On 10/07/16 19:08, Mark Cave-Ayland wrote:
> By default large amounts of DBDMA debugging are produced when often it is just
> 1 or 2 channels that are of interest. Introduce DEBUG_DBDMA_CHANMASK to allow
> the developer to select the channels of interest at compile time, and then
> further add the extra channel information to each debug statement where
> possible.
>
> Also clearly mark the start/end of DBDMA_run_bh to allow tracking the bottom
> half execution.
>
> Signed-off-by: Mark Cave-Ayland <mark.cave-ayl...@ilande.co.uk>
> Acked-by: Benjamin Herrenschmidt <b...@kernel.crashing.org>
> ---
> hw/misc/macio/mac_dbdma.c | 75
> +++++++++++++++++++++++++--------------------
> 1 file changed, 42 insertions(+), 33 deletions(-)
>
> diff --git a/hw/misc/macio/mac_dbdma.c b/hw/misc/macio/mac_dbdma.c
> index b6639f4..8e4b208 100644
> --- a/hw/misc/macio/mac_dbdma.c
> +++ b/hw/misc/macio/mac_dbdma.c
> @@ -46,6 +46,7 @@
>
> /* debug DBDMA */
> #define DEBUG_DBDMA 0
> +#define DEBUG_DBDMA_CHANMASK ((1ul << DBDMA_CHANNELS) - 1)
Someone flagged up that this doesn't work on 32-bit hosts because while
the channel is 0 to 31, the compiler shifts first and then subtracts
which gives warnings like this:
hw/misc/macio/mac_dbdma.c: In function ‘dbdma_cmdptr_load’:
hw/misc/macio/mac_dbdma.c:49:36: error: left shift count >= width of
type [-Werror=shift-count-overflow]
#define DEBUG_DBDMA_CHANMASK ((1ul << DBDMA_CHANNELS) - 1)
The fix here is to change the above to:
#define DEBUG_DBDMA_CHANMASK ((1ull << DBDMA_CHANNELS) - 1)
Here the unsigned long long will ensure that the shift is valid before
the subtraction to get the final 32-bit wide mask.
David, are you able to fix this up in your branch?
> #define DBDMA_DPRINTF(fmt, ...) do { \
> if (DEBUG_DBDMA) { \
> @@ -53,6 +54,14 @@
> } \
> } while (0);
>
> +#define DBDMA_DPRINTFCH(ch, fmt, ...) do { \
> + if (DEBUG_DBDMA) { \
> + if ((1ul << (ch)->channel) & DEBUG_DBDMA_CHANMASK) { \
> + printf("DBDMA[%02x]: " fmt , (ch)->channel, ## __VA_ARGS__); \
> + } \
> + } \
> +} while (0);
> +
This part is still okay for a channel range 0 to 31.
ATB,
Mark.
