Re: [Emc-users] Problem with the new hm2_pci write function
Peter C. Wallace wrote: > On Wed, 19 Nov 2008, Sebastian Kuzminsky wrote: > >> If you can't or don't want to compile from source, you can work around >> the bug by enabling one pwmgen instance in your config modparam. >> Assuming, of course, that you dont need those IO pins for GPIO... > > I can also make a stepper only HM2 config if that will sidestep the bug > (will it Sebastian?) Unfortunately not. -- Sebastian Kuzminsky how many no money boys are crazy how many boys are raw how many no money boys are rowdy how many start a war - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
On Wed, 19 Nov 2008, Sebastian Kuzminsky wrote:
> Date: Wed, 19 Nov 2008 20:29:31 -0600
> From: Sebastian Kuzminsky <[EMAIL PROTECTED]>
> Reply-To: "Enhanced Machine Controller (EMC)"
>
> To: "Enhanced Machine Controller (EMC)"
> Subject: Re: [Emc-users] Problem with the new hm2_pci write function
>
> Roberto Caminiti wrote:
>> Here, the .hal source file: http://www.robomatek.com/files/test.hal
>> and the dmesg response: http://www.robomatek.com/files/error.txt
>>
>> I think that in this file there are important clues to solve my problem.
>
> Yes, the excellent debugging files you provided definately show a bug in
> the driver. The bug only manifests when there are no pwmgens enabled,
> and I guess except for you, all the hostmot2 users are on servo machines.
>
> Hold on, I'll fix it... Ok, I just checked the bugfix into CVS (TRUNK
> and 2.2), if you can update and recompile that should fix it.
>
> If you can't or don't want to compile from source, you can work around
> the bug by enabling one pwmgen instance in your config modparam.
> Assuming, of course, that you dont need those IO pins for GPIO...
I can also make a stepper only HM2 config if that will sidestep the bug
(will it Sebastian?)
>
>
> Thanks for helping us find this! Your help is much appreciated.
>
>
> --
> Sebastian Kuzminsky
>
> -
Peter Wallace
Mesa Electronics
(\__/)
(='.'=) This is Bunny. Copy and paste bunny into your
(")_(") signature to help him gain world domination.
-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
Roberto Caminiti wrote: > Here, the .hal source file: http://www.robomatek.com/files/test.hal > and the dmesg response: http://www.robomatek.com/files/error.txt > > I think that in this file there are important clues to solve my problem. Yes, the excellent debugging files you provided definately show a bug in the driver. The bug only manifests when there are no pwmgens enabled, and I guess except for you, all the hostmot2 users are on servo machines. Hold on, I'll fix it... Ok, I just checked the bugfix into CVS (TRUNK and 2.2), if you can update and recompile that should fix it. If you can't or don't want to compile from source, you can work around the bug by enabling one pwmgen instance in your config modparam. Assuming, of course, that you dont need those IO pins for GPIO... Thanks for helping us find this! Your help is much appreciated. -- Sebastian Kuzminsky - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
On Thu, 2008-11-20 at 01:47 +0100, Roberto Caminiti wrote: > Hi, > I want to bring up to date you about my sleepless night and my .write > function. > Please read my lasts messages to understand my problem. > > First, I have commented the addf hm2_5i20.0.write servo-thread line in > the configuration file and > I have launched EMC2. All works right, but there's no write function > latched to servo-thread. > After I have executed in the terminal the command: > > halcmd show all (output very long and I can copy only the last part) > > dmesg (entire response) > > The output are available in my site: http://www.robomatek.com/files/test.txt to get the entire halcmd into a text file, do this: halcmd show all 2>&1 > output.txt or, to see the output while creating a text file: halcmd show all 2>&1 | tee output.txt The same can be use with dmesg. - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
Hi, I want to bring up to date you about my sleepless night and my .write function. Please read my lasts messages to understand my problem. First, I have commented the addf hm2_5i20.0.write servo-thread line in the configuration file and I have launched EMC2. All works right, but there's no write function latched to servo-thread. After I have executed in the terminal the command: halcmd show all (output very long and I can copy only the last part) dmesg (entire response) The output are available in my site: http://www.robomatek.com/files/test.txt After, I have made a proof... I think that to solve the problem it's better to split the part to understand which make problem. So I have made a .hal file that I have executed by terminal after start up the realtime. After loaded the source file, I have put "halcmd start" in the command line, and my PC crashed down!... Fortunately, I have saved the "dmesg" response in a file. Here, the .hal source file: http://www.robomatek.com/files/test.hal and the dmesg response: http://www.robomatek.com/files/error.txt I think that in this file there are important clues to solve my problem. Thank you! BYEE Thank you! - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
Roberto Caminiti wrote: > Hi, > After the upgrade to EMC 2.2.7, I have modified my hal configuration > in order to be compatible with hm2_pci.ko. After a lot of proofs, I > have discovered that if I latch the function hm2_5i20.0.write to > servo-thread, EMC 2.2.7 doesn't run! I don't know why... > After I have tryied an other thing: > > In the configuration I have loaded only hostmot2 and hm2_pci with the > correct configuration but without latch any hm2_pci function to > servo-thread. > > After I have launched EMC2 (correctly) and I have loaded the Hal > configuration from the EMC2 menu. > And by using the command line i have wrote: > > addf hm2_5i20.0.read servo-thread > (EVERYTHING OK) > addf hm2_5i20.0.read_gpio servo-thread > (EVERYTHING OK) > > addf hm2_5i20.0.write servo-thread > (THE PC IS LOCKED AND SOMETIMES IT RESTART ITSELF) > > Sincerely I don't know if this is an RTAI problem or not, because > sometimes there are RTAI errors. > How can I discover where is the problem? > I have already made the latency test with zero overruns. > (my base thread is 10 ns). There's no reason to run .read() and .read_gpio() in the same thread. .read() reads everything, including the GPIOs. The only reason you'd ever use .read_gpio() is if you wanted to read GPIOs faster than the servo thread, you could put .read_gpio() in a faster thread. I think there should be nothing in your base thread with hostmot2. The base thread is in the FPGA. If you meant that the servo thread is running at 100,000 ns, try slowing it down to 1,000,000 ns and see if that helps the lockups. Wouldnt help the reboots though, that's strange... If you can, get the machine to just before you add the .write(), and send us the output of dmesg and "halcmd show all". -- Sebastian Kuzminsky how many no money boys are crazy how many boys are raw how many no money boys are rowdy how many start a war - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Problem with the new hm2_pci write function
Roberto Caminiti wrote: >[snip] >Sincerely I don't know if this is an RTAI problem or not, because >sometimes there are RTAI errors. >How can I discover where is the problem? >I have already made the latency test with zero overruns. >(my base thread is 10 ns). > > Is this your base thread or your servo thread period? What is your servo thread period? You can see the maximum execution time (in CPU cycles I believe) of each function by looking at `halcmd show param *tmax`. You can see the same information for each thread in the show threads display. Try setting the servo period to something very largs, like 10 milliseconds (1000 ns), and look at the execution times. If the execution time of the write function is long, you may need to slow down your servo cycle. Incidentally, what motherboard, chipset, and CPU are you using? - Steve - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
