On Wed, Jan 11, 2012 at 1:20 PM, Peter Maydell <peter.mayd...@linaro.org> wrote: > On 11 January 2012 11:53, Javier Martinez Canillas <jav...@dowhile0.org> > wrote: >> Yes, the IGEPv2 and the Overo are indeed very similar machines so we >> base the hardware modeling on the Overo. >> >> The only differences here are which OMAP version is used to initialize >> the MPU (3630 instead of 3430) and the NAND memory used, Numonyx >> OneNAND instead of Micron NAND. > > That's pretty minor -- can you just use the overo model as is? > I bet nothing at the software level cares that much about the > OMAP version or NAND manufacturer... >
Yes, I guess we can use the Overo as is. >> We just add this device model for our clients so I thought it was a >> good idea to post the patch upstream. But feel free to ignore me if >> you think that it doesn't add any value to Linaro QEMU. > > I appreciate the patch. My pushback is really that every new board > we accept carries an implicit future cost for testing/maintenance/etc, > so there needs to be a reasonable set of users who would use this > and not the overo or beagle. > Yes, I totally understand. Then just feel free to just ignore the patch, sorry for the noise. >> Sorry about that, I didn't know that GPLv2 or later was the norm in >> the QEMU project. > > We've just changed recently to asking people to include the "or later" > clause for new code (this is why, as you note, there are some existing > bits of code which are v2-only). > > -- PMM Best regards, Javier
