Skip to content

mimxrt: Question about extending the test base #7560

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
robert-hh opened this issue Jul 22, 2021 · 5 comments
Closed

mimxrt: Question about extending the test base #7560

robert-hh opened this issue Jul 22, 2021 · 5 comments
Labels
port-mimxrt

Comments

@robert-hh
Copy link
Contributor

In out attempt to extend the test base we procured a MIMXRT1050 board. That caused problems, since the old board is not available any more, and the new board is by deafult configured for a hyperflash chip instead a QSPI flash. The layout is different, especially the page size is 512 byte instead of 256 byte. For LFS that means that progsize has to be set differently, and Python code must have a means to determine the page size. A iotcl call could used for that, but it has to be defined: Currently the is no code provided for that. Do you know a code foreseen for this function, or should we just grap one, like 7?
@robert-hh
Copy link
Contributor Author

As another option, I added a QSTR to that specific block device called PAGE_SIZE, which will be set board-specific.

@dpgeorge
Copy link
Member

especially the page size is 512 byte instead of 256 byte

Do you mean the minimum write size is now 512 bytes? But the erase size is still 4096? Can you remind me why progsize needs to match this value (or be larger than it, I guess)?

As another option, I added a QSTR to that specific block device called PAGE_SIZE, which will be set board-specific.

Yeah, that's not a bad way to do it... but a new ioctl might be more general.

@robert-hh
Copy link
Contributor Author

For that flash, the page size if 512 byte, the sector size is 256 kByte, and the full flash size is 64MB. As far as I understand it, writing to the flash has to be done in page size items. I tried keeping progsize=256, but that failed. So it seems that progsize has match the page size, but I can do further testing.
I changed the QSTR to a bound methed. That way I can refer to a value in the flash config block instead of a #define. So it's defined at a single place. A IOCTL call would do the same.

@robert-hh
Copy link
Contributor Author

robert-hh commented Jul 22, 2021 •
edited
Loading

The restriction may be caused by the simple writeblocks() implementation in this port. It assumes that both the offset and data size is a multiple of the flash page size. From what I noticed earlier, vfs requests writes for lfs with a multiple of progsize for offset and length (length was actually always progsize). For FAT the offset was a multiple of 512. So I could change the writeblocks() implementation allowing writes of partial pages. But that would double the size of managing data in lfs to about 32k. That may be worse than having larger progsize items.

@robert-hh
Copy link
Contributor Author

I changed the writeblocks() code to allow partial page writes. Therefore the is no change required for vfs or the block driver interface,

tannewt added a commit to tannewt/circuitpython that referenced this issue Apr 21, 2023
@tannewt
Add address_little_endian for displayio
7089ea4 
Add address_little_endian for epaper displays with little endian
(low byte first) addresses.

Also clears allocated display and display bus memory so it has a
known state. The acep member wasn't always set so it varied
accidentally.

Fixes micropython#7560. May fix micropython#7778. Fixes micropython#5119.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
port-mimxrt
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@dpgeorge @robert-hh