mkimg: make MBR and EBR honor alignment, and change default alignment to 4k

[jmg]

Summary

mkimg previously, even if you specified -P 4k, would not align
MBR images to a 4k boundary. This hasn't been necessary for MBR
for almost 20 years.

As 4k drives are very common, change mkimg to default to a 4k
alignment. I would be fine w/ uping this to 16k or 64k.
This gives images the benefit w/o having to track down all the
various scripts that make an image to fix them.

Test Plan

If you see the changes to the test program, it mounts and outputs
the gpart show info for the raw images. I used this to verify
that things are now properly aligned...

[marcel]

Would you mind splitting the changes into 2 reviews/commits. I don't oppose the changes, but they are logically unrelated/independent and it helps us with reversals and/or MFCs if they are separate commits.

[imp]

In D22068#482123, @marcel wrote:

Would you mind splitting the changes into 2 reviews/commits. I don't oppose the changes, but they are logically unrelated/independent and it helps us with reversals and/or MFCs if they are separate commits.

My suggestion as well.

Otherwise they look good to me.

[jmg]

In D22068#482123, @marcel wrote:

Would you mind splitting the changes into 2 reviews/commits. I don't oppose the changes, but they are logically unrelated/independent and it helps us with reversals and/or MFCs if they are separate commits.

What way do you want to split the two apart? Do you mean blksz default and mbr/ebr fixes? Or do you mean the unrelated man page changes (to make it pass igor)?

[marcel]

As relayed via email: the changes are good. Please make them using independent commits. No need to send new reviews out. Thanks for fixing up the man page!

[marcel]

Copy-pasted the email sent to mjg@, since it bounced:

Heh, I didn’t even realize there were man page changes :-)

Please split them all apart. I’ll approve the review. That way you only have to commit them separately. No need to send new reviews out.