[PATCH AUTOSEL 5.6 066/606] dt-bindings: dma: fsl-edma: fix ls1028a-edma compatible

[PATCH AUTOSEL 5.6 066/606] dt-bindings: dma: fsl-edma: fix ls1028a-edma compatible

[Sasha Levin]

From: Michael Walle <michael@walle.cc>

commit d94a05f87327143f94f67dd256932163ac2bcd65 upstream.

The bootloader will fix up the IOMMU entries only on nodes with the
compatible "fsl,vf610-edma". Thus make this compatible string mandatory
for the ls1028a-edma.

While at it, fix the "fsl,fsl," typo.

Signed-off-by: Michael Walle <michael@walle.cc>
Fixes: d8c1bdb5288d ("dt-bindings: dma: fsl-edma: add new fsl,fsl,ls1028a-edma")
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 Documentation/devicetree/bindings/dma/fsl-edma.txt | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/Documentation/devicetree/bindings/dma/fsl-edma.txt b/Documentation/devicetree/bindings/dma/fsl-edma.txt
index e77b08ebcd06..ee1754739b4b 100644
--- a/Documentation/devicetree/bindings/dma/fsl-edma.txt
+++ b/Documentation/devicetree/bindings/dma/fsl-edma.txt
@@ -10,7 +10,8 @@ Required properties:
 - compatible :
 	- "fsl,vf610-edma" for eDMA used similar to that on Vybrid vf610 SoC
 	- "fsl,imx7ulp-edma" for eDMA2 used similar to that on i.mx7ulp
-	- "fsl,fsl,ls1028a-edma" for eDMA used similar to that on Vybrid vf610 SoC
+	- "fsl,ls1028a-edma" followed by "fsl,vf610-edma" for eDMA used on the
+	  LS1028A SoC.
 - reg : Specifies base physical address(s) and size of the eDMA registers.
 	The 1st region is eDMA control register's address and size.
 	The 2nd and the 3rd regions are programmable channel multiplexing
-- 
2.25.1

[PATCH AUTOSEL 5.6 156/606] dmaengine: tegra210-adma: Fix an error handling path in 'tegra_adma_probe()'

[Sasha Levin]

From: Christophe JAILLET <christophe.jaillet@wanadoo.fr>

commit 3a5fd0dbd87853f8bd2ea275a5b3b41d6686e761 upstream.

Commit b53611fb1ce9 ("dmaengine: tegra210-adma: Fix crash during probe")
has moved some code in the probe function and reordered the error handling
path accordingly.
However, a goto has been missed.

Fix it and goto the right label if 'dma_async_device_register()' fails, so
that all resources are released.

Fixes: b53611fb1ce9 ("dmaengine: tegra210-adma: Fix crash during probe")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Acked-by: Thierry Reding <treding@nvidia.com>
Link: https://lore.kernel.org/r/20200516214205.276266-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/dma/tegra210-adma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/dma/tegra210-adma.c b/drivers/dma/tegra210-adma.c
index 6e1268552f74..914901a680c8 100644
--- a/drivers/dma/tegra210-adma.c
+++ b/drivers/dma/tegra210-adma.c
@@ -900,7 +900,7 @@ static int tegra_adma_probe(struct platform_device *pdev)
 	ret = dma_async_device_register(&tdma->dma_dev);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "ADMA registration failed: %d\n", ret);
-		goto irq_dispose;
+		goto rpm_put;
 	}
 
 	ret = of_dma_controller_register(pdev->dev.of_node,
-- 
2.25.1

[PATCH AUTOSEL 5.6 158/606] dmaengine: dmatest: Restore default for channel

[Sasha Levin]

From: Vladimir Murzin <vladimir.murzin@arm.com>

commit 6b41030fdc79086db5d673c5ed7169f3ee8c13b9 upstream.

In case of dmatest is built-in and no channel was configured test
doesn't run with:

dmatest: Could not start test, no channels configured

Even though description to "channel" parameter claims that default is
any.

Add default channel back as it used to be rather than reject test with
no channel configuration.

Fixes: d53513d5dc285d9a95a534fc41c5c08af6b60eac ("dmaengine: dmatest: Add support for multi channel testing)
Reported-by: Dijil Mohan <Dijil.Mohan@arm.com>
Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com>
Link: https://lore.kernel.org/r/20200429071522.58148-1-vladimir.murzin@arm.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/dma/dmatest.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index 364dd34799d4..0425984db118 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -1166,10 +1166,11 @@ static int dmatest_run_set(const char *val, const struct kernel_param *kp)
 		mutex_unlock(&info->lock);
 		return ret;
 	} else if (dmatest_run) {
-		if (is_threaded_test_pending(info))
-			start_threaded_tests(info);
-		else
-			pr_info("Could not start test, no channels configured\n");
+		if (!is_threaded_test_pending(info)) {
+			pr_info("No channels configured, continue with any\n");
+			add_threaded_test(info);
+		}
+		start_threaded_tests(info);
 	} else {
 		stop_threaded_test(info);
 	}
-- 
2.25.1

[PATCH AUTOSEL 5.6 159/606] dmaengine: idxd: fix interrupt completion after unmasking

[Sasha Levin]

From: Dave Jiang <dave.jiang@intel.com>

commit 4f302642b70c1348773fe7e3ded9fc315fa92990 upstream.

The current implementation may miss completions after we unmask the
interrupt. In order to make sure we process all competions, we need to:
1. Do an MMIO read from the device as a barrier to ensure that all PCI
   writes for completions have arrived.
2. Check for any additional completions that we missed.

Fixes: 8f47d1a5e545 ("dmaengine: idxd: connect idxd to dmaengine subsystem")

Reported-by: Sanjay Kumar <sanjay.k.kumar@intel.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/158834641769.35613.1341160109892008587.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/dma/idxd/device.c |  7 +++++++
 drivers/dma/idxd/irq.c    | 26 +++++++++++++++++++-------
 2 files changed, 26 insertions(+), 7 deletions(-)

diff --git a/drivers/dma/idxd/device.c b/drivers/dma/idxd/device.c
index f6f49f0f6fae..8d79a8787104 100644
--- a/drivers/dma/idxd/device.c
+++ b/drivers/dma/idxd/device.c
@@ -62,6 +62,13 @@ int idxd_unmask_msix_vector(struct idxd_device *idxd, int vec_id)
 	perm.ignore = 0;
 	iowrite32(perm.bits, idxd->reg_base + offset);
 
+	/*
+	 * A readback from the device ensures that any previously generated
+	 * completion record writes are visible to software based on PCI
+	 * ordering rules.
+	 */
+	perm.bits = ioread32(idxd->reg_base + offset);
+
 	return 0;
 }
 
diff --git a/drivers/dma/idxd/irq.c b/drivers/dma/idxd/irq.c
index d6fcd2e60103..6510791b9921 100644
--- a/drivers/dma/idxd/irq.c
+++ b/drivers/dma/idxd/irq.c
@@ -173,6 +173,7 @@ static int irq_process_pending_llist(struct idxd_irq_entry *irq_entry,
 	struct llist_node *head;
 	int queued = 0;
 
+	*processed = 0;
 	head = llist_del_all(&irq_entry->pending_llist);
 	if (!head)
 		return 0;
@@ -197,6 +198,7 @@ static int irq_process_work_list(struct idxd_irq_entry *irq_entry,
 	struct list_head *node, *next;
 	int queued = 0;
 
+	*processed = 0;
 	if (list_empty(&irq_entry->work_list))
 		return 0;
 
@@ -218,10 +220,9 @@ static int irq_process_work_list(struct idxd_irq_entry *irq_entry,
 	return queued;
 }
 
-irqreturn_t idxd_wq_thread(int irq, void *data)
+static int idxd_desc_process(struct idxd_irq_entry *irq_entry)
 {
-	struct idxd_irq_entry *irq_entry = data;
-	int rc, processed = 0, retry = 0;
+	int rc, processed, total = 0;
 
 	/*
 	 * There are two lists we are processing. The pending_llist is where
@@ -244,15 +245,26 @@ irqreturn_t idxd_wq_thread(int irq, void *data)
 	 */
 	do {
 		rc = irq_process_work_list(irq_entry, &processed);
-		if (rc != 0) {
-			retry++;
+		total += processed;
+		if (rc != 0)
 			continue;
-		}
 
 		rc = irq_process_pending_llist(irq_entry, &processed);
-	} while (rc != 0 && retry != 10);
+		total += processed;
+	} while (rc != 0);
+
+	return total;
+}
+
+irqreturn_t idxd_wq_thread(int irq, void *data)
+{
+	struct idxd_irq_entry *irq_entry = data;
+	int processed;
 
+	processed = idxd_desc_process(irq_entry);
 	idxd_unmask_msix_vector(irq_entry->idxd, irq_entry->id);
+	/* catch anything unprocessed after unmasking */
+	processed += idxd_desc_process(irq_entry);
 
 	if (processed == 0)
 		return IRQ_NONE;
-- 
2.25.1

[PATCH AUTOSEL 5.6 160/606] dmaengine: owl: Use correct lock in owl_dma_get_pchan()

[Sasha Levin]

From: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>

commit f8f482deb078389b42768b2193e050a81aae137d upstream.

When the kernel is built with lockdep support and the owl-dma driver is
used, the following message is shown:

[    2.496939] INFO: trying to register non-static key.
[    2.501889] the code is fine but needs lockdep annotation.
[    2.507357] turning off the locking correctness validator.
[    2.512834] CPU: 0 PID: 12 Comm: kworker/0:1 Not tainted 5.6.3+ #15
[    2.519084] Hardware name: Generic DT based system
[    2.523878] Workqueue: events_freezable mmc_rescan
[    2.528681] [<801127f0>] (unwind_backtrace) from [<8010da58>] (show_stack+0x10/0x14)
[    2.536420] [<8010da58>] (show_stack) from [<8080fbe8>] (dump_stack+0xb4/0xe0)
[    2.543645] [<8080fbe8>] (dump_stack) from [<8017efa4>] (register_lock_class+0x6f0/0x718)
[    2.551816] [<8017efa4>] (register_lock_class) from [<8017b7d0>] (__lock_acquire+0x78/0x25f0)
[    2.560330] [<8017b7d0>] (__lock_acquire) from [<8017e5e4>] (lock_acquire+0xd8/0x1f4)
[    2.568159] [<8017e5e4>] (lock_acquire) from [<80831fb0>] (_raw_spin_lock_irqsave+0x3c/0x50)
[    2.576589] [<80831fb0>] (_raw_spin_lock_irqsave) from [<8051b5fc>] (owl_dma_issue_pending+0xbc/0x120)
[    2.585884] [<8051b5fc>] (owl_dma_issue_pending) from [<80668cbc>] (owl_mmc_request+0x1b0/0x390)
[    2.594655] [<80668cbc>] (owl_mmc_request) from [<80650ce0>] (mmc_start_request+0x94/0xbc)
[    2.602906] [<80650ce0>] (mmc_start_request) from [<80650ec0>] (mmc_wait_for_req+0x64/0xd0)
[    2.611245] [<80650ec0>] (mmc_wait_for_req) from [<8065aa10>] (mmc_app_send_scr+0x10c/0x144)
[    2.619669] [<8065aa10>] (mmc_app_send_scr) from [<80659b3c>] (mmc_sd_setup_card+0x4c/0x318)
[    2.628092] [<80659b3c>] (mmc_sd_setup_card) from [<80659f0c>] (mmc_sd_init_card+0x104/0x430)
[    2.636601] [<80659f0c>] (mmc_sd_init_card) from [<8065a3e0>] (mmc_attach_sd+0xcc/0x16c)
[    2.644678] [<8065a3e0>] (mmc_attach_sd) from [<8065301c>] (mmc_rescan+0x3ac/0x40c)
[    2.652332] [<8065301c>] (mmc_rescan) from [<80143244>] (process_one_work+0x2d8/0x780)
[    2.660239] [<80143244>] (process_one_work) from [<80143730>] (worker_thread+0x44/0x598)
[    2.668323] [<80143730>] (worker_thread) from [<8014b5f8>] (kthread+0x148/0x150)
[    2.675708] [<8014b5f8>] (kthread) from [<801010b4>] (ret_from_fork+0x14/0x20)
[    2.682912] Exception stack(0xee8fdfb0 to 0xee8fdff8)
[    2.687954] dfa0:                                     00000000 00000000 00000000 00000000
[    2.696118] dfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[    2.704277] dfe0: 00000000 00000000 00000000 00000000 00000013 00000000

The obvious fix would be to use 'spin_lock_init()' on 'pchan->lock'
before attempting to call 'spin_lock_irqsave()' in 'owl_dma_get_pchan()'.

However, according to Manivannan Sadhasivam, 'pchan->lock' was supposed
to only protect 'pchan->vchan' while 'od->lock' does a similar job in
'owl_dma_terminate_pchan()'.

Therefore, this patch substitutes 'pchan->lock' with 'od->lock' and
removes the 'lock' attribute in 'owl_dma_pchan' struct.

Fixes: 47e20577c24d ("dmaengine: Add Actions Semi Owl family S900 DMA driver")
Signed-off-by: Cristian Ciocaltea <cristian.ciocaltea@gmail.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Acked-by: Andreas Färber <afaerber@suse.de>
Link: https://lore.kernel.org/r/c6e6cdaca252b5364bd294093673951036488cf0.1588439073.git.cristian.ciocaltea@gmail.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/dma/owl-dma.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/drivers/dma/owl-dma.c b/drivers/dma/owl-dma.c
index c683051257fd..66ef70b00ec0 100644
--- a/drivers/dma/owl-dma.c
+++ b/drivers/dma/owl-dma.c
@@ -175,13 +175,11 @@ struct owl_dma_txd {
  * @id: physical index to this channel
  * @base: virtual memory base for the dma channel
  * @vchan: the virtual channel currently being served by this physical channel
- * @lock: a lock to use when altering an instance of this struct
  */
 struct owl_dma_pchan {
 	u32			id;
 	void __iomem		*base;
 	struct owl_dma_vchan	*vchan;
-	spinlock_t		lock;
 };
 
 /**
@@ -437,14 +435,14 @@ static struct owl_dma_pchan *owl_dma_get_pchan(struct owl_dma *od,
 	for (i = 0; i < od->nr_pchans; i++) {
 		pchan = &od->pchans[i];
 
-		spin_lock_irqsave(&pchan->lock, flags);
+		spin_lock_irqsave(&od->lock, flags);
 		if (!pchan->vchan) {
 			pchan->vchan = vchan;
-			spin_unlock_irqrestore(&pchan->lock, flags);
+			spin_unlock_irqrestore(&od->lock, flags);
 			break;
 		}
 
-		spin_unlock_irqrestore(&pchan->lock, flags);
+		spin_unlock_irqrestore(&od->lock, flags);
 	}
 
 	return pchan;
-- 
2.25.1