All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-25 22:15 Jon Smirl
  2009-05-25 22:15   ` Jon Smirl
                   ` (6 more replies)
  0 siblings, 7 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

Version 4. Changed timeouts to use relax_cpu() instead of udelay(). Fixed locking to lock bestcomm instead of
registers. Code reorg as requested in the various comments. Removed sysfs over/underrun reporting. Removed 
reset retries. Removing the retries makes the driver fail to load about 5% of the time. An oscilliscope is 
needed to figure out what is going on with failed resets. Both pcm030 and Efika ocasionally fail to reset 
and they use different codec chips.

---

Jon Smirl (5):
      Fabric bindings for STAC9766 on the Efika
      Support for AC97 on Phytec pmc030 base board.
      AC97 driver for mpc5200
      Main rewite of the mpc5200 audio DMA code
      The macro spin_event_timeout() takes a condition and timeout value


 arch/powerpc/include/asm/delay.h    |   33 +++
 sound/soc/fsl/Kconfig               |   27 ++
 sound/soc/fsl/Makefile              |    5 
 sound/soc/fsl/efika-audio-fabric.c  |   90 +++++++
 sound/soc/fsl/mpc5200_dma.c         |  442 ++++++++++++++++++++++-------------
 sound/soc/fsl/mpc5200_dma.h         |   43 ++-
 sound/soc/fsl/mpc5200_psc_ac97.c    |  331 ++++++++++++++++++++++++++
 sound/soc/fsl/mpc5200_psc_ac97.h    |   15 +
 sound/soc/fsl/mpc5200_psc_i2s.c     |  247 +++-----------------
 sound/soc/fsl/mpc5200_psc_i2s.h     |   12 +
 sound/soc/fsl/pcm030-audio-fabric.c |   90 +++++++
 11 files changed, 946 insertions(+), 389 deletions(-)
 create mode 100644 sound/soc/fsl/efika-audio-fabric.c
 create mode 100644 sound/soc/fsl/mpc5200_psc_ac97.c
 create mode 100644 sound/soc/fsl/mpc5200_psc_ac97.h
 create mode 100644 sound/soc/fsl/mpc5200_psc_i2s.h
 create mode 100644 sound/soc/fsl/pcm030-audio-fabric.c

-- 
Jon Smirl
jonsmirl@gmail.com

^ permalink raw reply	[flat|nested] 39+ messages in thread

* [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
@ 2009-05-25 22:15   ` Jon Smirl
  2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
                     ` (5 subsequent siblings)
  6 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

(in microseconds) as parameters.  It spins until either the condition is true
or the timeout expires.  It returns the result of the condition when the loop
was terminated.

This primary purpose of this macro is to poll on a hardware register until a
status bit changes.  The timeout ensures that the loop still terminates if the
bit doesn't change as expected.  This macro makes it easier for driver
developers to perform this kind of operation properly.

Signed-off-by: Timur Tabi <timur@freescale.com>
---
 arch/powerpc/include/asm/delay.h |   33 +++++++++++++++++++++++++++++++++
 1 files changed, 33 insertions(+), 0 deletions(-)

diff --git a/arch/powerpc/include/asm/delay.h b/arch/powerpc/include/asm/delay.h
index f9200a6..fedf037 100644
--- a/arch/powerpc/include/asm/delay.h
+++ b/arch/powerpc/include/asm/delay.h
@@ -2,8 +2,11 @@
 #define _ASM_POWERPC_DELAY_H
 #ifdef __KERNEL__
 
+#include <asm/time.h>
+
 /*
  * Copyright 1996, Paul Mackerras.
+ * Copyright (C) 2009 Freescale Semiconductor, Inc. All rights reserved.
  *
  * This program is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License
@@ -30,5 +33,35 @@ extern void udelay(unsigned long usecs);
 #define mdelay(n)	udelay((n) * 1000)
 #endif
 
+/**
+ * spin_event_timeout - spin until a condition gets true or a timeout elapses
+ * @condition: a C expression to evalate
+ * @timeout: timeout, in microseconds
+ * @delay: the number of microseconds to delay between eache evaluation of
+ *         @condition
+ * @rc: the last value of the condition
+ *
+ * The process spins until the condition evaluates to true (non-zero) or the
+ * timeout elapses.  Upon exit, @rc contains the value of the condition. This
+ * allows you to test the condition without incurring any side effects.
+ *
+ * This primary purpose of this macro is to poll on a hardware register
+ * until a status bit changes.  The timeout ensures that the loop still
+ * terminates even if the bit never changes.  The delay is for devices that
+ * need a delay in between successive reads.
+ *
+ * gcc will optimize out the if-statement if @delay is a constant.
+ */
+#define spin_event_timeout(condition, timeout, delay, rc)                   \
+{                                                                           \
+	unsigned long __loops = tb_ticks_per_usec * timeout;                \
+	unsigned long __start = get_tbl();                                  \
+	while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
+		if (delay)                                                  \
+			udelay(delay);                                      \
+		else	                                                    \
+			cpu_relax();                                        \
+}
+
 #endif /* __KERNEL__ */
 #endif /* _ASM_POWERPC_DELAY_H */

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
@ 2009-05-25 22:15   ` Jon Smirl
  0 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

(in microseconds) as parameters.  It spins until either the condition is true
or the timeout expires.  It returns the result of the condition when the loop
was terminated.

This primary purpose of this macro is to poll on a hardware register until a
status bit changes.  The timeout ensures that the loop still terminates if the
bit doesn't change as expected.  This macro makes it easier for driver
developers to perform this kind of operation properly.

Signed-off-by: Timur Tabi <timur@freescale.com>
---
 arch/powerpc/include/asm/delay.h |   33 +++++++++++++++++++++++++++++++++
 1 files changed, 33 insertions(+), 0 deletions(-)

diff --git a/arch/powerpc/include/asm/delay.h b/arch/powerpc/include/asm/delay.h
index f9200a6..fedf037 100644
--- a/arch/powerpc/include/asm/delay.h
+++ b/arch/powerpc/include/asm/delay.h
@@ -2,8 +2,11 @@
 #define _ASM_POWERPC_DELAY_H
 #ifdef __KERNEL__
 
+#include <asm/time.h>
+
 /*
  * Copyright 1996, Paul Mackerras.
+ * Copyright (C) 2009 Freescale Semiconductor, Inc. All rights reserved.
  *
  * This program is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License
@@ -30,5 +33,35 @@ extern void udelay(unsigned long usecs);
 #define mdelay(n)	udelay((n) * 1000)
 #endif
 
+/**
+ * spin_event_timeout - spin until a condition gets true or a timeout elapses
+ * @condition: a C expression to evalate
+ * @timeout: timeout, in microseconds
+ * @delay: the number of microseconds to delay between eache evaluation of
+ *         @condition
+ * @rc: the last value of the condition
+ *
+ * The process spins until the condition evaluates to true (non-zero) or the
+ * timeout elapses.  Upon exit, @rc contains the value of the condition. This
+ * allows you to test the condition without incurring any side effects.
+ *
+ * This primary purpose of this macro is to poll on a hardware register
+ * until a status bit changes.  The timeout ensures that the loop still
+ * terminates even if the bit never changes.  The delay is for devices that
+ * need a delay in between successive reads.
+ *
+ * gcc will optimize out the if-statement if @delay is a constant.
+ */
+#define spin_event_timeout(condition, timeout, delay, rc)                   \
+{                                                                           \
+	unsigned long __loops = tb_ticks_per_usec * timeout;                \
+	unsigned long __start = get_tbl();                                  \
+	while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
+		if (delay)                                                  \
+			udelay(delay);                                      \
+		else	                                                    \
+			cpu_relax();                                        \
+}
+
 #endif /* __KERNEL__ */
 #endif /* _ASM_POWERPC_DELAY_H */

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
  2009-05-25 22:15   ` Jon Smirl
@ 2009-05-25 22:15 ` Jon Smirl
  2009-05-26 11:01   ` [alsa-devel] " Mark Brown
  2009-05-25 22:15 ` [PATCH V4 3/5] AC97 driver for mpc5200 Jon Smirl
                   ` (4 subsequent siblings)
  6 siblings, 1 reply; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

Rewrite the mpc5200 audio DMA code to support both I2S and AC97.

Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
---
 sound/soc/fsl/Kconfig           |    1 
 sound/soc/fsl/mpc5200_dma.c     |  442 ++++++++++++++++++++++++---------------
 sound/soc/fsl/mpc5200_dma.h     |   43 ++--
 sound/soc/fsl/mpc5200_psc_i2s.c |  247 ++++------------------
 sound/soc/fsl/mpc5200_psc_i2s.h |   12 +
 5 files changed, 356 insertions(+), 389 deletions(-)
 create mode 100644 sound/soc/fsl/mpc5200_psc_i2s.h

diff --git a/sound/soc/fsl/Kconfig b/sound/soc/fsl/Kconfig
index dc79bdf..1918c78 100644
--- a/sound/soc/fsl/Kconfig
+++ b/sound/soc/fsl/Kconfig
@@ -25,7 +25,6 @@ config SND_SOC_MPC8610_HPCD
 config SND_SOC_MPC5200_I2S
 	tristate "Freescale MPC5200 PSC in I2S mode driver"
 	depends on PPC_MPC52xx && PPC_BESTCOMM
-	select SND_SOC_OF_SIMPLE
 	select SND_MPC52xx_DMA
 	select PPC_BESTCOMM_GEN_BD
 	help
diff --git a/sound/soc/fsl/mpc5200_dma.c b/sound/soc/fsl/mpc5200_dma.c
index 6850392..efec33a 100644
--- a/sound/soc/fsl/mpc5200_dma.c
+++ b/sound/soc/fsl/mpc5200_dma.c
@@ -3,23 +3,13 @@
  * ALSA SoC Platform driver
  *
  * Copyright (C) 2008 Secret Lab Technologies Ltd.
+ * Copyright (C) 2009 Jon Smirl, Digispeaker
  */
 
-#include <linux/init.h>
 #include <linux/module.h>
-#include <linux/interrupt.h>
-#include <linux/device.h>
-#include <linux/delay.h>
 #include <linux/of_device.h>
-#include <linux/of_platform.h>
-#include <linux/dma-mapping.h>
 
-#include <sound/core.h>
-#include <sound/pcm.h>
-#include <sound/pcm_params.h>
-#include <sound/initval.h>
 #include <sound/soc.h>
-#include <sound/soc-of-simple.h>
 
 #include <sysdev/bestcomm/bestcomm.h>
 #include <sysdev/bestcomm/gen_bd.h>
@@ -27,10 +17,6 @@
 
 #include "mpc5200_dma.h"
 
-MODULE_AUTHOR("Grant Likely <grant.likely@secretlab.ca>");
-MODULE_DESCRIPTION("Freescale MPC5200 PSC in DMA mode ASoC Driver");
-MODULE_LICENSE("GPL");
-
 /*
  * Interrupt handlers
  */
@@ -50,7 +36,7 @@ static irqreturn_t psc_dma_status_irq(int irq, void *_psc_dma)
 	if (psc_dma->capture.active && (isr & MPC52xx_PSC_IMR_ORERR))
 		psc_dma->stats.overrun_count++;
 
-	out_8(&regs->command, 4 << 4);	/* reset the error status */
+	out_8(&regs->command, MPC52xx_PSC_RST_ERR_STAT);
 
 	return IRQ_HANDLED;
 }
@@ -81,21 +67,36 @@ static void psc_dma_bcom_enqueue_next_buffer(struct psc_dma_stream *s)
 		s->period_next_pt = s->period_start;
 }
 
+static void psc_dma_bcom_enqueue_tx(struct psc_dma_stream *s)
+{
+	while (s->appl_ptr < s->runtime->control->appl_ptr) {
+
+		if (bcom_queue_full(s->bcom_task))
+			return;
+
+		s->appl_ptr += s->period_size;
+
+		psc_dma_bcom_enqueue_next_buffer(s);
+	}
+}
+
 /* Bestcomm DMA irq handler */
-static irqreturn_t psc_dma_bcom_irq(int irq, void *_psc_dma_stream)
+static irqreturn_t psc_dma_bcom_irq_tx(int irq, void *_psc_dma_stream)
 {
 	struct psc_dma_stream *s = _psc_dma_stream;
 
+	spin_lock(&s->psc_dma->lock);
 	/* For each finished period, dequeue the completed period buffer
 	 * and enqueue a new one in it's place. */
 	while (bcom_buffer_done(s->bcom_task)) {
 		bcom_retrieve_buffer(s->bcom_task, NULL, NULL);
+
 		s->period_current_pt += s->period_bytes;
 		if (s->period_current_pt >= s->period_end)
 			s->period_current_pt = s->period_start;
-		psc_dma_bcom_enqueue_next_buffer(s);
-		bcom_enable(s->bcom_task);
 	}
+	psc_dma_bcom_enqueue_tx(s);
+	spin_unlock(&s->psc_dma->lock);
 
 	/* If the stream is active, then also inform the PCM middle layer
 	 * of the period finished event. */
@@ -105,49 +106,33 @@ static irqreturn_t psc_dma_bcom_irq(int irq, void *_psc_dma_stream)
 	return IRQ_HANDLED;
 }
 
-/**
- * psc_dma_startup: create a new substream
- *
- * This is the first function called when a stream is opened.
- *
- * If this is the first stream open, then grab the IRQ and program most of
- * the PSC registers.
- */
-int psc_dma_startup(struct snd_pcm_substream *substream,
-			   struct snd_soc_dai *dai)
+static irqreturn_t psc_dma_bcom_irq_rx(int irq, void *_psc_dma_stream)
 {
-	struct snd_soc_pcm_runtime *rtd = substream->private_data;
-	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
-	int rc;
+	struct psc_dma_stream *s = _psc_dma_stream;
 
-	dev_dbg(psc_dma->dev, "psc_dma_startup(substream=%p)\n", substream);
+	spin_lock(&s->psc_dma->lock);
+	/* For each finished period, dequeue the completed period buffer
+	 * and enqueue a new one in it's place. */
+	while (bcom_buffer_done(s->bcom_task)) {
+		bcom_retrieve_buffer(s->bcom_task, NULL, NULL);
 
-	if (!psc_dma->playback.active &&
-	    !psc_dma->capture.active) {
-		/* Setup the IRQs */
-		rc = request_irq(psc_dma->irq, &psc_dma_status_irq, IRQF_SHARED,
-				 "psc-dma-status", psc_dma);
-		rc |= request_irq(psc_dma->capture.irq,
-				  &psc_dma_bcom_irq, IRQF_SHARED,
-				  "psc-dma-capture", &psc_dma->capture);
-		rc |= request_irq(psc_dma->playback.irq,
-				  &psc_dma_bcom_irq, IRQF_SHARED,
-				  "psc-dma-playback", &psc_dma->playback);
-		if (rc) {
-			free_irq(psc_dma->irq, psc_dma);
-			free_irq(psc_dma->capture.irq,
-				 &psc_dma->capture);
-			free_irq(psc_dma->playback.irq,
-				 &psc_dma->playback);
-			return -ENODEV;
-		}
+		s->period_current_pt += s->period_bytes;
+		if (s->period_current_pt >= s->period_end)
+			s->period_current_pt = s->period_start;
+
+		psc_dma_bcom_enqueue_next_buffer(s);
 	}
+	spin_unlock(&s->psc_dma->lock);
 
-	return 0;
+	/* If the stream is active, then also inform the PCM middle layer
+	 * of the period finished event. */
+	if (s->active)
+		snd_pcm_period_elapsed(s->stream);
+
+	return IRQ_HANDLED;
 }
 
-int psc_dma_hw_free(struct snd_pcm_substream *substream,
-			   struct snd_soc_dai *dai)
+static int psc_dma_hw_free(struct snd_pcm_substream *substream)
 {
 	snd_pcm_set_runtime_buffer(substream, NULL);
 	return 0;
@@ -159,8 +144,7 @@ int psc_dma_hw_free(struct snd_pcm_substream *substream,
  * This function is called by ALSA to start, stop, pause, and resume the DMA
  * transfer of data.
  */
-int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
-			   struct snd_soc_dai *dai)
+static int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd)
 {
 	struct snd_soc_pcm_runtime *rtd = substream->private_data;
 	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
@@ -168,8 +152,8 @@ int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
 	struct psc_dma_stream *s;
 	struct mpc52xx_psc __iomem *regs = psc_dma->psc_regs;
 	u16 imr;
-	u8 psc_cmd;
 	unsigned long flags;
+	int i;
 
 	if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
 		s = &psc_dma->capture;
@@ -189,68 +173,48 @@ int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
 				(s->period_bytes * runtime->periods);
 		s->period_next_pt = s->period_start;
 		s->period_current_pt = s->period_start;
+		s->period_size = runtime->period_size;
 		s->active = 1;
 
-		/* First; reset everything */
+		/* track appl_ptr so that we have a better chance of detecting
+		 * end of stream and not over running it.
+		 */
+		s->runtime = runtime;
+		s->appl_ptr = s->runtime->control->appl_ptr -
+				(runtime->period_size * runtime->periods);
+
+		/* Fill up the bestcomm bd queue and enable DMA.
+		 * This will begin filling the PSC's fifo.
+		 */
+		spin_lock_irqsave(&psc_dma->lock, flags);
+
 		if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE) {
-			out_8(&regs->command, MPC52xx_PSC_RST_RX);
-			out_8(&regs->command, MPC52xx_PSC_RST_ERR_STAT);
+			bcom_gen_bd_rx_reset(s->bcom_task);
+			for (i = 0; i < runtime->periods; i++)
+				if (!bcom_queue_full(s->bcom_task))
+					psc_dma_bcom_enqueue_next_buffer(s);
 		} else {
-			out_8(&regs->command, MPC52xx_PSC_RST_TX);
-			out_8(&regs->command, MPC52xx_PSC_RST_ERR_STAT);
+			bcom_gen_bd_tx_reset(s->bcom_task);
+			psc_dma_bcom_enqueue_tx(s);
 		}
 
-		/* Next, fill up the bestcomm bd queue and enable DMA.
-		 * This will begin filling the PSC's fifo. */
-		if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
-			bcom_gen_bd_rx_reset(s->bcom_task);
-		else
-			bcom_gen_bd_tx_reset(s->bcom_task);
-		while (!bcom_queue_full(s->bcom_task))
-			psc_dma_bcom_enqueue_next_buffer(s);
 		bcom_enable(s->bcom_task);
-
-		/* Due to errata in the dma mode; need to line up enabling
-		 * the transmitter with a transition on the frame sync
-		 * line */
-
-		spin_lock_irqsave(&psc_dma->lock, flags);
-		/* first make sure it is low */
-		while ((in_8(&regs->ipcr_acr.ipcr) & 0x80) != 0)
-			;
-		/* then wait for the transition to high */
-		while ((in_8(&regs->ipcr_acr.ipcr) & 0x80) == 0)
-			;
-		/* Finally, enable the PSC.
-		 * Receiver must always be enabled; even when we only want
-		 * transmit.  (see 15.3.2.3 of MPC5200B User's Guide) */
-		psc_cmd = MPC52xx_PSC_RX_ENABLE;
-		if (substream->pstr->stream == SNDRV_PCM_STREAM_PLAYBACK)
-			psc_cmd |= MPC52xx_PSC_TX_ENABLE;
-		out_8(&regs->command, psc_cmd);
 		spin_unlock_irqrestore(&psc_dma->lock, flags);
 
+		out_8(&regs->command, MPC52xx_PSC_RST_ERR_STAT);
+
 		break;
 
 	case SNDRV_PCM_TRIGGER_STOP:
-		/* Turn off the PSC */
 		s->active = 0;
-		if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE) {
-			if (!psc_dma->playback.active) {
-				out_8(&regs->command, 2 << 4);	/* reset rx */
-				out_8(&regs->command, 3 << 4);	/* reset tx */
-				out_8(&regs->command, 4 << 4);	/* reset err */
-			}
-		} else {
-			out_8(&regs->command, 3 << 4);	/* reset tx */
-			out_8(&regs->command, 4 << 4);	/* reset err */
-			if (!psc_dma->capture.active)
-				out_8(&regs->command, 2 << 4);	/* reset rx */
-		}
 
+		spin_lock_irqsave(&psc_dma->lock, flags);
 		bcom_disable(s->bcom_task);
-		while (!bcom_queue_empty(s->bcom_task))
-			bcom_retrieve_buffer(s->bcom_task, NULL, NULL);
+		if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
+			bcom_gen_bd_rx_reset(s->bcom_task);
+		else
+			bcom_gen_bd_tx_reset(s->bcom_task);
+		spin_unlock_irqrestore(&psc_dma->lock, flags);
 
 		break;
 
@@ -265,44 +229,11 @@ int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
 		imr |= MPC52xx_PSC_IMR_TXEMP;
 	if (psc_dma->capture.active)
 		imr |= MPC52xx_PSC_IMR_ORERR;
-	out_be16(&regs->isr_imr.imr, imr);
+	out_be16(&regs->isr_imr.imr, psc_dma->imr | imr);
 
 	return 0;
 }
 
-/**
- * psc_dma_shutdown: shutdown the data transfer on a stream
- *
- * Shutdown the PSC if there are no other substreams open.
- */
-void psc_dma_shutdown(struct snd_pcm_substream *substream,
-			     struct snd_soc_dai *dai)
-{
-	struct snd_soc_pcm_runtime *rtd = substream->private_data;
-	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
-
-	dev_dbg(psc_dma->dev, "psc_dma_shutdown(substream=%p)\n", substream);
-
-	/*
-	 * If this is the last active substream, disable the PSC and release
-	 * the IRQ.
-	 */
-	if (!psc_dma->playback.active &&
-	    !psc_dma->capture.active) {
-
-		/* Disable all interrupts and reset the PSC */
-		out_be16(&psc_dma->psc_regs->isr_imr.imr, 0);
-		out_8(&psc_dma->psc_regs->command, 3 << 4); /* reset tx */
-		out_8(&psc_dma->psc_regs->command, 2 << 4); /* reset rx */
-		out_8(&psc_dma->psc_regs->command, 1 << 4); /* reset mode */
-		out_8(&psc_dma->psc_regs->command, 4 << 4); /* reset error */
-
-		/* Release irqs */
-		free_irq(psc_dma->irq, psc_dma);
-		free_irq(psc_dma->capture.irq, &psc_dma->capture);
-		free_irq(psc_dma->playback.irq, &psc_dma->playback);
-	}
-}
 
 /* ---------------------------------------------------------------------
  * The PSC DMA 'ASoC platform' driver
@@ -312,62 +243,78 @@ void psc_dma_shutdown(struct snd_pcm_substream *substream,
  * interaction with the attached codec
  */
 
-static const struct snd_pcm_hardware psc_dma_pcm_hardware = {
+static const struct snd_pcm_hardware psc_dma_hardware = {
 	.info = SNDRV_PCM_INFO_MMAP | SNDRV_PCM_INFO_MMAP_VALID |
 		SNDRV_PCM_INFO_INTERLEAVED | SNDRV_PCM_INFO_BLOCK_TRANSFER |
 		SNDRV_PCM_INFO_BATCH,
 	.formats = SNDRV_PCM_FMTBIT_S8 | SNDRV_PCM_FMTBIT_S16_BE |
-		   SNDRV_PCM_FMTBIT_S24_BE | SNDRV_PCM_FMTBIT_S32_BE,
+		SNDRV_PCM_FMTBIT_S24_BE | SNDRV_PCM_FMTBIT_S32_BE,
 	.rate_min = 8000,
 	.rate_max = 48000,
-	.channels_min = 2,
+	.channels_min = 1,
 	.channels_max = 2,
 	.period_bytes_max	= 1024 * 1024,
 	.period_bytes_min	= 32,
 	.periods_min		= 2,
 	.periods_max		= 256,
 	.buffer_bytes_max	= 2 * 1024 * 1024,
-	.fifo_size		= 0,
+	.fifo_size		= 512,
 };
 
-static int psc_dma_pcm_open(struct snd_pcm_substream *substream)
+static int psc_dma_open(struct snd_pcm_substream *substream)
 {
+	struct snd_pcm_runtime *runtime = substream->runtime;
 	struct snd_soc_pcm_runtime *rtd = substream->private_data;
 	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
 	struct psc_dma_stream *s;
+	int rc;
 
-	dev_dbg(psc_dma->dev, "psc_dma_pcm_open(substream=%p)\n", substream);
+	dev_dbg(psc_dma->dev, "psc_dma_open(substream=%p)\n", substream);
 
 	if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
 		s = &psc_dma->capture;
 	else
 		s = &psc_dma->playback;
 
-	snd_soc_set_runtime_hwparams(substream, &psc_dma_pcm_hardware);
+	snd_soc_set_runtime_hwparams(substream, &psc_dma_hardware);
+
+	rc = snd_pcm_hw_constraint_integer(runtime,
+		SNDRV_PCM_HW_PARAM_PERIODS);
+	if (rc < 0) {
+		dev_err(substream->pcm->card->dev, "invalid buffer size\n");
+		return rc;
+	}
 
 	s->stream = substream;
 	return 0;
 }
 
-static int psc_dma_pcm_close(struct snd_pcm_substream *substream)
+static int psc_dma_close(struct snd_pcm_substream *substream)
 {
 	struct snd_soc_pcm_runtime *rtd = substream->private_data;
 	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
 	struct psc_dma_stream *s;
 
-	dev_dbg(psc_dma->dev, "psc_dma_pcm_close(substream=%p)\n", substream);
+	dev_dbg(psc_dma->dev, "psc_dma_close(substream=%p)\n", substream);
 
 	if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
 		s = &psc_dma->capture;
 	else
 		s = &psc_dma->playback;
 
+	if (!psc_dma->playback.active &&
+	    !psc_dma->capture.active) {
+
+		/* Disable all interrupts and reset the PSC */
+		out_be16(&psc_dma->psc_regs->isr_imr.imr, psc_dma->imr);
+		out_8(&psc_dma->psc_regs->command, 4 << 4); /* reset error */
+	}
 	s->stream = NULL;
 	return 0;
 }
 
 static snd_pcm_uframes_t
-psc_dma_pcm_pointer(struct snd_pcm_substream *substream)
+psc_dma_pointer(struct snd_pcm_substream *substream)
 {
 	struct snd_soc_pcm_runtime *rtd = substream->private_data;
 	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
@@ -384,60 +331,78 @@ psc_dma_pcm_pointer(struct snd_pcm_substream *substream)
 	return bytes_to_frames(substream->runtime, count);
 }
 
-static struct snd_pcm_ops psc_dma_pcm_ops = {
-	.open		= psc_dma_pcm_open,
-	.close		= psc_dma_pcm_close,
+static int
+psc_dma_hw_params(struct snd_pcm_substream *substream,
+			 struct snd_pcm_hw_params *params)
+{
+	snd_pcm_set_runtime_buffer(substream, &substream->dma_buffer);
+
+	return 0;
+}
+
+static struct snd_pcm_ops psc_dma_ops = {
+	.open		= psc_dma_open,
+	.close		= psc_dma_close,
+	.hw_free	= psc_dma_hw_free,
 	.ioctl		= snd_pcm_lib_ioctl,
-	.pointer	= psc_dma_pcm_pointer,
+	.pointer	= psc_dma_pointer,
+	.trigger	= psc_dma_trigger,
+	.hw_params	= psc_dma_hw_params,
 };
 
-static u64 psc_dma_pcm_dmamask = 0xffffffff;
-static int psc_dma_pcm_new(struct snd_card *card, struct snd_soc_dai *dai,
+static u64 psc_dma_dmamask = 0xffffffff;
+static int psc_dma_new(struct snd_card *card, struct snd_soc_dai *dai,
 			   struct snd_pcm *pcm)
 {
 	struct snd_soc_pcm_runtime *rtd = pcm->private_data;
-	size_t size = psc_dma_pcm_hardware.buffer_bytes_max;
+	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
+	size_t size = psc_dma_hardware.buffer_bytes_max;
 	int rc = 0;
 
-	dev_dbg(rtd->socdev->dev, "psc_dma_pcm_new(card=%p, dai=%p, pcm=%p)\n",
+	dev_dbg(rtd->socdev->dev, "psc_dma_new(card=%p, dai=%p, pcm=%p)\n",
 		card, dai, pcm);
 
 	if (!card->dev->dma_mask)
-		card->dev->dma_mask = &psc_dma_pcm_dmamask;
+		card->dev->dma_mask = &psc_dma_dmamask;
 	if (!card->dev->coherent_dma_mask)
 		card->dev->coherent_dma_mask = 0xffffffff;
 
 	if (pcm->streams[0].substream) {
-		rc = snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, pcm->dev, size,
-					&pcm->streams[0].substream->dma_buffer);
+		rc = snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, pcm->card->dev,
+				size, &pcm->streams[0].substream->dma_buffer);
 		if (rc)
 			goto playback_alloc_err;
 	}
 
 	if (pcm->streams[1].substream) {
-		rc = snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, pcm->dev, size,
-					&pcm->streams[1].substream->dma_buffer);
+		rc = snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, pcm->card->dev,
+				size, &pcm->streams[1].substream->dma_buffer);
 		if (rc)
 			goto capture_alloc_err;
 	}
 
+	if (rtd->socdev->card->codec->ac97)
+		rtd->socdev->card->codec->ac97->private_data = psc_dma;
+
 	return 0;
 
  capture_alloc_err:
 	if (pcm->streams[0].substream)
 		snd_dma_free_pages(&pcm->streams[0].substream->dma_buffer);
+
  playback_alloc_err:
 	dev_err(card->dev, "Cannot allocate buffer(s)\n");
+
 	return -ENOMEM;
 }
 
-static void psc_dma_pcm_free(struct snd_pcm *pcm)
+static void psc_dma_free(struct snd_pcm *pcm)
 {
 	struct snd_soc_pcm_runtime *rtd = pcm->private_data;
 	struct snd_pcm_substream *substream;
 	int stream;
 
-	dev_dbg(rtd->socdev->dev, "psc_dma_pcm_free(pcm=%p)\n", pcm);
+	dev_dbg(rtd->socdev->dev, "psc_dma_free(pcm=%p)\n", pcm);
 
 	for (stream = 0; stream < 2; stream++) {
 		substream = pcm->streams[stream].substream;
@@ -449,10 +414,151 @@ static void psc_dma_pcm_free(struct snd_pcm *pcm)
 	}
 }
 
-struct snd_soc_platform psc_dma_pcm_soc_platform = {
+struct snd_soc_platform mpc5200_audio_dma_platform = {
 	.name		= "mpc5200-psc-audio",
-	.pcm_ops	= &psc_dma_pcm_ops,
-	.pcm_new	= &psc_dma_pcm_new,
-	.pcm_free	= &psc_dma_pcm_free,
+	.pcm_ops	= &psc_dma_ops,
+	.pcm_new	= &psc_dma_new,
+	.pcm_free	= &psc_dma_free,
 };
+EXPORT_SYMBOL_GPL(mpc5200_audio_dma_platform);
+
+int mpc5200_audio_dma_create(struct of_device *op)
+{
+	phys_addr_t fifo;
+	struct psc_dma *psc_dma;
+	struct resource res;
+	int size, irq, rc;
+	const __be32 *prop;
+	void __iomem *regs;
+
+	/* Fetch the registers and IRQ of the PSC */
+	irq = irq_of_parse_and_map(op->node, 0);
+	if (of_address_to_resource(op->node, 0, &res)) {
+		dev_err(&op->dev, "Missing reg property\n");
+		return -ENODEV;
+	}
+	regs = ioremap(res.start, 1 + res.end - res.start);
+	if (!regs) {
+		dev_err(&op->dev, "Could not map registers\n");
+		return -ENODEV;
+	}
+
+	/* Allocate and initialize the driver private data */
+	psc_dma = kzalloc(sizeof *psc_dma, GFP_KERNEL);
+	if (!psc_dma) {
+		iounmap(regs);
+		return -ENOMEM;
+	}
+
+	/* Get the PSC ID */
+	prop = of_get_property(op->node, "cell-index", &size);
+	if (!prop || size < sizeof *prop)
+		return -ENODEV;
+
+	spin_lock_init(&psc_dma->lock);
+	psc_dma->id = be32_to_cpu(*prop);
+	psc_dma->irq = irq;
+	psc_dma->psc_regs = regs;
+	psc_dma->fifo_regs = regs + sizeof *psc_dma->psc_regs;
+	psc_dma->dev = &op->dev;
+	psc_dma->playback.psc_dma = psc_dma;
+	psc_dma->capture.psc_dma = psc_dma;
+	snprintf(psc_dma->name, sizeof psc_dma->name, "PSC%u", psc_dma->id);
+
+	/* Find the address of the fifo data registers and setup the
+	 * DMA tasks */
+	fifo = res.start + offsetof(struct mpc52xx_psc, buffer.buffer_32);
+	psc_dma->capture.bcom_task =
+		bcom_psc_gen_bd_rx_init(psc_dma->id, 10, fifo, 512);
+	psc_dma->playback.bcom_task =
+		bcom_psc_gen_bd_tx_init(psc_dma->id, 10, fifo);
+	if (!psc_dma->capture.bcom_task ||
+	    !psc_dma->playback.bcom_task) {
+		dev_err(&op->dev, "Could not allocate bestcomm tasks\n");
+		iounmap(regs);
+		kfree(psc_dma);
+		return -ENODEV;
+	}
+
+	/* Disable all interrupts and reset the PSC */
+	out_be16(&psc_dma->psc_regs->isr_imr.imr, psc_dma->imr);
+	 /* reset receiver */
+	out_8(&psc_dma->psc_regs->command, MPC52xx_PSC_RST_RX);
+	 /* reset transmitter */
+	out_8(&psc_dma->psc_regs->command, MPC52xx_PSC_RST_TX);
+	 /* reset error */
+	out_8(&psc_dma->psc_regs->command, MPC52xx_PSC_RST_ERR_STAT);
+	 /* reset mode */
+	out_8(&psc_dma->psc_regs->command, MPC52xx_PSC_SEL_MODE_REG_1);
+
+	/* Set up mode register;
+	 * First write: RxRdy (FIFO Alarm) generates rx FIFO irq
+	 * Second write: register Normal mode for non loopback
+	 */
+	out_8(&psc_dma->psc_regs->mode, 0);
+	out_8(&psc_dma->psc_regs->mode, 0);
+
+	/* Set the TX and RX fifo alarm thresholds */
+	out_be16(&psc_dma->fifo_regs->rfalarm, 0x100);
+	out_8(&psc_dma->fifo_regs->rfcntl, 0x4);
+	out_be16(&psc_dma->fifo_regs->tfalarm, 0x100);
+	out_8(&psc_dma->fifo_regs->tfcntl, 0x7);
+
+	/* Lookup the IRQ numbers */
+	psc_dma->playback.irq =
+		bcom_get_task_irq(psc_dma->playback.bcom_task);
+	psc_dma->capture.irq =
+		bcom_get_task_irq(psc_dma->capture.bcom_task);
+
+	rc = request_irq(psc_dma->irq, &psc_dma_status_irq, IRQF_SHARED,
+			 "psc-dma-status", psc_dma);
+	rc |= request_irq(psc_dma->capture.irq,
+			  &psc_dma_bcom_irq_rx, IRQF_SHARED,
+			  "psc-dma-capture", &psc_dma->capture);
+	rc |= request_irq(psc_dma->playback.irq,
+			  &psc_dma_bcom_irq_tx, IRQF_SHARED,
+			  "psc-dma-playback", &psc_dma->playback);
+	if (rc) {
+		free_irq(psc_dma->irq, psc_dma);
+		free_irq(psc_dma->capture.irq,
+			 &psc_dma->capture);
+		free_irq(psc_dma->playback.irq,
+			 &psc_dma->playback);
+		return -ENODEV;
+	}
 
+	/* Save what we've done so it can be found again later */
+	dev_set_drvdata(&op->dev, psc_dma);
+
+	/* Tell the ASoC OF helpers about it */
+	return snd_soc_register_platform(&mpc5200_audio_dma_platform);
+}
+EXPORT_SYMBOL_GPL(mpc5200_audio_dma_create);
+
+int mpc5200_audio_dma_destroy(struct of_device *op)
+{
+	struct psc_dma *psc_dma = dev_get_drvdata(&op->dev);
+
+	dev_dbg(&op->dev, "mpc5200_audio_dma_destroy()\n");
+
+	snd_soc_unregister_platform(&mpc5200_audio_dma_platform);
+
+	bcom_gen_bd_rx_release(psc_dma->capture.bcom_task);
+	bcom_gen_bd_tx_release(psc_dma->playback.bcom_task);
+
+	/* Release irqs */
+	free_irq(psc_dma->irq, psc_dma);
+	free_irq(psc_dma->capture.irq, &psc_dma->capture);
+	free_irq(psc_dma->playback.irq, &psc_dma->playback);
+
+	iounmap(psc_dma->psc_regs);
+	kfree(psc_dma);
+	dev_set_drvdata(&op->dev, NULL);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(mpc5200_audio_dma_destroy);
+
+MODULE_AUTHOR("Grant Likely <grant.likely@secretlab.ca>");
+MODULE_DESCRIPTION("Freescale MPC5200 PSC in DMA mode ASoC Driver");
+MODULE_LICENSE("GPL");
diff --git a/sound/soc/fsl/mpc5200_dma.h b/sound/soc/fsl/mpc5200_dma.h
index a33232c..7414c6f 100644
--- a/sound/soc/fsl/mpc5200_dma.h
+++ b/sound/soc/fsl/mpc5200_dma.h
@@ -5,8 +5,10 @@
 #ifndef __SOUND_SOC_FSL_MPC5200_DMA_H__
 #define __SOUND_SOC_FSL_MPC5200_DMA_H__
 
+#define PSC_STREAM_NAME_LEN 32
+
 /**
- * psc_dma_stream - Data specific to a single stream (playback or capture)
+ * psc_ac97_stream - Data specific to a single stream (playback or capture)
  * @active:		flag indicating if the stream is active
  * @psc_dma:		pointer back to parent psc_dma data structure
  * @bcom_task:		bestcomm task structure
@@ -17,6 +19,9 @@
  * @period_bytes:	size of DMA period in bytes
  */
 struct psc_dma_stream {
+	struct snd_pcm_runtime *runtime;
+	snd_pcm_uframes_t appl_ptr;
+
 	int active;
 	struct psc_dma *psc_dma;
 	struct bcom_task *bcom_task;
@@ -27,6 +32,7 @@ struct psc_dma_stream {
 	dma_addr_t period_next_pt;
 	dma_addr_t period_current_pt;
 	int period_bytes;
+	int period_size;
 };
 
 /**
@@ -48,9 +54,12 @@ struct psc_dma {
 	struct mpc52xx_psc_fifo __iomem *fifo_regs;
 	unsigned int irq;
 	struct device *dev;
-	struct snd_soc_dai dai;
 	spinlock_t lock;
 	u32 sicr;
+	uint sysclk;
+	int imr;
+	int id;
+	unsigned int slots;
 
 	/* per-stream data */
 	struct psc_dma_stream playback;
@@ -58,24 +67,28 @@ struct psc_dma {
 
 	/* Statistics */
 	struct {
-		int overrun_count;
-		int underrun_count;
+		unsigned long overrun_count;
+		unsigned long underrun_count;
 	} stats;
 };
 
+int mpc5200_audio_dma_create(struct of_device *op);
+int mpc5200_audio_dma_destroy(struct of_device *op);
 
-int psc_dma_startup(struct snd_pcm_substream *substream,
-			   struct snd_soc_dai *dai);
-
-int psc_dma_hw_free(struct snd_pcm_substream *substream,
-			   struct snd_soc_dai *dai);
-
-void psc_dma_shutdown(struct snd_pcm_substream *substream,
-			     struct snd_soc_dai *dai);
+extern struct snd_soc_platform mpc5200_audio_dma_platform;
 
-int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
-			   struct snd_soc_dai *dai);
+/* whack this after Timur's patch is merged in to arch/powerpc/include/asm/delay.h */
+#define spin_event_timeout(condition, timeout, delay, rc)                   \
+{                                                                           \
+       unsigned long __loops = tb_ticks_per_usec * timeout;                \
+       unsigned long __start = get_tbl();                                  \
+       while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
+               if (delay)                                                  \
+                       udelay(delay);                                      \
+               else                                                        \
+                       cpu_relax();                                        \
+}
+/* whack this after Timur's patch is merged in to arch/powerpc/include/asm/delay.h */
 
-extern struct snd_soc_platform psc_dma_pcm_soc_platform;
 
 #endif /* __SOUND_SOC_FSL_MPC5200_DMA_H__ */
diff --git a/sound/soc/fsl/mpc5200_psc_i2s.c b/sound/soc/fsl/mpc5200_psc_i2s.c
index 12a7917..ce8de90 100644
--- a/sound/soc/fsl/mpc5200_psc_i2s.c
+++ b/sound/soc/fsl/mpc5200_psc_i2s.c
@@ -3,34 +3,22 @@
  * ALSA SoC Digital Audio Interface (DAI) driver
  *
  * Copyright (C) 2008 Secret Lab Technologies Ltd.
+ * Copyright (C) 2009 Jon Smirl, Digispeaker
  */
 
-#include <linux/init.h>
 #include <linux/module.h>
-#include <linux/interrupt.h>
-#include <linux/device.h>
-#include <linux/delay.h>
 #include <linux/of_device.h>
 #include <linux/of_platform.h>
-#include <linux/dma-mapping.h>
 
-#include <sound/core.h>
 #include <sound/pcm.h>
 #include <sound/pcm_params.h>
-#include <sound/initval.h>
 #include <sound/soc.h>
-#include <sound/soc-of-simple.h>
 
-#include <sysdev/bestcomm/bestcomm.h>
-#include <sysdev/bestcomm/gen_bd.h>
 #include <asm/mpc52xx_psc.h>
 
+#include "mpc5200_psc_i2s.h"
 #include "mpc5200_dma.h"
 
-MODULE_AUTHOR("Grant Likely <grant.likely@secretlab.ca>");
-MODULE_DESCRIPTION("Freescale MPC5200 PSC in I2S mode ASoC Driver");
-MODULE_LICENSE("GPL");
-
 /**
  * PSC_I2S_RATES: sample rates supported by the I2S
  *
@@ -46,8 +34,7 @@ MODULE_LICENSE("GPL");
  * PSC_I2S_FORMATS: audio formats supported by the PSC I2S mode
  */
 #define PSC_I2S_FORMATS (SNDRV_PCM_FMTBIT_S8 | SNDRV_PCM_FMTBIT_S16_BE | \
-			 SNDRV_PCM_FMTBIT_S24_BE | SNDRV_PCM_FMTBIT_S24_BE | \
-			 SNDRV_PCM_FMTBIT_S32_BE)
+			 SNDRV_PCM_FMTBIT_S24_BE | SNDRV_PCM_FMTBIT_S32_BE)
 
 static int psc_i2s_hw_params(struct snd_pcm_substream *substream,
 				 struct snd_pcm_hw_params *params,
@@ -82,8 +69,6 @@ static int psc_i2s_hw_params(struct snd_pcm_substream *substream,
 	}
 	out_be32(&psc_dma->psc_regs->sicr, psc_dma->sicr | mode);
 
-	snd_pcm_set_runtime_buffer(substream, &substream->dma_buffer);
-
 	return 0;
 }
 
@@ -140,16 +125,13 @@ static int psc_i2s_set_fmt(struct snd_soc_dai *cpu_dai, unsigned int format)
  * psc_i2s_dai_template: template CPU Digital Audio Interface
  */
 static struct snd_soc_dai_ops psc_i2s_dai_ops = {
-	.startup	= psc_dma_startup,
 	.hw_params	= psc_i2s_hw_params,
-	.hw_free	= psc_dma_hw_free,
-	.shutdown	= psc_dma_shutdown,
-	.trigger	= psc_dma_trigger,
 	.set_sysclk	= psc_i2s_set_sysclk,
 	.set_fmt	= psc_i2s_set_fmt,
 };
 
-static struct snd_soc_dai psc_i2s_dai_template = {
+struct snd_soc_dai psc_i2s_dai[] = {{
+	.name   = "I2S",
 	.playback = {
 		.channels_min = 2,
 		.channels_max = 2,
@@ -163,71 +145,8 @@ static struct snd_soc_dai psc_i2s_dai_template = {
 		.formats = PSC_I2S_FORMATS,
 	},
 	.ops = &psc_i2s_dai_ops,
-};
-
-/* ---------------------------------------------------------------------
- * Sysfs attributes for debugging
- */
-
-static ssize_t psc_i2s_status_show(struct device *dev,
-			   struct device_attribute *attr, char *buf)
-{
-	struct psc_dma *psc_dma = dev_get_drvdata(dev);
-
-	return sprintf(buf, "status=%.4x sicr=%.8x rfnum=%i rfstat=0x%.4x "
-			"tfnum=%i tfstat=0x%.4x\n",
-			in_be16(&psc_dma->psc_regs->sr_csr.status),
-			in_be32(&psc_dma->psc_regs->sicr),
-			in_be16(&psc_dma->fifo_regs->rfnum) & 0x1ff,
-			in_be16(&psc_dma->fifo_regs->rfstat),
-			in_be16(&psc_dma->fifo_regs->tfnum) & 0x1ff,
-			in_be16(&psc_dma->fifo_regs->tfstat));
-}
-
-static int *psc_i2s_get_stat_attr(struct psc_dma *psc_dma, const char *name)
-{
-	if (strcmp(name, "playback_underrun") == 0)
-		return &psc_dma->stats.underrun_count;
-	if (strcmp(name, "capture_overrun") == 0)
-		return &psc_dma->stats.overrun_count;
-
-	return NULL;
-}
-
-static ssize_t psc_i2s_stat_show(struct device *dev,
-				 struct device_attribute *attr, char *buf)
-{
-	struct psc_dma *psc_dma = dev_get_drvdata(dev);
-	int *attrib;
-
-	attrib = psc_i2s_get_stat_attr(psc_dma, attr->attr.name);
-	if (!attrib)
-		return 0;
-
-	return sprintf(buf, "%i\n", *attrib);
-}
-
-static ssize_t psc_i2s_stat_store(struct device *dev,
-				  struct device_attribute *attr,
-				  const char *buf,
-				  size_t count)
-{
-	struct psc_dma *psc_dma = dev_get_drvdata(dev);
-	int *attrib;
-
-	attrib = psc_i2s_get_stat_attr(psc_dma, attr->attr.name);
-	if (!attrib)
-		return 0;
-
-	*attrib = simple_strtoul(buf, NULL, 0);
-	return count;
-}
-
-static DEVICE_ATTR(status, 0644, psc_i2s_status_show, NULL);
-static DEVICE_ATTR(playback_underrun, 0644, psc_i2s_stat_show,
-			psc_i2s_stat_store);
-static DEVICE_ATTR(capture_overrun, 0644, psc_i2s_stat_show,
-			psc_i2s_stat_store);
+} };
+EXPORT_SYMBOL_GPL(psc_i2s_dai);
 
 /* ---------------------------------------------------------------------
  * OF platform bus binding code:
@@ -237,82 +156,26 @@ static DEVICE_ATTR(capture_overrun, 0644, psc_i2s_stat_show,
 static int __devinit psc_i2s_of_probe(struct of_device *op,
 				      const struct of_device_id *match)
 {
-	phys_addr_t fifo;
+	int rc;
 	struct psc_dma *psc_dma;
-	struct resource res;
-	int size, psc_id, irq, rc;
-	const __be32 *prop;
-	void __iomem *regs;
-
-	dev_dbg(&op->dev, "probing psc i2s device\n");
-
-	/* Get the PSC ID */
-	prop = of_get_property(op->node, "cell-index", &size);
-	if (!prop || size < sizeof *prop)
-		return -ENODEV;
-	psc_id = be32_to_cpu(*prop);
-
-	/* Fetch the registers and IRQ of the PSC */
-	irq = irq_of_parse_and_map(op->node, 0);
-	if (of_address_to_resource(op->node, 0, &res)) {
-		dev_err(&op->dev, "Missing reg property\n");
-		return -ENODEV;
-	}
-	regs = ioremap(res.start, 1 + res.end - res.start);
-	if (!regs) {
-		dev_err(&op->dev, "Could not map registers\n");
-		return -ENODEV;
-	}
+	struct mpc52xx_psc __iomem *regs;
 
-	/* Allocate and initialize the driver private data */
-	psc_dma = kzalloc(sizeof *psc_dma, GFP_KERNEL);
-	if (!psc_dma) {
-		iounmap(regs);
-		return -ENOMEM;
-	}
-	spin_lock_init(&psc_dma->lock);
-	psc_dma->irq = irq;
-	psc_dma->psc_regs = regs;
-	psc_dma->fifo_regs = regs + sizeof *psc_dma->psc_regs;
-	psc_dma->dev = &op->dev;
-	psc_dma->playback.psc_dma = psc_dma;
-	psc_dma->capture.psc_dma = psc_dma;
-	snprintf(psc_dma->name, sizeof psc_dma->name, "PSC%u", psc_id+1);
-
-	/* Fill out the CPU DAI structure */
-	memcpy(&psc_dma->dai, &psc_i2s_dai_template, sizeof psc_dma->dai);
-	psc_dma->dai.private_data = psc_dma;
-	psc_dma->dai.name = psc_dma->name;
-	psc_dma->dai.id = psc_id;
-
-	/* Find the address of the fifo data registers and setup the
-	 * DMA tasks */
-	fifo = res.start + offsetof(struct mpc52xx_psc, buffer.buffer_32);
-	psc_dma->capture.bcom_task =
-		bcom_psc_gen_bd_rx_init(psc_id, 10, fifo, 512);
-	psc_dma->playback.bcom_task =
-		bcom_psc_gen_bd_tx_init(psc_id, 10, fifo);
-	if (!psc_dma->capture.bcom_task ||
-	    !psc_dma->playback.bcom_task) {
-		dev_err(&op->dev, "Could not allocate bestcomm tasks\n");
-		iounmap(regs);
-		kfree(psc_dma);
-		return -ENODEV;
+	rc = mpc5200_audio_dma_create(op);
+	if (rc != 0)
+		return rc;
+
+	rc = snd_soc_register_dais(psc_i2s_dai, ARRAY_SIZE(psc_i2s_dai));
+	if (rc != 0) {
+		pr_err("Failed to register DAI\n");
+		return 0;
 	}
 
-	/* Disable all interrupts and reset the PSC */
-	out_be16(&psc_dma->psc_regs->isr_imr.imr, 0);
-	out_8(&psc_dma->psc_regs->command, 3 << 4); /* reset transmitter */
-	out_8(&psc_dma->psc_regs->command, 2 << 4); /* reset receiver */
-	out_8(&psc_dma->psc_regs->command, 1 << 4); /* reset mode */
-	out_8(&psc_dma->psc_regs->command, 4 << 4); /* reset error */
+	psc_dma = dev_get_drvdata(&op->dev);
+	regs = psc_dma->psc_regs;
 
 	/* Configure the serial interface mode; defaulting to CODEC8 mode */
 	psc_dma->sicr = MPC52xx_PSC_SICR_DTS1 | MPC52xx_PSC_SICR_I2S |
 			MPC52xx_PSC_SICR_CLKPOL;
-	if (of_get_property(op->node, "fsl,cellslave", NULL))
-		psc_dma->sicr |= MPC52xx_PSC_SICR_CELLSLAVE |
-				 MPC52xx_PSC_SICR_GENCLK;
 	out_be32(&psc_dma->psc_regs->sicr,
 		 psc_dma->sicr | MPC52xx_PSC_SICR_SIM_CODEC_8);
 
@@ -321,66 +184,37 @@ static int __devinit psc_i2s_of_probe(struct of_device *op,
 	if (!of_get_property(op->node, "codec-handle", NULL))
 		return 0;
 
-	/* Set up mode register;
-	 * First write: RxRdy (FIFO Alarm) generates rx FIFO irq
-	 * Second write: register Normal mode for non loopback
-	 */
-	out_8(&psc_dma->psc_regs->mode, 0);
-	out_8(&psc_dma->psc_regs->mode, 0);
-
-	/* Set the TX and RX fifo alarm thresholds */
-	out_be16(&psc_dma->fifo_regs->rfalarm, 0x100);
-	out_8(&psc_dma->fifo_regs->rfcntl, 0x4);
-	out_be16(&psc_dma->fifo_regs->tfalarm, 0x100);
-	out_8(&psc_dma->fifo_regs->tfcntl, 0x7);
-
-	/* Lookup the IRQ numbers */
-	psc_dma->playback.irq =
-		bcom_get_task_irq(psc_dma->playback.bcom_task);
-	psc_dma->capture.irq =
-		bcom_get_task_irq(psc_dma->capture.bcom_task);
-
-	/* Save what we've done so it can be found again later */
-	dev_set_drvdata(&op->dev, psc_dma);
-
-	/* Register the SYSFS files */
-	rc = device_create_file(psc_dma->dev, &dev_attr_status);
-	rc |= device_create_file(psc_dma->dev, &dev_attr_capture_overrun);
-	rc |= device_create_file(psc_dma->dev, &dev_attr_playback_underrun);
-	if (rc)
-		dev_info(psc_dma->dev, "error creating sysfs files\n");
-
-	snd_soc_register_platform(&psc_dma_pcm_soc_platform);
-
-	/* Tell the ASoC OF helpers about it */
-	of_snd_soc_register_platform(&psc_dma_pcm_soc_platform, op->node,
-				     &psc_dma->dai);
+	/* Due to errata in the dma mode; need to line up enabling
+	 * the transmitter with a transition on the frame sync
+	 * line */
+
+	/* first make sure it is low */
+	while ((in_8(&regs->ipcr_acr.ipcr) & 0x80) != 0)
+		;
+	/* then wait for the transition to high */
+	while ((in_8(&regs->ipcr_acr.ipcr) & 0x80) == 0)
+		;
+	/* Finally, enable the PSC.
+	 * Receiver must always be enabled; even when we only want
+	 * transmit.  (see 15.3.2.3 of MPC5200B User's Guide) */
+
+	/* Go */
+	out_8(&psc_dma->psc_regs->command,
+			MPC52xx_PSC_TX_ENABLE | MPC52xx_PSC_RX_ENABLE);
 
 	return 0;
+
 }
 
 static int __devexit psc_i2s_of_remove(struct of_device *op)
 {
-	struct psc_dma *psc_dma = dev_get_drvdata(&op->dev);
-
-	dev_dbg(&op->dev, "psc_i2s_remove()\n");
-
-	snd_soc_unregister_platform(&psc_dma_pcm_soc_platform);
-
-	bcom_gen_bd_rx_release(psc_dma->capture.bcom_task);
-	bcom_gen_bd_tx_release(psc_dma->playback.bcom_task);
-
-	iounmap(psc_dma->psc_regs);
-	iounmap(psc_dma->fifo_regs);
-	kfree(psc_dma);
-	dev_set_drvdata(&op->dev, NULL);
-
-	return 0;
+	return mpc5200_audio_dma_destroy(op);
 }
 
 /* Match table for of_platform binding */
 static struct of_device_id psc_i2s_match[] __devinitdata = {
 	{ .compatible = "fsl,mpc5200-psc-i2s", },
+	{ .compatible = "fsl,mpc5200b-psc-i2s", },
 	{}
 };
 MODULE_DEVICE_TABLE(of, psc_i2s_match);
@@ -411,4 +245,7 @@ static void __exit psc_i2s_exit(void)
 }
 module_exit(psc_i2s_exit);
 
+MODULE_AUTHOR("Grant Likely <grant.likely@secretlab.ca>");
+MODULE_DESCRIPTION("Freescale MPC5200 PSC in I2S mode ASoC Driver");
+MODULE_LICENSE("GPL");
 
diff --git a/sound/soc/fsl/mpc5200_psc_i2s.h b/sound/soc/fsl/mpc5200_psc_i2s.h
new file mode 100644
index 0000000..ce55e07
--- /dev/null
+++ b/sound/soc/fsl/mpc5200_psc_i2s.h
@@ -0,0 +1,12 @@
+/*
+ * Freescale MPC5200 PSC in I2S mode
+ * ALSA SoC Digital Audio Interface (DAI) driver
+ *
+ */
+
+#ifndef __SOUND_SOC_FSL_MPC52xx_PSC_I2S_H__
+#define __SOUND_SOC_FSL_MPC52xx_PSC_I2S_H__
+
+extern struct snd_soc_dai psc_i2s_dai[];
+
+#endif /* __SOUND_SOC_FSL_MPC52xx_PSC_I2S_H__ */

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* [PATCH V4 3/5] AC97 driver for mpc5200
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
  2009-05-25 22:15   ` Jon Smirl
  2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
@ 2009-05-25 22:15 ` Jon Smirl
  2009-05-26  3:41     ` [alsa-devel] " Timur Tabi
  2009-05-25 22:15 ` [PATCH V4 4/5] Support for AC97 on Phytec pmc030 base board Jon Smirl
                   ` (3 subsequent siblings)
  6 siblings, 1 reply; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

AC97 driver for mpc5200

I've implemented retries for when the AC97 hardware doesn't reset on
first try. About 10% of the time both the Efika and pcm030 AC97 codecs
don't reset on first try and need to be poked multiple times.  Failure
is indicated by not having the link clock start ticking. Every once in
a while even five pokes won't get the link started and I have to power
cycle.

Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
---
 sound/soc/fsl/Kconfig            |   11 +
 sound/soc/fsl/Makefile           |    1 
 sound/soc/fsl/mpc5200_psc_ac97.c |  331 ++++++++++++++++++++++++++++++++++++++
 sound/soc/fsl/mpc5200_psc_ac97.h |   15 ++
 4 files changed, 358 insertions(+), 0 deletions(-)
 create mode 100644 sound/soc/fsl/mpc5200_psc_ac97.c
 create mode 100644 sound/soc/fsl/mpc5200_psc_ac97.h

diff --git a/sound/soc/fsl/Kconfig b/sound/soc/fsl/Kconfig
index 1918c78..3bce952 100644
--- a/sound/soc/fsl/Kconfig
+++ b/sound/soc/fsl/Kconfig
@@ -29,3 +29,14 @@ config SND_SOC_MPC5200_I2S
 	select PPC_BESTCOMM_GEN_BD
 	help
 	  Say Y here to support the MPC5200 PSCs in I2S mode.
+
+config SND_SOC_MPC5200_AC97
+	tristate "Freescale MPC5200 PSC in AC97 mode driver"
+	depends on PPC_MPC52xx && PPC_BESTCOMM
+	select AC97_BUS
+	select SND_MPC52xx_DMA
+	select PPC_BESTCOMM_GEN_BD
+	help
+	  Say Y here to support the MPC5200 PSCs in AC97 mode.
+
+
diff --git a/sound/soc/fsl/Makefile b/sound/soc/fsl/Makefile
index 7731ef2..14631a1 100644
--- a/sound/soc/fsl/Makefile
+++ b/sound/soc/fsl/Makefile
@@ -13,4 +13,5 @@ obj-$(CONFIG_SND_SOC_MPC8610) += snd-soc-fsl-ssi.o snd-soc-fsl-dma.o
 # MPC5200 Platform Support
 obj-$(CONFIG_SND_MPC52xx_DMA) += mpc5200_dma.o
 obj-$(CONFIG_SND_SOC_MPC5200_I2S) += mpc5200_psc_i2s.o
+obj-$(CONFIG_SND_SOC_MPC5200_AC97) += mpc5200_psc_ac97.o
 
diff --git a/sound/soc/fsl/mpc5200_psc_ac97.c b/sound/soc/fsl/mpc5200_psc_ac97.c
new file mode 100644
index 0000000..3e6838c
--- /dev/null
+++ b/sound/soc/fsl/mpc5200_psc_ac97.c
@@ -0,0 +1,331 @@
+/*
+ * linux/sound/mpc5200-ac97.c -- AC97 support for the Freescale MPC52xx chip.
+ *
+ * Copyright (C) 2009 Jon Smirl, Digispeaker
+ * Author: Jon Smirl <jonsmirl@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/of_platform.h>
+
+#include <sound/pcm.h>
+#include <sound/pcm_params.h>
+#include <sound/soc.h>
+
+#include <asm/time.h>
+#include <asm/delay.h>
+#include <asm/mpc52xx_psc.h>
+
+#include "mpc5200_dma.h"
+#include "mpc5200_psc_ac97.h"
+
+#define DRV_NAME "mpc5200-psc-ac97"
+
+/* ALSA only supports a single AC97 device so static is recommend here */
+static struct psc_dma *psc_dma;
+
+static unsigned short psc_ac97_read(struct snd_ac97 *ac97, unsigned short reg)
+{
+	int rc;
+	unsigned int val;
+
+	/* Wait for command send status zero = ready */
+	spin_event_timeout((in_be16(&psc_dma->psc_regs->sr_csr.status) &
+				MPC52xx_PSC_SR_CMDSEND), 100, 0, rc);
+	if (rc != 0) {
+		pr_err("timeout on ac97 bus (rdy)\n");
+		return -ENODEV;
+	}
+	/* Send the read */
+	out_be32(&psc_dma->psc_regs->ac97_cmd, (1<<31) | ((reg & 0x7f) << 24));
+
+	/* Wait for the answer */
+	spin_event_timeout(!(in_be16(&psc_dma->psc_regs->sr_csr.status) &
+				MPC52xx_PSC_SR_DATA_VAL), 100, 0, rc);
+	if (rc != 0) {
+		pr_err("timeout on ac97 read (val) %x\n",
+				in_be16(&psc_dma->psc_regs->sr_csr.status));
+		return -ENODEV;
+	}
+	/* Get the data */
+	val = in_be32(&psc_dma->psc_regs->ac97_data);
+	if (((val >> 24) & 0x7f) != reg) {
+		pr_err("reg echo error on ac97 read\n");
+		return -ENODEV;
+	}
+	val = (val >> 8) & 0xffff;
+
+	return (unsigned short) val;
+}
+
+static void psc_ac97_write(struct snd_ac97 *ac97,
+				unsigned short reg, unsigned short val)
+{
+	int rc;
+
+	/* Wait for command status zero = ready */
+	spin_event_timeout((in_be16(&psc_dma->psc_regs->sr_csr.status) &
+				MPC52xx_PSC_SR_CMDSEND), 100, 0, rc);
+	if (rc != 0) {
+		pr_err("timeout on ac97 bus (write)\n");
+		return;
+	}
+	/* Write data */
+	out_be32(&psc_dma->psc_regs->ac97_cmd,
+			((reg & 0x7f) << 24) | (val << 8));
+}
+
+static void psc_ac97_warm_reset(struct snd_ac97 *ac97)
+{
+	int rc;
+	struct mpc52xx_psc __iomem *regs = psc_dma->psc_regs;
+
+	out_be32(&regs->sicr, psc_dma->sicr | MPC52xx_PSC_SICR_AWR);
+	spin_event_timeout(1, 3, 0, rc);
+	out_be32(&regs->sicr, psc_dma->sicr);
+}
+
+static void psc_ac97_cold_reset(struct snd_ac97 *ac97)
+{
+	int rc;
+	struct mpc52xx_psc __iomem *regs = psc_dma->psc_regs;
+
+	/* Do a cold reset */
+	out_8(&regs->op1, MPC52xx_PSC_OP_RES);
+	spin_event_timeout(1, 10, 0, rc);
+	out_8(&regs->op0, MPC52xx_PSC_OP_RES);
+	spin_event_timeout(1, 50, 0, rc);
+	psc_ac97_warm_reset(ac97);
+}
+
+struct snd_ac97_bus_ops soc_ac97_ops = {
+	.read		= psc_ac97_read,
+	.write		= psc_ac97_write,
+	.reset		= psc_ac97_cold_reset,
+	.warm_reset	= psc_ac97_warm_reset,
+};
+EXPORT_SYMBOL_GPL(soc_ac97_ops);
+
+static int psc_ac97_hw_analog_params(struct snd_pcm_substream *substream,
+				 struct snd_pcm_hw_params *params,
+				 struct snd_soc_dai *cpu_dai)
+{
+	struct psc_dma *psc_dma = cpu_dai->private_data;
+
+	dev_dbg(psc_dma->dev, "%s(substream=%p) p_size=%i p_bytes=%i"
+		" periods=%i buffer_size=%i  buffer_bytes=%i channels=%i"
+		" rate=%i format=%i\n",
+		__func__, substream, params_period_size(params),
+		params_period_bytes(params), params_periods(params),
+		params_buffer_size(params), params_buffer_bytes(params),
+		params_channels(params), params_rate(params),
+		params_format(params));
+
+
+	if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE) {
+		if (params_channels(params) == 1)
+			psc_dma->slots |= 0x00000100;
+		else
+			psc_dma->slots |= 0x00000300;
+	} else {
+		if (params_channels(params) == 1)
+			psc_dma->slots |= 0x01000000;
+		else
+			psc_dma->slots |= 0x03000000;
+	}
+	out_be32(&psc_dma->psc_regs->ac97_slots, psc_dma->slots);
+
+	return 0;
+}
+
+static int psc_ac97_hw_digital_params(struct snd_pcm_substream *substream,
+				 struct snd_pcm_hw_params *params,
+				 struct snd_soc_dai *cpu_dai)
+{
+	struct psc_dma *psc_dma = cpu_dai->private_data;
+
+	if (params_channels(params) == 1)
+		out_be32(&psc_dma->psc_regs->ac97_slots, 0x01000000);
+	else
+		out_be32(&psc_dma->psc_regs->ac97_slots, 0x03000000);
+
+	return 0;
+}
+
+static int psc_ac97_trigger(struct snd_pcm_substream *substream, int cmd,
+							struct snd_soc_dai *dai)
+{
+	struct snd_soc_pcm_runtime *rtd = substream->private_data;
+	struct psc_dma *psc_dma = rtd->dai->cpu_dai->private_data;
+
+	switch (cmd) {
+	case SNDRV_PCM_TRIGGER_STOP:
+		if (substream->pstr->stream == SNDRV_PCM_STREAM_CAPTURE)
+			psc_dma->slots &= 0xFFFF0000;
+		else
+			psc_dma->slots &= 0x0000FFFF;
+
+		out_be32(&psc_dma->psc_regs->ac97_slots, psc_dma->slots);
+		break;
+	}
+	return 0;
+}
+
+static int psc_ac97_probe(struct platform_device *pdev,
+					struct snd_soc_dai *cpu_dai)
+{
+	struct psc_dma *psc_dma = cpu_dai->private_data;
+	struct mpc52xx_psc __iomem *regs = psc_dma->psc_regs;
+
+	/* Go */
+	out_8(&regs->command, MPC52xx_PSC_TX_ENABLE | MPC52xx_PSC_RX_ENABLE);
+	return 0;
+}
+
+/* ---------------------------------------------------------------------
+ * ALSA SoC Bindings
+ *
+ * - Digital Audio Interface (DAI) template
+ * - create/destroy dai hooks
+ */
+
+/**
+ * psc_ac97_dai_template: template CPU Digital Audio Interface
+ */
+static struct snd_soc_dai_ops psc_ac97_analog_ops = {
+	.hw_params	= psc_ac97_hw_analog_params,
+	.trigger	= psc_ac97_trigger,
+};
+
+static struct snd_soc_dai_ops psc_ac97_digital_ops = {
+	.hw_params	= psc_ac97_hw_digital_params,
+};
+
+struct snd_soc_dai psc_ac97_dai[] = {
+{
+	.name   = "AC97",
+	.ac97_control = 1,
+	.probe	= psc_ac97_probe,
+	.playback = {
+		.channels_min   = 1,
+		.channels_max   = 6,
+		.rates          = SNDRV_PCM_RATE_8000_48000,
+		.formats = SNDRV_PCM_FMTBIT_S32_BE,
+	},
+	.capture = {
+		.channels_min   = 1,
+		.channels_max   = 2,
+		.rates          = SNDRV_PCM_RATE_8000_48000,
+		.formats = SNDRV_PCM_FMTBIT_S32_BE,
+	},
+	.ops = &psc_ac97_analog_ops,
+},
+{
+	.name   = "SPDIF",
+	.ac97_control = 1,
+	.playback = {
+		.channels_min   = 1,
+		.channels_max   = 2,
+		.rates          = SNDRV_PCM_RATE_32000 | \
+			SNDRV_PCM_RATE_44100 | SNDRV_PCM_RATE_48000,
+		.formats = SNDRV_PCM_FMTBIT_IEC958_SUBFRAME_BE,
+	},
+	.ops = &psc_ac97_digital_ops,
+} };
+EXPORT_SYMBOL_GPL(psc_ac97_dai);
+
+
+
+/* ---------------------------------------------------------------------
+ * OF platform bus binding code:
+ * - Probe/remove operations
+ * - OF device match table
+ */
+static int __devinit psc_ac97_of_probe(struct of_device *op,
+				      const struct of_device_id *match)
+{
+	int rc, i;
+	struct snd_ac97 ac97;
+	struct mpc52xx_psc __iomem *regs;
+
+	rc = mpc5200_audio_dma_create(op);
+	if (rc != 0)
+		return rc;
+
+	for (i = 0; i < ARRAY_SIZE(psc_ac97_dai); i++)
+		psc_ac97_dai[i].dev = &op->dev;
+
+	rc = snd_soc_register_dais(psc_ac97_dai, ARRAY_SIZE(psc_ac97_dai));
+	if (rc != 0) {
+		dev_err(&op->dev, "Failed to register DAI\n");
+		return rc;
+	}
+
+	psc_dma = dev_get_drvdata(&op->dev);
+	regs = psc_dma->psc_regs;
+	ac97.private_data = psc_dma;
+
+	for (i = 0; i < ARRAY_SIZE(psc_ac97_dai); i++)
+		psc_ac97_dai[i].private_data = psc_dma;
+
+	psc_dma->imr = 0;
+	out_be16(&psc_dma->psc_regs->isr_imr.imr, psc_dma->imr);
+
+	/* Configure the serial interface mode to AC97 */
+	psc_dma->sicr = MPC52xx_PSC_SICR_SIM_AC97 | MPC52xx_PSC_SICR_ENAC97;
+	out_be32(&regs->sicr, psc_dma->sicr);
+
+	/* No slots active */
+	out_be32(&regs->ac97_slots, 0x00000000);
+
+	return 0;
+}
+
+static int __devexit psc_ac97_of_remove(struct of_device *op)
+{
+	return mpc5200_audio_dma_destroy(op);
+}
+
+/* Match table for of_platform binding */
+static struct of_device_id psc_ac97_match[] __devinitdata = {
+	{ .compatible = "fsl,mpc5200-psc-ac97", },
+	{ .compatible = "fsl,mpc5200b-psc-ac97", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, psc_ac97_match);
+
+static struct of_platform_driver psc_ac97_driver = {
+	.match_table = psc_ac97_match,
+	.probe = psc_ac97_of_probe,
+	.remove = __devexit_p(psc_ac97_of_remove),
+	.driver = {
+		.name = "mpc5200-psc-ac97",
+		.owner = THIS_MODULE,
+	},
+};
+
+/* ---------------------------------------------------------------------
+ * Module setup and teardown; simply register the of_platform driver
+ * for the PSC in AC97 mode.
+ */
+static int __init psc_ac97_init(void)
+{
+	return of_register_platform_driver(&psc_ac97_driver);
+}
+module_init(psc_ac97_init);
+
+static void __exit psc_ac97_exit(void)
+{
+	of_unregister_platform_driver(&psc_ac97_driver);
+}
+module_exit(psc_ac97_exit);
+
+MODULE_AUTHOR("Jon Smirl <jonsmirl@gmail.com>");
+MODULE_DESCRIPTION("mpc5200 AC97 module");
+MODULE_LICENSE("GPL");
+
diff --git a/sound/soc/fsl/mpc5200_psc_ac97.h b/sound/soc/fsl/mpc5200_psc_ac97.h
new file mode 100644
index 0000000..4bc18c3
--- /dev/null
+++ b/sound/soc/fsl/mpc5200_psc_ac97.h
@@ -0,0 +1,15 @@
+/*
+ * Freescale MPC5200 PSC in AC97 mode
+ * ALSA SoC Digital Audio Interface (DAI) driver
+ *
+ */
+
+#ifndef __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__
+#define __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__
+
+extern struct snd_soc_dai psc_ac97_dai[];
+
+#define MPC5200_AC97_NORMAL 0
+#define MPC5200_AC97_SPDIF 1
+
+#endif /* __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__ */

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* [PATCH V4 4/5] Support for AC97 on Phytec pmc030 base board.
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
                   ` (2 preceding siblings ...)
  2009-05-25 22:15 ` [PATCH V4 3/5] AC97 driver for mpc5200 Jon Smirl
@ 2009-05-25 22:15 ` Jon Smirl
  2009-05-25 22:15 ` [PATCH V4 5/5] Fabric bindings for STAC9766 on the Efika Jon Smirl
                   ` (2 subsequent siblings)
  6 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

Support for AC97 on Phytec pmc030 base board. A wm9712 AC97 codec is used.

Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
---
 sound/soc/fsl/Kconfig               |    7 +++
 sound/soc/fsl/Makefile              |    3 +
 sound/soc/fsl/pcm030-audio-fabric.c |   90 +++++++++++++++++++++++++++++++++++
 3 files changed, 100 insertions(+), 0 deletions(-)
 create mode 100644 sound/soc/fsl/pcm030-audio-fabric.c

diff --git a/sound/soc/fsl/Kconfig b/sound/soc/fsl/Kconfig
index 3bce952..79579ae 100644
--- a/sound/soc/fsl/Kconfig
+++ b/sound/soc/fsl/Kconfig
@@ -39,4 +39,11 @@ config SND_SOC_MPC5200_AC97
 	help
 	  Say Y here to support the MPC5200 PSCs in AC97 mode.
 
+config SND_MPC52xx_SOC_PCM030
+	tristate "SoC AC97 Audio support for Phytec pcm030 and WM9712"
+	depends on PPC_MPC5200_SIMPLE
+	select SND_SOC_MPC5200_AC97
+	select SND_SOC_WM9712
+	help
+	  Say Y if you want to add support for sound on the Phytec pcm030 baseboard.
 
diff --git a/sound/soc/fsl/Makefile b/sound/soc/fsl/Makefile
index 14631a1..66d88c8 100644
--- a/sound/soc/fsl/Makefile
+++ b/sound/soc/fsl/Makefile
@@ -15,3 +15,6 @@ obj-$(CONFIG_SND_MPC52xx_DMA) += mpc5200_dma.o
 obj-$(CONFIG_SND_SOC_MPC5200_I2S) += mpc5200_psc_i2s.o
 obj-$(CONFIG_SND_SOC_MPC5200_AC97) += mpc5200_psc_ac97.o
 
+# MPC5200 Machine Support
+obj-$(CONFIG_SND_MPC52xx_SOC_PCM030) += pcm030-audio-fabric.o
+
diff --git a/sound/soc/fsl/pcm030-audio-fabric.c b/sound/soc/fsl/pcm030-audio-fabric.c
new file mode 100644
index 0000000..8766f7a
--- /dev/null
+++ b/sound/soc/fsl/pcm030-audio-fabric.c
@@ -0,0 +1,90 @@
+/*
+ * Phytec pcm030 driver for the PSC of the Freescale MPC52xx
+ * configured as AC97 interface
+ *
+ * Copyright 2008 Jon Smirl, Digispeaker
+ * Author: Jon Smirl <jonsmirl@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/of_device.h>
+#include <linux/of_platform.h>
+#include <linux/dma-mapping.h>
+
+#include <sound/core.h>
+#include <sound/pcm.h>
+#include <sound/pcm_params.h>
+#include <sound/initval.h>
+#include <sound/soc.h>
+#include <sound/soc-of-simple.h>
+
+#include "mpc5200_dma.h"
+#include "mpc5200_psc_ac97.h"
+#include "../codecs/wm9712.h"
+
+static struct snd_soc_device device;
+static struct snd_soc_card card;
+
+static struct snd_soc_dai_link pcm030_fabric_dai[] = {
+{
+	.name = "AC97",
+	.stream_name = "AC97 Analog",
+	.codec_dai = &wm9712_dai[WM9712_DAI_AC97_HIFI],
+	.cpu_dai = &psc_ac97_dai[MPC5200_AC97_NORMAL],
+},
+{
+	.name = "AC97",
+	.stream_name = "AC97 IEC958",
+	.codec_dai = &wm9712_dai[WM9712_DAI_AC97_AUX],
+	.cpu_dai = &psc_ac97_dai[MPC5200_AC97_SPDIF],
+},
+};
+
+static __init int pcm030_fabric_init(void)
+{
+	struct platform_device *pdev;
+	int rc;
+
+	if (!machine_is_compatible("phytec,pcm030"))
+		return -ENODEV;
+
+	card.platform = &mpc5200_audio_dma_platform;
+	card.name = "pcm030";
+	card.dai_link = pcm030_fabric_dai;
+	card.num_links = ARRAY_SIZE(pcm030_fabric_dai);
+
+	device.card = &card;
+	device.codec_dev = &soc_codec_dev_wm9712;
+
+	pdev = platform_device_alloc("soc-audio", 1);
+	if (!pdev) {
+		pr_err("pcm030_fabric_init: platform_device_alloc() failed\n");
+		return -ENODEV;
+	}
+
+	platform_set_drvdata(pdev, &device);
+	device.dev = &pdev->dev;
+
+	rc = platform_device_add(pdev);
+	if (rc) {
+		pr_err("pcm030_fabric_init: platform_device_add() failed\n");
+		return -ENODEV;
+	}
+	return 0;
+}
+
+module_init(pcm030_fabric_init);
+
+
+MODULE_AUTHOR("Jon Smirl <jonsmirl@gmail.com>");
+MODULE_DESCRIPTION(DRV_NAME ": mpc5200 pcm030 fabric driver");
+MODULE_LICENSE("GPL");
+

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* [PATCH V4 5/5] Fabric bindings for STAC9766 on the Efika
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
                   ` (3 preceding siblings ...)
  2009-05-25 22:15 ` [PATCH V4 4/5] Support for AC97 on Phytec pmc030 base board Jon Smirl
@ 2009-05-25 22:15 ` Jon Smirl
  2009-05-26  3:44   ` Timur Tabi
  2009-05-26 11:04   ` [alsa-devel] " Mark Brown
  6 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-25 22:15 UTC (permalink / raw)
  To: grant.likely, linuxppc-dev, alsa-devel, broonie, timur

Fabric bindings for STAC9766 AC97 codec on the Efika.

Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
---
 sound/soc/fsl/Kconfig              |    8 +++
 sound/soc/fsl/Makefile             |    1 
 sound/soc/fsl/efika-audio-fabric.c |   90 ++++++++++++++++++++++++++++++++++++
 3 files changed, 99 insertions(+), 0 deletions(-)
 create mode 100644 sound/soc/fsl/efika-audio-fabric.c

diff --git a/sound/soc/fsl/Kconfig b/sound/soc/fsl/Kconfig
index 79579ae..f571c6e 100644
--- a/sound/soc/fsl/Kconfig
+++ b/sound/soc/fsl/Kconfig
@@ -47,3 +47,11 @@ config SND_MPC52xx_SOC_PCM030
 	help
 	  Say Y if you want to add support for sound on the Phytec pcm030 baseboard.
 
+config SND_MPC52xx_SOC_EFIKA
+	tristate "SoC AC97 Audio support for bbplan Efika and STAC9766"
+	depends on PPC_EFIKA
+	select SND_SOC_MPC5200_AC97
+	select SND_SOC_STAC9766
+	help
+	  Say Y if you want to add support for sound on the Efika.
+
diff --git a/sound/soc/fsl/Makefile b/sound/soc/fsl/Makefile
index 66d88c8..a83a739 100644
--- a/sound/soc/fsl/Makefile
+++ b/sound/soc/fsl/Makefile
@@ -17,4 +17,5 @@ obj-$(CONFIG_SND_SOC_MPC5200_AC97) += mpc5200_psc_ac97.o
 
 # MPC5200 Machine Support
 obj-$(CONFIG_SND_MPC52xx_SOC_PCM030) += pcm030-audio-fabric.o
+obj-$(CONFIG_SND_MPC52xx_SOC_EFIKA) += efika-audio-fabric.o
 
diff --git a/sound/soc/fsl/efika-audio-fabric.c b/sound/soc/fsl/efika-audio-fabric.c
new file mode 100644
index 0000000..85b0e75
--- /dev/null
+++ b/sound/soc/fsl/efika-audio-fabric.c
@@ -0,0 +1,90 @@
+/*
+ * Efika driver for the PSC of the Freescale MPC52xx
+ * configured as AC97 interface
+ *
+ * Copyright 2008 Jon Smirl, Digispeaker
+ * Author: Jon Smirl <jonsmirl@gmail.com>
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/device.h>
+#include <linux/delay.h>
+#include <linux/of_device.h>
+#include <linux/of_platform.h>
+#include <linux/dma-mapping.h>
+
+#include <sound/core.h>
+#include <sound/pcm.h>
+#include <sound/pcm_params.h>
+#include <sound/initval.h>
+#include <sound/soc.h>
+#include <sound/soc-of-simple.h>
+
+#include "mpc5200_dma.h"
+#include "mpc5200_psc_ac97.h"
+#include "../codecs/stac9766.h"
+
+static struct snd_soc_device device;
+static struct snd_soc_card card;
+
+static struct snd_soc_dai_link efika_fabric_dai[] = {
+{
+	.name = "AC97",
+	.stream_name = "AC97 Analog",
+	.codec_dai = &stac9766_dai[STAC9766_DAI_AC97_ANALOG],
+	.cpu_dai = &psc_ac97_dai[MPC5200_AC97_NORMAL],
+},
+{
+	.name = "AC97",
+	.stream_name = "AC97 IEC958",
+	.codec_dai = &stac9766_dai[STAC9766_DAI_AC97_DIGITAL],
+	.cpu_dai = &psc_ac97_dai[MPC5200_AC97_SPDIF],
+},
+};
+
+static __init int efika_fabric_init(void)
+{
+	struct platform_device *pdev;
+	int rc;
+
+	if (!machine_is_compatible("bplan,efika"))
+		return -ENODEV;
+
+	card.platform = &mpc5200_audio_dma_platform;
+	card.name = "Efika";
+	card.dai_link = efika_fabric_dai;
+	card.num_links = ARRAY_SIZE(efika_fabric_dai);
+
+	device.card = &card;
+	device.codec_dev = &soc_codec_dev_stac9766;
+
+	pdev = platform_device_alloc("soc-audio", 1);
+	if (!pdev) {
+		pr_err("efika_fabric_init: platform_device_alloc() failed\n");
+		return -ENODEV;
+	}
+
+	platform_set_drvdata(pdev, &device);
+	device.dev = &pdev->dev;
+
+	rc = platform_device_add(pdev);
+	if (rc) {
+		pr_err("efika_fabric_init: platform_device_add() failed\n");
+		return -ENODEV;
+	}
+	return 0;
+}
+
+module_init(efika_fabric_init);
+
+
+MODULE_AUTHOR("Jon Smirl <jonsmirl@gmail.com>");
+MODULE_DESCRIPTION(DRV_NAME ": mpc5200 Efika fabric driver");
+MODULE_LICENSE("GPL");
+

^ permalink raw reply related	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 3/5] AC97 driver for mpc5200
  2009-05-25 22:15 ` [PATCH V4 3/5] AC97 driver for mpc5200 Jon Smirl
@ 2009-05-26  3:41     ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26  3:41 UTC (permalink / raw)
  To: Jon Smirl; +Cc: grant.likely, linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

> +       spin_event_timeout(1, 3, 0, rc);

Ugh, don't do this.  Just use udelay!

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 3/5] AC97 driver for mpc5200
@ 2009-05-26  3:41     ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26  3:41 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

> + =A0 =A0 =A0 spin_event_timeout(1, 3, 0, rc);

Ugh, don't do this.  Just use udelay!

--=20
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
@ 2009-05-26  3:44   ` Timur Tabi
  2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
                     ` (5 subsequent siblings)
  6 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26  3:44 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

>      The macro spin_event_timeout() takes a condition and timeout value

My patch is already set to go through the powerpc tree, so there's no
need to repost it here.  Please, you changed the title of the patch.

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26  3:44   ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26  3:44 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

> =A0 =A0 =A0The macro spin_event_timeout() takes a condition and timeout v=
alue

My patch is already set to go through the powerpc tree, so there's no
need to repost it here.  Please, you changed the title of the patch.

--=20
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 3/5] AC97 driver for mpc5200
  2009-05-26  3:41     ` [alsa-devel] " Timur Tabi
@ 2009-05-26  3:47       ` Grant Likely
  -1 siblings, 0 replies; 39+ messages in thread
From: Grant Likely @ 2009-05-26  3:47 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 9:41 PM, Timur Tabi <timur@freescale.com> wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>
>> +       spin_event_timeout(1, 3, 0, rc);
>
> Ugh, don't do this.  Just use udelay!

heh; indeed.  My objection to udelay is that it burns cycles it the
test loop that could be used to decide whether or not the loop is
complete.  Not an issue when the explicit goal is to burn a fixed
number of cycles.

g.

-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 3/5] AC97 driver for mpc5200
@ 2009-05-26  3:47       ` Grant Likely
  0 siblings, 0 replies; 39+ messages in thread
From: Grant Likely @ 2009-05-26  3:47 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel, broonie

On Mon, May 25, 2009 at 9:41 PM, Timur Tabi <timur@freescale.com> wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>
>> + =A0 =A0 =A0 spin_event_timeout(1, 3, 0, rc);
>
> Ugh, don't do this. =A0Just use udelay!

heh; indeed.  My objection to udelay is that it burns cycles it the
test loop that could be used to decide whether or not the loop is
complete.  Not an issue when the explicit goal is to burn a fixed
number of cycles.

g.

--=20
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-25 22:15   ` Jon Smirl
  (?)
@ 2009-05-26  7:29   ` Geert Uytterhoeven
  2009-05-26 11:32       ` Timur Tabi
  -1 siblings, 1 reply; 39+ messages in thread
From: Geert Uytterhoeven @ 2009-05-26  7:29 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, broonie, timur

On Mon, 25 May 2009, Jon Smirl wrote:
> (in microseconds) as parameters.  It spins until either the condition is true
> or the timeout expires.  It returns the result of the condition when the loop
> was terminated.
> 
> This primary purpose of this macro is to poll on a hardware register until a
> status bit changes.  The timeout ensures that the loop still terminates if the
> bit doesn't change as expected.  This macro makes it easier for driver
> developers to perform this kind of operation properly.
> 
> Signed-off-by: Timur Tabi <timur@freescale.com>
> ---
>  arch/powerpc/include/asm/delay.h |   33 +++++++++++++++++++++++++++++++++
>  1 files changed, 33 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/powerpc/include/asm/delay.h b/arch/powerpc/include/asm/delay.h
> index f9200a6..fedf037 100644
> --- a/arch/powerpc/include/asm/delay.h
> +++ b/arch/powerpc/include/asm/delay.h
> @@ -2,8 +2,11 @@
>  #define _ASM_POWERPC_DELAY_H
>  #ifdef __KERNEL__
>  
> +#include <asm/time.h>
> +
>  /*
>   * Copyright 1996, Paul Mackerras.
> + * Copyright (C) 2009 Freescale Semiconductor, Inc. All rights reserved.
>   *
>   * This program is free software; you can redistribute it and/or
>   * modify it under the terms of the GNU General Public License
> @@ -30,5 +33,35 @@ extern void udelay(unsigned long usecs);
>  #define mdelay(n)	udelay((n) * 1000)
>  #endif
>  
> +/**
> + * spin_event_timeout - spin until a condition gets true or a timeout elapses
> + * @condition: a C expression to evalate
> + * @timeout: timeout, in microseconds
> + * @delay: the number of microseconds to delay between eache evaluation of
> + *         @condition
> + * @rc: the last value of the condition
> + *
> + * The process spins until the condition evaluates to true (non-zero) or the
> + * timeout elapses.  Upon exit, @rc contains the value of the condition. This
> + * allows you to test the condition without incurring any side effects.
> + *
> + * This primary purpose of this macro is to poll on a hardware register
> + * until a status bit changes.  The timeout ensures that the loop still
> + * terminates even if the bit never changes.  The delay is for devices that
> + * need a delay in between successive reads.
> + *
> + * gcc will optimize out the if-statement if @delay is a constant.
> + */
> +#define spin_event_timeout(condition, timeout, delay, rc)                   \

static inline function, returning rc, instead of a macro?

> +{                                                                           \
> +	unsigned long __loops = tb_ticks_per_usec * timeout;                \
> +	unsigned long __start = get_tbl();                                  \
> +	while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
> +		if (delay)                                                  \
> +			udelay(delay);                                      \
> +		else	                                                    \
> +			cpu_relax();                                        \
> +}
> +
>  #endif /* __KERNEL__ */
>  #endif /* _ASM_POWERPC_DELAY_H */

With kind regards,

Geert Uytterhoeven
Software Architect
Techsoft Centre

Technology and Software Centre Europe
The Corporate Village · Da Vincilaan 7-D1 · B-1935 Zaventem · Belgium

Phone:    +32 (0)2 700 8453
Fax:      +32 (0)2 700 8622
E-mail:   Geert.Uytterhoeven@sonycom.com
Internet: http://www.sony-europe.com/

A division of Sony Europe (Belgium) N.V.
VAT BE 0413.825.160 · RPR Brussels
Fortis · BIC GEBABEBB · IBAN BE41293037680010

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-26  3:44   ` Timur Tabi
@ 2009-05-26  9:21     ` Mark Brown
  -1 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26  9:21 UTC (permalink / raw)
  To: Timur Tabi; +Cc: grant.likely, linuxppc-dev, alsa-devel

On Mon, May 25, 2009 at 10:44:12PM -0500, Timur Tabi wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

> > ? ? ?The macro spin_event_timeout() takes a condition and timeout value

> My patch is already set to go through the powerpc tree, so there's no
> need to repost it here.  Please, you changed the title of the patch.

Indeed.  Since this is a new driver I'm happy to let the code go in
without the new function and then sort itself out in the merge window.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26  9:21     ` Mark Brown
  0 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26  9:21 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel

On Mon, May 25, 2009 at 10:44:12PM -0500, Timur Tabi wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:

> > ? ? ?The macro spin_event_timeout() takes a condition and timeout value

> My patch is already set to go through the powerpc tree, so there's no
> need to repost it here.  Please, you changed the title of the patch.

Indeed.  Since this is a new driver I'm happy to let the code go in
without the new function and then sort itself out in the merge window.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
@ 2009-05-26 11:01   ` Mark Brown
  2009-05-26 16:51     ` Grant Likely
  0 siblings, 1 reply; 39+ messages in thread
From: Mark Brown @ 2009-05-26 11:01 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, timur

On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
> 
> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>

Grant, I'm OK with that if you are?

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
@ 2009-05-26 11:04   ` Mark Brown
  2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
                     ` (5 subsequent siblings)
  6 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26 11:04 UTC (permalink / raw)
  To: Jon Smirl; +Cc: grant.likely, linuxppc-dev, alsa-devel, timur

On Mon, May 25, 2009 at 06:15:05PM -0400, Jon Smirl wrote:

> reset retries. Removing the retries makes the driver fail to load about 5% of the time. An oscilliscope is 

Please also send a patch sorting this out in the STAC9766 driver as
requested.  Other than that and the udelay() thing it looks like we can
merge this now - there's the reset issue and probably other things to
clean up but it'd be better to do those as incremental patches.

> needed to figure out what is going on with failed resets. Both pcm030 and Efika ocasionally fail to reset 
> and they use different codec chips.

Like I say, it may be worth checking into the startup of the master
clock for the CODECs, especially if it's an output from the SoC on both
boards.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26 11:04   ` Mark Brown
  0 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26 11:04 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, timur

On Mon, May 25, 2009 at 06:15:05PM -0400, Jon Smirl wrote:

> reset retries. Removing the retries makes the driver fail to load about 5% of the time. An oscilliscope is 

Please also send a patch sorting this out in the STAC9766 driver as
requested.  Other than that and the udelay() thing it looks like we can
merge this now - there's the reset issue and probably other things to
clean up but it'd be better to do those as incremental patches.

> needed to figure out what is going on with failed resets. Both pcm030 and Efika ocasionally fail to reset 
> and they use different codec chips.

Like I say, it may be worth checking into the startup of the master
clock for the CODECs, especially if it's an output from the SoC on both
boards.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-26  3:44   ` Timur Tabi
@ 2009-05-26 11:08     ` Jon Smirl
  -1 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-26 11:08 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel, broonie, Geert Uytterhoeven

On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>
>>      The macro spin_event_timeout() takes a condition and timeout value
>
> My patch is already set to go through the powerpc tree, so there's no
> need to repost it here.  Please, you changed the title of the patch.

I'm changing the code, I removed the ! on the rc. Geert just asked to
have it changed to an in-line too.

I forgot to add the title line so it used the beginning of the comment.


>
> --
> Timur Tabi
> Linux kernel developer at Freescale
>



-- 
Jon Smirl
jonsmirl@gmail.com

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26 11:08     ` Jon Smirl
  0 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-26 11:08 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel, broonie, Geert Uytterhoeven

On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:
> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>
>> =A0 =A0 =A0The macro spin_event_timeout() takes a condition and timeout =
value
>
> My patch is already set to go through the powerpc tree, so there's no
> need to repost it here. =A0Please, you changed the title of the patch.

I'm changing the code, I removed the ! on the rc. Geert just asked to
have it changed to an in-line too.

I forgot to add the title line so it used the beginning of the comment.


>
> --
> Timur Tabi
> Linux kernel developer at Freescale
>



--=20
Jon Smirl
jonsmirl@gmail.com

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-26  7:29   ` Geert Uytterhoeven
@ 2009-05-26 11:32       ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 11:32 UTC (permalink / raw)
  To: Geert Uytterhoeven; +Cc: linuxppc-dev, alsa-devel, broonie

On Tue, May 26, 2009 at 2:29 AM, Geert Uytterhoeven
<Geert.Uytterhoeven@sonycom.com> wrote:

> static inline function, returning rc, instead of a macro?

It won't work as an inline function ...

>> +     unsigned long __loops = tb_ticks_per_usec * timeout;                \
>> +     unsigned long __start = get_tbl();                                  \
>> +     while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \

There's no way to pass a expression to an inline function.  It has to
be a macro.  Here, the loop evaluates "condition" at every pass.  If
this were an inline function, "condition" would be evaluated once when
the function were called, and never again.

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
@ 2009-05-26 11:32       ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 11:32 UTC (permalink / raw)
  To: Geert Uytterhoeven; +Cc: linuxppc-dev, alsa-devel, broonie

On Tue, May 26, 2009 at 2:29 AM, Geert Uytterhoeven
<Geert.Uytterhoeven@sonycom.com> wrote:

> static inline function, returning rc, instead of a macro?

It won't work as an inline function ...

>> + =A0 =A0 unsigned long __loops =3D tb_ticks_per_usec * timeout; =A0 =A0=
 =A0 =A0 =A0 =A0 =A0 =A0\
>> + =A0 =A0 unsigned long __start =3D get_tbl(); =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
>> + =A0 =A0 while ((rc =3D (condition)) && (tb_ticks_since(__start) <=3D _=
_loops)) \

There's no way to pass a expression to an inline function.  It has to
be a macro.  Here, the loop evaluates "condition" at every pass.  If
this were an inline function, "condition" would be evaluated once when
the function were called, and never again.

--=20
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-26 11:08     ` Jon Smirl
@ 2009-05-26 11:39       ` Timur Tabi
  -1 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 11:39 UTC (permalink / raw)
  To: Jon Smirl
  Cc: grant.likely, linuxppc-dev, alsa-devel, broonie, Geert Uytterhoeven

On Tue, May 26, 2009 at 6:08 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:
>> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>>
>>>      The macro spin_event_timeout() takes a condition and timeout value
>>
>> My patch is already set to go through the powerpc tree, so there's no
>> need to repost it here.  Please, you changed the title of the patch.
>
> I'm changing the code,

My code?  If you want to make your own version of spin_event_timeout,
then you should change the name and keep it local to your C files.

> I removed the ! on the rc. Geert just asked to
> have it changed to an in-line too.

And Geert's request won't work.

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26 11:39       ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 11:39 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, broonie, Geert Uytterhoeven

On Tue, May 26, 2009 at 6:08 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:
>> On Mon, May 25, 2009 at 5:15 PM, Jon Smirl <jonsmirl@gmail.com> wrote:
>>
>>> =A0 =A0 =A0The macro spin_event_timeout() takes a condition and timeout=
 value
>>
>> My patch is already set to go through the powerpc tree, so there's no
>> need to repost it here. =A0Please, you changed the title of the patch.
>
> I'm changing the code,

My code?  If you want to make your own version of spin_event_timeout,
then you should change the name and keep it local to your C files.

> I removed the ! on the rc. Geert just asked to
> have it changed to an in-line too.

And Geert's request won't work.

--=20
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 0/5] AC97 driver for mpc5200
  2009-05-26 11:08     ` Jon Smirl
@ 2009-05-26 11:43       ` Mark Brown
  -1 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26 11:43 UTC (permalink / raw)
  To: Jon Smirl
  Cc: grant.likely, linuxppc-dev, alsa-devel, Timur Tabi, Geert Uytterhoeven

On Tue, May 26, 2009 at 07:08:52AM -0400, Jon Smirl wrote:
> On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:

> > My patch is already set to go through the powerpc tree, so there's no
> > need to repost it here. ?Please, you changed the title of the patch.

> I'm changing the code, I removed the ! on the rc. Geert just asked to
> have it changed to an in-line too.

If you're doing that you need make it explicit and submit an incremental
patch against the PowerPC tree.  You'll also need to take care of any
affected users there.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200
@ 2009-05-26 11:43       ` Mark Brown
  0 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26 11:43 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, Timur Tabi, Geert Uytterhoeven

On Tue, May 26, 2009 at 07:08:52AM -0400, Jon Smirl wrote:
> On Mon, May 25, 2009 at 11:44 PM, Timur Tabi <timur@freescale.com> wrote:

> > My patch is already set to go through the powerpc tree, so there's no
> > need to repost it here. ?Please, you changed the title of the patch.

> I'm changing the code, I removed the ! on the rc. Geert just asked to
> have it changed to an in-line too.

If you're doing that you need make it explicit and submit an incremental
patch against the PowerPC tree.  You'll also need to take care of any
affected users there.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-26 11:32       ` Timur Tabi
  (?)
@ 2009-05-26 11:51       ` Geert Uytterhoeven
  2009-05-26 12:49         ` Arnd Bergmann
  -1 siblings, 1 reply; 39+ messages in thread
From: Geert Uytterhoeven @ 2009-05-26 11:51 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel, broonie

On Tue, 26 May 2009, Timur Tabi wrote:
> On Tue, May 26, 2009 at 2:29 AM, Geert Uytterhoeven
> <Geert.Uytterhoeven@sonycom.com> wrote:
> 
> > static inline function, returning rc, instead of a macro?
> 
> It won't work as an inline function ...
> 
> >> +     unsigned long __loops = tb_ticks_per_usec * timeout;                \
> >> +     unsigned long __start = get_tbl();                                  \
> >> +     while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
> 
> There's no way to pass a expression to an inline function.  It has to
> be a macro.  Here, the loop evaluates "condition" at every pass.  If
> this were an inline function, "condition" would be evaluated once when
> the function were called, and never again.

You're right, I missed that part. Sorry about that.

However, you can still improve useability by making the macro return the rc,
instead of letting the caller pass it, cfr. wait_event_timeout() and friends.

With kind regards,

Geert Uytterhoeven
Software Architect
Techsoft Centre

Technology and Software Centre Europe
The Corporate Village · Da Vincilaan 7-D1 · B-1935 Zaventem · Belgium

Phone:    +32 (0)2 700 8453
Fax:      +32 (0)2 700 8622
E-mail:   Geert.Uytterhoeven@sonycom.com
Internet: http://www.sony-europe.com/

A division of Sony Europe (Belgium) N.V.
VAT BE 0413.825.160 · RPR Brussels
Fortis · BIC GEBABEBB · IBAN BE41293037680010

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-26 11:51       ` Geert Uytterhoeven
@ 2009-05-26 12:49         ` Arnd Bergmann
  2009-05-26 14:20             ` Timur Tabi
  0 siblings, 1 reply; 39+ messages in thread
From: Arnd Bergmann @ 2009-05-26 12:49 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Geert Uytterhoeven, alsa-devel, Timur Tabi, broonie

On Tuesday 26 May 2009, Geert Uytterhoeven wrote:
> However, you can still improve useability by making the macro return the rc,
> instead of letting the caller pass it, cfr. wait_event_timeout() and friends.

Either that, or it should at least use the do { ... } while (0) construct
to make the macro a statement. All multi-line macros need to either
use ({ ... }) or do { ... } while (0) to make sure they behave well.

	Arnd <><

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
  2009-05-26 12:49         ` Arnd Bergmann
@ 2009-05-26 14:20             ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 14:20 UTC (permalink / raw)
  To: Arnd Bergmann
  Cc: alsa-devel, Benjamin Herrenschmidt, linuxppc-dev,
	Geert Uytterhoeven, broonie

Arnd Bergmann wrote:
> On Tuesday 26 May 2009, Geert Uytterhoeven wrote:
>> However, you can still improve useability by making the macro return the rc,
>> instead of letting the caller pass it, cfr. wait_event_timeout() and friends.

I had that originally, but somewhere during the seven revisions of my macro, it got lost.

I'll post a v9 soon.  This version will take 3 parameters and return __ret, so Jon will need to change his code (sorry).

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value
@ 2009-05-26 14:20             ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 14:20 UTC (permalink / raw)
  To: Arnd Bergmann; +Cc: alsa-devel, linuxppc-dev, Geert Uytterhoeven, broonie

Arnd Bergmann wrote:
> On Tuesday 26 May 2009, Geert Uytterhoeven wrote:
>> However, you can still improve useability by making the macro return the rc,
>> instead of letting the caller pass it, cfr. wait_event_timeout() and friends.

I had that originally, but somewhere during the seven revisions of my macro, it got lost.

I'll post a v9 soon.  This version will take 3 parameters and return __ret, so Jon will need to change his code (sorry).

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-26 11:01   ` [alsa-devel] " Mark Brown
@ 2009-05-26 16:51     ` Grant Likely
  2009-05-26 16:53         ` [alsa-devel] " Jon Smirl
  0 siblings, 1 reply; 39+ messages in thread
From: Grant Likely @ 2009-05-26 16:51 UTC (permalink / raw)
  To: Mark Brown; +Cc: linuxppc-dev, alsa-devel, timur

On Tue, May 26, 2009 at 5:01 AM, Mark Brown
<broonie@opensource.wolfsonmicro.com> wrote:
> On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
>> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
>>
>> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
>
> Grant, I'm OK with that if you are?

Acked-by: Grant Likely <grant.likely@secretlab.ca>




-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-26 16:51     ` Grant Likely
@ 2009-05-26 16:53         ` Jon Smirl
  0 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-26 16:53 UTC (permalink / raw)
  To: Grant Likely; +Cc: linuxppc-dev, alsa-devel, Mark Brown, timur

On Tue, May 26, 2009 at 12:51 PM, Grant Likely
<grant.likely@secretlab.ca> wrote:
> On Tue, May 26, 2009 at 5:01 AM, Mark Brown
> <broonie@opensource.wolfsonmicro.com> wrote:
>> On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
>>> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
>>>
>>> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
>>
>> Grant, I'm OK with that if you are?
>
> Acked-by: Grant Likely <grant.likely@secretlab.ca>

Put in the V5 version this one has this in the h file:

-int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
-                          struct snd_soc_dai *dai);
+/* whack this after Timur's patch is merged in to
arch/powerpc/include/asm/delay.h */
+#define spin_event_timeout(condition, timeout, delay, rc)                   \
+{                                                                           \
+       unsigned long __loops = tb_ticks_per_usec * timeout;                \
+       unsigned long __start = get_tbl();                                  \
+       while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
+               if (delay)                                                  \
+                       udelay(delay);                                      \
+               else                                                        \
+                       cpu_relax();                                        \
+}
+/* whack this after Timur's patch is merged in to
arch/powerpc/include/asm/delay.h */




>
>
>
>
> --
> Grant Likely, B.Sc., P.Eng.
> Secret Lab Technologies Ltd.
>



-- 
Jon Smirl
jonsmirl@gmail.com

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
@ 2009-05-26 16:53         ` Jon Smirl
  0 siblings, 0 replies; 39+ messages in thread
From: Jon Smirl @ 2009-05-26 16:53 UTC (permalink / raw)
  To: Grant Likely; +Cc: linuxppc-dev, alsa-devel, Mark Brown, timur

On Tue, May 26, 2009 at 12:51 PM, Grant Likely
<grant.likely@secretlab.ca> wrote:
> On Tue, May 26, 2009 at 5:01 AM, Mark Brown
> <broonie@opensource.wolfsonmicro.com> wrote:
>> On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
>>> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
>>>
>>> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
>>
>> Grant, I'm OK with that if you are?
>
> Acked-by: Grant Likely <grant.likely@secretlab.ca>

Put in the V5 version this one has this in the h file:

-int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
-                          struct snd_soc_dai *dai);
+/* whack this after Timur's patch is merged in to
arch/powerpc/include/asm/delay.h */
+#define spin_event_timeout(condition, timeout, delay, rc)                   \
+{                                                                           \
+       unsigned long __loops = tb_ticks_per_usec * timeout;                \
+       unsigned long __start = get_tbl();                                  \
+       while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
+               if (delay)                                                  \
+                       udelay(delay);                                      \
+               else                                                        \
+                       cpu_relax();                                        \
+}
+/* whack this after Timur's patch is merged in to
arch/powerpc/include/asm/delay.h */




>
>
>
>
> --
> Grant Likely, B.Sc., P.Eng.
> Secret Lab Technologies Ltd.
>



-- 
Jon Smirl
jonsmirl@gmail.com

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-26 16:53         ` [alsa-devel] " Jon Smirl
@ 2009-05-26 16:56           ` Grant Likely
  -1 siblings, 0 replies; 39+ messages in thread
From: Grant Likely @ 2009-05-26 16:56 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, Mark Brown, timur

On Tue, May 26, 2009 at 10:53 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> On Tue, May 26, 2009 at 12:51 PM, Grant Likely
> <grant.likely@secretlab.ca> wrote:
>> On Tue, May 26, 2009 at 5:01 AM, Mark Brown
>> <broonie@opensource.wolfsonmicro.com> wrote:
>>> On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
>>>> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
>>>>
>>>> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
>>>
>>> Grant, I'm OK with that if you are?
>>
>> Acked-by: Grant Likely <grant.likely@secretlab.ca>
>
> Put in the V5 version this one has this in the h file:
>
> -int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
> -                          struct snd_soc_dai *dai);
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */
> +#define spin_event_timeout(condition, timeout, delay, rc)                   \
> +{                                                                           \
> +       unsigned long __loops = tb_ticks_per_usec * timeout;                \
> +       unsigned long __start = get_tbl();                                  \
> +       while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
> +               if (delay)                                                  \
> +                       udelay(delay);                                      \
> +               else                                                        \
> +                       cpu_relax();                                        \
> +}
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */

sure, no problem if it will grease the merge.

g.




-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
@ 2009-05-26 16:56           ` Grant Likely
  0 siblings, 0 replies; 39+ messages in thread
From: Grant Likely @ 2009-05-26 16:56 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, Mark Brown, timur

On Tue, May 26, 2009 at 10:53 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> On Tue, May 26, 2009 at 12:51 PM, Grant Likely
> <grant.likely@secretlab.ca> wrote:
>> On Tue, May 26, 2009 at 5:01 AM, Mark Brown
>> <broonie@opensource.wolfsonmicro.com> wrote:
>>> On Mon, May 25, 2009 at 06:15:09PM -0400, Jon Smirl wrote:
>>>> Rewrite the mpc5200 audio DMA code to support both I2S and AC97.
>>>>
>>>> Signed-off-by: Jon Smirl <jonsmirl@gmail.com>
>>>
>>> Grant, I'm OK with that if you are?
>>
>> Acked-by: Grant Likely <grant.likely@secretlab.ca>
>
> Put in the V5 version this one has this in the h file:
>
> -int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
> - =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0struct snd_soc_dai *=
dai);
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */
> +#define spin_event_timeout(condition, timeout, delay, rc) =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 \
> +{ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 \
> + =A0 =A0 =A0 unsigned long __loops =3D tb_ticks_per_usec * timeout; =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 unsigned long __start =3D get_tbl(); =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 while ((rc =3D (condition)) && (tb_ticks_since(__start) <=
=3D __loops)) \
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (delay) =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 udelay(delay); =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 else =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 cpu_relax(); =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> +}
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */

sure, no problem if it will grease the merge.

g.




--=20
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-26 16:53         ` [alsa-devel] " Jon Smirl
@ 2009-05-26 17:08           ` Timur Tabi
  -1 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 17:08 UTC (permalink / raw)
  To: Jon Smirl; +Cc: Grant Likely, linuxppc-dev, alsa-devel, Mark Brown

On Tue, May 26, 2009 at 11:53 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> Put in the V5 version this one has this in the h file:
>
> -int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
> -                          struct snd_soc_dai *dai);
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */
> +#define spin_event_timeout(condition, timeout, delay, rc)                   \
> +{                                                                           \
> +       unsigned long __loops = tb_ticks_per_usec * timeout;                \
> +       unsigned long __start = get_tbl();                                  \
> +       while ((rc = (condition)) && (tb_ticks_since(__start) <= __loops)) \
> +               if (delay)                                                  \
> +                       udelay(delay);                                      \
> +               else                                                        \
> +                       cpu_relax();                                        \
> +}
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */

It won't be that simple.  V9 of my patch changes the number of
parameters, so not only will you need to whack this copy of the macro,
you'll also need to change the callers.

-- 
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
@ 2009-05-26 17:08           ` Timur Tabi
  0 siblings, 0 replies; 39+ messages in thread
From: Timur Tabi @ 2009-05-26 17:08 UTC (permalink / raw)
  To: Jon Smirl; +Cc: linuxppc-dev, alsa-devel, Mark Brown

On Tue, May 26, 2009 at 11:53 AM, Jon Smirl <jonsmirl@gmail.com> wrote:
> Put in the V5 version this one has this in the h file:
>
> -int psc_dma_trigger(struct snd_pcm_substream *substream, int cmd,
> - =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0struct snd_soc_dai *=
dai);
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */
> +#define spin_event_timeout(condition, timeout, delay, rc) =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 \
> +{ =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 \
> + =A0 =A0 =A0 unsigned long __loops =3D tb_ticks_per_usec * timeout; =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 unsigned long __start =3D get_tbl(); =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 while ((rc =3D (condition)) && (tb_ticks_since(__start) <=
=3D __loops)) \
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (delay) =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 udelay(delay); =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 else =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
\
> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 cpu_relax(); =A0 =A0 =A0 =
=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0\
> +}
> +/* whack this after Timur's patch is merged in to
> arch/powerpc/include/asm/delay.h */

It won't be that simple.  V9 of my patch changes the number of
parameters, so not only will you need to whack this copy of the macro,
you'll also need to change the callers.

--=20
Timur Tabi
Linux kernel developer at Freescale

^ permalink raw reply	[flat|nested] 39+ messages in thread

* Re: [alsa-devel] [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code
  2009-05-26 17:08           ` [alsa-devel] " Timur Tabi
  (?)
@ 2009-05-26 20:21           ` Mark Brown
  -1 siblings, 0 replies; 39+ messages in thread
From: Mark Brown @ 2009-05-26 20:21 UTC (permalink / raw)
  To: Timur Tabi; +Cc: linuxppc-dev, alsa-devel

On Tue, May 26, 2009 at 12:08:01PM -0500, Timur Tabi wrote:

> It won't be that simple.  V9 of my patch changes the number of
> parameters, so not only will you need to whack this copy of the macro,
> you'll also need to change the callers.

OK, in the interests of cutting down on the amount of review here I've
applied v5 of the series and then marked the AC97 machine drivers as
BROKEN for the time being - once everything lands in mainline we can
make sure that the driver is in sync with the final spin_event_timeout()
API and remove the BROKEN marking.

^ permalink raw reply	[flat|nested] 39+ messages in thread

end of thread, other threads:[~2009-05-26 20:21 UTC | newest]

Thread overview: 39+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-05-25 22:15 [PATCH V4 0/5] AC97 driver for mpc5200 Jon Smirl
2009-05-25 22:15 ` [PATCH V4 1/5] The macro spin_event_timeout() takes a condition and timeout value Jon Smirl
2009-05-25 22:15   ` Jon Smirl
2009-05-26  7:29   ` Geert Uytterhoeven
2009-05-26 11:32     ` Timur Tabi
2009-05-26 11:32       ` Timur Tabi
2009-05-26 11:51       ` Geert Uytterhoeven
2009-05-26 12:49         ` Arnd Bergmann
2009-05-26 14:20           ` Timur Tabi
2009-05-26 14:20             ` Timur Tabi
2009-05-25 22:15 ` [PATCH V4 2/5] Main rewite of the mpc5200 audio DMA code Jon Smirl
2009-05-26 11:01   ` [alsa-devel] " Mark Brown
2009-05-26 16:51     ` Grant Likely
2009-05-26 16:53       ` Jon Smirl
2009-05-26 16:53         ` [alsa-devel] " Jon Smirl
2009-05-26 16:56         ` Grant Likely
2009-05-26 16:56           ` [alsa-devel] " Grant Likely
2009-05-26 17:08         ` Timur Tabi
2009-05-26 17:08           ` [alsa-devel] " Timur Tabi
2009-05-26 20:21           ` Mark Brown
2009-05-25 22:15 ` [PATCH V4 3/5] AC97 driver for mpc5200 Jon Smirl
2009-05-26  3:41   ` Timur Tabi
2009-05-26  3:41     ` [alsa-devel] " Timur Tabi
2009-05-26  3:47     ` Grant Likely
2009-05-26  3:47       ` [alsa-devel] " Grant Likely
2009-05-25 22:15 ` [PATCH V4 4/5] Support for AC97 on Phytec pmc030 base board Jon Smirl
2009-05-25 22:15 ` [PATCH V4 5/5] Fabric bindings for STAC9766 on the Efika Jon Smirl
2009-05-26  3:44 ` [alsa-devel] [PATCH V4 0/5] AC97 driver for mpc5200 Timur Tabi
2009-05-26  3:44   ` Timur Tabi
2009-05-26  9:21   ` Mark Brown
2009-05-26  9:21     ` [alsa-devel] " Mark Brown
2009-05-26 11:08   ` Jon Smirl
2009-05-26 11:08     ` Jon Smirl
2009-05-26 11:39     ` Timur Tabi
2009-05-26 11:39       ` [alsa-devel] " Timur Tabi
2009-05-26 11:43     ` Mark Brown
2009-05-26 11:43       ` [alsa-devel] " Mark Brown
2009-05-26 11:04 ` Mark Brown
2009-05-26 11:04   ` [alsa-devel] " Mark Brown

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.