Add FlashSafeDma wrapper for eh traits which ensure correct DMA transfers from flash

[MabezDev]

An alternative solution to #674, using a wrapper type instead of adding the checks to the current dma operations. This is only implemented for Spi traits at the moment, as that's the only peripheral we have dma support for and e-h traits.

Benefits to this approach:

• You only pay for it if you need it
• We can add different e-h impls for different peripherals but reuse this struct
• The end use can control the scratch buffer size

Negatives to this approach:

• You have to run into the issue first, and then try to find this wrapper, instead of it being handled seemlessly
• Instead of more logic in e-h impls, we now have more e-h impls (though most just pass through to the inner impl, except for writes usually)
• Added type complexity in the end user spi type

Thank you for your contribution.
Please make sure that your submission includes the following:

Must

• The code compiles without errors or warnings.
• All examples work.
• cargo fmt was run.
• Your changes were added to the CHANGELOG.md in the proper section.
• You updated existing examples or added examples (if applicable).
• Added examples are checked in CI

Nice to have

• You add a description of your work to this PR.
• You added proper docs for your newly added features and code.

[bjoernQ]

I like this approach much better than the other PR

[MabezDev]

I like this approach much better than the other PR

Thanks for the feedback! I'll look into fleshing out this approach.

One thing I could use some help with is what to call this thing; I plucked FlashSafeDma out of the air, but its not actually that descriptive. FlashSafeDma only intercepts tx transfers where the source memory is not RAM, so can anyone think of a better name? I suppose we can always embellish the docs to describe its precise functionality.

[bugadani]

FlashSafeDma sounds good enough to me. It should be documented why it's necessary, but otherwise, I find it descriptive enough without being overly verbose.

Edit: maaaaaaybe some users could get confused and think it can also receive into flash (i.e. disambiguating into FlashReadSafeDma may be better, maybe?)?

[MabezDev]

FlashSafeDma sounds good enough to me. It should be documented why it's necessary, but otherwise, I find it descriptive enough without being overly verbose.

Edit: maaaaaaybe some users could get confused and think it can also receive into flash (i.e. disambiguating into FlashReadSafeDma may be better, maybe?)?

I've kept it as FlashSafeDma but added some docs to explain what it actually does.

[jessebraham]

Sorry for the delay in reviewing this, the notification must have gotten buried in my inbox because I missed it at some point 😅 I left one small suggestion but overall LGTM, thanks for this!