 2015 09:57 by rip1980. Reason: details
Please Log in or Create an account to join the conversation.
04 Jan 2015 12:02 #54540
by PCW
Replied by PCW on topic Parport didn't survive upgrade to 2.6.5
Are you sure a driver is needed?
Linuxcnc doesnt really need a linux driver for the parallel port, _unless_
the parallel port card requires some non-standard poking to be enabled.
(and if this driver is required its only used for setup so is not related to RTAI at all)
What symptom do you have?
Linuxcnc doesnt really need a linux driver for the parallel port, _unless_
the parallel port card requires some non-standard poking to be enabled.
(and if this driver is required its only used for setup so is not related to RTAI at all)
What symptom do you have?
The following user(s) said Thank You: rip1980
Please Log in or Create an account to join the conversation.
04 Jan 2015 13:16 #54541
by rip1980
Replied by rip1980 on topic Parport didn't survive upgrade to 2.6.5
I use a gecko G540 can can't talk to it anymore.
I was able, in the past, to use address 0xd010 and irq=7.
One reason I think it's messed up is this seems new to me:
I/O ports at d010 [disabled]
I/O ports at d000 [disabled]
Disabled?
I was under the impression that parport was needed reading the wiki (admittedly, somewhat dated.) somewhat dated But reading further here , guess I'm wrong (happens a few times a year, off to a great start.)
The NetMos chunk of the wiki talks about [disabled] near the very end. I have not been able to make that work yet.
So anyway, simply, it doesn't seem to be talking anymore. I have lots of test equipment, I'll put a scope on it and see if maybe it's talking gibberish or something as sanity check.
I was able, in the past, to use address 0xd010 and irq=7.
One reason I think it's messed up is this seems new to me:
I/O ports at d010 [disabled]
I/O ports at d000 [disabled]
Disabled?
I was under the impression that parport was needed reading the wiki (admittedly, somewhat dated.) somewhat dated But reading further here , guess I'm wrong (happens a few times a year, off to a great start.)
The NetMos chunk of the wiki talks about [disabled] near the very end. I have not been able to make that work yet.
So anyway, simply, it doesn't seem to be talking anymore. I have lots of test equipment, I'll put a scope on it and see if maybe it's talking gibberish or something as sanity check.
Please Log in or Create an account to join the conversation.
04 Jan 2015 18:38 #54543
by ArcEye
Replied by ArcEye on topic Parport didn't survive upgrade to 2.6.5
Hi
As PCW says, you do not need a driver for a NetMOS card, it will only interfere with Linuxcnc.
What you don't say is what distro you are using? Is it Ubuntu or Debian?
I have had a few instances of unexpected parport behavior since switching to Debian, albeit the opposite to yours.
If it is not blacklisted, it should
Start up your computer and do a lsmod | grep parport in a terminal.
You should have the low level driver parport listed, without it nothing can access the port
The other possibilities are
lp
ppdev and
parport_pc
The nub of the wiki you linked to is that sometimes a parport card comes up with ports shown as disabled.
These can often be enabled by loading the normal printing drivers and then removing them before starting Linuxcnc
So assuming you have the parport driver and nothing else (if not load it and that is your problem )
modprobe lp
modprobe ppdev
modprobe parport_pc
Then lspci -vv | grep Parallel or similar and see if the ports are now enabled
rmmod the modules in the reverse order leaving parport loaded and start Linuxcnc
Let us know how you get on
regards
As PCW says, you do not need a driver for a NetMOS card, it will only interfere with Linuxcnc.
What you don't say is what distro you are using? Is it Ubuntu or Debian?
I have had a few instances of unexpected parport behavior since switching to Debian, albeit the opposite to yours.
The last line in LSPCI should be Kernel driver in use: (Probably parport_pc).
If it is not blacklisted, it should
Start up your computer and do a lsmod | grep parport in a terminal.
You should have the low level driver parport listed, without it nothing can access the port
The other possibilities are
lp
ppdev and
parport_pc
The nub of the wiki you linked to is that sometimes a parport card comes up with ports shown as disabled.
These can often be enabled by loading the normal printing drivers and then removing them before starting Linuxcnc
So assuming you have the parport driver and nothing else (if not load it and that is your problem )
modprobe lp
modprobe ppdev
modprobe parport_pc
Then lspci -vv | grep Parallel or similar and see if the ports are now enabled
rmmod the modules in the reverse order leaving parport loaded and start Linuxcnc
Let us know how you get on
regards
Please Log in or Create an account to join the conversation.
05 Jan 2015 08:29 - 05 Jan 2015 08:38 #54563
by rip1980
Replied by rip1980 on topic Parport didn't survive upgrade to 2.6.5
Thank for your replies and help.
I don't know exactly what it was. I'm thinking a BIOS gremlin?
I've seen weirdness before (I manage >50 racks of equipment in real life) and so thinking that something could be corrupt in NVRAM, did a CMOS CLEAR, new battery and reloaded the whole thing....
No more disabled. Donno know why, I made sure there were no conflicts, etc.
So it's talking again through stepconf. LinuxCNC, just starting to mess with it but appears ok.
Thank you both. Silly parallel port was conspiring against me.
I don't know exactly what it was. I'm thinking a BIOS gremlin?
I've seen weirdness before (I manage >50 racks of equipment in real life) and so thinking that something could be corrupt in NVRAM, did a CMOS CLEAR, new battery and reloaded the whole thing....
No more disabled. Donno know why, I made sure there were no conflicts, etc.
So it's talking again through stepconf. LinuxCNC, just starting to mess with it but appears ok.
Thank you both. Silly parallel port was conspiring against me.
Last edit: 05 Jan 2015 08:38 by rip1980. Reason: detail
Please Log in or Create an account to join the conversation.
Time to create page: 0.073 seconds