Merge tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Don't destroy user event fields when removal fails

   User event fields are destroyed before the event is removed from
   visibility. But that can fail leaving the still visible event with no
   fields. Move the destroying of the fields to after the event is
   successfully removed from visibility.

 - Initialize function graph state is fork before calling
   copy_exec_state()

   For non-CLONE_VM forks, copy_exec_state() allocates a new
   task_exec_state. If that allocation fails, ftrace_graph_exit_task()
   will free the tasks ret_stack pointer. Since that pointer is still
   using the parent's ret_stack, it mistakenly frees the parent's
   pointer too.

   Call ftrace_graph_init() on the task first which will NULL out the
   new tasks's ret_stack and if the copy fails, it will not free
   anything.

 - Remove FGRAPH_MAX_INDEX

   The macro FGRAPH_MAX_INDEX was added but never used. Remove it.

 - Save ent_size in function graph printing of nested functions

   The function graph tracer needs to look at the next event to see if
   the next event is the return of the current function entry. If it is,
   it prints a single line:

	ktime_get();

   Otherwise it prints it like a nested function:

	tick_nohz_irq_exit() {
	    ktime_get();
	    kcpustat_irq_exit();
	}

   In order to look at the next event, it must save the current event so
   that it has the information to print from it. It saves the event in
   the iterator descriptor called "ent". What it doesn't save is the
   ent_size of the event which is now used to know if the function graph
   arguments are to be printed. The peek doesn't save the size so the
   size used happens to be that of the size of the last event that was
   seen.

   Save the entry event size in the iterator descriptor so that the
   correct size is used.

 - Fix several errors with freeing data in the histogram code

   The histogram code had a lot of leaked or or incorrect accounting
   when failures happen. Correct them.

 - Fix histogram regression of .percent and .graph modifiers

   Up until 6.3 histogram values could have "percent" or "graph"
   modifiers that changed how they were printed. But a change that added
   restricting histograms values from being strings, stack traces and
   other modifiers inadvertently prevented them from using the percent
   and graph modifiers, which were legal use cases for values.

   Put back the percent and graph modifiers.

 - Fix various typos in the comments

 - Set the trace_clock before initializing a histogram with clock
   argument

   The histogram API allows the user to specific which trace clock to
   use via a "clock=" string. The histogram is set up first before the
   clock is checked. If the passed in clock is not valid, it exits
   without fully fixing up the histogram leaving it on the list and a
   use-after-free can trigger.

   Update the clock argument first and if it fails then exit gracefully
   before the histogram trigger is placed on any lists.

 - Restore :mod: trailer after parsing in ftrace_set_clr_event

   The function ftrace_set_clr_event() modifies the parse string and
   needs to put it back to what was passed in. It searches for ":mod:"
   via a strsep() but fails to put back the first ':' in the string.

   Add back the ':' in the passed in string.

 - Take trace_array reference when opening a tracer options file

   The options files are dynamically created and some tracers add their
   own options. When a tracer adds their own list of options, the
   trace_array holding them has an array to hold the list of options for
   each tracer. This array increases in size via a krealloc(), and the
   new entry gets a newly allocated array to hold the options of the new
   tracer being added.

   The element in each entry of the tracer's option array holds a
   pointer back to the trace_array, a pointer to the tracer it is
   associated to, a pointer to the flags of the option.

   The issue is that these arrays are freed when the trace_array is
   freed when its instance it represents is removed from the instances
   directory. There's a race that an open of one of these options files
   can happen when the instance is being removed.

   Add a new helper function to be called by the open function of the
   options file to iterate all existing trace_arrays under a lock and
   find the one that has the given option element in one of it's tracer
   arrays. If found, then update the associated trace_array's reference
   counter to keep it from being freed. If not found, have the open call
   return -ENODEV.

 - Disable interrupts when acquiring the lock in rb_wake_up_waiters()

   The function rb_wake_up_waiters() assumes it will be called in
   interrupt context and does not disable irqs when taking
   cpu_buffer->reader_lock, which can be called in hard interrupt
   context. The issue is in PREEMPT_RT, this function is called in
   thread context leaving this lock open to a deadlock.

   Take the lock with interrupts disabled.

 - Use rcu_assign_pointer() for tmp_ops filter hash

   The tmp_ops used in update_ftrace_direct_mod() assigns its
   filter_hash field directly, but that field is annotated as __rcu and
   sparse complains. Assign it with rcu_assign_pointer()

 - Fix use-after-free in enable_trigger_private_data_free()

   The trace_event_call is accessed through the event_trigger_data's
   trace_event_file pointer to put the trace_event_call on freeing. The
   issue is that the trace_event_file data may have been freed already
   causing a use-after-free. Add a field to the event_trigger_data that
   points directly to the trace_event_call so that it can decrement its
   reference directly without needing to go through the
   trace_event_file.

 - Fix accounting of buffer data remote headers

   trace_buffer_desc_size() and trace_remote_alloc_buffer() undercount
   the number of pages is needed for the asked for size as it doesn't
   take into account the meta data on each page. Add a helper function
   to do the calculation properly and use that in these functions.

 - Catch nr_page_va overflow in ring_buffer_desc sizing

   The number of pages per remote ring buffer is capped by
   ring_buffer_desc::nr_page_va (32 bits). A buffer_size large enough to
   overflow that field would silently allocate a descriptor smaller than
   what was asked for.

 - Do not resize the subbuf order if any per_cpu buffer is disabled

   The mmapping of ring buffers disables resizing the subbuffers, but it
   is done per-cpu whereas the subbuf size change is done for all the
   per_cpu buffers under the buffer->mutex. It could change the size of
   some while the mapping is happening on others. Have the resize of the
   subbuf order check all the per_cpu buffers under the lock to see if
   any of them is disabled before starting and causing an inconsistency
   between buffers that are being mapped.

* tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: (25 commits)
  ring-buffer: Check resize_disabled before publishing the new subbuf order
  tracing/remotes: Catch nr_page_va overflow in ring_buffer_desc sizing
  tracing/remotes: Account for ring buffer page header in size calculation
  tracing: Don't dereference trace_event_file in deferred trigger free
  ftrace: Use rcu_assign_pointer() for tmp_ops filter hash
  ring-buffer: Acquire the lock with irqsave in rb_wake_up_waiters()
  tracing: Take trace_array reference when opening a tracer options file
  tracing: Fix ring_buffer_read_page_size() kernel-doc
  tracing: Restore :mod: trailer after parsing in ftrace_set_clr_event()
  tracing: Fix memory corruption from a "STACKTRACE" histogram key
  tracing: Fix memory corruption from the histogram stacktrace modifier
  tracing: Undo the registration when enabling the histogram trigger fails
  tracing: Take the reference before publishing the named histogram trigger
  tracing: Set the trace clock before registering the histogram trigger
  tracing: Fix typo "preceeded" in comment
  tracing: Fix typo "availabe" in comment
  tracing: Let histogram values keep the percent and graph modifiers
  tracing: Keep the entry count when the histogram stats allocation fails
  tracing: Free histogram the field rejected for a bad modifier
  tracing: Free histogram the var ref when its initialization fails
  ...

1 files changed, 1123 insertions, 0 deletions

diff --git a/drivers/spi/spi-loopback-test.c b/drivers/spi/spi-loopback-test.c
new file mode 100644
index 000000000..066386aba
--- /dev/null
+++ b/drivers/spi/spi-loopback-test.c
@@ -0,0 +1,1123 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *  linux/drivers/spi/spi-loopback-test.c
+ *
+ *  (c) Martin Sperl <kernel@martin.sperl.org>
+ *
+ *  Loopback test driver to test several typical spi_message conditions
+ *  that a spi_master driver may encounter
+ *  this can also get used for regression testing
+ */
+
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/ktime.h>
+#include <linux/list.h>
+#include <linux/list_sort.h>
+#include <linux/module.h>
+#include <linux/printk.h>
+#include <linux/vmalloc.h>
+#include <linux/spi/spi.h>
+
+#include "spi-test.h"
+
+/* flag to only simulate transfers */
+static int simulate_only;
+module_param(simulate_only, int, 0);
+MODULE_PARM_DESC(simulate_only, "if not 0 do not execute the spi message");
+
+/* dump spi messages */
+static int dump_messages;
+module_param(dump_messages, int, 0);
+MODULE_PARM_DESC(dump_messages,
+		 "=1 dump the basic spi_message_structure, " \
+		 "=2 dump the spi_message_structure including data, " \
+		 "=3 dump the spi_message structure before and after execution");
+/* the device is jumpered for loopback - enabling some rx_buf tests */
+static int loopback;
+module_param(loopback, int, 0);
+MODULE_PARM_DESC(loopback,
+		 "if set enable loopback mode, where the rx_buf "	\
+		 "is checked to match tx_buf after the spi_message "	\
+		 "is executed");
+
+static int loop_req;
+module_param(loop_req, int, 0);
+MODULE_PARM_DESC(loop_req,
+		 "if set controller will be asked to enable test loop mode. " \
+		 "If controller supported it, MISO and MOSI will be connected");
+
+static int no_cs;
+module_param(no_cs, int, 0);
+MODULE_PARM_DESC(no_cs,
+		 "if set Chip Select (CS) will not be used");
+
+/* run tests only for a specific length */
+static int run_only_iter_len = -1;
+module_param(run_only_iter_len, int, 0);
+MODULE_PARM_DESC(run_only_iter_len,
+		 "only run tests for a length of this number in iterate_len list");
+
+/* run only a specific test */
+static int run_only_test = -1;
+module_param(run_only_test, int, 0);
+MODULE_PARM_DESC(run_only_test,
+		 "only run the test with this number (0-based !)");
+
+/* use vmalloc'ed buffers */
+static int use_vmalloc;
+module_param(use_vmalloc, int, 0644);
+MODULE_PARM_DESC(use_vmalloc,
+		 "use vmalloc'ed buffers instead of kmalloc'ed");
+
+/* check rx ranges */
+static int check_ranges = 1;
+module_param(check_ranges, int, 0644);
+MODULE_PARM_DESC(check_ranges,
+		 "checks rx_buffer pattern are valid");
+
+static unsigned int delay_ms = 100;
+module_param(delay_ms, uint, 0644);
+MODULE_PARM_DESC(delay_ms,
+		 "delay between tests, in milliseconds (default: 100)");
+
+/* the actual tests to execute */
+static struct spi_test spi_tests[] = {
+	{
+		.description	= "tx/rx-transfer - start of page",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "tx/rx-transfer - crossing PAGE_SIZE",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(PAGE_SIZE - 4),
+				.rx_buf = RX(PAGE_SIZE - 4),
+			},
+		},
+	},
+	{
+		.description	= "tx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "rx-transfer - only",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_rx_align = ITERATE_ALIGN,
+		.transfer_count = 1,
+		.transfers		= {
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				/* this is why we cant use ITERATE_MAX_LEN */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(64),
+			},
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx-transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 16,
+				.tx_buf = TX(0),
+			},
+			{
+				.tx_buf = TX(64),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter tx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+			},
+			{
+				.len = 1,
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two transfers tx then rx - alter rx",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+			},
+			{
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter both",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				/* making sure we align without overwrite
+				 * the reason we can not use ITERATE_MAX_LEN
+				 */
+				.tx_buf = TX(SPI_TEST_MAX_SIZE_HALF),
+				.rx_buf = RX(SPI_TEST_MAX_SIZE_HALF),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter first",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(0),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+			{
+				.len = 1,
+				/* making sure we align without overwrite */
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - alter second",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_tx_align = ITERATE_ALIGN,
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				/* making sure we align without overwrite */
+				.tx_buf = TX(1024),
+				.rx_buf = RX(1024),
+			},
+		},
+	},
+	{
+		.description	= "two tx+rx transfers - delay after transfer",
+		.fill_option	= FILL_COUNT_8,
+		.iterate_len    = { ITERATE_MAX_LEN },
+		.iterate_transfer_mask = BIT(0) | BIT(1),
+		.transfer_count = 2,
+		.transfers		= {
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+				.delay = {
+					.value = 1000,
+					.unit = SPI_DELAY_UNIT_USECS,
+				},
+			},
+			{
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+				.delay = {
+					.value = 1000,
+					.unit = SPI_DELAY_UNIT_USECS,
+				},
+			},
+		},
+	},
+	{
+		.description	= "three tx+rx transfers with overlapping cache lines",
+		.fill_option	= FILL_COUNT_8,
+		/*
+		 * This should be large enough for the controller driver to
+		 * choose to transfer it with DMA.
+		 */
+		.iterate_len    = { 512, -1 },
+		.iterate_transfer_mask = BIT(1),
+		.transfer_count = 3,
+		.transfers		= {
+			{
+				.len = 1,
+				.tx_buf = TX(0),
+				.rx_buf = RX(0),
+			},
+			{
+				.tx_buf = TX(1),
+				.rx_buf = RX(1),
+			},
+			{
+				.len = 1,
+				.tx_buf = TX(513),
+				.rx_buf = RX(513),
+			},
+		},
+	},
+
+	{ /* end of tests sequence */ }
+};
+
+static int spi_loopback_test_probe(struct spi_device *spi)
+{
+	int ret;
+
+	if (loop_req || no_cs) {
+		spi->mode |= loop_req ? SPI_LOOP : 0;
+		spi->mode |= no_cs ? SPI_NO_CS : 0;
+		ret = spi_setup(spi);
+		if (ret) {
+			dev_err(&spi->dev, "SPI setup with SPI_LOOP or SPI_NO_CS failed (%d)\n",
+				ret);
+			return ret;
+		}
+	}
+
+	dev_info(&spi->dev, "Executing spi-loopback-tests\n");
+
+	ret = spi_test_run_tests(spi, spi_tests);
+
+	dev_info(&spi->dev, "Finished spi-loopback-tests with return: %i\n",
+		 ret);
+
+	return ret;
+}
+
+/* non const match table to permit to change via a module parameter */
+static struct of_device_id spi_loopback_test_of_match[] = {
+	{ .compatible	= "linux,spi-loopback-test", },
+	{ }
+};
+
+/* allow to override the compatible string via a module_parameter */
+module_param_string(compatible, spi_loopback_test_of_match[0].compatible,
+		    sizeof(spi_loopback_test_of_match[0].compatible),
+		    0000);
+
+MODULE_DEVICE_TABLE(of, spi_loopback_test_of_match);
+
+static struct spi_driver spi_loopback_test_driver = {
+	.driver = {
+		.name = "spi-loopback-test",
+		.of_match_table = spi_loopback_test_of_match,
+	},
+	.probe = spi_loopback_test_probe,
+};
+
+module_spi_driver(spi_loopback_test_driver);
+
+MODULE_AUTHOR("Martin Sperl <kernel@martin.sperl.org>");
+MODULE_DESCRIPTION("test spi_driver to check core functionality");
+MODULE_LICENSE("GPL");
+
+/*-------------------------------------------------------------------------*/
+
+/* spi_test implementation */
+
+#define RANGE_CHECK(ptr, plen, start, slen) \
+	((ptr >= start) && (ptr + plen <= start + slen))
+
+/* we allocate one page more, to allow for offsets */
+#define SPI_TEST_MAX_SIZE_PLUS (SPI_TEST_MAX_SIZE + PAGE_SIZE)
+
+static void spi_test_print_hex_dump(char *pre, const void *ptr, size_t len)
+{
+	/* limit the hex_dump */
+	if (len <= 1024) {
+		print_hex_dump(KERN_INFO, pre,
+			       DUMP_PREFIX_OFFSET, 16, 1,
+			       ptr, len, 0);
+		return;
+	}
+	/* print head */
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr, 512, 0);
+	/* print tail */
+	pr_info("%s truncated - continuing at offset %04zx\n",
+		pre, len - 512);
+	print_hex_dump(KERN_INFO, pre,
+		       DUMP_PREFIX_OFFSET, 16, 1,
+		       ptr + (len - 512), 512, 0);
+}
+
+static void spi_test_dump_message(struct spi_device *spi,
+				  struct spi_message *msg,
+				  bool dump_data)
+{
+	struct spi_transfer *xfer;
+	int i;
+	u8 b;
+
+	dev_info(&spi->dev, "  spi_msg@%p\n", msg);
+	if (msg->status)
+		dev_info(&spi->dev, "    status:        %i\n",
+			 msg->status);
+	dev_info(&spi->dev, "    frame_length:  %i\n",
+		 msg->frame_length);
+	dev_info(&spi->dev, "    actual_length: %i\n",
+		 msg->actual_length);
+
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		dev_info(&spi->dev, "    spi_transfer@%p\n", xfer);
+		dev_info(&spi->dev, "      len:    %i\n", xfer->len);
+		dev_info(&spi->dev, "      tx_buf: %p\n", xfer->tx_buf);
+		if (dump_data && xfer->tx_buf)
+			spi_test_print_hex_dump("          TX: ",
+						xfer->tx_buf,
+						xfer->len);
+
+		dev_info(&spi->dev, "      rx_buf: %p\n", xfer->rx_buf);
+		if (dump_data && xfer->rx_buf)
+			spi_test_print_hex_dump("          RX: ",
+						xfer->rx_buf,
+						xfer->len);
+		/* check for unwritten test pattern on rx_buf */
+		if (xfer->rx_buf) {
+			for (i = 0 ; i < xfer->len ; i++) {
+				b = ((u8 *)xfer->rx_buf)[xfer->len - 1 - i];
+				if (b != SPI_TEST_PATTERN_UNWRITTEN)
+					break;
+			}
+			if (i)
+				dev_info(&spi->dev,
+					 "      rx_buf filled with %02x starts at offset: %i\n",
+					 SPI_TEST_PATTERN_UNWRITTEN,
+					 xfer->len - i);
+		}
+	}
+}
+
+struct rx_ranges {
+	struct list_head list;
+	u8 *start;
+	u8 *end;
+};
+
+static int rx_ranges_cmp(void *priv, const struct list_head *a,
+			 const struct list_head *b)
+{
+	const struct rx_ranges *rx_a = list_entry(a, struct rx_ranges, list);
+	const struct rx_ranges *rx_b = list_entry(b, struct rx_ranges, list);
+
+	if (rx_a->start > rx_b->start)
+		return 1;
+	if (rx_a->start < rx_b->start)
+		return -1;
+	return 0;
+}
+
+static int spi_check_rx_ranges(struct spi_device *spi,
+			       struct spi_message *msg,
+			       void *rx)
+{
+	struct spi_transfer *xfer;
+	struct rx_ranges ranges[SPI_TEST_MAX_TRANSFERS], *r;
+	int i = 0;
+	LIST_HEAD(ranges_list);
+	u8 *addr;
+	int ret = 0;
+
+	/* loop over all transfers to fill in the rx_ranges */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->rx_buf)
+			continue;
+		/* fill in the rx_range */
+		if (RANGE_CHECK(xfer->rx_buf, xfer->len,
+				rx, SPI_TEST_MAX_SIZE_PLUS)) {
+			ranges[i].start = xfer->rx_buf;
+			ranges[i].end = xfer->rx_buf + xfer->len;
+			list_add(&ranges[i].list, &ranges_list);
+			i++;
+		}
+	}
+
+	/* if no ranges, then we can return and avoid the checks...*/
+	if (!i)
+		return 0;
+
+	/* sort the list */
+	list_sort(NULL, &ranges_list, rx_ranges_cmp);
+
+	/* and iterate over all the rx addresses */
+	for (addr = rx; addr < (u8 *)rx + SPI_TEST_MAX_SIZE_PLUS; addr++) {
+		/* if we are the DO not write pattern,
+		 * then continue with the loop...
+		 */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* check if we are inside a range */
+		list_for_each_entry(r, &ranges_list, list) {
+			/* if so then set to end... */
+			if ((addr >= r->start) && (addr < r->end))
+				addr = r->end;
+		}
+		/* second test after a (hopefull) translation */
+		if (*addr == SPI_TEST_PATTERN_DO_NOT_WRITE)
+			continue;
+
+		/* if still not found then something has modified too much */
+		/* we could list the "closest" transfer here... */
+		dev_err(&spi->dev,
+			"loopback strangeness - rx changed outside of allowed range at: %p\n",
+			addr);
+		/* do not return, only set ret,
+		 * so that we list all addresses
+		 */
+		ret = -ERANGE;
+	}
+
+	return ret;
+}
+
+static int spi_test_check_elapsed_time(struct spi_device *spi,
+				       struct spi_test *test)
+{
+	int i;
+	unsigned long long estimated_time = 0;
+	unsigned long long delay_usecs = 0;
+
+	for (i = 0; i < test->transfer_count; i++) {
+		struct spi_transfer *xfer = test->transfers + i;
+		unsigned long long nbits = (unsigned long long)BITS_PER_BYTE *
+					   xfer->len;
+
+		delay_usecs += xfer->delay.value;
+		if (!xfer->speed_hz)
+			continue;
+		estimated_time += div_u64(nbits * NSEC_PER_SEC, xfer->speed_hz);
+	}
+
+	estimated_time += delay_usecs * NSEC_PER_USEC;
+	if (test->elapsed_time < estimated_time) {
+		dev_err(&spi->dev,
+			"elapsed time %lld ns is shorter than minimum estimated time %lld ns\n",
+			test->elapsed_time, estimated_time);
+
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int spi_test_check_loopback_result(struct spi_device *spi,
+					  struct spi_message *msg,
+					  void *tx, void *rx)
+{
+	struct spi_transfer *xfer;
+	u8 rxb, txb;
+	size_t i;
+	int ret;
+
+	/* checks rx_buffer pattern are valid with loopback or without */
+	if (check_ranges) {
+		ret = spi_check_rx_ranges(spi, msg, rx);
+		if (ret)
+			return ret;
+	}
+
+	/* if we run without loopback, then return now */
+	if (!loopback)
+		return 0;
+
+	/* if applicable to transfer check that rx_buf is equal to tx_buf */
+	list_for_each_entry(xfer, &msg->transfers, transfer_list) {
+		/* if there is no rx, then no check is needed */
+		if (!xfer->len || !xfer->rx_buf)
+			continue;
+		/* so depending on tx_buf we need to handle things */
+		if (xfer->tx_buf) {
+			for (i = 0; i < xfer->len; i++) {
+				txb = ((u8 *)xfer->tx_buf)[i];
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (txb != rxb)
+					goto mismatch_error;
+			}
+		} else {
+			/* first byte received */
+			txb = ((u8 *)xfer->rx_buf)[0];
+			/* first byte may be 0 or 0xff */
+			if (txb != 0 && txb != 0xff) {
+				dev_err(&spi->dev,
+					"loopback strangeness - we expect 0x00 or 0xff, but not 0x%02x\n",
+					txb);
+				return -EINVAL;
+			}
+			/* check that all bytes are identical */
+			for (i = 1; i < xfer->len; i++) {
+				rxb = ((u8 *)xfer->rx_buf)[i];
+				if (rxb != txb)
+					goto mismatch_error;
+			}
+		}
+	}
+
+	return 0;
+
+mismatch_error:
+	dev_err(&spi->dev,
+		"loopback strangeness - transfer mismatch on byte %04zx - expected 0x%02x, but got 0x%02x\n",
+		i, txb, rxb);
+
+	return -EINVAL;
+}
+
+static int spi_test_translate(struct spi_device *spi,
+			      void **ptr, size_t len,
+			      void *tx, void *rx)
+{
+	size_t off;
+
+	/* return on null */
+	if (!*ptr)
+		return 0;
+
+	/* in the MAX_SIZE_HALF case modify the pointer */
+	if (((size_t)*ptr) & SPI_TEST_MAX_SIZE_HALF)
+		/* move the pointer to the correct range */
+		*ptr += (SPI_TEST_MAX_SIZE_PLUS / 2) -
+			SPI_TEST_MAX_SIZE_HALF;
+
+	/* RX range
+	 * - we check against MAX_SIZE_PLUS to allow for automated alignment
+	 */
+	if (RANGE_CHECK(*ptr, len,  RX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - RX(0);
+		*ptr = rx + off;
+
+		return 0;
+	}
+
+	/* TX range */
+	if (RANGE_CHECK(*ptr, len,  TX(0), SPI_TEST_MAX_SIZE_PLUS)) {
+		off = *ptr - TX(0);
+		*ptr = tx + off;
+
+		return 0;
+	}
+
+	dev_err(&spi->dev,
+		"PointerRange [%p:%p[ not in range [%p:%p[ or [%p:%p[\n",
+		*ptr, *ptr + len,
+		RX(0), RX(SPI_TEST_MAX_SIZE),
+		TX(0), TX(SPI_TEST_MAX_SIZE));
+
+	return -EINVAL;
+}
+
+static int spi_test_fill_pattern(struct spi_device *spi,
+				 struct spi_test *test)
+{
+	struct spi_transfer *xfers = test->transfers;
+	u8 *tx_buf;
+	size_t count = 0;
+	int i, j;
+
+#ifdef __BIG_ENDIAN
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (bytes - 1 - count % bytes)))
+#else
+#define GET_VALUE_BYTE(value, index, bytes) \
+	(value >> (8 * (count % bytes)))
+#endif
+
+	/* fill all transfers with the pattern requested */
+	for (i = 0; i < test->transfer_count; i++) {
+		/* fill rx_buf with SPI_TEST_PATTERN_UNWRITTEN */
+		if (xfers[i].rx_buf)
+			memset(xfers[i].rx_buf, SPI_TEST_PATTERN_UNWRITTEN,
+			       xfers[i].len);
+		/* if tx_buf is NULL then skip */
+		tx_buf = (u8 *)xfers[i].tx_buf;
+		if (!tx_buf)
+			continue;
+		/* modify all the transfers */
+		for (j = 0; j < xfers[i].len; j++, tx_buf++, count++) {
+			/* fill tx */
+			switch (test->fill_option) {
+			case FILL_MEMSET_8:
+				*tx_buf = test->fill_pattern;
+				break;
+			case FILL_MEMSET_16:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 2);
+				break;
+			case FILL_MEMSET_24:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 3);
+				break;
+			case FILL_MEMSET_32:
+				*tx_buf = GET_VALUE_BYTE(test->fill_pattern,
+							 count, 4);
+				break;
+			case FILL_COUNT_8:
+				*tx_buf = count;
+				break;
+			case FILL_COUNT_16:
+				*tx_buf = GET_VALUE_BYTE(count, count, 2);
+				break;
+			case FILL_COUNT_24:
+				*tx_buf = GET_VALUE_BYTE(count, count, 3);
+				break;
+			case FILL_COUNT_32:
+				*tx_buf = GET_VALUE_BYTE(count, count, 4);
+				break;
+			case FILL_TRANSFER_BYTE_8:
+				*tx_buf = j;
+				break;
+			case FILL_TRANSFER_BYTE_16:
+				*tx_buf = GET_VALUE_BYTE(j, j, 2);
+				break;
+			case FILL_TRANSFER_BYTE_24:
+				*tx_buf = GET_VALUE_BYTE(j, j, 3);
+				break;
+			case FILL_TRANSFER_BYTE_32:
+				*tx_buf = GET_VALUE_BYTE(j, j, 4);
+				break;
+			case FILL_TRANSFER_NUM:
+				*tx_buf = i;
+				break;
+			default:
+				dev_err(&spi->dev,
+					"unsupported fill_option: %i\n",
+					test->fill_option);
+				return -EINVAL;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int _spi_test_run_iter(struct spi_device *spi,
+			      struct spi_test *test,
+			      void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	struct spi_transfer *x;
+	int i, ret;
+
+	/* initialize message - zero-filled via static initialization */
+	spi_message_init_no_memset(msg);
+
+	/* fill rx with the DO_NOT_WRITE pattern */
+	memset(rx, SPI_TEST_PATTERN_DO_NOT_WRITE, SPI_TEST_MAX_SIZE_PLUS);
+
+	/* add the individual transfers */
+	for (i = 0; i < test->transfer_count; i++) {
+		x = &test->transfers[i];
+
+		/* patch the values of tx_buf */
+		ret = spi_test_translate(spi, (void **)&x->tx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* patch the values of rx_buf */
+		ret = spi_test_translate(spi, &x->rx_buf, x->len,
+					 (void *)tx, rx);
+		if (ret)
+			return ret;
+
+		/* and add it to the list */
+		spi_message_add_tail(x, msg);
+	}
+
+	/* fill in the transfer buffers with pattern */
+	ret = spi_test_fill_pattern(spi, test);
+	if (ret)
+		return ret;
+
+	/* and execute */
+	if (test->execute_msg)
+		ret = test->execute_msg(spi, test, tx, rx);
+	else
+		ret = spi_test_execute_msg(spi, test, tx, rx);
+
+	/* handle result */
+	if (ret == test->expected_return)
+		return 0;
+
+	dev_err(&spi->dev,
+		"test failed - test returned %i, but we expect %i\n",
+		ret, test->expected_return);
+
+	if (ret)
+		return ret;
+
+	/* if it is 0, as we expected something else,
+	 * then return something special
+	 */
+	return -EFAULT;
+}
+
+static int spi_test_run_iter(struct spi_device *spi,
+			     const struct spi_test *testtemplate,
+			     void *tx, void *rx,
+			     size_t len,
+			     size_t tx_off,
+			     size_t rx_off
+	)
+{
+	struct spi_test test;
+	int i, tx_count, rx_count;
+
+	/* copy the test template to test */
+	memcpy(&test, testtemplate, sizeof(test));
+
+	/* if iterate_transfer_mask is not set,
+	 * then set it to first transfer only
+	 */
+	if (!(test.iterate_transfer_mask & (BIT(test.transfer_count) - 1)))
+		test.iterate_transfer_mask = 1;
+
+	/* count number of transfers with tx/rx_buf != NULL */
+	rx_count = tx_count = 0;
+	for (i = 0; i < test.transfer_count; i++) {
+		if (test.transfers[i].tx_buf)
+			tx_count++;
+		if (test.transfers[i].rx_buf)
+			rx_count++;
+	}
+
+	/* in some iteration cases warn and exit early,
+	 * as there is nothing to do, that has not been tested already...
+	 */
+	if (tx_off && (!tx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_tx_off configured with tx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+	if (rx_off && (!rx_count)) {
+		dev_warn_once(&spi->dev,
+			      "%s: iterate_rx_off configured with rx_buf==NULL - ignoring\n",
+			      test.description);
+		return 0;
+	}
+
+	/* write out info */
+	if (!(len || tx_off || rx_off)) {
+		dev_info(&spi->dev, "Running test %s\n", test.description);
+	} else {
+		dev_info(&spi->dev,
+			 "  with iteration values: len = %zu, tx_off = %zu, rx_off = %zu\n",
+			 len, tx_off, rx_off);
+	}
+
+	/* update in the values from iteration values */
+	for (i = 0; i < test.transfer_count; i++) {
+		/* only when bit in transfer mask is set */
+		if (!(test.iterate_transfer_mask & BIT(i)))
+			continue;
+		test.transfers[i].len = len;
+		if (test.transfers[i].tx_buf)
+			test.transfers[i].tx_buf += tx_off;
+		if (test.transfers[i].rx_buf)
+			test.transfers[i].rx_buf += rx_off;
+	}
+
+	/* and execute */
+	return _spi_test_run_iter(spi, &test, tx, rx);
+}
+
+/**
+ * spi_test_execute_msg - default implementation to run a test
+ *
+ * @spi: @spi_device on which to run the @spi_message
+ * @test: the test to execute, which already contains @msg
+ * @tx:   the tx buffer allocated for the test sequence
+ * @rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: error code of spi_sync as well as basic error checking
+ */
+int spi_test_execute_msg(struct spi_device *spi, struct spi_test *test,
+			 void *tx, void *rx)
+{
+	struct spi_message *msg = &test->msg;
+	int ret = 0;
+	int i;
+
+	/* only if we do not simulate */
+	if (!simulate_only) {
+		ktime_t start;
+
+		/* dump the complete message before and after the transfer */
+		if (dump_messages == 3)
+			spi_test_dump_message(spi, msg, true);
+
+		start = ktime_get();
+		/* run spi message */
+		ret = spi_sync(spi, msg);
+		test->elapsed_time = ktime_to_ns(ktime_sub(ktime_get(), start));
+		if (ret == -ETIMEDOUT) {
+			dev_info(&spi->dev,
+				 "spi-message timed out - rerunning...\n");
+			/* rerun after a few explicit schedules */
+			for (i = 0; i < 16; i++)
+				schedule();
+			ret = spi_sync(spi, msg);
+		}
+		if (ret) {
+			dev_err(&spi->dev,
+				"Failed to execute spi_message: %i\n",
+				ret);
+			goto exit;
+		}
+
+		/* do some extra error checks */
+		if (msg->frame_length != msg->actual_length) {
+			dev_err(&spi->dev,
+				"actual length differs from expected\n");
+			ret = -EIO;
+			goto exit;
+		}
+
+		/* run rx-buffer tests */
+		ret = spi_test_check_loopback_result(spi, msg, tx, rx);
+		if (ret)
+			goto exit;
+
+		ret = spi_test_check_elapsed_time(spi, test);
+	}
+
+	/* if requested or on error dump message (including data) */
+exit:
+	if (dump_messages || ret)
+		spi_test_dump_message(spi, msg,
+				      (dump_messages >= 2) || (ret));
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_execute_msg);
+
+/**
+ * spi_test_run_test - run an individual spi_test
+ *                     including all the relevant iterations on:
+ *                     length and buffer alignment
+ *
+ * @spi:  the spi_device to send the messages to
+ * @test: the test which we need to execute
+ * @tx:   the tx buffer allocated for the test sequence
+ * @rx:   the rx buffer allocated for the test sequence
+ *
+ * Returns: status code of spi_sync or other failures
+ */
+
+int spi_test_run_test(struct spi_device *spi, const struct spi_test *test,
+		      void *tx, void *rx)
+{
+	int idx_len;
+	size_t len;
+	size_t tx_align, rx_align;
+	int ret;
+
+	/* test for transfer limits */
+	if (test->transfer_count >= SPI_TEST_MAX_TRANSFERS) {
+		dev_err(&spi->dev,
+			"%s: Exceeded max number of transfers with %i\n",
+			test->description, test->transfer_count);
+		return -E2BIG;
+	}
+
+	/* setting up some values in spi_message
+	 * based on some settings in spi_master
+	 * some of this can also get done in the run() method
+	 */
+
+	/* iterate over all the iterable values using macros
+	 * (to make it a bit more readable...
+	 */
+#define FOR_EACH_ALIGNMENT(var)						\
+	for (var = 0;							\
+	    var < (test->iterate_##var ?				\
+			(spi->controller->dma_alignment ?		\
+			 spi->controller->dma_alignment :		\
+			 test->iterate_##var) :				\
+			1);						\
+	    var++)
+
+	for (idx_len = 0; idx_len < SPI_TEST_MAX_ITERATE &&
+	     (len = test->iterate_len[idx_len]) != -1; idx_len++) {
+		if ((run_only_iter_len > -1) && len != run_only_iter_len)
+			continue;
+		FOR_EACH_ALIGNMENT(tx_align) {
+			FOR_EACH_ALIGNMENT(rx_align) {
+				/* and run the iteration */
+				ret = spi_test_run_iter(spi, test,
+							tx, rx,
+							len,
+							tx_align,
+							rx_align);
+				if (ret)
+					return ret;
+			}
+		}
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_test);
+
+/**
+ * spi_test_run_tests - run an array of spi_messages tests
+ * @spi: the spi device on which to run the tests
+ * @tests: NULL-terminated array of @spi_test
+ *
+ * Returns: status errors as per @spi_test_run_test()
+ */
+
+int spi_test_run_tests(struct spi_device *spi,
+		       struct spi_test *tests)
+{
+	char *rx = NULL, *tx = NULL;
+	int ret = 0, count = 0;
+	struct spi_test *test;
+
+	/* allocate rx/tx buffers of 128kB size without devm
+	 * in the hope that is on a page boundary
+	 */
+	if (use_vmalloc)
+		rx = vmalloc(SPI_TEST_MAX_SIZE_PLUS);
+	else
+		rx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!rx)
+		return -ENOMEM;
+
+
+	if (use_vmalloc)
+		tx = vmalloc(SPI_TEST_MAX_SIZE_PLUS);
+	else
+		tx = kzalloc(SPI_TEST_MAX_SIZE_PLUS, GFP_KERNEL);
+	if (!tx) {
+		ret = -ENOMEM;
+		goto err_tx;
+	}
+
+	/* now run the individual tests in the table */
+	for (test = tests, count = 0; test->description[0];
+	     test++, count++) {
+		/* only run test if requested */
+		if ((run_only_test > -1) && (count != run_only_test))
+			continue;
+		/* run custom implementation */
+		if (test->run_test)
+			ret = test->run_test(spi, test, tx, rx);
+		else
+			ret = spi_test_run_test(spi, test, tx, rx);
+		if (ret)
+			goto out;
+		/* add some delays so that we can easily
+		 * detect the individual tests when using a logic analyzer
+		 * we also add scheduling to avoid potential spi_timeouts...
+		 */
+		if (delay_ms)
+			mdelay(delay_ms);
+		schedule();
+	}
+
+out:
+	kvfree(tx);
+err_tx:
+	kvfree(rx);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(spi_test_run_tests);