Merge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/async_tx

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/async_tx:
  drivers/dma: Correct use after free
  drivers/dma: drop unnecesary memset
  ioat2,3: put channel hardware in known state at init
  async_tx: expand async raid6 test to cover ioatdma corner case
  ioat3: fix p-disabled q-continuation
  sh: fix DMA driver's descriptor chaining and cookie assignment
  dma: at_hdmac: correct incompatible type for argument 1 of 'spin_lock_bh'

12 files changed, 341 insertions, 159 deletions

diff --git a/crypto/async_tx/raid6test.c b/crypto/async_tx/raid6test.c
index 3ec27c7e62ea..f84f6b4301d9 100644
--- a/crypto/async_tx/raid6test.c
+++ b/crypto/async_tx/raid6test.c
@@ -214,6 +214,13 @@ static int raid6_test(void)
 		err += test(4, &tests);
 	if (NDISKS > 5)
 		err += test(5, &tests);
+	/* the 11 and 12 disk cases are special for ioatdma (p-disabled
+	 * q-continuation without extended descriptor)
+	 */
+	if (NDISKS > 12) {
+		err += test(11, &tests);
+		err += test(12, &tests);
+	}
 	err += test(NDISKS, &tests);
 
 	pr("\n");
diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
index f15112569c1d..efc1a61ca231 100644
--- a/drivers/dma/at_hdmac.c
+++ b/drivers/dma/at_hdmac.c
@@ -815,7 +815,7 @@ atc_is_tx_complete(struct dma_chan *chan,
 	dev_vdbg(chan2dev(chan), "is_tx_complete: %d (d%d, u%d)\n",
 			cookie, done ? *done : 0, used ? *used : 0);
 
-	spin_lock_bh(atchan->lock);
+	spin_lock_bh(&atchan->lock);
 
 	last_complete = atchan->completed_cookie;
 	last_used = chan->cookie;
@@ -830,7 +830,7 @@ atc_is_tx_complete(struct dma_chan *chan,
 		ret = dma_async_is_complete(cookie, last_complete, last_used);
 	}
 
-	spin_unlock_bh(atchan->lock);
+	spin_unlock_bh(&atchan->lock);
 
 	if (done)
 		*done = last_complete;
diff --git a/drivers/dma/coh901318.c b/drivers/dma/coh901318.c
index 4a99cd94536b..b5f2ee0f8e2c 100644
--- a/drivers/dma/coh901318.c
+++ b/drivers/dma/coh901318.c
@@ -1294,8 +1294,8 @@ static int __exit coh901318_remove(struct platform_device *pdev)
 	dma_async_device_unregister(&base->dma_slave);
 	coh901318_pool_destroy(&base->pool);
 	free_irq(platform_get_irq(pdev, 0), base);
-	kfree(base);
 	iounmap(base->virtbase);
+	kfree(base);
 	release_mem_region(pdev->resource->start,
 			   resource_size(pdev->resource));
 	return 0;
diff --git a/drivers/dma/dw_dmac.c b/drivers/dma/dw_dmac.c
index 285bed0fe17b..d28369f7afd2 100644
--- a/drivers/dma/dw_dmac.c
+++ b/drivers/dma/dw_dmac.c
@@ -1270,8 +1270,6 @@ static int __init dw_probe(struct platform_device *pdev)
 		goto err_kfree;
 	}
 
-	memset(dw, 0, sizeof *dw);
-
 	dw->regs = ioremap(io->start, DW_REGLEN);
 	if (!dw->regs) {
 		err = -ENOMEM;
diff --git a/drivers/dma/ioat/dma.c b/drivers/dma/ioat/dma.c
index c524d36d3c2e..dcc4ab78b32b 100644
--- a/drivers/dma/ioat/dma.c
+++ b/drivers/dma/ioat/dma.c
@@ -1032,7 +1032,7 @@ int __devinit ioat_probe(struct ioatdma_device *device)
 	dma->dev = &pdev->dev;
 
 	if (!dma->chancnt) {
-		dev_err(dev, "zero channels detected\n");
+		dev_err(dev, "channel enumeration error\n");
 		goto err_setup_interrupts;
 	}
 
diff --git a/drivers/dma/ioat/dma.h b/drivers/dma/ioat/dma.h
index 45edde996480..bbc3e78ef333 100644
--- a/drivers/dma/ioat/dma.h
+++ b/drivers/dma/ioat/dma.h
@@ -60,6 +60,7 @@
  * @dca: direct cache access context
  * @intr_quirk: interrupt setup quirk (for ioat_v1 devices)
  * @enumerate_channels: hw version specific channel enumeration
+ * @reset_hw: hw version specific channel (re)initialization
  * @cleanup_tasklet: select between the v2 and v3 cleanup routines
  * @timer_fn: select between the v2 and v3 timer watchdog routines
  * @self_test: hardware version specific self test for each supported op type
@@ -78,6 +79,7 @@ struct ioatdma_device {
 	struct dca_provider *dca;
 	void (*intr_quirk)(struct ioatdma_device *device);
 	int (*enumerate_channels)(struct ioatdma_device *device);
+	int (*reset_hw)(struct ioat_chan_common *chan);
 	void (*cleanup_tasklet)(unsigned long data);
 	void (*timer_fn)(unsigned long data);
 	int (*self_test)(struct ioatdma_device *device);
@@ -264,6 +266,22 @@ static inline void ioat_suspend(struct ioat_chan_common *chan)
 	writeb(IOAT_CHANCMD_SUSPEND, chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
 }
 
+static inline void ioat_reset(struct ioat_chan_common *chan)
+{
+	u8 ver = chan->device->version;
+
+	writeb(IOAT_CHANCMD_RESET, chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
+}
+
+static inline bool ioat_reset_pending(struct ioat_chan_common *chan)
+{
+	u8 ver = chan->device->version;
+	u8 cmd;
+
+	cmd = readb(chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
+	return (cmd & IOAT_CHANCMD_RESET) == IOAT_CHANCMD_RESET;
+}
+
 static inline void ioat_set_chainaddr(struct ioat_dma_chan *ioat, u64 addr)
 {
 	struct ioat_chan_common *chan = &ioat->base;
diff --git a/drivers/dma/ioat/dma_v2.c b/drivers/dma/ioat/dma_v2.c
index 8f1f7f05deaa..5f7a500e18d0 100644
--- a/drivers/dma/ioat/dma_v2.c
+++ b/drivers/dma/ioat/dma_v2.c
@@ -239,20 +239,50 @@ void __ioat2_restart_chan(struct ioat2_dma_chan *ioat)
 		__ioat2_start_null_desc(ioat);
 }
 
-static void ioat2_restart_channel(struct ioat2_dma_chan *ioat)
+int ioat2_quiesce(struct ioat_chan_common *chan, unsigned long tmo)
 {
-	struct ioat_chan_common *chan = &ioat->base;
-	unsigned long phys_complete;
+	unsigned long end = jiffies + tmo;
+	int err = 0;
 	u32 status;
 
 	status = ioat_chansts(chan);
 	if (is_ioat_active(status) || is_ioat_idle(status))
 		ioat_suspend(chan);
 	while (is_ioat_active(status) || is_ioat_idle(status)) {
+		if (end && time_after(jiffies, end)) {
+			err = -ETIMEDOUT;
+			break;
+		}
 		status = ioat_chansts(chan);
 		cpu_relax();
 	}
 
+	return err;
+}
+
+int ioat2_reset_sync(struct ioat_chan_common *chan, unsigned long tmo)
+{
+	unsigned long end = jiffies + tmo;
+	int err = 0;
+
+	ioat_reset(chan);
+	while (ioat_reset_pending(chan)) {
+		if (end && time_after(jiffies, end)) {
+			err = -ETIMEDOUT;
+			break;
+		}
+		cpu_relax();
+	}
+
+	return err;
+}
+
+static void ioat2_restart_channel(struct ioat2_dma_chan *ioat)
+{
+	struct ioat_chan_common *chan = &ioat->base;
+	unsigned long phys_complete;
+
+	ioat2_quiesce(chan, 0);
 	if (ioat_cleanup_preamble(chan, &phys_complete))
 		__cleanup(ioat, phys_complete);
 
@@ -318,6 +348,19 @@ void ioat2_timer_event(unsigned long data)
 	spin_unlock_bh(&chan->cleanup_lock);
 }
 
+static int ioat2_reset_hw(struct ioat_chan_common *chan)
+{
+	/* throw away whatever the channel was doing and get it initialized */
+	u32 chanerr;
+
+	ioat2_quiesce(chan, msecs_to_jiffies(100));
+
+	chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET);
+	writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET);
+
+	return ioat2_reset_sync(chan, msecs_to_jiffies(200));
+}
+
 /**
  * ioat2_enumerate_channels - find and initialize the device's channels
  * @device: the device to be enumerated
@@ -360,6 +403,10 @@ int ioat2_enumerate_channels(struct ioatdma_device *device)
 				  (unsigned long) ioat);
 		ioat->xfercap_log = xfercap_log;
 		spin_lock_init(&ioat->ring_lock);
+		if (device->reset_hw(&ioat->base)) {
+			i = 0;
+			break;
+		}
 	}
 	dma->chancnt = i;
 	return i;
@@ -467,7 +514,6 @@ int ioat2_alloc_chan_resources(struct dma_chan *c)
 	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
 	struct ioat_chan_common *chan = &ioat->base;
 	struct ioat_ring_ent **ring;
-	u32 chanerr;
 	int order;
 
 	/* have we already been set up? */
@@ -477,12 +523,6 @@ int ioat2_alloc_chan_resources(struct dma_chan *c)
 	/* Setup register to interrupt and write completion status on error */
 	writew(IOAT_CHANCTRL_RUN, chan->reg_base + IOAT_CHANCTRL_OFFSET);
 
-	chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET);
-	if (chanerr) {
-		dev_err(to_dev(chan), "CHANERR = %x, clearing\n", chanerr);
-		writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET);
-	}
-
 	/* allocate a completion writeback area */
 	/* doing 2 32bit writes to mmio since 1 64b write doesn't work */
 	chan->completion = pci_pool_alloc(chan->device->completion_pool,
@@ -746,13 +786,7 @@ void ioat2_free_chan_resources(struct dma_chan *c)
 	tasklet_disable(&chan->cleanup_task);
 	del_timer_sync(&chan->timer);
 	device->cleanup_tasklet((unsigned long) ioat);
-
-	/* Delay 100ms after reset to allow internal DMA logic to quiesce
-	 * before removing DMA descriptor resources.
-	 */
-	writeb(IOAT_CHANCMD_RESET,
-	       chan->reg_base + IOAT_CHANCMD_OFFSET(chan->device->version));
-	mdelay(100);
+	device->reset_hw(chan);
 
 	spin_lock_bh(&ioat->ring_lock);
 	descs = ioat2_ring_space(ioat);
@@ -839,6 +873,7 @@ int __devinit ioat2_dma_probe(struct ioatdma_device *device, int dca)
 	int err;
 
 	device->enumerate_channels = ioat2_enumerate_channels;
+	device->reset_hw = ioat2_reset_hw;
 	device->cleanup_tasklet = ioat2_cleanup_tasklet;
 	device->timer_fn = ioat2_timer_event;
 	device->self_test = ioat_dma_self_test;
diff --git a/drivers/dma/ioat/dma_v2.h b/drivers/dma/ioat/dma_v2.h
index 1d849ef74d5f..3afad8da43cc 100644
--- a/drivers/dma/ioat/dma_v2.h
+++ b/drivers/dma/ioat/dma_v2.h
@@ -185,6 +185,8 @@ bool reshape_ring(struct ioat2_dma_chan *ioat, int order);
 void __ioat2_issue_pending(struct ioat2_dma_chan *ioat);
 void ioat2_cleanup_tasklet(unsigned long data);
 void ioat2_timer_event(unsigned long data);
+int ioat2_quiesce(struct ioat_chan_common *chan, unsigned long tmo);
+int ioat2_reset_sync(struct ioat_chan_common *chan, unsigned long tmo);
 extern struct kobj_type ioat2_ktype;
 extern struct kmem_cache *ioat2_cache;
 #endif /* IOATDMA_V2_H */
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index 42f6f10fb0cc..9908c9e94b2d 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -650,9 +650,11 @@ __ioat3_prep_pq_lock(struct dma_chan *c, enum sum_check_flags *result,
 
 	num_descs = ioat2_xferlen_to_descs(ioat, len);
 	/* we need 2x the number of descriptors to cover greater than 3
-	 * sources
+	 * sources (we need 1 extra source in the q-only continuation
+	 * case and 3 extra sources in the p+q continuation case.
 	 */
-	if (src_cnt > 3 || flags & DMA_PREP_CONTINUE) {
+	if (src_cnt + dmaf_p_disabled_continue(flags) > 3 ||
+	    (dmaf_continue(flags) && !dmaf_p_disabled_continue(flags))) {
 		with_ext = 1;
 		num_descs *= 2;
 	} else
@@ -1128,6 +1130,45 @@ static int __devinit ioat3_dma_self_test(struct ioatdma_device *device)
 	return 0;
 }
 
+static int ioat3_reset_hw(struct ioat_chan_common *chan)
+{
+	/* throw away whatever the channel was doing and get it
+	 * initialized, with ioat3 specific workarounds
+	 */
+	struct ioatdma_device *device = chan->device;
+	struct pci_dev *pdev = device->pdev;
+	u32 chanerr;
+	u16 dev_id;
+	int err;
+
+	ioat2_quiesce(chan, msecs_to_jiffies(100));
+
+	chanerr = readl(chan->reg_base + IOAT_CHANERR_OFFSET);
+	writel(chanerr, chan->reg_base + IOAT_CHANERR_OFFSET);
+
+	/* -= IOAT ver.3 workarounds =- */
+	/* Write CHANERRMSK_INT with 3E07h to mask out the errors
+	 * that can cause stability issues for IOAT ver.3, and clear any
+	 * pending errors
+	 */
+	pci_write_config_dword(pdev, IOAT_PCI_CHANERRMASK_INT_OFFSET, 0x3e07);
+	err = pci_read_config_dword(pdev, IOAT_PCI_CHANERR_INT_OFFSET, &chanerr);
+	if (err) {
+		dev_err(&pdev->dev, "channel error register unreachable\n");
+		return err;
+	}
+	pci_write_config_dword(pdev, IOAT_PCI_CHANERR_INT_OFFSET, chanerr);
+
+	/* Clear DMAUNCERRSTS Cfg-Reg Parity Error status bit
+	 * (workaround for spurious config parity error after restart)
+	 */
+	pci_read_config_word(pdev, IOAT_PCI_DEVICE_ID_OFFSET, &dev_id);
+	if (dev_id == PCI_DEVICE_ID_INTEL_IOAT_TBG0)
+		pci_write_config_dword(pdev, IOAT_PCI_DMAUNCERRSTS_OFFSET, 0x10);
+
+	return ioat2_reset_sync(chan, msecs_to_jiffies(200));
+}
+
 int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca)
 {
 	struct pci_dev *pdev = device->pdev;
@@ -1137,10 +1178,10 @@ int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	struct ioat_chan_common *chan;
 	bool is_raid_device = false;
 	int err;
-	u16 dev_id;
 	u32 cap;
 
 	device->enumerate_channels = ioat2_enumerate_channels;
+	device->reset_hw = ioat3_reset_hw;
 	device->self_test = ioat3_dma_self_test;
 	dma = &device->common;
 	dma->device_prep_dma_memcpy = ioat2_dma_prep_memcpy_lock;
@@ -1216,19 +1257,6 @@ int __devinit ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	dma->device_prep_dma_xor_val = NULL;
 	#endif
 
-	/* -= IOAT ver.3 workarounds =- */
-	/* Write CHANERRMSK_INT with 3E07h to mask out the errors
-	 * that can cause stability issues for IOAT ver.3
-	 */
-	pci_write_config_dword(pdev, IOAT_PCI_CHANERRMASK_INT_OFFSET, 0x3e07);
-
-	/* Clear DMAUNCERRSTS Cfg-Reg Parity Error status bit
-	 * (workaround for spurious config parity error after restart)
-	 */
-	pci_read_config_word(pdev, IOAT_PCI_DEVICE_ID_OFFSET, &dev_id);
-	if (dev_id == PCI_DEVICE_ID_INTEL_IOAT_TBG0)
-		pci_write_config_dword(pdev, IOAT_PCI_DMAUNCERRSTS_OFFSET, 0x10);
-
 	err = ioat_probe(device);
 	if (err)
 		return err;
diff --git a/drivers/dma/ioat/registers.h b/drivers/dma/ioat/registers.h
index f015ec196700..e8ae63baf588 100644
--- a/drivers/dma/ioat/registers.h
+++ b/drivers/dma/ioat/registers.h
@@ -27,6 +27,7 @@
 
 #define IOAT_PCI_DEVICE_ID_OFFSET		0x02
 #define IOAT_PCI_DMAUNCERRSTS_OFFSET		0x148
+#define IOAT_PCI_CHANERR_INT_OFFSET		0x180
 #define IOAT_PCI_CHANERRMASK_INT_OFFSET		0x184
 
 /* MMIO Device Registers */
diff --git a/drivers/dma/shdma.c b/drivers/dma/shdma.c
index 2e4a54c8afeb..d10cc899c460 100644
--- a/drivers/dma/shdma.c
+++ b/drivers/dma/shdma.c
@@ -23,16 +23,19 @@
 #include <linux/dmaengine.h>
 #include <linux/delay.h>
 #include <linux/dma-mapping.h>
-#include <linux/dmapool.h>
 #include <linux/platform_device.h>
 #include <cpu/dma.h>
 #include <asm/dma-sh.h>
 #include "shdma.h"
 
 /* DMA descriptor control */
-#define DESC_LAST	(-1)
-#define DESC_COMP	(1)
-#define DESC_NCOMP	(0)
+enum sh_dmae_desc_status {
+	DESC_IDLE,
+	DESC_PREPARED,
+	DESC_SUBMITTED,
+	DESC_COMPLETED,	/* completed, have to call callback */
+	DESC_WAITING,	/* callback called, waiting for ack / re-submit */
+};
 
 #define NR_DESCS_PER_CHANNEL 32
 /*
@@ -45,6 +48,8 @@
  */
 #define RS_DEFAULT  (RS_DUAL)
 
+static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all);
+
 #define SH_DMAC_CHAN_BASE(id) (dma_base_addr[id])
 static void sh_dmae_writel(struct sh_dmae_chan *sh_dc, u32 data, u32 reg)
 {
@@ -106,11 +111,11 @@ static inline unsigned int calc_xmit_shift(struct sh_dmae_chan *sh_chan)
 	return ts_shift[(chcr & CHCR_TS_MASK) >> CHCR_TS_SHIFT];
 }
 
-static void dmae_set_reg(struct sh_dmae_chan *sh_chan, struct sh_dmae_regs hw)
+static void dmae_set_reg(struct sh_dmae_chan *sh_chan, struct sh_dmae_regs *hw)
 {
-	sh_dmae_writel(sh_chan, hw.sar, SAR);
-	sh_dmae_writel(sh_chan, hw.dar, DAR);
-	sh_dmae_writel(sh_chan, hw.tcr >> calc_xmit_shift(sh_chan), TCR);
+	sh_dmae_writel(sh_chan, hw->sar, SAR);
+	sh_dmae_writel(sh_chan, hw->dar, DAR);
+	sh_dmae_writel(sh_chan, hw->tcr >> calc_xmit_shift(sh_chan), TCR);
 }
 
 static void dmae_start(struct sh_dmae_chan *sh_chan)
@@ -184,8 +189,9 @@ static int dmae_set_dmars(struct sh_dmae_chan *sh_chan, u16 val)
 
 static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx)
 {
-	struct sh_desc *desc = tx_to_sh_desc(tx);
+	struct sh_desc *desc = tx_to_sh_desc(tx), *chunk, *last = desc, *c;
 	struct sh_dmae_chan *sh_chan = to_sh_chan(tx->chan);
+	dma_async_tx_callback callback = tx->callback;
 	dma_cookie_t cookie;
 
 	spin_lock_bh(&sh_chan->desc_lock);
@@ -195,45 +201,53 @@ static dma_cookie_t sh_dmae_tx_submit(struct dma_async_tx_descriptor *tx)
 	if (cookie < 0)
 		cookie = 1;
 
-	/* If desc only in the case of 1 */
-	if (desc->async_tx.cookie != -EBUSY)
-		desc->async_tx.cookie = cookie;
-	sh_chan->common.cookie = desc->async_tx.cookie;
+	sh_chan->common.cookie = cookie;
+	tx->cookie = cookie;
+
+	/* Mark all chunks of this descriptor as submitted, move to the queue */
+	list_for_each_entry_safe(chunk, c, desc->node.prev, node) {
+		/*
+		 * All chunks are on the global ld_free, so, we have to find
+		 * the end of the chain ourselves
+		 */
+		if (chunk != desc && (chunk->mark == DESC_IDLE ||
+				      chunk->async_tx.cookie > 0 ||
+				      chunk->async_tx.cookie == -EBUSY ||
+				      &chunk->node == &sh_chan->ld_free))
+			break;
+		chunk->mark = DESC_SUBMITTED;
+		/* Callback goes to the last chunk */
+		chunk->async_tx.callback = NULL;
+		chunk->cookie = cookie;
+		list_move_tail(&chunk->node, &sh_chan->ld_queue);
+		last = chunk;
+	}
+
+	last->async_tx.callback = callback;
+	last->async_tx.callback_param = tx->callback_param;
 
-	list_splice_init(&desc->tx_list, sh_chan->ld_queue.prev);
+	dev_dbg(sh_chan->dev, "submit #%d@%p on %d: %x[%d] -> %x\n",
+		tx->cookie, &last->async_tx, sh_chan->id,
+		desc->hw.sar, desc->hw.tcr, desc->hw.dar);
 
 	spin_unlock_bh(&sh_chan->desc_lock);
 
 	return cookie;
 }
 
+/* Called with desc_lock held */
 static struct sh_desc *sh_dmae_get_desc(struct sh_dmae_chan *sh_chan)
 {
-	struct sh_desc *desc, *_desc, *ret = NULL;
+	struct sh_desc *desc;
 
-	spin_lock_bh(&sh_chan->desc_lock);
-	list_for_each_entry_safe(desc, _desc, &sh_chan->ld_free, node) {
-		if (async_tx_test_ack(&desc->async_tx)) {
+	list_for_each_entry(desc, &sh_chan->ld_free, node)
+		if (desc->mark != DESC_PREPARED) {
+			BUG_ON(desc->mark != DESC_IDLE);
 			list_del(&desc->node);
-			ret = desc;
-			break;
+			return desc;
 		}
-	}
-	spin_unlock_bh(&sh_chan->desc_lock);
-
-	return ret;
-}
-
-static void sh_dmae_put_desc(struct sh_dmae_chan *sh_chan, struct sh_desc *desc)
-{
-	if (desc) {
-		spin_lock_bh(&sh_chan->desc_lock);
-
-		list_splice_init(&desc->tx_list, &sh_chan->ld_free);
-		list_add(&desc->node, &sh_chan->ld_free);
 
-		spin_unlock_bh(&sh_chan->desc_lock);
-	}
+	return NULL;
 }
 
 static int sh_dmae_alloc_chan_resources(struct dma_chan *chan)
@@ -252,11 +266,10 @@ static int sh_dmae_alloc_chan_resources(struct dma_chan *chan)
 		dma_async_tx_descriptor_init(&desc->async_tx,
 					&sh_chan->common);
 		desc->async_tx.tx_submit = sh_dmae_tx_submit;
-		desc->async_tx.flags = DMA_CTRL_ACK;
-		INIT_LIST_HEAD(&desc->tx_list);
-		sh_dmae_put_desc(sh_chan, desc);
+		desc->mark = DESC_IDLE;
 
 		spin_lock_bh(&sh_chan->desc_lock);
+		list_add(&desc->node, &sh_chan->ld_free);
 		sh_chan->descs_allocated++;
 	}
 	spin_unlock_bh(&sh_chan->desc_lock);
@@ -273,7 +286,10 @@ static void sh_dmae_free_chan_resources(struct dma_chan *chan)
 	struct sh_desc *desc, *_desc;
 	LIST_HEAD(list);
 
-	BUG_ON(!list_empty(&sh_chan->ld_queue));
+	/* Prepared and not submitted descriptors can still be on the queue */
+	if (!list_empty(&sh_chan->ld_queue))
+		sh_dmae_chan_ld_cleanup(sh_chan, true);
+
 	spin_lock_bh(&sh_chan->desc_lock);
 
 	list_splice_init(&sh_chan->ld_free, &list);
@@ -292,6 +308,8 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
 	struct sh_dmae_chan *sh_chan;
 	struct sh_desc *first = NULL, *prev = NULL, *new;
 	size_t copy_size;
+	LIST_HEAD(tx_list);
+	int chunks = (len + SH_DMA_TCR_MAX) / (SH_DMA_TCR_MAX + 1);
 
 	if (!chan)
 		return NULL;
@@ -301,108 +319,189 @@ static struct dma_async_tx_descriptor *sh_dmae_prep_memcpy(
 
 	sh_chan = to_sh_chan(chan);
 
+	/* Have to lock the whole loop to protect against concurrent release */
+	spin_lock_bh(&sh_chan->desc_lock);
+
+	/*
+	 * Chaining:
+	 * first descriptor is what user is dealing with in all API calls, its
+	 *	cookie is at first set to -EBUSY, at tx-submit to a positive
+	 *	number
+	 * if more than one chunk is needed further chunks have cookie = -EINVAL
+	 * the last chunk, if not equal to the first, has cookie = -ENOSPC
+	 * all chunks are linked onto the tx_list head with their .node heads
+	 *	only during this function, then they are immediately spliced
+	 *	back onto the free list in form of a chain
+	 */
 	do {
-		/* Allocate the link descriptor from DMA pool */
+		/* Allocate the link descriptor from the free list */
 		new = sh_dmae_get_desc(sh_chan);
 		if (!new) {
 			dev_err(sh_chan->dev,
 				"No free memory for link descriptor\n");
-			goto err_get_desc;
+			list_for_each_entry(new, &tx_list, node)
+				new->mark = DESC_IDLE;
+			list_splice(&tx_list, &sh_chan->ld_free);
+			spin_unlock_bh(&sh_chan->desc_lock);
+			return NULL;
 		}
 
-		copy_size = min(len, (size_t)SH_DMA_TCR_MAX);
+		copy_size = min(len, (size_t)SH_DMA_TCR_MAX + 1);
 
 		new->hw.sar = dma_src;
 		new->hw.dar = dma_dest;
 		new->hw.tcr = copy_size;
-		if (!first)
+		if (!first) {
+			/* First desc */
+			new->async_tx.cookie = -EBUSY;
 			first = new;
+		} else {
+			/* Other desc - invisible to the user */
+			new->async_tx.cookie = -EINVAL;
+		}
 
-		new->mark = DESC_NCOMP;
-		async_tx_ack(&new->async_tx);
+		dev_dbg(sh_chan->dev,
+			"chaining %u of %u with %p, dst %x, cookie %d\n",
+			copy_size, len, &new->async_tx, dma_dest,
+			new->async_tx.cookie);
+
+		new->mark = DESC_PREPARED;
+		new->async_tx.flags = flags;
+		new->chunks = chunks--;
 
 		prev = new;
 		len -= copy_size;
 		dma_src += copy_size;
 		dma_dest += copy_size;
 		/* Insert the link descriptor to the LD ring */
-		list_add_tail(&new->node, &first->tx_list);
+		list_add_tail(&new->node, &tx_list);
 	} while (len);
 
-	new->async_tx.flags = flags; /* client is in control of this ack */
-	new->async_tx.cookie = -EBUSY; /* Last desc */
+	if (new != first)
+		new->async_tx.cookie = -ENOSPC;
 
-	return &first->async_tx;
+	/* Put them back on the free list, so, they don't get lost */
+	list_splice_tail(&tx_list, &sh_chan->ld_free);
 
-err_get_desc:
-	sh_dmae_put_desc(sh_chan, first);
-	return NULL;
+	spin_unlock_bh(&sh_chan->desc_lock);
 
+	return &first->async_tx;
 }
 
-/*
- * sh_chan_ld_cleanup - Clean up link descriptors
- *
- * This function clean up the ld_queue of DMA channel.
- */
-static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan)
+static dma_async_tx_callback __ld_cleanup(struct sh_dmae_chan *sh_chan, bool all)
 {
 	struct sh_desc *desc, *_desc;
+	/* Is the "exposed" head of a chain acked? */
+	bool head_acked = false;
+	dma_cookie_t cookie = 0;
+	dma_async_tx_callback callback = NULL;
+	void *param = NULL;
 
 	spin_lock_bh(&sh_chan->desc_lock);
 	list_for_each_entry_safe(desc, _desc, &sh_chan->ld_queue, node) {
-		dma_async_tx_callback callback;
-		void *callback_param;
-
-		/* non send data */
-		if (desc->mark == DESC_NCOMP)
+		struct dma_async_tx_descriptor *tx = &desc->async_tx;
+
+		BUG_ON(tx->cookie > 0 && tx->cookie != desc->cookie);
+		BUG_ON(desc->mark != DESC_SUBMITTED &&
+		       desc->mark != DESC_COMPLETED &&
+		       desc->mark != DESC_WAITING);
+
+		/*
+		 * queue is ordered, and we use this loop to (1) clean up all
+		 * completed descriptors, and to (2) update descriptor flags of
+		 * any chunks in a (partially) completed chain
+		 */
+		if (!all && desc->mark == DESC_SUBMITTED &&
+		    desc->cookie != cookie)
 			break;
 
-		/* send data sesc */
-		callback = desc->async_tx.callback;
-		callback_param = desc->async_tx.callback_param;
+		if (tx->cookie > 0)
+			cookie = tx->cookie;
 
-		/* Remove from ld_queue list */
-		list_splice_init(&desc->tx_list, &sh_chan->ld_free);
+		if (desc->mark == DESC_COMPLETED && desc->chunks == 1) {
+			BUG_ON(sh_chan->completed_cookie != desc->cookie - 1);
+			sh_chan->completed_cookie = desc->cookie;
+		}
 
-		dev_dbg(sh_chan->dev, "link descriptor %p will be recycle.\n",
-				desc);
+		/* Call callback on the last chunk */
+		if (desc->mark == DESC_COMPLETED && tx->callback) {
+			desc->mark = DESC_WAITING;
+			callback = tx->callback;
+			param = tx->callback_param;
+			dev_dbg(sh_chan->dev, "descriptor #%d@%p on %d callback\n",
+				tx->cookie, tx, sh_chan->id);
+			BUG_ON(desc->chunks != 1);
+			break;
+		}
 
-		list_move(&desc->node, &sh_chan->ld_free);
-		/* Run the link descriptor callback function */
-		if (callback) {
-			spin_unlock_bh(&sh_chan->desc_lock);
-			dev_dbg(sh_chan->dev, "link descriptor %p callback\n",
-					desc);
-			callback(callback_param);
-			spin_lock_bh(&sh_chan->desc_lock);
+		if (tx->cookie > 0 || tx->cookie == -EBUSY) {
+			if (desc->mark == DESC_COMPLETED) {
+				BUG_ON(tx->cookie < 0);
+				desc->mark = DESC_WAITING;
+			}
+			head_acked = async_tx_test_ack(tx);
+		} else {
+			switch (desc->mark) {
+			case DESC_COMPLETED:
+				desc->mark = DESC_WAITING;
+				/* Fall through */
+			case DESC_WAITING:
+				if (head_acked)
+					async_tx_ack(&desc->async_tx);
+			}
+		}
+
+		dev_dbg(sh_chan->dev, "descriptor %p #%d completed.\n",
+			tx, tx->cookie);
+
+		if (((desc->mark == DESC_COMPLETED ||
+		      desc->mark == DESC_WAITING) &&
+		     async_tx_test_ack(&desc->async_tx)) || all) {
+			/* Remove from ld_queue list */
+			desc->mark = DESC_IDLE;
+			list_move(&desc->node, &sh_chan->ld_free);
 		}
 	}
 	spin_unlock_bh(&sh_chan->desc_lock);
+
+	if (callback)
+		callback(param);
+
+	return callback;
+}
+
+/*
+ * sh_chan_ld_cleanup - Clean up link descriptors
+ *
+ * This function cleans up the ld_queue of DMA channel.
+ */
+static void sh_dmae_chan_ld_cleanup(struct sh_dmae_chan *sh_chan, bool all)
+{
+	while (__ld_cleanup(sh_chan, all))
+		;
 }
 
 static void sh_chan_xfer_ld_queue(struct sh_dmae_chan *sh_chan)
 {
-	struct list_head *ld_node;
-	struct sh_dmae_regs hw;
+	struct sh_desc *sd;
 
+	spin_lock_bh(&sh_chan->desc_lock);
 	/* DMA work check */
-	if (dmae_is_busy(sh_chan))
+	if (dmae_is_busy(sh_chan)) {
+		spin_unlock_bh(&sh_chan->desc_lock);
 		return;
+	}
 
 	/* Find the first un-transfer desciptor */
-	for (ld_node = sh_chan->ld_queue.next;
-		(ld_node != &sh_chan->ld_queue)
-			&& (to_sh_desc(ld_node)->mark == DESC_COMP);
-		ld_node = ld_node->next)
-		cpu_relax();
-
-	if (ld_node != &sh_chan->ld_queue) {
-		/* Get the ld start address from ld_queue */
-		hw = to_sh_desc(ld_node)->hw;
-		dmae_set_reg(sh_chan, hw);
-		dmae_start(sh_chan);
-	}
+	list_for_each_entry(sd, &sh_chan->ld_queue, node)
+		if (sd->mark == DESC_SUBMITTED) {
+			/* Get the ld start address from ld_queue */
+			dmae_set_reg(sh_chan, &sd->hw);
+			dmae_start(sh_chan);
+			break;
+		}
+
+	spin_unlock_bh(&sh_chan->desc_lock);
 }
 
 static void sh_dmae_memcpy_issue_pending(struct dma_chan *chan)
@@ -420,12 +519,11 @@ static enum dma_status sh_dmae_is_complete(struct dma_chan *chan,
 	dma_cookie_t last_used;
 	dma_cookie_t last_complete;
 
-	sh_dmae_chan_ld_cleanup(sh_chan);
+	sh_dmae_chan_ld_cleanup(sh_chan, false);
 
 	last_used = chan->cookie;
 	last_complete = sh_chan->completed_cookie;
-	if (last_complete == -EBUSY)
-		last_complete = last_used;
+	BUG_ON(last_complete < 0);
 
 	if (done)
 		*done = last_complete;
@@ -480,11 +578,13 @@ static irqreturn_t sh_dmae_err(int irq, void *data)
 		err = sh_dmae_rst(0);
 		if (err)
 			return err;
+#ifdef SH_DMAC_BASE1
 		if (shdev->pdata.mode & SHDMA_DMAOR1) {
 			err = sh_dmae_rst(1);
 			if (err)
 				return err;
 		}
+#endif
 		disable_irq(irq);
 		return IRQ_HANDLED;
 	}
@@ -494,35 +594,25 @@ static irqreturn_t sh_dmae_err(int irq, void *data)
 static void dmae_do_tasklet(unsigned long data)
 {
 	struct sh_dmae_chan *sh_chan = (struct sh_dmae_chan *)data;
-	struct sh_desc *desc, *_desc, *cur_desc = NULL;
+	struct sh_desc *desc;
 	u32 sar_buf = sh_dmae_readl(sh_chan, SAR);
 
-	list_for_each_entry_safe(desc, _desc,
-				 &sh_chan->ld_queue, node) {
-		if ((desc->hw.sar + desc->hw.tcr) == sar_buf) {
-			cur_desc = desc;
+	spin_lock(&sh_chan->desc_lock);
+	list_for_each_entry(desc, &sh_chan->ld_queue, node) {
+		if ((desc->hw.sar + desc->hw.tcr) == sar_buf &&
+		    desc->mark == DESC_SUBMITTED) {
+			dev_dbg(sh_chan->dev, "done #%d@%p dst %u\n",
+				desc->async_tx.cookie, &desc->async_tx,
+				desc->hw.dar);
+			desc->mark = DESC_COMPLETED;
 			break;
 		}
 	}
+	spin_unlock(&sh_chan->desc_lock);
 
-	if (cur_desc) {
-		switch (cur_desc->async_tx.cookie) {
-		case 0: /* other desc data */
-			break;
-		case -EBUSY: /* last desc */
-		sh_chan->completed_cookie =
-				cur_desc->async_tx.cookie;
-			break;
-		default: /* first desc ( 0 < )*/
-			sh_chan->completed_cookie =
-				cur_desc->async_tx.cookie - 1;
-			break;
-		}
-		cur_desc->mark = DESC_COMP;
-	}
 	/* Next desc */
 	sh_chan_xfer_ld_queue(sh_chan);
-	sh_dmae_chan_ld_cleanup(sh_chan);
+	sh_dmae_chan_ld_cleanup(sh_chan, false);
 }
 
 static unsigned int get_dmae_irq(unsigned int id)
diff --git a/drivers/dma/shdma.h b/drivers/dma/shdma.h
index 60b81e529b42..108f1cffb6f5 100644
--- a/drivers/dma/shdma.h
+++ b/drivers/dma/shdma.h
@@ -13,9 +13,9 @@
 #ifndef __DMA_SHDMA_H
 #define __DMA_SHDMA_H
 
-#include <linux/device.h>
-#include <linux/dmapool.h>
 #include <linux/dmaengine.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
 
 #define SH_DMA_TCR_MAX 0x00FFFFFF	/* 16MB */
 
@@ -26,13 +26,16 @@ struct sh_dmae_regs {
 };
 
 struct sh_desc {
-	struct list_head tx_list;
 	struct sh_dmae_regs hw;
 	struct list_head node;
 	struct dma_async_tx_descriptor async_tx;
+	dma_cookie_t cookie;
+	int chunks;
 	int mark;
 };
 
+struct device;
+
 struct sh_dmae_chan {
 	dma_cookie_t completed_cookie;	/* The maximum cookie completed */
 	spinlock_t desc_lock;		/* Descriptor operation lock */