Merge tag 'dmaengine-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/dmaengine

Pull dmaengine update from Dan Williams:
 "Collection of random updates to the core and some end-driver fixups
  for ioatdma and mv_xor:
   - NUMA aware channel allocation
   - Cleanup dmatest debugfs interface
   - ioat: make raid-support Atom only
   - mv_xor: big endian

  Aside from the top three commits these have all had some soak time in
  -next.  The top commit fixes a recent build breakage.

  It has been a long while since my last pull request, hopefully it does
  not show.  Thanks to Vinod for keeping an eye on drivers/dma/ this
  past year"

* tag 'dmaengine-3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/djbw/dmaengine:
  dmaengine: dma_sync_wait and dma_find_channel undefined
  MAINTAINERS: update email for Dan Williams
  dma: mv_xor: Fix incorrect error path
  ioatdma: silence GCC warnings
  dmaengine: make dma_channel_rebalance() NUMA aware
  dmaengine: make dma_submit_error() return an error code
  ioatdma: disable RAID on non-Atom platforms and reenable unaligned copies
  mv_xor: support big endian systems using descriptor swap feature
  mv_xor: use {readl, writel}_relaxed instead of __raw_{readl, writel}
  dmatest: print message on debug level in case of no error
  dmatest: remove IS_ERR_OR_NULL checks of debugfs calls
  dmatest: make module parameters writable

8 files changed, 146 insertions, 248 deletions

diff --git a/Documentation/dmatest.txt b/Documentation/dmatest.txt
index 132a094c7bc3..a2b5663eae26 100644
--- a/Documentation/dmatest.txt
+++ b/Documentation/dmatest.txt
@@ -16,15 +16,16 @@ be built as module or inside kernel. Let's consider those cases.
 	Part 2 - When dmatest is built as a module...
 
 After mounting debugfs and loading the module, the /sys/kernel/debug/dmatest
-folder with nodes will be created. They are the same as module parameters with
-addition of the 'run' node that controls run and stop phases of the test.
+folder with nodes will be created. There are two important files located. First
+is the 'run' node that controls run and stop phases of the test, and the second
+one, 'results', is used to get the test case results.
 
 Note that in this case test will not run on load automatically.
 
 Example of usage:
-	% echo dma0chan0 > /sys/kernel/debug/dmatest/channel
-	% echo 2000 > /sys/kernel/debug/dmatest/timeout
-	% echo 1 > /sys/kernel/debug/dmatest/iterations
+	% echo dma0chan0 > /sys/module/dmatest/parameters/channel
+	% echo 2000 > /sys/module/dmatest/parameters/timeout
+	% echo 1 > /sys/module/dmatest/parameters/iterations
 	% echo 1 > /sys/kernel/debug/dmatest/run
 
 Hint: available channel list could be extracted by running the following
@@ -55,8 +56,8 @@ for the first performed test. After user gets a control, the test could be
 re-run with the same or different parameters. For the details see the above
 section "Part 2 - When dmatest is built as a module..."
 
-In both cases the module parameters are used as initial values for the test case.
-You always could check them at run-time by running
+In both cases the module parameters are used as the actual values for the test
+case. You always could check them at run-time by running
 	% grep -H . /sys/module/dmatest/parameters/*
 
 	Part 4 - Gathering the test results
diff --git a/MAINTAINERS b/MAINTAINERS
index 7a78b9f2aae2..87efa1f5c7f3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -933,24 +933,24 @@ F:	arch/arm/mach-pxa/colibri-pxa270-income.c
 
 ARM/INTEL IOP32X ARM ARCHITECTURE
 M:	Lennert Buytenhek <kernel@wantstofly.org>
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 
 ARM/INTEL IOP33X ARM ARCHITECTURE
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 
 ARM/INTEL IOP13XX ARM ARCHITECTURE
 M:	Lennert Buytenhek <kernel@wantstofly.org>
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 
 ARM/INTEL IQ81342EX MACHINE SUPPORT
 M:	Lennert Buytenhek <kernel@wantstofly.org>
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 
@@ -975,7 +975,7 @@ F:	drivers/pcmcia/pxa2xx_stargate2.c
 
 ARM/INTEL XSC3 (MANZANO) ARM CORE
 M:	Lennert Buytenhek <kernel@wantstofly.org>
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained
 
@@ -1386,7 +1386,7 @@ F:	drivers/platform/x86/asus*.c
 F:	drivers/platform/x86/eeepc*.c
 
 ASYNCHRONOUS TRANSFERS/TRANSFORMS (IOAT) API
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 W:	http://sourceforge.net/projects/xscaleiop
 S:	Maintained
 F:	Documentation/crypto/async-tx-api.txt
@@ -2691,7 +2691,7 @@ T:	git git://git.linaro.org/people/sumitsemwal/linux-dma-buf.git
 
 DMA GENERIC OFFLOAD ENGINE SUBSYSTEM
 M:	Vinod Koul <vinod.koul@intel.com>
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 S:	Supported
 F:	drivers/dma/
 F:	include/linux/dma*
@@ -4323,7 +4323,7 @@ F:	arch/x86/kernel/microcode_core.c
 F:	arch/x86/kernel/microcode_intel.c
 
 INTEL I/OAT DMA DRIVER
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 S:	Maintained
 F:	drivers/dma/ioat*
 
@@ -4336,7 +4336,7 @@ F:	drivers/iommu/intel-iommu.c
 F:	include/linux/intel-iommu.h
 
 INTEL IOP-ADMA DMA DRIVER
-M:	Dan Williams <djbw@fb.com>
+M:	Dan Williams <dan.j.williams@intel.com>
 S:	Odd fixes
 F:	drivers/dma/iop-adma.c
 
diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c
index 99af4db5948b..eee16b01fa89 100644
--- a/drivers/dma/dmaengine.c
+++ b/drivers/dma/dmaengine.c
@@ -382,20 +382,30 @@ void dma_issue_pending_all(void)
 EXPORT_SYMBOL(dma_issue_pending_all);
 
 /**
- * nth_chan - returns the nth channel of the given capability
+ * dma_chan_is_local - returns true if the channel is in the same numa-node as the cpu
+ */
+static bool dma_chan_is_local(struct dma_chan *chan, int cpu)
+{
+	int node = dev_to_node(chan->device->dev);
+	return node == -1 || cpumask_test_cpu(cpu, cpumask_of_node(node));
+}
+
+/**
+ * min_chan - returns the channel with min count and in the same numa-node as the cpu
  * @cap: capability to match
- * @n: nth channel desired
+ * @cpu: cpu index which the channel should be close to
  *
- * Defaults to returning the channel with the desired capability and the
- * lowest reference count when 'n' cannot be satisfied.  Must be called
- * under dma_list_mutex.
+ * If some channels are close to the given cpu, the one with the lowest
+ * reference count is returned. Otherwise, cpu is ignored and only the
+ * reference count is taken into account.
+ * Must be called under dma_list_mutex.
  */
-static struct dma_chan *nth_chan(enum dma_transaction_type cap, int n)
+static struct dma_chan *min_chan(enum dma_transaction_type cap, int cpu)
 {
 	struct dma_device *device;
 	struct dma_chan *chan;
-	struct dma_chan *ret = NULL;
 	struct dma_chan *min = NULL;
+	struct dma_chan *localmin = NULL;
 
 	list_for_each_entry(device, &dma_device_list, global_node) {
 		if (!dma_has_cap(cap, device->cap_mask) ||
@@ -404,27 +414,22 @@ static struct dma_chan *nth_chan(enum dma_transaction_type cap, int n)
 		list_for_each_entry(chan, &device->channels, device_node) {
 			if (!chan->client_count)
 				continue;
-			if (!min)
-				min = chan;
-			else if (chan->table_count < min->table_count)
+			if (!min || chan->table_count < min->table_count)
 				min = chan;
 
-			if (n-- == 0) {
-				ret = chan;
-				break; /* done */
-			}
+			if (dma_chan_is_local(chan, cpu))
+				if (!localmin ||
+				    chan->table_count < localmin->table_count)
+					localmin = chan;
 		}
-		if (ret)
-			break; /* done */
 	}
 
-	if (!ret)
-		ret = min;
+	chan = localmin ? localmin : min;
 
-	if (ret)
-		ret->table_count++;
+	if (chan)
+		chan->table_count++;
 
-	return ret;
+	return chan;
 }
 
 /**
@@ -441,7 +446,6 @@ static void dma_channel_rebalance(void)
 	struct dma_device *device;
 	int cpu;
 	int cap;
-	int n;
 
 	/* undo the last distribution */
 	for_each_dma_cap_mask(cap, dma_cap_mask_all)
@@ -460,14 +464,9 @@ static void dma_channel_rebalance(void)
 		return;
 
 	/* redistribute available channels */
-	n = 0;
 	for_each_dma_cap_mask(cap, dma_cap_mask_all)
 		for_each_online_cpu(cpu) {
-			if (num_possible_cpus() > 1)
-				chan = nth_chan(cap, n++);
-			else
-				chan = nth_chan(cap, -1);
-
+			chan = min_chan(cap, cpu);
 			per_cpu_ptr(channel_table[cap], cpu)->chan = chan;
 		}
 }
diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
index e88ded2c8d2f..92f796cdc6ab 100644
--- a/drivers/dma/dmatest.c
+++ b/drivers/dma/dmatest.c
@@ -25,44 +25,46 @@
 #include <linux/seq_file.h>
 
 static unsigned int test_buf_size = 16384;
-module_param(test_buf_size, uint, S_IRUGO);
+module_param(test_buf_size, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(test_buf_size, "Size of the memcpy test buffer");
 
 static char test_channel[20];
-module_param_string(channel, test_channel, sizeof(test_channel), S_IRUGO);
+module_param_string(channel, test_channel, sizeof(test_channel),
+		S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(channel, "Bus ID of the channel to test (default: any)");
 
 static char test_device[20];
-module_param_string(device, test_device, sizeof(test_device), S_IRUGO);
+module_param_string(device, test_device, sizeof(test_device),
+		S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(device, "Bus ID of the DMA Engine to test (default: any)");
 
 static unsigned int threads_per_chan = 1;
-module_param(threads_per_chan, uint, S_IRUGO);
+module_param(threads_per_chan, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(threads_per_chan,
 		"Number of threads to start per channel (default: 1)");
 
 static unsigned int max_channels;
-module_param(max_channels, uint, S_IRUGO);
+module_param(max_channels, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(max_channels,
 		"Maximum number of channels to use (default: all)");
 
 static unsigned int iterations;
-module_param(iterations, uint, S_IRUGO);
+module_param(iterations, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(iterations,
 		"Iterations before stopping test (default: infinite)");
 
 static unsigned int xor_sources = 3;
-module_param(xor_sources, uint, S_IRUGO);
+module_param(xor_sources, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(xor_sources,
 		"Number of xor source buffers (default: 3)");
 
 static unsigned int pq_sources = 3;
-module_param(pq_sources, uint, S_IRUGO);
+module_param(pq_sources, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(pq_sources,
 		"Number of p+q source buffers (default: 3)");
 
 static int timeout = 3000;
-module_param(timeout, uint, S_IRUGO);
+module_param(timeout, uint, S_IRUGO | S_IWUSR);
 MODULE_PARM_DESC(timeout, "Transfer Timeout in msec (default: 3000), "
 		 "Pass -1 for infinite timeout");
 
@@ -193,7 +195,6 @@ struct dmatest_info {
 
 	/* debugfs related stuff */
 	struct dentry		*root;
-	struct dmatest_params	dbgfs_params;
 
 	/* Test results */
 	struct list_head	results;
@@ -406,7 +407,11 @@ static int thread_result_add(struct dmatest_info *info,
 	list_add_tail(&tr->node, &r->results);
 	mutex_unlock(&info->results_lock);
 
-	pr_warn("%s\n", thread_result_get(r->name, tr));
+	if (tr->type == DMATEST_ET_OK)
+		pr_debug("%s\n", thread_result_get(r->name, tr));
+	else
+		pr_warn("%s\n", thread_result_get(r->name, tr));
+
 	return 0;
 }
 
@@ -1007,7 +1012,15 @@ static int __restart_threaded_test(struct dmatest_info *info, bool run)
 	result_free(info, NULL);
 
 	/* Copy test parameters */
-	memcpy(params, &info->dbgfs_params, sizeof(*params));
+	params->buf_size = test_buf_size;
+	strlcpy(params->channel, strim(test_channel), sizeof(params->channel));
+	strlcpy(params->device, strim(test_device), sizeof(params->device));
+	params->threads_per_chan = threads_per_chan;
+	params->max_channels = max_channels;
+	params->iterations = iterations;
+	params->xor_sources = xor_sources;
+	params->pq_sources = pq_sources;
+	params->timeout = timeout;
 
 	/* Run test with new parameters */
 	return __run_threaded_test(info);
@@ -1029,71 +1042,6 @@ static bool __is_threaded_test_run(struct dmatest_info *info)
 	return false;
 }
 
-static ssize_t dtf_write_string(void *to, size_t available, loff_t *ppos,
-		const void __user *from, size_t count)
-{
-	char tmp[20];
-	ssize_t len;
-
-	len = simple_write_to_buffer(tmp, sizeof(tmp) - 1, ppos, from, count);
-	if (len >= 0) {
-		tmp[len] = '\0';
-		strlcpy(to, strim(tmp), available);
-	}
-
-	return len;
-}
-
-static ssize_t dtf_read_channel(struct file *file, char __user *buf,
-		size_t count, loff_t *ppos)
-{
-	struct dmatest_info *info = file->private_data;
-	return simple_read_from_buffer(buf, count, ppos,
-			info->dbgfs_params.channel,
-			strlen(info->dbgfs_params.channel));
-}
-
-static ssize_t dtf_write_channel(struct file *file, const char __user *buf,
-		size_t size, loff_t *ppos)
-{
-	struct dmatest_info *info = file->private_data;
-	return dtf_write_string(info->dbgfs_params.channel,
-				sizeof(info->dbgfs_params.channel),
-				ppos, buf, size);
-}
-
-static const struct file_operations dtf_channel_fops = {
-	.read	= dtf_read_channel,
-	.write	= dtf_write_channel,
-	.open	= simple_open,
-	.llseek	= default_llseek,
-};
-
-static ssize_t dtf_read_device(struct file *file, char __user *buf,
-		size_t count, loff_t *ppos)
-{
-	struct dmatest_info *info = file->private_data;
-	return simple_read_from_buffer(buf, count, ppos,
-			info->dbgfs_params.device,
-			strlen(info->dbgfs_params.device));
-}
-
-static ssize_t dtf_write_device(struct file *file, const char __user *buf,
-		size_t size, loff_t *ppos)
-{
-	struct dmatest_info *info = file->private_data;
-	return dtf_write_string(info->dbgfs_params.device,
-				sizeof(info->dbgfs_params.device),
-				ppos, buf, size);
-}
-
-static const struct file_operations dtf_device_fops = {
-	.read	= dtf_read_device,
-	.write	= dtf_write_device,
-	.open	= simple_open,
-	.llseek	= default_llseek,
-};
-
 static ssize_t dtf_read_run(struct file *file, char __user *user_buf,
 		size_t count, loff_t *ppos)
 {
@@ -1187,8 +1135,6 @@ static const struct file_operations dtf_results_fops = {
 static int dmatest_register_dbgfs(struct dmatest_info *info)
 {
 	struct dentry *d;
-	struct dmatest_params *params = &info->dbgfs_params;
-	int ret = -ENOMEM;
 
 	d = debugfs_create_dir("dmatest", NULL);
 	if (IS_ERR(d))
@@ -1198,81 +1144,24 @@ static int dmatest_register_dbgfs(struct dmatest_info *info)
 
 	info->root = d;
 
-	/* Copy initial values */
-	memcpy(params, &info->params, sizeof(*params));
-
-	/* Test parameters */
-
-	d = debugfs_create_u32("test_buf_size", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->buf_size);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_file("channel", S_IRUGO | S_IWUSR, info->root,
-				info, &dtf_channel_fops);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_file("device", S_IRUGO | S_IWUSR, info->root,
-				info, &dtf_device_fops);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("threads_per_chan", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->threads_per_chan);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("max_channels", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->max_channels);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("iterations", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->iterations);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("xor_sources", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->xor_sources);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("pq_sources", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->pq_sources);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
-	d = debugfs_create_u32("timeout", S_IWUSR | S_IRUGO, info->root,
-			       (u32 *)&params->timeout);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
-
 	/* Run or stop threaded test */
-	d = debugfs_create_file("run", S_IWUSR | S_IRUGO, info->root,
-				info, &dtf_run_fops);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
+	debugfs_create_file("run", S_IWUSR | S_IRUGO, info->root, info,
+			    &dtf_run_fops);
 
 	/* Results of test in progress */
-	d = debugfs_create_file("results", S_IRUGO, info->root, info,
-				&dtf_results_fops);
-	if (IS_ERR_OR_NULL(d))
-		goto err_node;
+	debugfs_create_file("results", S_IRUGO, info->root, info,
+			    &dtf_results_fops);
 
 	return 0;
 
-err_node:
-	debugfs_remove_recursive(info->root);
 err_root:
 	pr_err("dmatest: Failed to initialize debugfs\n");
-	return ret;
+	return -ENOMEM;
 }
 
 static int __init dmatest_init(void)
 {
 	struct dmatest_info *info = &test_info;
-	struct dmatest_params *params = &info->params;
 	int ret;
 
 	memset(info, 0, sizeof(*info));
@@ -1283,17 +1172,6 @@ static int __init dmatest_init(void)
 	mutex_init(&info->results_lock);
 	INIT_LIST_HEAD(&info->results);
 
-	/* Set default parameters */
-	params->buf_size = test_buf_size;
-	strlcpy(params->channel, test_channel, sizeof(params->channel));
-	strlcpy(params->device, test_device, sizeof(params->device));
-	params->threads_per_chan = threads_per_chan;
-	params->max_channels = max_channels;
-	params->iterations = iterations;
-	params->xor_sources = xor_sources;
-	params->pq_sources = pq_sources;
-	params->timeout = timeout;
-
 	ret = dmatest_register_dbgfs(info);
 	if (ret)
 		return ret;
diff --git a/drivers/dma/ioat/dma_v3.c b/drivers/dma/ioat/dma_v3.c
index b642e035579b..d8ececaf1b57 100644
--- a/drivers/dma/ioat/dma_v3.c
+++ b/drivers/dma/ioat/dma_v3.c
@@ -251,7 +251,7 @@ static bool is_bwd_noraid(struct pci_dev *pdev)
 }
 
 static void pq16_set_src(struct ioat_raw_descriptor *desc[3],
-			dma_addr_t addr, u32 offset, u8 coef, int idx)
+			dma_addr_t addr, u32 offset, u8 coef, unsigned idx)
 {
 	struct ioat_pq_descriptor *pq = (struct ioat_pq_descriptor *)desc[0];
 	struct ioat_pq16a_descriptor *pq16 =
@@ -1775,15 +1775,12 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	dma->device_alloc_chan_resources = ioat2_alloc_chan_resources;
 	dma->device_free_chan_resources = ioat2_free_chan_resources;
 
-	if (is_xeon_cb32(pdev))
-		dma->copy_align = 6;
-
 	dma_cap_set(DMA_INTERRUPT, dma->cap_mask);
 	dma->device_prep_dma_interrupt = ioat3_prep_interrupt_lock;
 
 	device->cap = readl(device->reg_base + IOAT_DMA_CAP_OFFSET);
 
-	if (is_bwd_noraid(pdev))
+	if (is_xeon_cb32(pdev) || is_bwd_noraid(pdev))
 		device->cap &= ~(IOAT_CAP_XOR | IOAT_CAP_PQ | IOAT_CAP_RAID16SS);
 
 	/* dca is incompatible with raid operations */
@@ -1793,7 +1790,6 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	if (device->cap & IOAT_CAP_XOR) {
 		is_raid_device = true;
 		dma->max_xor = 8;
-		dma->xor_align = 6;
 
 		dma_cap_set(DMA_XOR, dma->cap_mask);
 		dma->device_prep_dma_xor = ioat3_prep_xor;
@@ -1812,13 +1808,8 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 
 		if (device->cap & IOAT_CAP_RAID16SS) {
 			dma_set_maxpq(dma, 16, 0);
-			dma->pq_align = 0;
 		} else {
 			dma_set_maxpq(dma, 8, 0);
-			if (is_xeon_cb32(pdev))
-				dma->pq_align = 6;
-			else
-				dma->pq_align = 0;
 		}
 
 		if (!(device->cap & IOAT_CAP_XOR)) {
@@ -1829,13 +1820,8 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 
 			if (device->cap & IOAT_CAP_RAID16SS) {
 				dma->max_xor = 16;
-				dma->xor_align = 0;
 			} else {
 				dma->max_xor = 8;
-				if (is_xeon_cb32(pdev))
-					dma->xor_align = 6;
-				else
-					dma->xor_align = 0;
 			}
 		}
 	}
@@ -1844,14 +1830,6 @@ int ioat3_dma_probe(struct ioatdma_device *device, int dca)
 	device->cleanup_fn = ioat3_cleanup_event;
 	device->timer_fn = ioat3_timer_event;
 
-	if (is_xeon_cb32(pdev)) {
-		dma_cap_clear(DMA_XOR_VAL, dma->cap_mask);
-		dma->device_prep_dma_xor_val = NULL;
-
-		dma_cap_clear(DMA_PQ_VAL, dma->cap_mask);
-		dma->device_prep_dma_pq_val = NULL;
-	}
-
 	/* starting with CB3.3 super extended descriptors are supported */
 	if (device->cap & IOAT_CAP_RAID16SS) {
 		char pool_name[14];
diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c
index 200f1a3c9a44..0ec086d2b6a0 100644
--- a/drivers/dma/mv_xor.c
+++ b/drivers/dma/mv_xor.c
@@ -64,7 +64,7 @@ static u32 mv_desc_get_src_addr(struct mv_xor_desc_slot *desc,
 				int src_idx)
 {
 	struct mv_xor_desc *hw_desc = desc->hw_desc;
-	return hw_desc->phy_src_addr[src_idx];
+	return hw_desc->phy_src_addr[mv_phy_src_idx(src_idx)];
 }
 
 
@@ -107,32 +107,32 @@ static void mv_desc_set_src_addr(struct mv_xor_desc_slot *desc,
 				 int index, dma_addr_t addr)
 {
 	struct mv_xor_desc *hw_desc = desc->hw_desc;
-	hw_desc->phy_src_addr[index] = addr;
+	hw_desc->phy_src_addr[mv_phy_src_idx(index)] = addr;
 	if (desc->type == DMA_XOR)
 		hw_desc->desc_command |= (1 << index);
 }
 
 static u32 mv_chan_get_current_desc(struct mv_xor_chan *chan)
 {
-	return __raw_readl(XOR_CURR_DESC(chan));
+	return readl_relaxed(XOR_CURR_DESC(chan));
 }
 
 static void mv_chan_set_next_descriptor(struct mv_xor_chan *chan,
 					u32 next_desc_addr)
 {
-	__raw_writel(next_desc_addr, XOR_NEXT_DESC(chan));
+	writel_relaxed(next_desc_addr, XOR_NEXT_DESC(chan));
 }
 
 static void mv_chan_unmask_interrupts(struct mv_xor_chan *chan)
 {
-	u32 val = __raw_readl(XOR_INTR_MASK(chan));
+	u32 val = readl_relaxed(XOR_INTR_MASK(chan));
 	val |= XOR_INTR_MASK_VALUE << (chan->idx * 16);
-	__raw_writel(val, XOR_INTR_MASK(chan));
+	writel_relaxed(val, XOR_INTR_MASK(chan));
 }
 
 static u32 mv_chan_get_intr_cause(struct mv_xor_chan *chan)
 {
-	u32 intr_cause = __raw_readl(XOR_INTR_CAUSE(chan));
+	u32 intr_cause = readl_relaxed(XOR_INTR_CAUSE(chan));
 	intr_cause = (intr_cause >> (chan->idx * 16)) & 0xFFFF;
 	return intr_cause;
 }
@@ -149,13 +149,13 @@ static void mv_xor_device_clear_eoc_cause(struct mv_xor_chan *chan)
 {
 	u32 val = ~(1 << (chan->idx * 16));
 	dev_dbg(mv_chan_to_devp(chan), "%s, val 0x%08x\n", __func__, val);
-	__raw_writel(val, XOR_INTR_CAUSE(chan));
+	writel_relaxed(val, XOR_INTR_CAUSE(chan));
 }
 
 static void mv_xor_device_clear_err_status(struct mv_xor_chan *chan)
 {
 	u32 val = 0xFFFF0000 >> (chan->idx * 16);
-	__raw_writel(val, XOR_INTR_CAUSE(chan));
+	writel_relaxed(val, XOR_INTR_CAUSE(chan));
 }
 
 static int mv_can_chain(struct mv_xor_desc_slot *desc)
@@ -173,7 +173,7 @@ static void mv_set_mode(struct mv_xor_chan *chan,
 			       enum dma_transaction_type type)
 {
 	u32 op_mode;
-	u32 config = __raw_readl(XOR_CONFIG(chan));
+	u32 config = readl_relaxed(XOR_CONFIG(chan));
 
 	switch (type) {
 	case DMA_XOR:
@@ -192,7 +192,14 @@ static void mv_set_mode(struct mv_xor_chan *chan,
 
 	config &= ~0x7;
 	config |= op_mode;
-	__raw_writel(config, XOR_CONFIG(chan));
+
+#if defined(__BIG_ENDIAN)
+	config |= XOR_DESCRIPTOR_SWAP;
+#else
+	config &= ~XOR_DESCRIPTOR_SWAP;
+#endif
+
+	writel_relaxed(config, XOR_CONFIG(chan));
 	chan->current_type = type;
 }
 
@@ -201,14 +208,14 @@ static void mv_chan_activate(struct mv_xor_chan *chan)
 	u32 activation;
 
 	dev_dbg(mv_chan_to_devp(chan), " activate chan.\n");
-	activation = __raw_readl(XOR_ACTIVATION(chan));
+	activation = readl_relaxed(XOR_ACTIVATION(chan));
 	activation |= 0x1;
-	__raw_writel(activation, XOR_ACTIVATION(chan));
+	writel_relaxed(activation, XOR_ACTIVATION(chan));
 }
 
 static char mv_chan_is_busy(struct mv_xor_chan *chan)
 {
-	u32 state = __raw_readl(XOR_ACTIVATION(chan));
+	u32 state = readl_relaxed(XOR_ACTIVATION(chan));
 
 	state = (state >> 4) & 0x3;
 
@@ -755,22 +762,22 @@ static void mv_dump_xor_regs(struct mv_xor_chan *chan)
 {
 	u32 val;
 
-	val = __raw_readl(XOR_CONFIG(chan));
+	val = readl_relaxed(XOR_CONFIG(chan));
 	dev_err(mv_chan_to_devp(chan), "config       0x%08x\n", val);
 
-	val = __raw_readl(XOR_ACTIVATION(chan));
+	val = readl_relaxed(XOR_ACTIVATION(chan));
 	dev_err(mv_chan_to_devp(chan), "activation   0x%08x\n", val);
 
-	val = __raw_readl(XOR_INTR_CAUSE(chan));
+	val = readl_relaxed(XOR_INTR_CAUSE(chan));
 	dev_err(mv_chan_to_devp(chan), "intr cause   0x%08x\n", val);
 
-	val = __raw_readl(XOR_INTR_MASK(chan));
+	val = readl_relaxed(XOR_INTR_MASK(chan));
 	dev_err(mv_chan_to_devp(chan), "intr mask    0x%08x\n", val);
 
-	val = __raw_readl(XOR_ERROR_CAUSE(chan));
+	val = readl_relaxed(XOR_ERROR_CAUSE(chan));
 	dev_err(mv_chan_to_devp(chan), "error cause  0x%08x\n", val);
 
-	val = __raw_readl(XOR_ERROR_ADDR(chan));
+	val = readl_relaxed(XOR_ERROR_ADDR(chan));
 	dev_err(mv_chan_to_devp(chan), "error addr   0x%08x\n", val);
 }
 
@@ -1029,10 +1036,8 @@ mv_xor_channel_add(struct mv_xor_device *xordev,
 	struct dma_device *dma_dev;
 
 	mv_chan = devm_kzalloc(&pdev->dev, sizeof(*mv_chan), GFP_KERNEL);
-	if (!mv_chan) {
-		ret = -ENOMEM;
-		goto err_free_dma;
-	}
+	if (!mv_chan)
+		return ERR_PTR(-ENOMEM);
 
 	mv_chan->idx = idx;
 	mv_chan->irq = irq;
diff --git a/drivers/dma/mv_xor.h b/drivers/dma/mv_xor.h
index c619359cb7fe..06b067f24c9b 100644
--- a/drivers/dma/mv_xor.h
+++ b/drivers/dma/mv_xor.h
@@ -29,8 +29,10 @@
 #define MV_XOR_THRESHOLD		1
 #define MV_XOR_MAX_CHANNELS             2
 
+/* Values for the XOR_CONFIG register */
 #define XOR_OPERATION_MODE_XOR		0
 #define XOR_OPERATION_MODE_MEMCPY	2
+#define XOR_DESCRIPTOR_SWAP		BIT(14)
 
 #define XOR_CURR_DESC(chan)	(chan->mmr_base + 0x210 + (chan->idx * 4))
 #define XOR_NEXT_DESC(chan)	(chan->mmr_base + 0x200 + (chan->idx * 4))
@@ -143,7 +145,16 @@ struct mv_xor_desc_slot {
 #endif
 };
 
-/* This structure describes XOR descriptor size 64bytes	*/
+/*
+ * This structure describes XOR descriptor size 64bytes. The
+ * mv_phy_src_idx() macro must be used when indexing the values of the
+ * phy_src_addr[] array. This is due to the fact that the 'descriptor
+ * swap' feature, used on big endian systems, swaps descriptors data
+ * within blocks of 8 bytes. So two consecutive values of the
+ * phy_src_addr[] array are actually swapped in big-endian, which
+ * explains the different mv_phy_src_idx() implementation.
+ */
+#if defined(__LITTLE_ENDIAN)
 struct mv_xor_desc {
 	u32 status;		/* descriptor execution status */
 	u32 crc32_result;	/* result of CRC-32 calculation */
@@ -155,6 +166,21 @@ struct mv_xor_desc {
 	u32 reserved0;
 	u32 reserved1;
 };
+#define mv_phy_src_idx(src_idx) (src_idx)
+#else
+struct mv_xor_desc {
+	u32 crc32_result;	/* result of CRC-32 calculation */
+	u32 status;		/* descriptor execution status */
+	u32 phy_next_desc;	/* next descriptor address pointer */
+	u32 desc_command;	/* type of operation to be carried out */
+	u32 phy_dest_addr;	/* destination block address */
+	u32 byte_count;		/* size of src/dst blocks in bytes */
+	u32 phy_src_addr[8];	/* source block addresses */
+	u32 reserved1;
+	u32 reserved0;
+};
+#define mv_phy_src_idx(src_idx) (src_idx ^ 1)
+#endif
 
 #define to_mv_sw_desc(addr_hw_desc)		\
 	container_of(addr_hw_desc, struct mv_xor_desc_slot, hw_desc)
diff --git a/include/linux/dmaengine.h b/include/linux/dmaengine.h
index cb286b1acdb6..0c72b89a172c 100644
--- a/include/linux/dmaengine.h
+++ b/include/linux/dmaengine.h
@@ -38,7 +38,10 @@ typedef s32 dma_cookie_t;
 #define DMA_MIN_COOKIE	1
 #define DMA_MAX_COOKIE	INT_MAX
 
-#define dma_submit_error(cookie) ((cookie) < 0 ? 1 : 0)
+static inline int dma_submit_error(dma_cookie_t cookie)
+{
+	return cookie < 0 ? cookie : 0;
+}
 
 /**
  * enum dma_status - DMA transaction status
@@ -958,8 +961,9 @@ dma_set_tx_state(struct dma_tx_state *st, dma_cookie_t last, dma_cookie_t used,
 	}
 }
 
-enum dma_status dma_sync_wait(struct dma_chan *chan, dma_cookie_t cookie);
 #ifdef CONFIG_DMA_ENGINE
+struct dma_chan *dma_find_channel(enum dma_transaction_type tx_type);
+enum dma_status dma_sync_wait(struct dma_chan *chan, dma_cookie_t cookie);
 enum dma_status dma_wait_for_async_tx(struct dma_async_tx_descriptor *tx);
 void dma_issue_pending_all(void);
 struct dma_chan *__dma_request_channel(const dma_cap_mask_t *mask,
@@ -967,6 +971,14 @@ struct dma_chan *__dma_request_channel(const dma_cap_mask_t *mask,
 struct dma_chan *dma_request_slave_channel(struct device *dev, const char *name);
 void dma_release_channel(struct dma_chan *chan);
 #else
+static inline struct dma_chan *dma_find_channel(enum dma_transaction_type tx_type)
+{
+	return NULL;
+}
+static inline enum dma_status dma_sync_wait(struct dma_chan *chan, dma_cookie_t cookie)
+{
+	return DMA_SUCCESS;
+}
 static inline enum dma_status dma_wait_for_async_tx(struct dma_async_tx_descriptor *tx)
 {
 	return DMA_SUCCESS;
@@ -994,7 +1006,6 @@ static inline void dma_release_channel(struct dma_chan *chan)
 int dma_async_device_register(struct dma_device *device);
 void dma_async_device_unregister(struct dma_device *device);
 void dma_run_dependencies(struct dma_async_tx_descriptor *tx);
-struct dma_chan *dma_find_channel(enum dma_transaction_type tx_type);
 struct dma_chan *net_dma_find_channel(void);
 #define dma_request_channel(mask, x, y) __dma_request_channel(&(mask), x, y)
 #define dma_request_slave_channel_compat(mask, x, y, dev, name) \