Hello Uwe, On Monday, 31. May 2010 16:01:18 Uwe Bonnes wrote: > with appended code and a revised verilog code, I can reliable read blocks > of 16 Bytes of date generated about each Microsecond, like running for > about 500 seconds: > > 8 errors of 375293955 blocks (2.131662e-08), 9 (2.398120e-08) blocks > skipped > > From time to time some one or two blocks are skipped, which is something > to expect, as I only have about 128 byte of additional FIFO beside the > FT2232H internal fifo. First test showed that a larger FIFO lead to some > more skips, either by some flaw in my logic or some strange behaviour of > the FT2232H chip. > > Writing the generated data to disk leads to skips for several to several > ten microseconds each several to several seconds. Probably the BLK (big > kernel lock) hits here. > > The patch only adds some definition to ftdi.h and the additional file > ftdi_stream.c, with Micah Dowty okay to dual license his code. so I hope > it is okay to apply. Should the verilog code be inclided in libftdi too?
As always, thanks for your patch :) Could you be so nice and comment on the possible time jump issue I mentioned last time? Cheers, Thomas -- libftdi - see http://www.intra2net.com/en/developer/libftdi for details. To unsubscribe send a mail to [email protected]
