Hi, On Wed, Jan 25, 2012 at 10:54:56AM -0800, Paul Zimmerman wrote: > On Wed, 25 Jan 2012, Felipe Balbi wrote: > > > > > You probably don't want a module parameter. Maybe we could enable this depending > > > > on the PCI ID, or from platform data? > > > > > > Platform data will be vanishing soon, but we can have a devicetree > > > attribute to select that, for sure ;-) > > > > > > Will hack that up quickly and send a revised patch. > > > here's the differential diff to add a flag for fifo resizing: > > <snip> > > Thanks Felipe, looks good. > > One question, though. I am completely ignorant about the device tree stuff. > Will a standard PC platform have a device tree for the PCI devices, and if > so, where will this option flag get set? I'm not sure how devicetree is implemented for the PCI stuff, maybe the guys in the devicetree mailing list can help. Any tips guys ? Anyway, when I actually figure out a nice way to allocate fifo space based on enabled endpoints, I believe most users will just start setting that flag because it's far better to use the FIFO for enabled endpoints then not allocate it at all (image a situation where we have 4 IN eps and 1024 bytes FIFO space for each one of them, if we're running with only 1 IN ep enabled, we could allocate all 4x 1024 to the only enabled endpoint). -- balbi
Attachment:
signature.asc
Description: Digital signature
