[PATCH v5 00/10] dmaengine: dw: serialize completion and termination

[PATCH v5 00/10] dmaengine: dw: serialize completion and termination

[James Hilliard]

Keep client callbacks in the controller tasklet while preserving polled
completion. Serialize descriptor ownership across completion, error
recovery and termination, and wait for callbacks before a client reuses
its resources.

The series first serializes completion, error recovery and cookie
accounting with termination. Separate patches extract hardware-LLP
residue calculation, move it into the status getter, simplify completion
handling and report queued residue. Channel release and
device_synchronize then gain tasklet draining before the final patch
adds deferred notifications and shared completed-descriptor cleanup.

The series adapts Serge Semin's channel-restart and callback fixes [0-2].
Polling continues to advance cookies and software-LLP blocks, while client
notifications run from the controller tasklet. Synchronization preserves
pending tasklet runs from other channels.

Based on dmaengine/fixes at 6628e7d4c6b2.

[0] https://lore.kernel.org/r/20240911184710.4207-1-fancer.lancer@gmail.com
[1] https://lore.kernel.org/r/20240911184710.4207-2-fancer.lancer@gmail.com
[2] https://lore.kernel.org/r/20240911184710.4207-3-fancer.lancer@gmail.com

Assisted-by: Codex:gpt-6-astra
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---
Changes in v5:
- Rebase onto dmaengine/fixes at 6628e7d4c6b2, using the shared
  dmaengine_chan_dev() helper for channel logging.
- Split residue handling into a behavior-preserving extraction, status
  calculation, scanner simplification and queued-residue reporting,
  following Frank's review.
- Share completed-descriptor freeing between the tasklet and channel
  release, retaining callbacks for normal completion and suppressing
  them during teardown. Name the helper dwc_free_completed().
- Carry Frank's additional review tags and put the submitter sign-off
  after the review tags.
- Link to v4: https://patch.msgid.link/20261029-submit-dw-dma-synchronize-v1-v4-0-fea447d9f285@gmail.com

Changes in v4:
- Split v3's callback-deferral patch into cookie accounting, residue
  calculation, channel-release draining and callback-deferral patches,
  following Frank's review. The resulting seven-patch series preserves
  v3's final implementation.
- Place device_synchronize before callback deferral so clients can drain
  deferred notifications when the completion list is introduced.
- Carry Frank's Reviewed-by tags on the unchanged channel-completion,
  error-recovery and device_synchronize patches.
- Link to v3: https://patch.msgid.link/20261023-submit-dw-dma-synchronize-v1-v3-0-c62bf048f5ce@gmail.com

Changes in v3:
- Order channel-completion and error-recovery locking before callback
  deferral, retaining four patches.
- Require a locked RAW.XFER event before retiring a transfer; zero source
  residue alone cannot establish destination completion.
- Preserve polled completion and software-LLP progress in tx_status(),
  queuing client notifications for the controller tasklet.
- Serialize cookie updates with termination and report queued residue.
- Drain deferred notifications on channel release and let a pending
  tasklet return without MMIO after the last channel is released.
- Adapt Serge's two patches and retain his authorship and revision notes.
  Drop the Tested-by tags that covered their earlier implementations.
- Identify the introduction of the synchronization API in the final
  patch's Fixes tag.
- Link to v2: https://patch.msgid.link/20261022-submit-dw-dma-synchronize-v1-v2-0-5244e94dca8c@gmail.com

Changes in v2:
- Carry Serge Semin's two posted fixes before the synchronization change,
  preserving his authorship and revision notes. Add the requested stable
  Cc trailers and Andy Shevchenko's Tested-by tags to those unchanged fixes.
- Add a separate error-recovery fix for stale error status, empty channel
  lists and software-LLP state when terminating or restarting a channel.
- Use tasklet_disable()/tasklet_enable() instead of tasklet_kill() so
  synchronization preserves scheduling requests from other channels.
- Explain descriptor ownership during termination and acknowledge the
  shared-tasklet scheduling race reported by Sashiko.
- Link to v1: https://patch.msgid.link/20261021-submit-dw-dma-synchronize-v1-v1-1-6af85c2d6f69@gmail.com

Revision history from Serge's series:

Link: https://lore.kernel.org/dmaengine/20240802080756.7415-1-fancer.lancer@gmail.com/
Changelog RFC:
- Add a new patch:
  [PATCH 2/2] dmaengine: dw: Fix XFER bit set, but channel not idle error
  fixing the "XFER bit set, but channel not idle" error.
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.

---
James Hilliard (6):
      dmaengine: dw: serialize error recovery with channel termination
      dmaengine: dw: complete cookies before releasing the channel lock
      dmaengine: dw: simplify transfer-complete handling
      dmaengine: dw: report residue for queued transfers
      dmaengine: dw: drain the tasklet before releasing channel resources
      dmaengine: dw: synchronize completion callbacks

Serge Semin (4):
      dmaengine: dw: Fix XFER bit set, but channel not idle error
      dmaengine: dw: extract hardware-LLP residue calculation
      dmaengine: dw: calculate residue in the status getter
      dmaengine: dw: Prevent tx-status calling DMA-desc callback

 drivers/dma/dw/core.c | 301 +++++++++++++++++++++++++-------------------------
 drivers/dma/dw/regs.h |   1 +
 2 files changed, 154 insertions(+), 148 deletions(-)
---
base-commit: 6628e7d4c6b25c8d18f1345e3a20d21829aae7e4
change-id: 20261021-submit-dw-dma-synchronize-v1-e550f47fd1da

Best regards,
--  
James Hilliard <james.hilliard1@gmail.com>

[PATCH v5 01/10] dmaengine: dw: Fix XFER bit set, but channel not idle error

[James Hilliard]

From: Serge Semin <fancer.lancer@gmail.com>

If a client driver gets to use the DW DMAC engine device tougher
than usual, with occasional DMA-transfers termination and restart, then
the next error can be randomly spotted in the system log:

> dma dma0chan0: BUG: XFER bit set, but channel not idle!

For instance that happens in case of the 8250 UART port driver handling
the looped back high-speed traffic (in my case > 1.5Mbaud) by means of the
DMA-engine interface.

The error happens due to the two-staged nature of the DW DMAC IRQs
handling procedure and due to the critical section break in the meantime.
In particular in case if the DMA-transfer is terminated and restarted:
1. after the IRQ-handler submitted the tasklet but before the tasklet
   started handling the DMA-descriptors in dwc_scan_descriptors();
2. after the XFER completion flag was detected in the
   dwc_scan_descriptors() method, but before the dwc_complete_all() method
   is called
the error denoted above is printed due to the overlap of the last transfer
completion and the new transfer execution stages.

Clear the channel interrupt status when stopping it, and keep the XFER
check, descriptor detachment and queued-transfer restart in one locked
section. An empty active list needs no completion processing.

When the locked XFER recheck finds no event, only update residue. A stale
tasklet snapshot may belong to a terminated transfer, and a zero residue
only means the source has been read, not that destination writes have
finished. Requiring RAW.XFER also prevents the fallback scan from freeing
a descriptor still linked on active_list.

Fixes: 69cea5a00d31 ("dmaengine/dw_dmac: Replace spin_lock* with irqsave variants and enable submission from callback")
Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Signed-off-by: Serge Semin <fancer.lancer@gmail.com>
[james: require a locked XFER event and handle an empty active list]
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 drivers/dma/dw/core.c | 70 ++++++++++++++++++++-------------------------------
 1 file changed, 27 insertions(+), 43 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 575b7bb6af9a..2a9db9e39f5f 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -136,6 +136,12 @@ static inline void dwc_chan_disable(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	channel_clear_bit(dw, CH_EN, dwc->mask);
 	while (dma_readl(dw, CH_EN) & dwc->mask)
 		cpu_relax();
+
+	dma_writel(dw, CLEAR.XFER, dwc->mask);
+	dma_writel(dw, CLEAR.BLOCK, dwc->mask);
+	dma_writel(dw, CLEAR.SRC_TRAN, dwc->mask);
+	dma_writel(dw, CLEAR.DST_TRAN, dwc->mask);
+	dma_writel(dw, CLEAR.ERROR, dwc->mask);
 }
 
 /*----------------------------------------------------------------------*/
@@ -252,34 +258,6 @@ dwc_descriptor_complete(struct dw_dma_chan *dwc, struct dw_desc *desc,
 	dmaengine_desc_callback_invoke(&cb, NULL);
 }
 
-static void dwc_complete_all(struct dw_dma *dw, struct dw_dma_chan *dwc)
-{
-	struct dw_desc *desc, *_desc;
-	LIST_HEAD(list);
-	unsigned long flags;
-
-	spin_lock_irqsave(&dwc->lock, flags);
-	if (dma_readl(dw, CH_EN) & dwc->mask) {
-		dev_err(dmaengine_chan_dev(&dwc->chan),
-			"BUG: XFER bit set, but channel not idle!\n");
-
-		/* Try to continue after resetting the channel... */
-		dwc_chan_disable(dw, dwc);
-	}
-
-	/*
-	 * Submit queued descriptors ASAP, i.e. before we go through
-	 * the completed ones.
-	 */
-	list_splice_init(&dwc->active_list, &list);
-	dwc_dostart_first_queued(dwc);
-
-	spin_unlock_irqrestore(&dwc->lock, flags);
-
-	list_for_each_entry_safe(desc, _desc, &list, desc_node)
-		dwc_descriptor_complete(dwc, desc, true);
-}
-
 /* Returns how many bytes were already received from source */
 static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
 {
@@ -297,6 +275,7 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	struct dw_desc *child;
 	u32 status_xfer;
 	unsigned long flags;
+	LIST_HEAD(list);
 
 	spin_lock_irqsave(&dwc->lock, flags);
 	llp = channel_readl(dwc, LLP);
@@ -305,6 +284,10 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	if (status_xfer & dwc->mask) {
 		/* Everything we've submitted is done */
 		dma_writel(dw, CLEAR.XFER, dwc->mask);
+		if (list_empty(&dwc->active_list)) {
+			spin_unlock_irqrestore(&dwc->lock, flags);
+			return;
+		}
 
 		if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
 			struct list_head *head, *active = dwc->tx_node_active;
@@ -336,9 +319,23 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 			clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
 		}
 
+		if (dma_readl(dw, CH_EN) & dwc->mask) {
+			dev_err(dmaengine_chan_dev(&dwc->chan),
+				"BUG: XFER bit set, but channel not idle!\n");
+
+			/* Try to continue after resetting the channel... */
+			dwc_chan_disable(dw, dwc);
+		}
+
+		/* Detach the completed transfer before starting queued work. */
+		list_splice_init(&dwc->active_list, &list);
+		dwc_dostart_first_queued(dwc);
 		spin_unlock_irqrestore(&dwc->lock, flags);
 
-		dwc_complete_all(dw, dwc);
+		list_for_each_entry_safe(desc, _desc, &list, desc_node) {
+			list_del_init(&desc->desc_node);
+			dwc_descriptor_complete(dwc, desc, true);
+		}
 		return;
 	}
 
@@ -384,22 +381,9 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 			desc->residue -= child->len;
 		}
 
-		/*
-		 * No descriptors so far seem to be in progress, i.e.
-		 * this one must be done.
-		 */
-		spin_unlock_irqrestore(&dwc->lock, flags);
-		dwc_descriptor_complete(dwc, desc, true);
-		spin_lock_irqsave(&dwc->lock, flags);
 	}
 
-	dev_err(dmaengine_chan_dev(&dwc->chan),
-		"BUG: All descriptors done, but channel not idle!\n");
-
-	/* Try to continue after resetting the channel... */
-	dwc_chan_disable(dw, dwc);
-
-	dwc_dostart_first_queued(dwc);
+	/* Residue describes source reads; only RAW.XFER completes a transfer. */
 	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 

-- 
2.53.0

[PATCH v5 02/10] dmaengine: dw: serialize error recovery with channel termination

[James Hilliard]

The tasklet samples RAW.ERROR before taking the channel lock. A
concurrent termination can empty the active list before dwc_handle_error()
uses that cached error to select a descriptor. The list head is then
interpreted as a descriptor. Even with an active descriptor, moving
queue.next unconditionally can put an empty queue's head on the active
list and pass it to dwc_dostart().

Recheck the error status under the channel lock and stop the channel
before selecting its failed descriptor. Only one transaction is issued
at a time, so take that descriptor directly instead of scanning for
completions first. The scan can otherwise complete it and start queued
work before the error handler has identified the failed transaction.

Detach the failed descriptor under the same lock and use the queued-start
helper, which leaves an empty queue alone. Clear the software-LLP flag
before restarting so a failed multiblock transfer does not prevent the
next transaction from starting. Keep the completion callback outside the
lock, with the detached descriptor owned by the tasklet.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Fixes: f336e42f73d9 ("dw_dmac: Move single descriptor from dwc->queue to dwc->active_list in dwc_complete_all")
Cc: stable@vger.kernel.org
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 drivers/dma/dw/core.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 2a9db9e39f5f..f7ca9da5af90 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -403,10 +403,17 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	struct dw_desc *child;
 	unsigned long flags;
 
-	dwc_scan_descriptors(dw, dwc);
-
 	spin_lock_irqsave(&dwc->lock, flags);
 
+	/* Termination may have cleared the error since the tasklet sampled it. */
+	if (!(dma_readl(dw, RAW.ERROR) & dwc->mask))
+		goto unlock;
+
+	dwc_chan_disable(dw, dwc);
+	clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
+	if (list_empty(&dwc->active_list))
+		goto unlock;
+
 	/*
 	 * The descriptor currently at the head of the active list is
 	 * borked. Since we don't have any way to report errors, we'll
@@ -414,12 +421,7 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	 */
 	bad_desc = dwc_first_active(dwc);
 	list_del_init(&bad_desc->desc_node);
-	list_move(dwc->queue.next, dwc->active_list.prev);
-
-	/* Clear the error flag and try to restart the controller */
-	dma_writel(dw, CLEAR.ERROR, dwc->mask);
-	if (!list_empty(&dwc->active_list))
-		dwc_dostart(dwc, dwc_first_active(dwc));
+	dwc_dostart_first_queued(dwc);
 
 	/*
 	 * WARN may seem harsh, but since this only happens
@@ -438,6 +440,10 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 
 	/* Pretend the descriptor completed successfully */
 	dwc_descriptor_complete(dwc, bad_desc, true);
+	return;
+
+unlock:
+	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 
 static void dw_dma_tasklet(struct tasklet_struct *t)

-- 
2.53.0

[PATCH v5 03/10] dmaengine: dw: complete cookies before releasing the channel lock

[James Hilliard]

Completion detaches a descriptor from the active list under the channel
lock, but updates its cookie later when freeing it. Termination can take
the lock in between and complete newer descriptors. Completing the older
cookie afterwards moves completed_cookie backwards and makes an already
terminated transfer appear outstanding again.

Complete cookies while detaching descriptors in the completion, error
recovery and termination paths. Keep the active-before-queued ordering
when terminating so completed_cookie advances in submission order.
Descriptor freeing and client notification no longer update cookies.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 drivers/dma/dw/core.c | 24 +++++++++++++++---------
 1 file changed, 15 insertions(+), 9 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index f7ca9da5af90..2d69fab06601 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -231,18 +231,17 @@ static void dwc_dostart_first_queued(struct dw_dma_chan *dwc)
 /*----------------------------------------------------------------------*/
 
 static void
-dwc_descriptor_complete(struct dw_dma_chan *dwc, struct dw_desc *desc,
-		bool callback_required)
+dwc_descriptor_free(struct dw_dma_chan *dwc, struct dw_desc *desc,
+		    bool callback_required)
 {
 	struct dma_async_tx_descriptor	*txd = &desc->txd;
 	struct dw_desc			*child;
 	unsigned long			flags;
 	struct dmaengine_desc_callback	cb;
 
-	dev_vdbg(dmaengine_chan_dev(&dwc->chan), "descriptor %u complete\n", txd->cookie);
+	dev_vdbg(dmaengine_chan_dev(&dwc->chan), "freeing descriptor %p\n", desc);
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	dma_cookie_complete(txd);
 	if (callback_required)
 		dmaengine_desc_get_callback(txd, &cb);
 	else
@@ -329,12 +328,14 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 
 		/* Detach the completed transfer before starting queued work. */
 		list_splice_init(&dwc->active_list, &list);
+		list_for_each_entry(desc, &list, desc_node)
+			dma_cookie_complete(&desc->txd);
 		dwc_dostart_first_queued(dwc);
 		spin_unlock_irqrestore(&dwc->lock, flags);
 
 		list_for_each_entry_safe(desc, _desc, &list, desc_node) {
 			list_del_init(&desc->desc_node);
-			dwc_descriptor_complete(dwc, desc, true);
+			dwc_descriptor_free(dwc, desc, true);
 		}
 		return;
 	}
@@ -436,10 +437,11 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	list_for_each_entry(child, &bad_desc->tx_list, desc_node)
 		dwc_dump_lli(dwc, child);
 
+	/* Pretend the descriptor completed successfully */
+	dma_cookie_complete(&bad_desc->txd);
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
-	/* Pretend the descriptor completed successfully */
-	dwc_descriptor_complete(dwc, bad_desc, true);
+	dwc_descriptor_free(dwc, bad_desc, true);
 	return;
 
 unlock:
@@ -940,12 +942,16 @@ static int dwc_terminate_all(struct dma_chan *chan)
 	/* active_list entries will end up before queued entries */
 	list_splice_init(&dwc->queue, &list);
 	list_splice_init(&dwc->active_list, &list);
+	list_for_each_entry(desc, &list, desc_node)
+		dma_cookie_complete(&desc->txd);
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
 	/* Flush all pending and queued descriptors */
-	list_for_each_entry_safe(desc, _desc, &list, desc_node)
-		dwc_descriptor_complete(dwc, desc, false);
+	list_for_each_entry_safe(desc, _desc, &list, desc_node) {
+		list_del_init(&desc->desc_node);
+		dwc_descriptor_free(dwc, desc, false);
+	}
 
 	return 0;
 }

-- 
2.53.0

[PATCH v5 04/10] dmaengine: dw: extract hardware-LLP residue calculation

[James Hilliard]

From: Serge Semin <fancer.lancer@gmail.com>

Move the hardware-LLP residue calculation into
dwc_get_hard_llp_desc_residue() so it can be reused independently of
completion handling.

Keep the scanner's LLP snapshot, descriptor updates and early returns
unchanged. The helper reports whether it found the current descriptor,
including when that descriptor has zero residue, rather than using the
residue value as a completion indication.

Signed-off-by: Serge Semin <fancer.lancer@gmail.com>
[james: extract the calculation without changing scanner behavior]
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---

Changelog RFC:
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.
---
 drivers/dma/dw/core.c | 58 +++++++++++++++++++++++++++++----------------------
 1 file changed, 33 insertions(+), 25 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 2d69fab06601..5a1caca7ad68 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -267,6 +267,38 @@ static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
 	return dw->block2bytes(dwc, ctlhi, ctllo >> 4 & 7);
 }
 
+static bool dwc_get_hard_llp_desc_residue(struct dw_dma_chan *dwc,
+				       struct dw_desc *desc, dma_addr_t llp)
+{
+	struct dw_desc *child;
+
+	/* Initial residue value */
+	desc->residue = desc->total_len;
+
+	/* Check first descriptors addr */
+	if (desc->txd.phys == DWC_LLP_LOC(llp))
+		return true;
+
+	/* Check first descriptors llp */
+	if (lli_read(desc, llp) == llp) {
+		/* This one is currently in progress */
+		desc->residue -= dwc_get_sent(dwc);
+		return true;
+	}
+
+	desc->residue -= desc->len;
+	list_for_each_entry(child, &desc->tx_list, desc_node) {
+		if (lli_read(child, llp) == llp) {
+			/* Currently in progress */
+			desc->residue -= dwc_get_sent(dwc);
+			return true;
+		}
+		desc->residue -= child->len;
+	}
+
+	return false;
+}
+
 static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 {
 	dma_addr_t llp;
@@ -354,34 +386,10 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	dev_vdbg(dmaengine_chan_dev(&dwc->chan), "%s: llp=%pad\n", __func__, &llp);
 
 	list_for_each_entry_safe(desc, _desc, &dwc->active_list, desc_node) {
-		/* Initial residue value */
-		desc->residue = desc->total_len;
-
-		/* Check first descriptors addr */
-		if (desc->txd.phys == DWC_LLP_LOC(llp)) {
-			spin_unlock_irqrestore(&dwc->lock, flags);
-			return;
-		}
-
-		/* Check first descriptors llp */
-		if (lli_read(desc, llp) == llp) {
-			/* This one is currently in progress */
-			desc->residue -= dwc_get_sent(dwc);
+		if (dwc_get_hard_llp_desc_residue(dwc, desc, llp)) {
 			spin_unlock_irqrestore(&dwc->lock, flags);
 			return;
 		}
-
-		desc->residue -= desc->len;
-		list_for_each_entry(child, &desc->tx_list, desc_node) {
-			if (lli_read(child, llp) == llp) {
-				/* Currently in progress */
-				desc->residue -= dwc_get_sent(dwc);
-				spin_unlock_irqrestore(&dwc->lock, flags);
-				return;
-			}
-			desc->residue -= child->len;
-		}
-
 	}
 
 	/* Residue describes source reads; only RAW.XFER completes a transfer. */

-- 
2.53.0

[PATCH v5 05/10] dmaengine: dw: calculate residue in the status getter

[James Hilliard]

From: Serge Semin <fancer.lancer@gmail.com>

Calculate hardware-LLP residue while looking up the requested cookie under
the channel lock. The scanner and the status getter currently take that
lock separately, allowing a descriptor to start between the scanner's
residue update and the status query.

Have dwc_get_hard_llp_desc_residue() return a local residue value from a
fresh LLP snapshot. Keep the persistent residue field for software-LLP
block accounting, subtracting the current block's progress when reporting
status. The descriptor scanner continues to advance software-LLP blocks
and complete cookies from RAW.XFER, including for polled transfers.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Signed-off-by: Serge Semin <fancer.lancer@gmail.com>
[james: retain polled completion and software-LLP progress]
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---

Changelog RFC:
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.
---
 drivers/dma/dw/core.c | 88 ++++++++++++++++++---------------------------------
 1 file changed, 31 insertions(+), 57 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 5a1caca7ad68..0d6b5bd8ea6b 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -267,41 +267,8 @@ static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
 	return dw->block2bytes(dwc, ctlhi, ctllo >> 4 & 7);
 }
 
-static bool dwc_get_hard_llp_desc_residue(struct dw_dma_chan *dwc,
-				       struct dw_desc *desc, dma_addr_t llp)
-{
-	struct dw_desc *child;
-
-	/* Initial residue value */
-	desc->residue = desc->total_len;
-
-	/* Check first descriptors addr */
-	if (desc->txd.phys == DWC_LLP_LOC(llp))
-		return true;
-
-	/* Check first descriptors llp */
-	if (lli_read(desc, llp) == llp) {
-		/* This one is currently in progress */
-		desc->residue -= dwc_get_sent(dwc);
-		return true;
-	}
-
-	desc->residue -= desc->len;
-	list_for_each_entry(child, &desc->tx_list, desc_node) {
-		if (lli_read(child, llp) == llp) {
-			/* Currently in progress */
-			desc->residue -= dwc_get_sent(dwc);
-			return true;
-		}
-		desc->residue -= child->len;
-	}
-
-	return false;
-}
-
 static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 {
-	dma_addr_t llp;
 	struct dw_desc *desc, *_desc;
 	struct dw_desc *child;
 	u32 status_xfer;
@@ -309,7 +276,6 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	LIST_HEAD(list);
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	llp = channel_readl(dwc, LLP);
 	status_xfer = dma_readl(dw, RAW.XFER);
 
 	if (status_xfer & dwc->mask) {
@@ -372,27 +338,6 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 		return;
 	}
 
-	if (list_empty(&dwc->active_list)) {
-		spin_unlock_irqrestore(&dwc->lock, flags);
-		return;
-	}
-
-	if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
-		dev_vdbg(dmaengine_chan_dev(&dwc->chan), "%s: soft LLP mode\n", __func__);
-		spin_unlock_irqrestore(&dwc->lock, flags);
-		return;
-	}
-
-	dev_vdbg(dmaengine_chan_dev(&dwc->chan), "%s: llp=%pad\n", __func__, &llp);
-
-	list_for_each_entry_safe(desc, _desc, &dwc->active_list, desc_node) {
-		if (dwc_get_hard_llp_desc_residue(dwc, desc, llp)) {
-			spin_unlock_irqrestore(&dwc->lock, flags);
-			return;
-		}
-	}
-
-	/* Residue describes source reads; only RAW.XFER completes a transfer. */
 	spin_unlock_irqrestore(&dwc->lock, flags);
 }
 
@@ -975,6 +920,32 @@ static struct dw_desc *dwc_find_desc(struct dw_dma_chan *dwc, dma_cookie_t c)
 	return NULL;
 }
 
+static u32 dwc_get_hard_llp_desc_residue(struct dw_dma_chan *dwc, struct dw_desc *desc)
+{
+	u32 residue = desc->total_len;
+	struct dw_desc *child;
+	dma_addr_t llp;
+
+	llp = channel_readl(dwc, LLP);
+
+	/* The first descriptor may still be waiting to be fetched. */
+	if (desc->txd.phys == DWC_LLP_LOC(llp))
+		return residue;
+
+	if (lli_read(desc, llp) == llp)
+		return residue - dwc_get_sent(dwc);
+
+	residue -= desc->len;
+	list_for_each_entry(child, &desc->tx_list, desc_node) {
+		if (lli_read(child, llp) == llp)
+			return residue - dwc_get_sent(dwc);
+
+		residue -= child->len;
+	}
+
+	return residue;
+}
+
 static u32 dwc_get_residue_and_status(struct dw_dma_chan *dwc, dma_cookie_t cookie,
 				      enum dma_status *status)
 {
@@ -987,9 +958,12 @@ static u32 dwc_get_residue_and_status(struct dw_dma_chan *dwc, dma_cookie_t cook
 	desc = dwc_find_desc(dwc, cookie);
 	if (desc) {
 		if (desc == dwc_first_active(dwc)) {
-			residue = desc->residue;
-			if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags) && residue)
+			if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
+				residue = desc->residue;
 				residue -= dwc_get_sent(dwc);
+			} else {
+				residue = dwc_get_hard_llp_desc_residue(dwc, desc);
+			}
 			if (test_bit(DW_DMA_IS_PAUSED, &dwc->flags))
 				*status = DMA_PAUSED;
 		} else {

-- 
2.53.0

[PATCH v5 06/10] dmaengine: dw: simplify transfer-complete handling

[James Hilliard]

With hardware-LLP residue handled by the status getter, the descriptor
scanner only needs to process transfer-complete events. Return through a
common unlock path when no event or active descriptor remains, and flatten
the completion handling.

Only one transaction is active on a channel. Detach that descriptor before
starting the next queued transfer, keeping cookie completion under the
channel lock and client notification outside it. Preserve software-LLP
block advancement before completing the transaction.

Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---

Changelog RFC:
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.
---
 drivers/dma/dw/core.c | 92 +++++++++++++++++++++------------------------------
 1 file changed, 38 insertions(+), 54 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 0d6b5bd8ea6b..e198c9f45f0b 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -271,74 +271,58 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 {
 	struct dw_desc *desc, *_desc;
 	struct dw_desc *child;
-	u32 status_xfer;
 	unsigned long flags;
 	LIST_HEAD(list);
 
 	spin_lock_irqsave(&dwc->lock, flags);
-	status_xfer = dma_readl(dw, RAW.XFER);
-
-	if (status_xfer & dwc->mask) {
-		/* Everything we've submitted is done */
-		dma_writel(dw, CLEAR.XFER, dwc->mask);
-		if (list_empty(&dwc->active_list)) {
-			spin_unlock_irqrestore(&dwc->lock, flags);
-			return;
-		}
-
-		if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
-			struct list_head *head, *active = dwc->tx_node_active;
-
-			/*
-			 * We are inside first active descriptor.
-			 * Otherwise something is really wrong.
-			 */
-			desc = dwc_first_active(dwc);
-
-			head = &desc->tx_list;
-			if (active != head) {
-				/* Update residue to reflect last sent descriptor */
-				if (active == head->next)
-					desc->residue -= desc->len;
-				else
-					desc->residue -= to_dw_desc(active->prev)->len;
 
-				child = to_dw_desc(active);
-
-				/* Submit next block */
-				dwc_do_single_block(dwc, child);
+	if (!(dma_readl(dw, RAW.XFER) & dwc->mask))
+		goto unlock;
 
-				spin_unlock_irqrestore(&dwc->lock, flags);
-				return;
-			}
+	dma_writel(dw, CLEAR.XFER, dwc->mask);
+	if (list_empty(&dwc->active_list))
+		goto unlock;
 
-			/* We are done here */
-			clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
+	desc = dwc_first_active(dwc);
+	if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
+		struct list_head *head = &desc->tx_list;
+		struct list_head *active = dwc->tx_node_active;
+
+		if (active != head) {
+			/* Update residue to reflect last sent descriptor. */
+			if (active == head->next)
+				desc->residue -= desc->len;
+			else
+				desc->residue -= to_dw_desc(active->prev)->len;
+
+			child = to_dw_desc(active);
+
+			/* Submit next block, including for polled transfers. */
+			dwc_do_single_block(dwc, child);
+			goto unlock;
 		}
 
-		if (dma_readl(dw, CH_EN) & dwc->mask) {
-			dev_err(dmaengine_chan_dev(&dwc->chan),
-				"BUG: XFER bit set, but channel not idle!\n");
-
-			/* Try to continue after resetting the channel... */
-			dwc_chan_disable(dw, dwc);
-		}
+		clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
+	}
 
-		/* Detach the completed transfer before starting queued work. */
-		list_splice_init(&dwc->active_list, &list);
-		list_for_each_entry(desc, &list, desc_node)
-			dma_cookie_complete(&desc->txd);
-		dwc_dostart_first_queued(dwc);
-		spin_unlock_irqrestore(&dwc->lock, flags);
+	if (dma_readl(dw, CH_EN) & dwc->mask) {
+		dev_err(dmaengine_chan_dev(&dwc->chan),
+			"BUG: XFER bit set, but channel not idle!\n");
 
-		list_for_each_entry_safe(desc, _desc, &list, desc_node) {
-			list_del_init(&desc->desc_node);
-			dwc_descriptor_free(dwc, desc, true);
-		}
-		return;
+		/* Try to continue after resetting the channel... */
+		dwc_chan_disable(dw, dwc);
 	}
 
+	dma_cookie_complete(&desc->txd);
+	list_move_tail(&desc->desc_node, &list);
+	dwc_dostart_first_queued(dwc);
+unlock:
 	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	list_for_each_entry_safe(desc, _desc, &list, desc_node) {
+		list_del_init(&desc->desc_node);
+		dwc_descriptor_free(dwc, desc, true);
+	}
 }
 
 static inline void dwc_dump_lli(struct dw_dma_chan *dwc, struct dw_desc *desc)

-- 
2.53.0

[PATCH v5 07/10] dmaengine: dw: report residue for queued transfers

[James Hilliard]

A submitted transfer may still be queued behind the active transaction.
dwc_find_desc() only searches the active list, so a status query reports
zero residue for the queued cookie even though none of its data has been
transferred.

Search the queue as well. The status getter already reports total_len for
a descriptor that is not the active transaction, giving queued transfers
their full remaining length.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---

Changelog RFC:
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.
---
 drivers/dma/dw/core.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index e198c9f45f0b..0fd49ff9257f 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -901,6 +901,10 @@ static struct dw_desc *dwc_find_desc(struct dw_dma_chan *dwc, dma_cookie_t c)
 		if (desc->txd.cookie == c)
 			return desc;
 
+	list_for_each_entry(desc, &dwc->queue, desc_node)
+		if (desc->txd.cookie == c)
+			return desc;
+
 	return NULL;
 }
 

-- 
2.53.0

[PATCH v5 08/10] dmaengine: dw: drain the tasklet before releasing channel resources

[James Hilliard]

The controller tasklet invokes callbacks after dropping the channel lock.
Releasing a channel can therefore clear its configuration and disable the
controller while that tasklet run is still using them.

Disable the tasklet before clearing channel resources and re-enable it
after updating the controller's in-use mask. This waits for an executing
run while preserving pending schedules for the other channels. Have a
pending run return before accessing registers when the last channel has
already been released.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 drivers/dma/dw/core.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 0fd49ff9257f..339ca85952fa 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -393,6 +393,10 @@ static void dw_dma_tasklet(struct tasklet_struct *t)
 	u32 status_err;
 	unsigned int i;
 
+	/* Channel release may have drained a polled completion before this run. */
+	if (!dw->in_use)
+		return;
+
 	status_xfer = dma_readl(dw, RAW.XFER);
 	status_err = dma_readl(dw, RAW.ERROR);
 
@@ -1062,6 +1066,8 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	struct dw_dma		*dw = to_dw_dma(chan->device);
 	unsigned long		flags;
 
+	tasklet_disable(&dw->tasklet);
+
 	dev_dbg(dmaengine_chan_dev(chan), "%s: descs allocated=%u\n", __func__,
 			dwc->descs_allocated);
 
@@ -1087,6 +1093,8 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	if (!dw->in_use)
 		do_dw_dma_off(dw);
 
+	tasklet_enable(&dw->tasklet);
+
 	dev_vdbg(dmaengine_chan_dev(chan), "%s: done\n", __func__);
 }
 

-- 
2.53.0

[PATCH v5 09/10] dmaengine: dw: synchronize completion callbacks

[James Hilliard]

The controller tasklet can detach completed descriptors and invoke their
callbacks outside the channel lock. Terminating the channel does not wait
for those callbacks, and there is no device_synchronize hook. A client can
therefore reuse callback state or release its buffers after
dmaengine_synchronize() while an old callback still uses them.

Disable and re-enable the controller tasklet from device_synchronize,
outside the channel lock. Disabling waits for an executing tasklet,
including descriptors it already removed from the channel lists. After
termination, a later tasklet run cannot acquire another descriptor from
the terminated channel. Clients must continue to exclude new submissions
on that channel until synchronization finishes.

The tasklet is shared, so other channels may schedule it while this wait
is in progress. Preserve those pending runs: tasklet_kill() temporarily
owns the scheduling bit and can consume a concurrent scheduling request,
leaving controller interrupts masked. The disable/enable pair leaves the
scheduling bit and pending tasklet queue intact. Other channels' hardware
transfers continue, though their callbacks may be briefly deferred.

Fixes: b36f09c3c441 ("dmaengine: Add transfer termination synchronization support")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Closes: https://lore.kernel.org/r/sashiko-outbox-180707@kernel.org
Cc: stable@vger.kernel.org
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
---
 drivers/dma/dw/core.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 339ca85952fa..6371b9156882 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -897,6 +897,19 @@ static int dwc_terminate_all(struct dma_chan *chan)
 	return 0;
 }
 
+static void dwc_synchronize(struct dma_chan *chan)
+{
+	struct dw_dma *dw = to_dw_dma(chan->device);
+
+	/*
+	 * Termination removes the channel's descriptors from the lists, but a
+	 * running tasklet may still own a completed descriptor and its callback.
+	 * Wait for that run without losing schedules from the other channels.
+	 */
+	tasklet_disable(&dw->tasklet);
+	tasklet_enable(&dw->tasklet);
+}
+
 static struct dw_desc *dwc_find_desc(struct dw_dma_chan *dwc, dma_cookie_t c)
 {
 	struct dw_desc *desc;
@@ -1290,6 +1303,7 @@ int do_dma_probe(struct dw_dma_chip *chip)
 	dw->dma.device_pause = dwc_pause;
 	dw->dma.device_resume = dwc_resume;
 	dw->dma.device_terminate_all = dwc_terminate_all;
+	dw->dma.device_synchronize = dwc_synchronize;
 
 	dw->dma.device_tx_status = dwc_tx_status;
 	dw->dma.device_issue_pending = dwc_issue_pending;

-- 
2.53.0

[PATCH v5 10/10] dmaengine: dw: Prevent tx-status calling DMA-desc callback

[James Hilliard]

From: Serge Semin <fancer.lancer@gmail.com>

The DMA transfer status getter can invoke a completion callback from
dwc_scan_descriptors(). An 8250 RX timeout queries that status while
holding the UART port lock, and its DMA callback takes the same lock:

  serial8250_handle_irq()
    uart_port_lock_irqsave(port)
    serial8250_rx_dma_flush()
      dmaengine_tx_status()
        dwc_scan_descriptors()
          dwc_descriptor_free()
            dma_rx_complete()
              uart_port_lock_irqsave(port)

Queue completed descriptors on a channel completion list under the channel
lock. Let the controller tasklet detach that list and invoke callbacks
outside the lock. Termination discards notifications still on the shared
list; the tasklet owns descriptors it has already detached, and the
synchronization hook waits for that run.

Keep handling raw completion events from tx_status(). Clients such as
catpt poll transfers without DMA_PREP_INTERRUPT and can mask the controller
interrupt externally. Polling must still advance software-emulated LLP
blocks and complete cookies without waiting for tasklet execution. Hold
the channel lock across event handling and status calculation, then
schedule the tasklet for deferred notifications.

On channel release, discard queued notifications after draining the
tasklet, before clearing the controller's in-use state.

Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
Cc: stable@vger.kernel.org
Signed-off-by: Serge Semin <fancer.lancer@gmail.com>
[james: defer notifications while preserving polled completion and cleanup]
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>

---

Changelog RFC:
- Instead of just dropping the dwc_scan_descriptors() method invocation
  calculate the residue in the Tx-status getter.
---
 drivers/dma/dw/core.c | 105 +++++++++++++++++++++++++++++---------------------
 drivers/dma/dw/regs.h |   1 +
 2 files changed, 62 insertions(+), 44 deletions(-)

diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
index 6371b9156882..e02082f47701 100644
--- a/drivers/dma/dw/core.c
+++ b/drivers/dma/dw/core.c
@@ -257,6 +257,22 @@ dwc_descriptor_free(struct dw_dma_chan *dwc, struct dw_desc *desc,
 	dmaengine_desc_callback_invoke(&cb, NULL);
 }
 
+static void dwc_free_completed(struct dw_dma_chan *dwc, bool callback_required)
+{
+	struct dw_desc *desc, *_desc;
+	unsigned long flags;
+	LIST_HEAD(list);
+
+	spin_lock_irqsave(&dwc->lock, flags);
+	list_splice_tail_init(&dwc->completed, &list);
+	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	list_for_each_entry_safe(desc, _desc, &list, desc_node) {
+		list_del_init(&desc->desc_node);
+		dwc_descriptor_free(dwc, desc, callback_required);
+	}
+}
+
 /* Returns how many bytes were already received from source */
 static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
 {
@@ -269,19 +285,17 @@ static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
 
 static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 {
-	struct dw_desc *desc, *_desc;
+	struct dw_desc *desc;
 	struct dw_desc *child;
-	unsigned long flags;
-	LIST_HEAD(list);
 
-	spin_lock_irqsave(&dwc->lock, flags);
+	lockdep_assert_held(&dwc->lock);
 
 	if (!(dma_readl(dw, RAW.XFER) & dwc->mask))
-		goto unlock;
+		return;
 
 	dma_writel(dw, CLEAR.XFER, dwc->mask);
 	if (list_empty(&dwc->active_list))
-		goto unlock;
+		return;
 
 	desc = dwc_first_active(dwc);
 	if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
@@ -299,7 +313,7 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 
 			/* Submit next block, including for polled transfers. */
 			dwc_do_single_block(dwc, child);
-			goto unlock;
+			return;
 		}
 
 		clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
@@ -314,15 +328,8 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
 	}
 
 	dma_cookie_complete(&desc->txd);
-	list_move_tail(&desc->desc_node, &list);
+	list_move_tail(&desc->desc_node, &dwc->completed);
 	dwc_dostart_first_queued(dwc);
-unlock:
-	spin_unlock_irqrestore(&dwc->lock, flags);
-
-	list_for_each_entry_safe(desc, _desc, &list, desc_node) {
-		list_del_init(&desc->desc_node);
-		dwc_descriptor_free(dwc, desc, true);
-	}
 }
 
 static inline void dwc_dump_lli(struct dw_dma_chan *dwc, struct dw_desc *desc)
@@ -339,18 +346,17 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 {
 	struct dw_desc *bad_desc;
 	struct dw_desc *child;
-	unsigned long flags;
 
-	spin_lock_irqsave(&dwc->lock, flags);
+	lockdep_assert_held(&dwc->lock);
 
 	/* Termination may have cleared the error since the tasklet sampled it. */
 	if (!(dma_readl(dw, RAW.ERROR) & dwc->mask))
-		goto unlock;
+		return;
 
 	dwc_chan_disable(dw, dwc);
 	clear_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags);
 	if (list_empty(&dwc->active_list))
-		goto unlock;
+		return;
 
 	/*
 	 * The descriptor currently at the head of the active list is
@@ -376,40 +382,41 @@ static void dwc_handle_error(struct dw_dma *dw, struct dw_dma_chan *dwc)
 
 	/* Pretend the descriptor completed successfully */
 	dma_cookie_complete(&bad_desc->txd);
-	spin_unlock_irqrestore(&dwc->lock, flags);
+	list_add_tail(&bad_desc->desc_node, &dwc->completed);
+}
 
-	dwc_descriptor_free(dwc, bad_desc, true);
-	return;
+static void dwc_handle_channel(struct dw_dma *dw, struct dw_dma_chan *dwc)
+{
+	lockdep_assert_held(&dwc->lock);
 
-unlock:
-	spin_unlock_irqrestore(&dwc->lock, flags);
+	if (dma_readl(dw, RAW.ERROR) & dwc->mask)
+		dwc_handle_error(dw, dwc);
+	else
+		dwc_scan_descriptors(dw, dwc);
 }
 
 static void dw_dma_tasklet(struct tasklet_struct *t)
 {
 	struct dw_dma *dw = from_tasklet(dw, t, tasklet);
 	struct dw_dma_chan *dwc;
-	u32 status_xfer;
-	u32 status_err;
+	unsigned long flags;
 	unsigned int i;
+	u32 status;
 
 	/* Channel release may have drained a polled completion before this run. */
 	if (!dw->in_use)
 		return;
 
-	status_xfer = dma_readl(dw, RAW.XFER);
-	status_err = dma_readl(dw, RAW.ERROR);
-
-	dev_vdbg(dw->dma.dev, "%s: status_err=%x\n", __func__, status_err);
+	status = dma_readl(dw, RAW.XFER) | dma_readl(dw, RAW.ERROR);
 
 	for (i = 0; i < dw->dma.chancnt; i++) {
 		dwc = &dw->chan[i];
-		if (test_bit(DW_DMA_IS_CYCLIC, &dwc->flags))
-			dev_vdbg(dw->dma.dev, "Cyclic xfer is not implemented\n");
-		else if (status_err & (1 << i))
-			dwc_handle_error(dw, dwc);
-		else if (status_xfer & (1 << i))
-			dwc_scan_descriptors(dw, dwc);
+		spin_lock_irqsave(&dwc->lock, flags);
+		/* The snapshot selects work; recheck events under the channel lock. */
+		if (status & dwc->mask)
+			dwc_handle_channel(dw, dwc);
+		spin_unlock_irqrestore(&dwc->lock, flags);
+		dwc_free_completed(dwc, true);
 	}
 
 	/* Re-enable interrupts */
@@ -885,6 +892,7 @@ static int dwc_terminate_all(struct dma_chan *chan)
 	list_splice_init(&dwc->active_list, &list);
 	list_for_each_entry(desc, &list, desc_node)
 		dma_cookie_complete(&desc->txd);
+	list_splice_init(&dwc->completed, &list);
 
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
@@ -955,10 +963,9 @@ static u32 dwc_get_residue_and_status(struct dw_dma_chan *dwc, dma_cookie_t cook
 				      enum dma_status *status)
 {
 	struct dw_desc *desc;
-	unsigned long flags;
 	u32 residue;
 
-	spin_lock_irqsave(&dwc->lock, flags);
+	lockdep_assert_held(&dwc->lock);
 
 	desc = dwc_find_desc(dwc, cookie);
 	if (desc) {
@@ -978,7 +985,6 @@ static u32 dwc_get_residue_and_status(struct dw_dma_chan *dwc, dma_cookie_t cook
 		residue = 0;
 	}
 
-	spin_unlock_irqrestore(&dwc->lock, flags);
 	return residue;
 }
 
@@ -988,19 +994,26 @@ dwc_tx_status(struct dma_chan *chan,
 	      struct dma_tx_state *txstate)
 {
 	struct dw_dma_chan	*dwc = to_dw_dma_chan(chan);
+	struct dw_dma		*dw = to_dw_dma(chan->device);
 	enum dma_status		ret;
+	unsigned long		flags;
 
+	spin_lock_irqsave(&dwc->lock, flags);
 	ret = dma_cookie_status(chan, cookie, txstate);
 	if (ret == DMA_COMPLETE)
-		return ret;
+		goto unlock;
 
-	dwc_scan_descriptors(to_dw_dma(chan->device), dwc);
+	/* Poll completion even when the client has masked the controller IRQ. */
+	dwc_handle_channel(dw, dwc);
+	if (!list_empty(&dwc->completed))
+		tasklet_schedule(&dw->tasklet);
 
 	ret = dma_cookie_status(chan, cookie, txstate);
-	if (ret == DMA_COMPLETE)
-		return ret;
+	if (ret != DMA_COMPLETE)
+		dma_set_residue(txstate, dwc_get_residue_and_status(dwc, cookie, &ret));
 
-	dma_set_residue(txstate, dwc_get_residue_and_status(dwc, cookie, &ret));
+unlock:
+	spin_unlock_irqrestore(&dwc->lock, flags);
 	return ret;
 }
 
@@ -1079,6 +1092,7 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	struct dw_dma		*dw = to_dw_dma(chan->device);
 	unsigned long		flags;
 
+	/* Polled completion can leave descriptors waiting for the tasklet. */
 	tasklet_disable(&dw->tasklet);
 
 	dev_dbg(dmaengine_chan_dev(chan), "%s: descs allocated=%u\n", __func__,
@@ -1089,6 +1103,8 @@ static void dwc_free_chan_resources(struct dma_chan *chan)
 	BUG_ON(!list_empty(&dwc->queue));
 	BUG_ON(dma_readl(to_dw_dma(chan->device), CH_EN) & dwc->mask);
 
+	dwc_free_completed(dwc, false);
+
 	spin_lock_irqsave(&dwc->lock, flags);
 
 	/* Clear custom channel configuration */
@@ -1238,6 +1254,7 @@ int do_dma_probe(struct dw_dma_chip *chip)
 
 		INIT_LIST_HEAD(&dwc->active_list);
 		INIT_LIST_HEAD(&dwc->queue);
+		INIT_LIST_HEAD(&dwc->completed);
 
 		channel_clear_bit(dw, CH_EN, dwc->mask);
 
diff --git a/drivers/dma/dw/regs.h b/drivers/dma/dw/regs.h
index 5969d9cc8d7a..c41f18281efe 100644
--- a/drivers/dma/dw/regs.h
+++ b/drivers/dma/dw/regs.h
@@ -279,6 +279,7 @@ struct dw_dma_chan {
 	unsigned long		flags;
 	struct list_head	active_list;
 	struct list_head	queue;
+	struct list_head	completed;
 
 	unsigned int		descs_allocated;
 

-- 
2.53.0

Re: [PATCH v5 05/10] dmaengine: dw: calculate residue in the status getter

[Frank Li]

On Mon, Nov 02, 2026 at 12:34:17PM -0700, James Hilliard wrote:
> From: Serge Semin <fancer.lancer@gmail.com>
>
> Calculate hardware-LLP residue while looking up the requested cookie under
> the channel lock. The scanner and the status getter currently take that
> lock separately, allowing a descriptor to start between the scanner's
> residue update and the status query.
>
> Have dwc_get_hard_llp_desc_residue() return a local residue value from a
> fresh LLP snapshot. Keep the persistent residue field for software-LLP
> block accounting, subtracting the current block's progress when reporting
> status. The descriptor scanner continues to advance software-LLP blocks
> and complete cookies from RAW.XFER, including for polled transfers.
>
> Fixes: 3bfb1d20b547 ("dmaengine: Driver for the Synopsys DesignWare DMA controller")
> Cc: stable@vger.kernel.org
> Signed-off-by: Serge Semin <fancer.lancer@gmail.com>
> [james: retain polled completion and software-LLP progress]
> Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
>
> ---
>
> Changelog RFC:
> - Instead of just dropping the dwc_scan_descriptors() method invocation
>   calculate the residue in the Tx-status getter.
> ---
>  drivers/dma/dw/core.c | 88 ++++++++++++++++++---------------------------------
>  1 file changed, 31 insertions(+), 57 deletions(-)
>
> diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c
> index 5a1caca7ad68..0d6b5bd8ea6b 100644
> --- a/drivers/dma/dw/core.c
> +++ b/drivers/dma/dw/core.c
> @@ -267,41 +267,8 @@ static inline u32 dwc_get_sent(struct dw_dma_chan *dwc)
>  	return dw->block2bytes(dwc, ctlhi, ctllo >> 4 & 7);
>  }
>
> -static bool dwc_get_hard_llp_desc_residue(struct dw_dma_chan *dwc,
> -				       struct dw_desc *desc, dma_addr_t llp)
> -{
> -	struct dw_desc *child;
> -
> -	/* Initial residue value */
> -	desc->residue = desc->total_len;
> -
> -	/* Check first descriptors addr */
> -	if (desc->txd.phys == DWC_LLP_LOC(llp))
> -		return true;
> -
> -	/* Check first descriptors llp */
> -	if (lli_read(desc, llp) == llp) {
> -		/* This one is currently in progress */
> -		desc->residue -= dwc_get_sent(dwc);
> -		return true;
> -	}
> -
> -	desc->residue -= desc->len;
> -	list_for_each_entry(child, &desc->tx_list, desc_node) {
> -		if (lli_read(child, llp) == llp) {
> -			/* Currently in progress */
> -			desc->residue -= dwc_get_sent(dwc);
> -			return true;
> -		}
> -		desc->residue -= child->len;
> -	}
> -
> -	return false;
> -}
> -
>  static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
>  {
> -	dma_addr_t llp;
>  	struct dw_desc *desc, *_desc;
>  	struct dw_desc *child;
>  	u32 status_xfer;
> @@ -309,7 +276,6 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
>  	LIST_HEAD(list);
>
>  	spin_lock_irqsave(&dwc->lock, flags);
> -	llp = channel_readl(dwc, LLP);
>  	status_xfer = dma_readl(dw, RAW.XFER);
>
>  	if (status_xfer & dwc->mask) {
> @@ -372,27 +338,6 @@ static void dwc_scan_descriptors(struct dw_dma *dw, struct dw_dma_chan *dwc)
>  		return;
>  	}
>
> -	if (list_empty(&dwc->active_list)) {
> -		spin_unlock_irqrestore(&dwc->lock, flags);
> -		return;
> -	}
> -
> -	if (test_bit(DW_DMA_IS_SOFT_LLP, &dwc->flags)) {
> -		dev_vdbg(dmaengine_chan_dev(&dwc->chan), "%s: soft LLP mode\n", __func__);
> -		spin_unlock_irqrestore(&dwc->lock, flags);
> -		return;
> -	}
> -
> -	dev_vdbg(dmaengine_chan_dev(&dwc->chan), "%s: llp=%pad\n", __func__, &llp);
> -
> -	list_for_each_entry_safe(desc, _desc, &dwc->active_list, desc_node) {
> -		if (dwc_get_hard_llp_desc_residue(dwc, desc, llp)) {
> -			spin_unlock_irqrestore(&dwc->lock, flags);
> -			return;
> -		}
> -	}
> -
> -	/* Residue describes source reads; only RAW.XFER completes a transfer. */
>  	spin_unlock_irqrestore(&dwc->lock, flags);
>  }
>
> @@ -975,6 +920,32 @@ static struct dw_desc *dwc_find_desc(struct dw_dma_chan *dwc, dma_cookie_t c)
>  	return NULL;
>  }
>
> +static u32 dwc_get_hard_llp_desc_residue(struct dw_dma_chan *dwc, struct dw_desc *desc)
> +{

Don't move this function, just change to return value boot to u32 to reduce
changes.


Frank