Attention is currently required from: Edward O'Callaghan, Jonathon Hall, Nikolai Artemiev, Peter Marheine.
4 comments:
Patchset:
Moving the map/unmap functions out of the programmer_entry struct is a good thing. But Imo it then should go into the *_master struct, rather than into an extra struct, getting registered at the same time. If I got the code right, only par_masters, especially the par_master_internal, need to be able to map/unmap memory depending on the chip size. So for the long term it could be better to modify the chip probing such that a mapping function is only called on par_masters.
I'm open for anything that solves this problem sustainably.
File atapromise.c:
Patch Set #2, Line 133: atapromise_map
This may not be needed any longer, because `atapromise_map` has the same implementation as `fallback_map`.
CB:67655
File atavia.c:
Patch Set #2, Line 147: atavia_map
This may also not be needed, CB:67656
But I have to test this on real hardware :/
File ichspi.c:
Patch Set #2, Line 1865: &mapper_phys
I don't think mapping is used here. As far as I know it is only used for the par_master in internal.c
Nico, do you know more?
To view, visit change 67695. To unsubscribe, or for help writing mail filters, visit settings.