Revert "Ensure page_size is always at least 1"

[umbynos]

This reverts commit da0e437.

Reason for this revert ElTangas/jtag2updi#16 (comment)
We use the jtag2udpi fw on the SAMD11 to be able to program the ATMega4809. That commit was breaking the upload with AVRdude

Related conversation with @MCUdude umbynos@9ae854d#commitcomment-73386063

Co-authored-by: Martino Facchin [email protected]

[MCUdude]

By looking at this PR, it's not obvious to me why the config_gram.y has to be modified.
With this PR it just blindly accepts whatever $3->value.number holds, instead of checking for <= 0 and printing a warning.

As for the removed line in jtagmkii.c I don't know if this will have any consequences for other hardware that uses this protocol. @dl8dtl do you have any thoughts on this one? AFAIK, page_size should be defined for all memory types for all targets anyways.

[dl8dtl]

I'm against reverting that change – after all, it has been made for a reason (fixing a bug or crash).
There must be a better solution, but I didn't look into the details so far.

[mcuee]

In this case, we can close this one.

[umbynos]

Unfortunately, I don't know the reasons behind da0e437. I understand this is not a fix but only a quick way to solve ElTangas/jtag2updi#16 (comment). For sure, there is a better way to fix this, but I don't know sufficiently about avrdude to do it.

[mcuee]

@umbynos Can you ask the affected user to try out latest avrdude 7.0 release or even git master to see if the issue has been fixed? I think there are a few improbements done on this area.

[MCUdude]

I can try this with my Nano Every later today!

[mcuee]

Re-open this topic, just want to make sure if the has already been fixed in avrdude 7.0 release and git.

[mcuee]

I do not have issues using avrdude-7.0 and git with Arduino Nano Every (with SAM D11 as the JTAG2UPDI controller). Another user also reported the susccess story in another issue.
Ref:

• avrdude with Arduino Nano Every #980
• avrdude with Arduino Nano Every: conf file with Arduino vs conf file from avrdude #991

[mcuee]

@umbynos

In this case, I think this issue can really be closed. Any objections?

[stefanrueger]

Looks like AVRDUDE 7.0 does not expose the issue that this PR tries to solve, so not reverting.