hda_controller.c

/*
 *
 *  Implementation of primary alsa driver code base for Intel HD Audio.
 *
 *  Copyright(c) 2004 Intel Corporation. All rights reserved.
 *
 *  Copyright (c) 2004 Takashi Iwai <tiwai@suse.de>
 *                     PeiSen Hou <pshou@realtek.com.tw>
 *
 *  This program is free software; you can redistribute it and/or modify it
 *  under the terms of the GNU General Public License as published by the Free
 *  Software Foundation; either version 2 of the License, or (at your option)
 *  any later version.
 *
 *  This program is distributed in the hope that it will be useful, but WITHOUT
 *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 *  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 *  more details.
 *
 *
 */

#include <linux/clocksource.h>
#include <linux/delay.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <sound/core.h>
#include <sound/initval.h>
#include "hda_priv.h"
#include "hda_controller.h"

#define CREATE_TRACE_POINTS
#include "hda_intel_trace.h"

/*
 * AZX stream operations.
 */

/* start a stream */
void azx_stream_start(struct azx *chip, struct azx_dev *azx_dev)
{
	/*
	 * Before stream start, initialize parameter
	 */
	azx_dev->insufficient = 1;

	/* enable SIE */
	azx_writel(chip, INTCTL,
		   azx_readl(chip, INTCTL) | (1 << azx_dev->index));
	/* set DMA start and interrupt mask */
	azx_sd_writeb(chip, azx_dev, SD_CTL,
		      azx_sd_readb(chip, azx_dev, SD_CTL) |
		      SD_CTL_DMA_START | SD_INT_MASK);
}
EXPORT_SYMBOL_GPL(azx_stream_start);

/* stop DMA */
static void azx_stream_clear(struct azx *chip, struct azx_dev *azx_dev)
{
	azx_sd_writeb(chip, azx_dev, SD_CTL,
		      azx_sd_readb(chip, azx_dev, SD_CTL) &
		      ~(SD_CTL_DMA_START | SD_INT_MASK));
	azx_sd_writeb(chip, azx_dev, SD_STS, SD_INT_MASK); /* to be sure */
}

/* stop a stream */
void azx_stream_stop(struct azx *chip, struct azx_dev *azx_dev)
{
	azx_stream_clear(chip, azx_dev);
	/* disable SIE */
	azx_writel(chip, INTCTL,
		   azx_readl(chip, INTCTL) & ~(1 << azx_dev->index));
}
EXPORT_SYMBOL_GPL(azx_stream_stop);

/* reset stream */
void azx_stream_reset(struct azx *chip, struct azx_dev *azx_dev)
{
	unsigned char val;
	int timeout;

	azx_stream_clear(chip, azx_dev);

	azx_sd_writeb(chip, azx_dev, SD_CTL,
		      azx_sd_readb(chip, azx_dev, SD_CTL) |
		      SD_CTL_STREAM_RESET);
	udelay(3);
	timeout = 300;
	while (!((val = azx_sd_readb(chip, azx_dev, SD_CTL)) &
		 SD_CTL_STREAM_RESET) && --timeout)
		;
	val &= ~SD_CTL_STREAM_RESET;
	azx_sd_writeb(chip, azx_dev, SD_CTL, val);
	udelay(3);

	timeout = 300;
	/* waiting for hardware to report that the stream is out of reset */
	while (((val = azx_sd_readb(chip, azx_dev, SD_CTL)) &
		SD_CTL_STREAM_RESET) && --timeout)
		;

	/* reset first position - may not be synced with hw at this time */
	*azx_dev->posbuf = 0;
}
EXPORT_SYMBOL_GPL(azx_stream_reset);

/*
 * set up the SD for streaming
 */
int azx_setup_controller(struct azx *chip, struct azx_dev *azx_dev)
{
	unsigned int val;
	/* make sure the run bit is zero for SD */
	azx_stream_clear(chip, azx_dev);
	/* program the stream_tag */
	val = azx_sd_readl(chip, azx_dev, SD_CTL);
	val = (val & ~SD_CTL_STREAM_TAG_MASK) |
		(azx_dev->stream_tag << SD_CTL_STREAM_TAG_SHIFT);
	if (!azx_snoop(chip))
		val |= SD_CTL_TRAFFIC_PRIO;
	azx_sd_writel(chip, azx_dev, SD_CTL, val);

	/* program the length of samples in cyclic buffer */
	azx_sd_writel(chip, azx_dev, SD_CBL, azx_dev->bufsize);

	/* program the stream format */
	/* this value needs to be the same as the one programmed */
	azx_sd_writew(chip, azx_dev, SD_FORMAT, azx_dev->format_val);

	/* program the stream LVI (last valid index) of the BDL */
	azx_sd_writew(chip, azx_dev, SD_LVI, azx_dev->frags - 1);

	/* program the BDL address */
	/* lower BDL address */
	azx_sd_writel(chip, azx_dev, SD_BDLPL, (u32)azx_dev->bdl.addr);
	/* upper BDL address */
	azx_sd_writel(chip, azx_dev, SD_BDLPU,
		      upper_32_bits(azx_dev->bdl.addr));

	/* enable the position buffer */
	if (chip->position_fix[0] != POS_FIX_LPIB ||
	    chip->position_fix[1] != POS_FIX_LPIB) {
		if (!(azx_readl(chip, DPLBASE) & ICH6_DPLBASE_ENABLE))
			azx_writel(chip, DPLBASE,
				(u32)chip->posbuf.addr | ICH6_DPLBASE_ENABLE);
	}

	/* set the interrupt enable bits in the descriptor control register */
	azx_sd_writel(chip, azx_dev, SD_CTL,
		      azx_sd_readl(chip, azx_dev, SD_CTL) | SD_INT_MASK);

	return 0;
}
EXPORT_SYMBOL_GPL(azx_setup_controller);

/* assign a stream for the PCM */
static inline struct azx_dev *
azx_assign_device(struct azx *chip, struct snd_pcm_substream *substream)
{
	int dev, i, nums;
	struct azx_dev *res = NULL;
	/* make a non-zero unique key for the substream */
	int key = (substream->pcm->device << 16) | (substream->number << 2) |
		(substream->stream + 1);

	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
		dev = chip->playback_index_offset;
		nums = chip->playback_streams;
	} else {
		dev = chip->capture_index_offset;
		nums = chip->capture_streams;
	}
	for (i = 0; i < nums; i++, dev++) {
		struct azx_dev *azx_dev = &chip->azx_dev[dev];
		dsp_lock(azx_dev);
		if (!azx_dev->opened && !dsp_is_locked(azx_dev)) {
			res = azx_dev;
			if (res->assigned_key == key) {
				res->opened = 1;
				res->assigned_key = key;
				dsp_unlock(azx_dev);
				return azx_dev;
			}
		}
		dsp_unlock(azx_dev);
	}
	if (res) {
		dsp_lock(res);
		res->opened = 1;
		res->assigned_key = key;
		dsp_unlock(res);
	}
	return res;
}

/* release the assigned stream */
static inline void azx_release_device(struct azx_dev *azx_dev)
{
	azx_dev->opened = 0;
}

static cycle_t azx_cc_read(const struct cyclecounter *cc)
{
	struct azx_dev *azx_dev = container_of(cc, struct azx_dev, azx_cc);
	struct snd_pcm_substream *substream = azx_dev->substream;
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;

	return azx_readl(chip, WALLCLK);
}

static void azx_timecounter_init(struct snd_pcm_substream *substream,
				bool force, cycle_t last)
{
	struct azx_dev *azx_dev = get_azx_dev(substream);
	struct timecounter *tc = &azx_dev->azx_tc;
	struct cyclecounter *cc = &azx_dev->azx_cc;
	u64 nsec;

	cc->read = azx_cc_read;
	cc->mask = CLOCKSOURCE_MASK(32);

	/*
	 * Converting from 24 MHz to ns means applying a 125/3 factor.
	 * To avoid any saturation issues in intermediate operations,
	 * the 125 factor is applied first. The division is applied
	 * last after reading the timecounter value.
	 * Applying the 1/3 factor as part of the multiplication
	 * requires at least 20 bits for a decent precision, however
	 * overflows occur after about 4 hours or less, not a option.
	 */

	cc->mult = 125; /* saturation after 195 years */
	cc->shift = 0;

	nsec = 0; /* audio time is elapsed time since trigger */
	timecounter_init(tc, cc, nsec);
	if (force)
		/*
		 * force timecounter to use predefined value,
		 * used for synchronized starts
		 */
		tc->cycle_last = last;
}

static u64 azx_adjust_codec_delay(struct snd_pcm_substream *substream,
				u64 nsec)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct hda_pcm_stream *hinfo = apcm->hinfo[substream->stream];
	u64 codec_frames, codec_nsecs;

	if (!hinfo->ops.get_delay)
		return nsec;

	codec_frames = hinfo->ops.get_delay(hinfo, apcm->codec, substream);
	codec_nsecs = div_u64(codec_frames * 1000000000LL,
			      substream->runtime->rate);

	if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
		return nsec + codec_nsecs;

	return (nsec > codec_nsecs) ? nsec - codec_nsecs : 0;
}

/*
 * set up a BDL entry
 */
int setup_bdle(struct azx *chip,
	       struct snd_dma_buffer *dmab,
	       struct azx_dev *azx_dev, u32 **bdlp,
	       int ofs, int size, int with_ioc)
{
	u32 *bdl = *bdlp;

	while (size > 0) {
		dma_addr_t addr;
		int chunk;

		if (azx_dev->frags >= AZX_MAX_BDL_ENTRIES)
			return -EINVAL;

		addr = snd_sgbuf_get_addr(dmab, ofs);
		/* program the address field of the BDL entry */
		bdl[0] = cpu_to_le32((u32)addr);
		bdl[1] = cpu_to_le32(upper_32_bits(addr));
		/* program the size field of the BDL entry */
		chunk = snd_sgbuf_get_chunk_size(dmab, ofs, size);
		/* one BDLE cannot cross 4K boundary on CTHDA chips */
		if (chip->driver_caps & AZX_DCAPS_4K_BDLE_BOUNDARY) {
			u32 remain = 0x1000 - (ofs & 0xfff);
			if (chunk > remain)
				chunk = remain;
		}
		bdl[2] = cpu_to_le32(chunk);
		/* program the IOC to enable interrupt
		 * only when the whole fragment is processed
		 */
		size -= chunk;
		bdl[3] = (size || !with_ioc) ? 0 : cpu_to_le32(0x01);
		bdl += 4;
		azx_dev->frags++;
		ofs += chunk;
	}
	*bdlp = bdl;
	return ofs;
}
EXPORT_SYMBOL_GPL(setup_bdle);

/*
 * set up BDL entries
 */
static int azx_setup_periods(struct azx *chip,
			     struct snd_pcm_substream *substream,
			     struct azx_dev *azx_dev)
{
	u32 *bdl;
	int i, ofs, periods, period_bytes;
	int pos_adj = 0;

	/* reset BDL address */
	azx_sd_writel(chip, azx_dev, SD_BDLPL, 0);
	azx_sd_writel(chip, azx_dev, SD_BDLPU, 0);

	period_bytes = azx_dev->period_bytes;
	periods = azx_dev->bufsize / period_bytes;

	/* program the initial BDL entries */
	bdl = (u32 *)azx_dev->bdl.area;
	ofs = 0;
	azx_dev->frags = 0;

	if (chip->bdl_pos_adj)
		pos_adj = chip->bdl_pos_adj[chip->dev_index];
	if (!azx_dev->no_period_wakeup && pos_adj > 0) {
		struct snd_pcm_runtime *runtime = substream->runtime;
		int pos_align = pos_adj;
		pos_adj = (pos_adj * runtime->rate + 47999) / 48000;
		if (!pos_adj)
			pos_adj = pos_align;
		else
			pos_adj = ((pos_adj + pos_align - 1) / pos_align) *
				pos_align;
		pos_adj = frames_to_bytes(runtime, pos_adj);
		if (pos_adj >= period_bytes) {
			dev_warn(chip->card->dev,"Too big adjustment %d\n",
				 pos_adj);
			pos_adj = 0;
		} else {
			ofs = setup_bdle(chip, snd_pcm_get_dma_buf(substream),
					 azx_dev,
					 &bdl, ofs, pos_adj, true);
			if (ofs < 0)
				goto error;
		}
	} else
		pos_adj = 0;

	for (i = 0; i < periods; i++) {
		if (i == periods - 1 && pos_adj)
			ofs = setup_bdle(chip, snd_pcm_get_dma_buf(substream),
					 azx_dev, &bdl, ofs,
					 period_bytes - pos_adj, 0);
		else
			ofs = setup_bdle(chip, snd_pcm_get_dma_buf(substream),
					 azx_dev, &bdl, ofs,
					 period_bytes,
					 !azx_dev->no_period_wakeup);
		if (ofs < 0)
			goto error;
	}
	return 0;

 error:
	dev_err(chip->card->dev, "Too many BDL entries: buffer=%d, period=%d\n",
		azx_dev->bufsize, period_bytes);
	return -EINVAL;
}

/*
 * PCM ops
 */

static int azx_pcm_close(struct snd_pcm_substream *substream)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct hda_pcm_stream *hinfo = apcm->hinfo[substream->stream];
	struct azx *chip = apcm->chip;
	struct azx_dev *azx_dev = get_azx_dev(substream);
	unsigned long flags;

	mutex_lock(&chip->open_mutex);
	spin_lock_irqsave(&chip->reg_lock, flags);
	azx_dev->substream = NULL;
	azx_dev->running = 0;
	spin_unlock_irqrestore(&chip->reg_lock, flags);
	azx_release_device(azx_dev);
	hinfo->ops.close(hinfo, apcm->codec, substream);
	snd_hda_power_down(apcm->codec);
	mutex_unlock(&chip->open_mutex);
	return 0;
}

static int azx_pcm_hw_params(struct snd_pcm_substream *substream,
			     struct snd_pcm_hw_params *hw_params)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;
	int ret;

	dsp_lock(get_azx_dev(substream));
	if (dsp_is_locked(get_azx_dev(substream))) {
		ret = -EBUSY;
		goto unlock;
	}

	ret = chip->ops->substream_alloc_pages(chip, substream,
					  params_buffer_bytes(hw_params));
unlock:
	dsp_unlock(get_azx_dev(substream));
	return ret;
}

static int azx_pcm_hw_free(struct snd_pcm_substream *substream)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx_dev *azx_dev = get_azx_dev(substream);
	struct azx *chip = apcm->chip;
	struct hda_pcm_stream *hinfo = apcm->hinfo[substream->stream];
	int err;

	/* reset BDL address */
	dsp_lock(azx_dev);
	if (!dsp_is_locked(azx_dev)) {
		azx_sd_writel(chip, azx_dev, SD_BDLPL, 0);
		azx_sd_writel(chip, azx_dev, SD_BDLPU, 0);
		azx_sd_writel(chip, azx_dev, SD_CTL, 0);
		azx_dev->bufsize = 0;
		azx_dev->period_bytes = 0;
		azx_dev->format_val = 0;
	}

	snd_hda_codec_cleanup(apcm->codec, hinfo, substream);

	err = chip->ops->substream_free_pages(chip, substream);
	azx_dev->prepared = 0;
	dsp_unlock(azx_dev);
	return err;
}

static int azx_pcm_prepare(struct snd_pcm_substream *substream)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;
	struct azx_dev *azx_dev = get_azx_dev(substream);
	struct hda_pcm_stream *hinfo = apcm->hinfo[substream->stream];
	struct snd_pcm_runtime *runtime = substream->runtime;
	unsigned int bufsize, period_bytes, format_val, stream_tag;
	int err;
	struct hda_spdif_out *spdif =
		snd_hda_spdif_out_of_nid(apcm->codec, hinfo->nid);
	unsigned short ctls = spdif ? spdif->ctls : 0;

	dsp_lock(azx_dev);
	if (dsp_is_locked(azx_dev)) {
		err = -EBUSY;
		goto unlock;
	}

	azx_stream_reset(chip, azx_dev);
	format_val = snd_hda_calc_stream_format(runtime->rate,
						runtime->channels,
						runtime->format,
						hinfo->maxbps,
						ctls);
	if (!format_val) {
		dev_err(chip->card->dev,
			"invalid format_val, rate=%d, ch=%d, format=%d\n",
			runtime->rate, runtime->channels, runtime->format);
		err = -EINVAL;
		goto unlock;
	}

	bufsize = snd_pcm_lib_buffer_bytes(substream);
	period_bytes = snd_pcm_lib_period_bytes(substream);

	dev_dbg(chip->card->dev, "azx_pcm_prepare: bufsize=0x%x, format=0x%x\n",
		bufsize, format_val);

	if (bufsize != azx_dev->bufsize ||
	    period_bytes != azx_dev->period_bytes ||
	    format_val != azx_dev->format_val ||
	    runtime->no_period_wakeup != azx_dev->no_period_wakeup) {
		azx_dev->bufsize = bufsize;
		azx_dev->period_bytes = period_bytes;
		azx_dev->format_val = format_val;
		azx_dev->no_period_wakeup = runtime->no_period_wakeup;
		err = azx_setup_periods(chip, substream, azx_dev);
		if (err < 0)
			goto unlock;
	}

	/* when LPIB delay correction gives a small negative value,
	 * we ignore it; currently set the threshold statically to
	 * 64 frames
	 */
	if (runtime->period_size > 64)
		azx_dev->delay_negative_threshold = -frames_to_bytes(runtime, 64);
	else
		azx_dev->delay_negative_threshold = 0;

	/* wallclk has 24Mhz clock source */
	azx_dev->period_wallclk = (((runtime->period_size * 24000) /
						runtime->rate) * 1000);
	azx_setup_controller(chip, azx_dev);
	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
		azx_dev->fifo_size =
			azx_sd_readw(chip, azx_dev, SD_FIFOSIZE) + 1;
	else
		azx_dev->fifo_size = 0;

	stream_tag = azx_dev->stream_tag;
	/* CA-IBG chips need the playback stream starting from 1 */
	if ((chip->driver_caps & AZX_DCAPS_CTX_WORKAROUND) &&
	    stream_tag > chip->capture_streams)
		stream_tag -= chip->capture_streams;
	err = snd_hda_codec_prepare(apcm->codec, hinfo, stream_tag,
				     azx_dev->format_val, substream);

 unlock:
	if (!err)
		azx_dev->prepared = 1;
	dsp_unlock(azx_dev);
	return err;
}

static int azx_pcm_trigger(struct snd_pcm_substream *substream, int cmd)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;
	struct azx_dev *azx_dev;
	struct snd_pcm_substream *s;
	int rstart = 0, start, nsync = 0, sbits = 0;
	int nwait, timeout;

	azx_dev = get_azx_dev(substream);
	trace_azx_pcm_trigger(chip, azx_dev, cmd);

	if (dsp_is_locked(azx_dev) || !azx_dev->prepared)
		return -EPIPE;

	switch (cmd) {
	case SNDRV_PCM_TRIGGER_START:
		rstart = 1;
	case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
	case SNDRV_PCM_TRIGGER_RESUME:
		start = 1;
		break;
	case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
	case SNDRV_PCM_TRIGGER_SUSPEND:
	case SNDRV_PCM_TRIGGER_STOP:
		start = 0;
		break;
	default:
		return -EINVAL;
	}

	snd_pcm_group_for_each_entry(s, substream) {
		if (s->pcm->card != substream->pcm->card)
			continue;
		azx_dev = get_azx_dev(s);
		sbits |= 1 << azx_dev->index;
		nsync++;
		snd_pcm_trigger_done(s, substream);
	}

	spin_lock(&chip->reg_lock);

	/* first, set SYNC bits of corresponding streams */
	if (chip->driver_caps & AZX_DCAPS_OLD_SSYNC)
		azx_writel(chip, OLD_SSYNC,
			azx_readl(chip, OLD_SSYNC) | sbits);
	else
		azx_writel(chip, SSYNC, azx_readl(chip, SSYNC) | sbits);

	snd_pcm_group_for_each_entry(s, substream) {
		if (s->pcm->card != substream->pcm->card)
			continue;
		azx_dev = get_azx_dev(s);
		if (start) {
			azx_dev->start_wallclk = azx_readl(chip, WALLCLK);
			if (!rstart)
				azx_dev->start_wallclk -=
						azx_dev->period_wallclk;
			azx_stream_start(chip, azx_dev);
		} else {
			azx_stream_stop(chip, azx_dev);
		}
		azx_dev->running = start;
	}
	spin_unlock(&chip->reg_lock);
	if (start) {
		/* wait until all FIFOs get ready */
		for (timeout = 5000; timeout; timeout--) {
			nwait = 0;
			snd_pcm_group_for_each_entry(s, substream) {
				if (s->pcm->card != substream->pcm->card)
					continue;
				azx_dev = get_azx_dev(s);
				if (!(azx_sd_readb(chip, azx_dev, SD_STS) &
				      SD_STS_FIFO_READY))
					nwait++;
			}
			if (!nwait)
				break;
			cpu_relax();
		}
	} else {
		/* wait until all RUN bits are cleared */
		for (timeout = 5000; timeout; timeout--) {
			nwait = 0;
			snd_pcm_group_for_each_entry(s, substream) {
				if (s->pcm->card != substream->pcm->card)
					continue;
				azx_dev = get_azx_dev(s);
				if (azx_sd_readb(chip, azx_dev, SD_CTL) &
				    SD_CTL_DMA_START)
					nwait++;
			}
			if (!nwait)
				break;
			cpu_relax();
		}
	}
	spin_lock(&chip->reg_lock);
	/* reset SYNC bits */
	if (chip->driver_caps & AZX_DCAPS_OLD_SSYNC)
		azx_writel(chip, OLD_SSYNC,
			azx_readl(chip, OLD_SSYNC) & ~sbits);
	else
		azx_writel(chip, SSYNC, azx_readl(chip, SSYNC) & ~sbits);
	if (start) {
		azx_timecounter_init(substream, 0, 0);
		if (nsync > 1) {
			cycle_t cycle_last;

			/* same start cycle for master and group */
			azx_dev = get_azx_dev(substream);
			cycle_last = azx_dev->azx_tc.cycle_last;

			snd_pcm_group_for_each_entry(s, substream) {
				if (s->pcm->card != substream->pcm->card)
					continue;
				azx_timecounter_init(s, 1, cycle_last);
			}
		}
	}
	spin_unlock(&chip->reg_lock);
	return 0;
}

/* get the current DMA position with correction on VIA chips */
static unsigned int azx_via_get_position(struct azx *chip,
					 struct azx_dev *azx_dev)
{
	unsigned int link_pos, mini_pos, bound_pos;
	unsigned int mod_link_pos, mod_dma_pos, mod_mini_pos;
	unsigned int fifo_size;

	link_pos = azx_sd_readl(chip, azx_dev, SD_LPIB);
	if (azx_dev->substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
		/* Playback, no problem using link position */
		return link_pos;
	}

	/* Capture */
	/* For new chipset,
	 * use mod to get the DMA position just like old chipset
	 */
	mod_dma_pos = le32_to_cpu(*azx_dev->posbuf);
	mod_dma_pos %= azx_dev->period_bytes;

	/* azx_dev->fifo_size can't get FIFO size of in stream.
	 * Get from base address + offset.
	 */
	fifo_size = readw(chip->remap_addr + VIA_IN_STREAM0_FIFO_SIZE_OFFSET);

	if (azx_dev->insufficient) {
		/* Link position never gather than FIFO size */
		if (link_pos <= fifo_size)
			return 0;

		azx_dev->insufficient = 0;
	}

	if (link_pos <= fifo_size)
		mini_pos = azx_dev->bufsize + link_pos - fifo_size;
	else
		mini_pos = link_pos - fifo_size;

	/* Find nearest previous boudary */
	mod_mini_pos = mini_pos % azx_dev->period_bytes;
	mod_link_pos = link_pos % azx_dev->period_bytes;
	if (mod_link_pos >= fifo_size)
		bound_pos = link_pos - mod_link_pos;
	else if (mod_dma_pos >= mod_mini_pos)
		bound_pos = mini_pos - mod_mini_pos;
	else {
		bound_pos = mini_pos - mod_mini_pos + azx_dev->period_bytes;
		if (bound_pos >= azx_dev->bufsize)
			bound_pos = 0;
	}

	/* Calculate real DMA position we want */
	return bound_pos + mod_dma_pos;
}

unsigned int azx_get_position(struct azx *chip,
			      struct azx_dev *azx_dev,
			      bool with_check)
{
	struct snd_pcm_substream *substream = azx_dev->substream;
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	unsigned int pos;
	int stream = substream->stream;
	struct hda_pcm_stream *hinfo = apcm->hinfo[stream];
	int delay = 0;

	switch (chip->position_fix[stream]) {
	case POS_FIX_LPIB:
		/* read LPIB */
		pos = azx_sd_readl(chip, azx_dev, SD_LPIB);
		break;
	case POS_FIX_VIACOMBO:
		pos = azx_via_get_position(chip, azx_dev);
		break;
	default:
		/* use the position buffer */
		pos = le32_to_cpu(*azx_dev->posbuf);
		if (with_check && chip->position_fix[stream] == POS_FIX_AUTO) {
			if (!pos || pos == (u32)-1) {
				dev_info(chip->card->dev,
					 "Invalid position buffer, using LPIB read method instead.\n");
				chip->position_fix[stream] = POS_FIX_LPIB;
				pos = azx_sd_readl(chip, azx_dev, SD_LPIB);
			} else
				chip->position_fix[stream] = POS_FIX_POSBUF;
		}
		break;
	}

	if (pos >= azx_dev->bufsize)
		pos = 0;

	/* calculate runtime delay from LPIB */
	if (substream->runtime &&
	    chip->position_fix[stream] == POS_FIX_POSBUF &&
	    (chip->driver_caps & AZX_DCAPS_COUNT_LPIB_DELAY)) {
		unsigned int lpib_pos = azx_sd_readl(chip, azx_dev, SD_LPIB);
		if (stream == SNDRV_PCM_STREAM_PLAYBACK)
			delay = pos - lpib_pos;
		else
			delay = lpib_pos - pos;
		if (delay < 0) {
			if (delay >= azx_dev->delay_negative_threshold)
				delay = 0;
			else
				delay += azx_dev->bufsize;
		}
		if (delay >= azx_dev->period_bytes) {
			dev_info(chip->card->dev,
				 "Unstable LPIB (%d >= %d); disabling LPIB delay counting\n",
				 delay, azx_dev->period_bytes);
			delay = 0;
			chip->driver_caps &= ~AZX_DCAPS_COUNT_LPIB_DELAY;
		}
		delay = bytes_to_frames(substream->runtime, delay);
	}

	if (substream->runtime) {
		if (hinfo->ops.get_delay)
			delay += hinfo->ops.get_delay(hinfo, apcm->codec,
						      substream);
		substream->runtime->delay = delay;
	}

	trace_azx_get_position(chip, azx_dev, pos, delay);
	return pos;
}
EXPORT_SYMBOL_GPL(azx_get_position);

static snd_pcm_uframes_t azx_pcm_pointer(struct snd_pcm_substream *substream)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;
	struct azx_dev *azx_dev = get_azx_dev(substream);
	return bytes_to_frames(substream->runtime,
			       azx_get_position(chip, azx_dev, false));
}

static int azx_get_wallclock_tstamp(struct snd_pcm_substream *substream,
				struct timespec *ts)
{
	struct azx_dev *azx_dev = get_azx_dev(substream);
	u64 nsec;

	nsec = timecounter_read(&azx_dev->azx_tc);
	nsec = div_u64(nsec, 3); /* can be optimized */
	nsec = azx_adjust_codec_delay(substream, nsec);

	*ts = ns_to_timespec(nsec);

	return 0;
}

static struct snd_pcm_hardware azx_pcm_hw = {
	.info =			(SNDRV_PCM_INFO_MMAP |
				 SNDRV_PCM_INFO_INTERLEAVED |
				 SNDRV_PCM_INFO_BLOCK_TRANSFER |
				 SNDRV_PCM_INFO_MMAP_VALID |
				 /* No full-resume yet implemented */
				 /* SNDRV_PCM_INFO_RESUME |*/
				 SNDRV_PCM_INFO_PAUSE |
				 SNDRV_PCM_INFO_SYNC_START |
				 SNDRV_PCM_INFO_HAS_WALL_CLOCK |
				 SNDRV_PCM_INFO_NO_PERIOD_WAKEUP),
	.formats =		SNDRV_PCM_FMTBIT_S16_LE,
	.rates =		SNDRV_PCM_RATE_48000,
	.rate_min =		48000,
	.rate_max =		48000,
	.channels_min =		2,
	.channels_max =		2,
	.buffer_bytes_max =	AZX_MAX_BUF_SIZE,
	.period_bytes_min =	128,
	.period_bytes_max =	AZX_MAX_BUF_SIZE / 2,
	.periods_min =		2,
	.periods_max =		AZX_MAX_FRAG,
	.fifo_size =		0,
};

static int azx_pcm_open(struct snd_pcm_substream *substream)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct hda_pcm_stream *hinfo = apcm->hinfo[substream->stream];
	struct azx *chip = apcm->chip;
	struct azx_dev *azx_dev;
	struct snd_pcm_runtime *runtime = substream->runtime;
	unsigned long flags;
	int err;
	int buff_step;

	mutex_lock(&chip->open_mutex);
	azx_dev = azx_assign_device(chip, substream);
	if (azx_dev == NULL) {
		mutex_unlock(&chip->open_mutex);
		return -EBUSY;
	}
	runtime->hw = azx_pcm_hw;
	runtime->hw.channels_min = hinfo->channels_min;
	runtime->hw.channels_max = hinfo->channels_max;
	runtime->hw.formats = hinfo->formats;
	runtime->hw.rates = hinfo->rates;
	snd_pcm_limit_hw_rates(runtime);
	snd_pcm_hw_constraint_integer(runtime, SNDRV_PCM_HW_PARAM_PERIODS);

	/* avoid wrap-around with wall-clock */
	snd_pcm_hw_constraint_minmax(runtime, SNDRV_PCM_HW_PARAM_BUFFER_TIME,
				     20,
				     178000000);

	if (chip->align_buffer_size)
		/* constrain buffer sizes to be multiple of 128
		   bytes. This is more efficient in terms of memory
		   access but isn't required by the HDA spec and
		   prevents users from specifying exact period/buffer
		   sizes. For example for 44.1kHz, a period size set
		   to 20ms will be rounded to 19.59ms. */
		buff_step = 128;
	else
		/* Don't enforce steps on buffer sizes, still need to
		   be multiple of 4 bytes (HDA spec). Tested on Intel
		   HDA controllers, may not work on all devices where
		   option needs to be disabled */
		buff_step = 4;

	snd_pcm_hw_constraint_step(runtime, 0, SNDRV_PCM_HW_PARAM_BUFFER_BYTES,
				   buff_step);
	snd_pcm_hw_constraint_step(runtime, 0, SNDRV_PCM_HW_PARAM_PERIOD_BYTES,
				   buff_step);
	snd_hda_power_up_d3wait(apcm->codec);
	err = hinfo->ops.open(hinfo, apcm->codec, substream);
	if (err < 0) {
		azx_release_device(azx_dev);
		snd_hda_power_down(apcm->codec);
		mutex_unlock(&chip->open_mutex);
		return err;
	}
	snd_pcm_limit_hw_rates(runtime);
	/* sanity check */
	if (snd_BUG_ON(!runtime->hw.channels_min) ||
	    snd_BUG_ON(!runtime->hw.channels_max) ||
	    snd_BUG_ON(!runtime->hw.formats) ||
	    snd_BUG_ON(!runtime->hw.rates)) {
		azx_release_device(azx_dev);
		hinfo->ops.close(hinfo, apcm->codec, substream);
		snd_hda_power_down(apcm->codec);
		mutex_unlock(&chip->open_mutex);
		return -EINVAL;
	}

	/* disable WALLCLOCK timestamps for capture streams
	   until we figure out how to handle digital inputs */
	if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
		runtime->hw.info &= ~SNDRV_PCM_INFO_HAS_WALL_CLOCK;

	spin_lock_irqsave(&chip->reg_lock, flags);
	azx_dev->substream = substream;
	azx_dev->running = 0;
	spin_unlock_irqrestore(&chip->reg_lock, flags);

	runtime->private_data = azx_dev;
	snd_pcm_set_sync(substream);
	mutex_unlock(&chip->open_mutex);
	return 0;
}

static int azx_pcm_mmap(struct snd_pcm_substream *substream,
			struct vm_area_struct *area)
{
	struct azx_pcm *apcm = snd_pcm_substream_chip(substream);
	struct azx *chip = apcm->chip;
	if (chip->ops->pcm_mmap_prepare)
		chip->ops->pcm_mmap_prepare(substream, area);
	return snd_pcm_lib_default_mmap(substream, area);
}

static struct snd_pcm_ops azx_pcm_ops = {
	.open = azx_pcm_open,
	.close = azx_pcm_close,
	.ioctl = snd_pcm_lib_ioctl,
	.hw_params = azx_pcm_hw_params,
	.hw_free = azx_pcm_hw_free,
	.prepare = azx_pcm_prepare,
	.trigger = azx_pcm_trigger,
	.pointer = azx_pcm_pointer,
	.wall_clock =  azx_get_wallclock_tstamp,
	.mmap = azx_pcm_mmap,
	.page = snd_pcm_sgbuf_ops_page,
};

static void azx_pcm_free(struct snd_pcm *pcm)
{
	struct azx_pcm *apcm = pcm->private_data;
	if (apcm) {
		list_del(&apcm->list);
		kfree(apcm);
	}
}

#define MAX_PREALLOC_SIZE	(32 * 1024 * 1024)

int azx_attach_pcm_stream(struct hda_bus *bus, struct hda_codec *codec,
			  struct hda_pcm *cpcm)
{
	struct azx *chip = bus->private_data;
	struct snd_pcm *pcm;
	struct azx_pcm *apcm;
	int pcm_dev = cpcm->device;
	unsigned int size;
	int s, err;

	list_for_each_entry(apcm, &chip->pcm_list, list) {
		if (apcm->pcm->device == pcm_dev) {
			dev_err(chip->card->dev, "PCM %d already exists\n",
				pcm_dev);
			return -EBUSY;
		}
	}
	err = snd_pcm_new(chip->card, cpcm->name, pcm_dev,
			  cpcm->stream[SNDRV_PCM_STREAM_PLAYBACK].substreams,
			  cpcm->stream[SNDRV_PCM_STREAM_CAPTURE].substreams,
			  &pcm);
	if (err < 0)
		return err;
	strlcpy(pcm->name, cpcm->name, sizeof(pcm->name));
	apcm = kzalloc(sizeof(*apcm), GFP_KERNEL);
	if (apcm == NULL)
		return -ENOMEM;
	apcm->chip = chip;
	apcm->pcm = pcm;
	apcm->codec = codec;
	pcm->private_data = apcm;
	pcm->private_free = azx_pcm_free;
	if (cpcm->pcm_type == HDA_PCM_TYPE_MODEM)
		pcm->dev_class = SNDRV_PCM_CLASS_MODEM;
	list_add_tail(&apcm->list, &chip->pcm_list);
	cpcm->pcm = pcm;
	for (s = 0; s < 2; s++) {
		apcm->hinfo[s] = &cpcm->stream[s];
		if (cpcm->stream[s].substreams)
			snd_pcm_set_ops(pcm, s, &azx_pcm_ops);
	}
	/* buffer pre-allocation */
	size = CONFIG_SND_HDA_PREALLOC_SIZE * 1024;
	if (size > MAX_PREALLOC_SIZE)
		size = MAX_PREALLOC_SIZE;
	snd_pcm_lib_preallocate_pages_for_all(pcm, SNDRV_DMA_TYPE_DEV_SG,
					      chip->card->dev,
					      size, MAX_PREALLOC_SIZE);
	/* link to codec */
	pcm->dev = &codec->dev;
	return 0;
}
EXPORT_SYMBOL_GPL(azx_attach_pcm_stream);

MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("Common HDA driver funcitons");