11:29 <d1b2> <brainstorm> Thanks @whitequark , let's see if I understand this... from what I'm reading, other than merging the classes, modifying the _fvfyi function is also needed so that isdata.data gets formatted into valid JED format?: https://github.com/GlasgowEmbedded/glasgow/blob/2439f1c4cf6ab366aab8303a45cdef396623e3ce/software/glasgow/applet/program/xc9500xl/__init__.py#L354-L372

13:45 <kbeckmann> memory-25x question. just want to check if i am using the command right. i want to erase-program with 64kB sector size which is supported by the connected flash. is this command line correct? `glasgow run memory-25x --port A -V 1.8 -f 4000 erase-program -S 65536 -P 256 -f dump.bin 0`

13:46 <kbeckmann> the reason I ask is that the data that i read out does not match. if i change to -S 4096, it works fine.

13:53 <kbeckmann> the binary is 4MB exactly, so using 64kB sectors should be fine.

13:58 <miek> what's the part number of the flash?

13:58 <whitequark> I think a part of your messages got lost

13:58 <whitequark> the question you asked, specifically

14:13 <miek> "memory-25x question. just want to check if i am using the command right. i want to erase-program with 64kB sector size which is supported by the connected flash. is this command line correct? `glasgow run memory-25x --port A -V 1.8 -f 4000 erase-program -S 65536 -P 256 -f dump.bin 0`"

14:14 <kbeckmann> ah thanks for reposting.

14:15 <kbeckmann> the output of identify: https://allg.one/2bvQ

14:16 <kbeckmann> the chip is MX25U3232F

14:17 <kbeckmann> the documentation for PP tells that "One to 256 bytes can be sent" so -P 256 should be fine right?

14:21 <miek> as far as i can see, the applet always uses opcode 0x20 (4kB sector erase) in erase-program. i'm not sure how -S is intended to work with that

14:22 <kbeckmann> oh interesting

14:23 <kbeckmann> i am sorry for not reading the docs properly. i realize now that this would never work unless glasgow picks a different command based on identify and sector size.

14:24 <kbeckmann> i somehow thought that because 64k was supported it would magically choose the right command but yeah. sorry for the noise. would be a cool feature though!

14:26 <whitequark> I misunderstood how SPI flash commands work

14:26 <whitequark> this needs to be fixed