FS#11580 - Unify 32mb and 64mb ipod video builds

Attached to Project: Rockbox
Opened by Frank Gevaerts (fg) - Sunday, 29 August 2010, 13:47 GMT
Last edited by Frank Gevaerts (fg) - Tuesday, 31 August 2010, 19:07 GMT
Task Type Patches
Category Operating System/Drivers
Status Closed
Assigned To No-one
Operating System iPod 5G
Severity Low
Priority Normal
Reported Version Daily build (which?)
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No


This quick and dirty patch makes use of the fact that on a 32MB ipod video the RAM is mirrored in the address space, so the only thing actually needed to make a 64MB build not do nasty things on a 32MB board is adjusting audiobufend.

A quick test on my 32MB 5G with a 64MB build seems to work fine. A bit more testing is needed though, especially to see if a 64MB board uses all available RAM.

Also, firmware/target/arm/thread-pp.c and firmware/drivers/ata.c seem to do something different on 64MB builds. I don't know what should be done with those.

This *isn't* ready for commit yet. it probably doesn't even compile for most targets
This task depends upon

Closed by  Frank Gevaerts (fg)
Tuesday, 31 August 2010, 19:07 GMT
Reason for closing:  Accepted
Additional comments about closing:  Committed as r27960
Comment by Frank Gevaerts (fg) - Sunday, 29 August 2010, 15:16 GMT
Slightly improved version. Functionally exactly the same, but this one doesn't break other builds. It also sets the default memory amount in configure to 64.
Comment by Frank Gevaerts (fg) - Sunday, 29 August 2010, 21:05 GMT
This version also handles the different current measurements in debug_menu.c and unifies the BATTERY_CAPACITY settings.

We might want to set BATTERY_CAPACITY_DEFAULT differently depending on detected RAM. I'm not sure of how to do that cleanly, and I'm not convinced we actually want that (due to boards being routinely swapped by repair centres and refurbishers). Apart from that, I think this patch is complete.
Comment by Frank Gevaerts (fg) - Sunday, 29 August 2010, 21:21 GMT
Final patch for the day: this one resolves the last remaining issue using the best solution for all thorny issues: document it in the manual.

I think this is ready for commit, assuming people don't find unexpected problems.