Hi Wolfgang, Hi Stefan,
thanks for your comments. I tried to consider them in my updated patch series. As a consequence of the lists size limit (40k) and our little discussion on irc I banned cogito from my system an started with plain git which seems to be slightly more powerful:-)
To fool the list's size limit I did two things: first I split the huge patch (3/3) into three single patches. Even they do not really make sense without the others, they will not break U-Boot's buildsystem at any time.
The second very cool thing is the very powerful git option for creating patch that contain 1:1 copies of existing files. The PMC440 board uses sequoia's sdram.c and sdram.h. Together these two files would explode the 40k size limit. So it's a good idea to generate patch by calling:
format-patch -5 -C -C -n
or (which has the same effect):
format-patch -5 -C --find-copies-harder -n
This is how the patch summary looks like in this case:
board/{amcc/sequoia => esd/pmc440}/Makefile | 4 +- board/{amcc/sequoia => esd/pmc440}/config.mk | 0 board/{amcc/sequoia => esd/pmc440}/init.S | 16 +- .../sequoia/sequoia.c => esd/pmc440/pmc440.c} | 512 +++++++++++++++----- board/esd/pmc440/pmc440.h | 148 ++++++ board/{amcc/sequoia => esd/pmc440}/sdram.c | 0 board/{amcc/sequoia => esd/pmc440}/sdram.h | 0 board/{amcc/sequoia => esd/pmc440}/u-boot-nand.lds | 0 board/{amcc/sequoia => esd/pmc440}/u-boot.lds | 0 9 files changed, 545 insertions(+), 135 deletions(-) copy board/{amcc/sequoia => esd/pmc440}/Makefile (95%) copy board/{amcc/sequoia => esd/pmc440}/config.mk (100%) copy board/{amcc/sequoia => esd/pmc440}/init.S (89%) copy board/{amcc/sequoia/sequoia.c => esd/pmc440/pmc440.c} (59%) create mode 100644 board/esd/pmc440/pmc440.h copy board/{amcc/sequoia => esd/pmc440}/sdram.c (100%) copy board/{amcc/sequoia => esd/pmc440}/sdram.h (100%) copy board/{amcc/sequoia => esd/pmc440}/u-boot-nand.lds (100%) copy board/{amcc/sequoia => esd/pmc440}/u-boot.lds (100%)
Please notice that the following fife patches replace all of my previous three patches. They are generated against Stefan's ppc4xx/for-1.3.1 brach because of the 4xx DCACHE rework.
Matthias