On Monday 28 December 2009, x wrote: > I'm fooling with a cable modem that has a MIPS 4kc processor and cfi > type flash chip from ST: M28W160CT. Flash configuration is as follows: > > flash bank cfi 0x90000000 0X200000 2 2 $_TARGETNAME > > openocd-0.2.0 has no problem detecting the flash with "flash_probe"
I've lost track of the status here ... was this the case where the flash needed to be treated as "non-CFI" and then it behaved? That would of course not address the endianness issues noted below. - Dave > Some other differences I've noted between version 2 and the last two > releases is in regards to endian ness. > Dump_image in version 2 produces an output file where the ascii text > (within the dump file) is in human readable order. The dumps with > version 3 and 4 produce an image where the text is rotated or flipped. > I've not been able to "load_image" with any version successfully or > "flash write_image". It does, however erase the flash in version two. > > To solve my dilemma, I modified the source code to "HairyDairyMaid's" > router debrick utility, which interestingly, some of the MIPS_4kc > (openocd) source is based on. With the debrick utility, I was able to > detect, read, erase, and write to the modem I was "tampering" with :). > I WISH I could use Openocd for this task as I could use some of the > additional TCL scripting features to automate testing and debugging. > > I've tried a few modifications by compairing the source from version 2 > to 3 and have not had success. I'm reasonably certain my config file > is not the issue, but I could be wrong. > > Regards, > > Mr. Smith > _______________________________________________ > Openocd-development mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/openocd-development > > _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
