On Thu, 9 Feb 2012 01:06:52 +0100 Stefan Tauner stefan.tauner@student.tuwien.ac.at wrote:
attached patch is untested. i will do that tomorrow, but wanted to give you the chance to reply early. one thing i remember to be missing is the 4k handling. apart from that and the loop error handling i hope we are quite finished. \o/
this one now is somewhat tested and not THAT broken as the one from last night... :)
i have (at least)... - removed the sfdp_fetch_pt function and integrated the hexdump-like code into the loop. it fits there better imo. - corrected the overflow check to use a 24b boundary instead of 16b because SFDP has 24b-addressing just like normal SPI flash space. - refined the output verbosity of probe_spi_sfdp (again). - renamed the struct flash's model from "SFDP device" to "SFDP-capable chip".