 | | LNP checksum optimizations
|
|
In a previous post I said that some optimizations can be made in LNP. Now I have the patch ready and uploaded it on sourceforge. Sadly I made it against a custom version of the kernel, so it may need a manual intervention to apply (fuzzyness (...) (24 years ago, 31-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: legOS lcd management and thoughts on velocities
|
|
(...) The timeout is really "expected time till next rotation tick + 50 %". And the velocity will not drop to 0, because my code assumes, a new rotation could happen any moment. If you check for speed < 7 ticks/sec (or 142 msec/tick) you should get (...) (24 years ago, 29-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: legOS lcd management and thoughts on velocities
|
|
Thank you Jochen for your answer. If I well understand the ds_rotation_handler code, the timeout is set to 1 second. So the velocity may drop to zero 1 second after wheels really stop. In the specific case of my robot (I call it R2D1), this will be (...) (24 years ago, 27-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: legOS lcd management and thoughts on velocities
|
|
(...) Yes, I have written that support. First you have to enable it, by removing the comments from the line in legOS/boot/config.h. Then recompile the kernel and all apps. After you have programmed the rotation sensor in the usual way you can read (...) (24 years ago, 26-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: sleep efficientcy
|
|
(...) value (...) There are two issues: 1 do you want to do anything while waiting, 2 how quick do you want to respond to the event. In both cases using an event is better. You can do something else until the event happens (synchronization issuses (...) (24 years ago, 25-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: sleep efficientcy
|
|
(...) value (...) Ross, I can't speak about efficiency - I dont know enogh about embedded programming, but I've had better results using wait_event. I've tried both msleep() and yield() in a loop, and just calling wait_event with a function that (...) (24 years ago, 25-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | sleep efficientcy
|
|
Hi again, i would like to know which is more efficient in code? having a while loop which is checking for a condition to be true and sleeping for an arbitary value while it is not, or using a wait_event statement passing a function pointer to a (...) (24 years ago, 25-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: Parallel sensor problem
|
|
(...) That isn't the problem. I reversed the sensor on one RCX, and they both showed it as being constantly pressed, so that must be the wrong polarity. But I also tried moving it from input 2 to 3. Works fine. Same programs, everything, it works in (...) (24 years ago, 24-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: LegOS 0.2.5 and LNP
|
|
(...) after (...) old (...) That's true. I've noticed it. I was speaking of PC implementation. (...) semaphore (...) in a (...) instruction (...) wait (...) of the (...) shouldn't (...) Ok, perfect. I vote for the semaphore implementation. I also (...) (24 years ago, 24-Jan-01, to lugnet.robotics.rcx.legos)
|
| |
| | Re: C++ OverHead
|
|
Hi, The answer here depends on your coding style. If you are using non-derived concrete classes, with simple member data and the default constructor/destructor, then it costs ~10-20 bytes/class. If you are using non-derived concrete classes, with (...) (24 years ago, 24-Jan-01, to lugnet.robotics.rcx.legos)
|
