You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If running the -a, open-au, and find-fat tests could be automated, that would be very helpful for getting new users over the hump of using flashbench and reporting results.
For example, the flow would be:
Run -a test, if erase block size is easy to see, output it and continue.
Run open-au tests, without random, to find open-au. Probably need to have the user specify a max number of open-au if they want to test cards with possibly big open-au (Samsung 32 GB essentials, etc). If open-au, non-random, is successful in finding a likely value, continue.
Run random open-au tests. Similar caveats to 2.
Run find-fat tests. Maybe do this even if open-au tests fail, so long as we have a likely erase block size from 1?
The text was updated successfully, but these errors were encountered:
Follow ons to this could include some nice features with low overhead:
Koen's idea of outputting sfdisk commands, given some input on partition desires, to format the card with ideal partition boundaries based on erase block sizes.
Automated submission of results to a server to collect data. This would, with user acceptance, potentially collect much more data on cards.
This would still be very nice. I've tried to use this tool to figure out erase block size on several microsd cards, and eventually gave up on all of them. I just have no clue.
Some automagic probability mechanism would make this actually useful.
If running the -a, open-au, and find-fat tests could be automated, that would be very helpful for getting new users over the hump of using flashbench and reporting results.
For example, the flow would be:
The text was updated successfully, but these errors were encountered: