Hi Daniel,
On 17.06.2010 13:56, Daniel Flinkmann wrote:
Am 17.06.2010 um 13:08 schrieb Carl-Daniel Hailfinger:
By the way. The first reading tests I have done ( svn1048, 1048+3xBP-patch and 1048+3xBP-patch+no_delay-patch) resulted in three identical binary files. I just double checked that with an md5sum.
Good. Did those images contain only 0xff bytes or was there real content inside (you can use hexdump or xxd to check).
Just 0xff bytes (sorry forget to say that). As you eventually remember, I had made erase tests before the read tests.
Ah. I thought you had tried read after the aborted write attempt. Sorry for the confusion.
Side note: My AAI patch on the list might result in a faster write if you change the write function for your chip in flashchips.c. It does not support progres printing, though, so you might want to wait for an updated progress printing patch.
Mh. I expect that an unpatched Bus Pirate driver will have the same problems. Which Bus Pirate is this, and which firmware are you running on it? AFAIK some firmware versions have known bugs which might affect us. If your firmware is indeed one of the broken versions, I hope we can add a blacklist to flashrom and refuse to communicate unless the Bus Pirate is upgraded.
I have a Bus-Pirate V3 with Firmware v4.5. Thats the most updated stable version. I had refreshed the firmware (and bootloader) just before I started flashing the spi flash chip. If you know about a firmware which causes the problem, please give me a link to the source of that info.
I think I read somewhere on dangerousprototypes.com that OpenOCD support was broken in some firmware versions. We use essentially the same interface. Anyway, I reported your problem here: http://dangerousprototypes.com/forum/index.php?topic=651.0
Regards, Carl-Daniel