>>>>> "Werner" == Werner Almesberger <[email protected]> writes:
[..] > At the moment, the timing works basically like this: > start_timer; while (timer != deadline); /* fine-tuning follows > */ > The timer runs at 112 MHz (PCLK), but I can access it only at about 13 > MHz (APB is slow). So this is the limit of the accuracy of this timing > loop. But doesn't reading from the APB synchronize the CPU to the APB clock (i.e. execution of the read finishes always after the same constant time relative to a APB clock edge)? Then if we had a timer which is an integer multiple 'm' of the APB clock, you'd be able to get as accurate as that timer? I.e. you'd only every be able read timer values m*N, but once read, the CPU would magically synchronize in a way that makes m*N really an accurate value to the last bit. David -- GnuPG public key: http://user.cs.tu-berlin.de/~dvdkhlng/dk.gpg Fingerprint: B17A DC95 D293 657B 4205 D016 7DEF 5323 C174 7D40
pgpxRGbXhMEhx.pgp
Description: PGP signature
_______________________________________________ Qi Hardware Discussion List Mail to list (members only): [email protected] Subscribe or Unsubscribe: http://lists.en.qi-hardware.com/mailman/listinfo/discussion
