ALSA: dice: optimize reading of consecutive registers

Instead of reading two consecutive register with two quadlet requests,
use one block read request.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>

sound/firewire/dice.c

@@ -374,8 +374,8 @@ static int dice_open(struct snd_pcm_substream *substream)
 	};
 	struct dice *dice = substream->private_data;
 	struct snd_pcm_runtime *runtime = substream->runtime;
-	__be32 clock_sel, number_audio, number_midi;
-	unsigned int rate_index, rate;
+	__be32 clock_sel, data[2];
+	unsigned int rate_index, number_audio, number_midi;
 	int err;
 
 	err = dice_try_lock(dice);
@@ -393,30 +393,25 @@ static int dice_open(struct snd_pcm_substream *substream)
 		err = -ENXIO;
 		goto err_lock;
 	}
-	rate = dice_rates[rate_index];
 
-	err = snd_fw_transaction(dice->unit, TCODE_READ_QUADLET_REQUEST,
+	err = snd_fw_transaction(dice->unit, TCODE_READ_BLOCK_REQUEST,
 				 rx_address(dice, RX_NUMBER_AUDIO),
-				 &number_audio, 4);
-	if (err < 0)
-		goto err_lock;
-	err = snd_fw_transaction(dice->unit, TCODE_READ_QUADLET_REQUEST,
-				 rx_address(dice, RX_NUMBER_MIDI),
-				 &number_midi, 4);
+				 data, 2 * 4);
 	if (err < 0)
 		goto err_lock;
+	number_audio = be32_to_cpu(data[0]);
+	number_midi = be32_to_cpu(data[1]);
 
 	runtime->hw = hardware;
 
-	runtime->hw.rates = snd_pcm_rate_to_rate_bit(rate);
+	runtime->hw.rates = snd_pcm_rate_to_rate_bit(dice_rates[rate_index]);
 	snd_pcm_limit_hw_rates(runtime);
 
-	runtime->hw.channels_min = be32_to_cpu(number_audio);
-	runtime->hw.channels_max = be32_to_cpu(number_audio);
+	runtime->hw.channels_min = number_audio;
+	runtime->hw.channels_max = number_audio;
 
-	amdtp_out_stream_set_parameters(&dice->stream, rate,
-					be32_to_cpu(number_audio),
-					be32_to_cpu(number_midi));
+	amdtp_out_stream_set_parameters(&dice->stream, dice_rates[rate_index],
+					number_audio, number_midi);
 
 	err = snd_pcm_hw_constraint_step(runtime, 0,
 					 SNDRV_PCM_HW_PARAM_PERIOD_SIZE,