Mark Brown <[email protected]> writes: > On Mon, Jul 13, 2009 at 04:10:21PM -0700, Kevin Hilman wrote: >> Mark Brown <[email protected]> writes: > >> > The other option is that the clocks are associated with the DAI device >> > for the platform device for the DAI and the clk_get() is done on the DAI >> > platform device with a fixed clock name. The ASoC code is almost all of >> > the way there, it's already got the struct device right and just needs >> > the clock name making fixed. > >> Do you have these queued up someplace? or can you describe in mor >> detail where the clk_get should be placed? > > git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound-2.6.git for-2.6.32 > > which should also be turning up in -next ATM. The driver should look > like a normal platform device driver, doing resource grabbing on probe() > and then registering with the ASoC core once that's happened. > >> Being somewhat ignorant of ASoC internals, I'm not sure what you mean >> by this proposal. I don't see any DAI setup done in platform code. > > There *should* be devices with names "davinci-asp" and "davinci-mcasp" > being set up. If there aren't the audio will never come up.
Somehow I missed seeing these. Using these devices, we can simply rely on the 'struct device' and have no need to pass in the clock names. The arch/arm/mach-davinci/* code can setup the clkdev nodes with the "davinci-asp" and "davinci-mcasp" names and then there would be no need to pass the clock names. Patches coming soon... Kevin _______________________________________________ Davinci-linux-open-source mailing list [email protected] http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
