On Thu, May 28, 2015 at 02:57:51PM +0100, Mark Brown wrote:
On Thu, May 28, 2015 at 02:45:14PM +0100, Richard Fitzgerald wrote:
+#define ADSP2_SCRATCH0 0x40 +#define ADSP2_SCRATCH1 0x41 +#define ADSP2_SCRATCH2 0x42 +#define ADSP2_SCRATCH3 0x43
May as well dump the other scratches while you're at it?
Fair enough
- unsigned int scratch1 = 0xFFFFFFFF;
Why is this being initialized?
It was done a way to avoid having two shutdown messages to cover the rare case that we somehow failed to read the register value, since the firmware could never write this value (it as 31 bits set).
adsp_dbg(dsp, "Shutdown complete\n");
adsp_info(dsp, "Shutdown complete (SCRATCH1:0x%x)\n", scratch1);That seems a bit loud for a diagnostic message, why raise the severity?
It's a diagnostic of the firmware, not the driver, and we can't assume that people trying to use a firmware have the ability to build and flash a kernel with a debug version of the driver