Jonathan Corbet wrote: > Andres Salomon <[EMAIL PROTECTED]> wrote: > > >> The question is now, why wouldn't we be seeing interrupts? >> > > A good question. > > Anyway...possible reasons for not getting interrupts: > > - Something changed in the pre-b3 board such that the interrupt line > from the camera controller no longer reaches the CPU. Seems > unlikely. >
Actually, this is exactly what happened. The CaFe INT pin was routed to a different GPIO on the 5536 in order to make room for a new INT from the LX chip itself. We found and fixed this yesterday. > - Something changed in the Cafe startup dance with the result that > interrupts are no longer being enabled in the controller. That > ritual has changed before...have there been Cafe changes this time > around? > > - The interrupt line is being masked in the CPU somehow. Are there any > "nobody cared!" messages in the logs? What else is sharing the > camera interrupt line on this board? > > - This board has the faster processor, right? It's possible that there > is a timing-related issue somewhere - the controller needs a delay > that wasn't required before. > > jon > > _______________________________________________ Devel mailing list [email protected] http://mailman.laptop.org/mailman/listinfo/devel
