Attention is currently required from: Simon Buhrow, Edward O'Callaghan, Samir Ibradžić. Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/40477 )
Change subject: ft2232_spi.c: Implement spi_send_multicommand() ......................................................................
Patch Set 32:
(1 comment)
Patchset:
PS32:
Not sure if I'm doing something wrong. The patch works so far, but […]
I wondered what device was used for testing so far and stumbled over this picture[1] mentioned in the original commit message. There's some 1.36MHz shown in it, so I tried again with such a low speed (divisor=44) and, tada, it makes a difference. No idea what is going on inside that chip. My numbers[2] even suggest that sending the commands in two packets (original chunksize) is faster *in my setup*.
[1] https://ibb.co/0c1J25d [2] https://paste.flashrom.org/view.php?id=3472
NB. What would really increase speed in the tested scenario (writing random data, hence everything needs to be erased) is to choose a bigger erase block size. I have this on my todo list since my very first days with flashrom. If I drop 4K and 32K blocks from `flashchips.c`, it's about twice as fast. But as this is not a common use case, I did not see it as a priority. OTOH, it always bugs me, for instance when I switch back and forth between coreboot and another firmware.
-
Nico Huber (Code Review)