Simeon Simeonov wrote: > Hi Manu, > > I was able to resolve my rotor problem by making the following changes in the stb6100, stb0899 and mantis sources: > u<X> replaced by unit<X>_t.. Some points: 1. The change to stb6100 should in no way matter for the rotor. 2. u* is kernel data types and uint*_t are POSIX data types. In fact this shouldn't matter AFAICT, since both are explicit datatypes. The only difference in there that can make a difference would be "long" I think, greping for long, the only thing that comes to my mind is stb0899_wait_diseqc_fifo_empty(), where you will see the error as stb0899_wait_diseqc_fifo_empty: timed out !! (do you see this error, in your logs ?) But that also shouldn't matter, since there the timeout what's provided there isn't quite large Also this doesn't come into picture, with your u* to uint* change. .. thinking .. > Since I am running the drivers on 64-bit machine my guess is that is a 64-bit pointer size thing. > Unfortunately it is painful to restore my rotor settings after loosing them and I cannot tell you at the > moment which change made the difference. Also, considering that since almost everyone else has this working, maybe it has to be specific to your setup/machine. Regards, Manu > > Thanks for your help, > Simeon > > ----- Original Message ---- > From: Manu Abraham <abraham.manu@xxxxxxxxx> > To: Simeon Simeonov <simeonov_2000@xxxxxxxxx> > Cc: linux-dvb@xxxxxxxxxxx > Sent: Monday, March 3, 2008 3:52:29 AM > Subject: Re: STB0899 users, please verify results was Re: TechniSat SkyStar HD: Problems scaning and zaping > > Simeon Simeonov wrote: >> I am using mythtv and here is the sequence of commands: >> >> Without repeat: >> DiSEqCDevTree: Changing LNB voltage to 13V >> DiSEqCDevTree: Rotor - Goto Stored Position 2 >> DiSEqCDevTree: Sending DiSEqC Command: e0 31 6b 2 >> DiSEqCDevTree: Changing to DiSEqC switch port 1/4 >> DiSEqCDevTree: Sending DiSEqC Command: e0 10 38 f0 >> >> With repeat: >> DiSEqCDevTree: Changing LNB voltage to 13V >> DiSEqCDevTree: Rotor - Goto Stored Position 2 >> DiSEqCDevTree: Sending DiSEqC Command: e0 31 6b 2 >> DiSEqCDevTree: Changing to DiSEqC switch port 1/4 >> DiSEqCDevTree: Sending DiSEqC Command: e0 10 38 f0 >> DiSEqCDevTree: Repeat DiSEqC Command: e1 10 38 f0 > > > In fact, looks really confusing to me. The diseqc commands just go out > as it is, > through the FIFO. I don't see anything that which changes the commands, but > if it was working with another card, it should be working with this card > too (i > had some doubts, but when you said it worked with some other card, then i > don't see the issue in here with the diseqc part. The FiFO either works > or not) > > > >> ----- Original Message ---- >> From: Manu Abraham <abraham.manu@xxxxxxxxx> >> To: Simeon Simeonov <simeonov_2000@xxxxxxxxx> >> Cc: linux-dvb@xxxxxxxxxxx >> Sent: Sunday, March 2, 2008 2:38:51 PM >> Subject: Re: STB0899 users, please verify results was Re: TechniSat SkyStar HD: Problems scaning and zaping >> >> Simeon Simeonov wrote: >>> Hi Manu, >>> >>> I am attaching two gzipped logs. They are supposed to tune to the same frequency using the tip >>> of Mantis tree. The difference between the two are that in the GOOD log no repeat command is used >>> and in the BAD log one repeat for the switch is issued. The initial position of my rotor is about 20 deg >>> east from the target rotor position. >>> Using the tunning without the repeats the rotor goes all the way through and tunes successfully - GOOD log. >>> When repeat command is included in the diseqc sequence the rotor goes about 10 degrees to the west and stops as if it has reached the desired position. The BAD log corresponds to that. >>> When I tried to move to any other rotor stored position I find that that all of the memorized in the rotor positions are shifted. My guess is that for some reason the rotor stops, stores current position as the target one and then >>> re-calculates all of them. But I do not see anything like that in the log file. The only thing I see is that >>> after the third byte in the diseqc repeat command fifo get full and sending the next byte has to >>> wait for one cycle. >>> The same sequences work just fine with my 102g card and the v4l drivers. >> Can you please try to get the DiSEqC strings that you are sending >> (from the application) in these 2 cases ? >> >> * Without repeat >> * With repeat >> >> Regards, >> Manu > > > > > > > ____________________________________________________________________________________ > Looking for last minute shopping deals? > Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping > _______________________________________________ linux-dvb mailing list linux-dvb@xxxxxxxxxxx http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
