On Sun, Feb 07, 2016 at 05:35:33PM -0800, Marc MERLIN wrote: > On Sun, Feb 07, 2016 at 05:04:04PM -0800, Greg Kroah-Hartman wrote: > > On Sun, Feb 07, 2016 at 04:50:00PM -0800, Marc MERLIN wrote: > > > > > > Is the way sync/async is treated inside usb-storage, some other USB > > > module, or outside of the usb stack altogether? > > > > Outside the USB completely, again, the USB storage driver is a very > > dumb, and tiny, scsi driver. > > Thanks. > > So, would you say my idea/suggestion is worth pursuing with the scsi > layer folks (assuming that's where the sync support is), or it's kind of > a crack idea, and not likely to be welcomed or implemented? :) I think you are going to get the same reaction from them, as they are going to point to the block layer above them that sends all of the data to them. Then the block layer is going to point to the file system, and they are going to point to the vfs layer :) So where you were to try to implement something like this, I don't know, I think Alan's original statement of "do it in userspace" makes the most sense. That way it works for any type of device, on any type of filesystem, with no kernel changes needed at all. sorry, the kernel is just the "dumb" pass-through here... greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
