When interrupted, wait_event_interruptible_timeout() returns
-ERESTARTSYS, and the SPI transfer in progress will fail, as expected:

    m25p80 spi0.0: SPI transfer failed: -512
    spi_master spi0: failed to transfer one message from queue

However, as the underlying DMA transfers may not have completed, all
subsequent SPI transfers may start to fail:

    spi_master spi0: receive timeout
    qspi_transfer_out_in() returned -110
    m25p80 spi0.0: SPI transfer failed: -110
    spi_master spi0: failed to transfer one message from queue

Fix this by calling dmaengine_terminate_all() not only for timeouts, but
also for errors.

This can be reproduced on r8a7991/koelsch, using "hd /dev/mtd0" followed
by CTRL-C.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org

If the SPI queue is running during system suspend, the system may lock
up.

Fix this by stopping/restarting the queue during system suspend/resume,
by calling spi_master_suspend()/spi_master_resume() from the PM
callbacks.  In-kernel users will receive an -ESHUTDOWN error while
system suspend/resume is in progress.

Based on a patch for sh-msiof by Gaku Inami.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org

Use enum dma_transfer_direction as required by dmaengine_prep_slave_sg
instead of enum dma_data_direction. This won't change behavior in
practice as the enum values are equivalent.

This fixes two warnings when building with clang:
  drivers/spi/spi-rspi.c:538:26: warning: implicit conversion from enumeration
      type 'enum dma_data_direction' to different enumeration type
      'enum dma_transfer_direction' [-Wenum-conversion]
                                        rx->sgl, rx->nents, DMA_FROM_DEVICE,
                                                            ^~~~~~~~~~~~~~~
  drivers/spi/spi-rspi.c:558:26: warning: implicit conversion from enumeration
      type 'enum dma_data_direction' to different enumeration type
      'enum dma_transfer_direction' [-Wenum-conversion]
                                        tx->sgl, tx->nents, DMA_TO_DEVICE,
                                                            ^~~~~~~~~~~~~
Signed-off-by: NStefan Agner <stefan@agner.ch>
Signed-off-by: NMark Brown <broonie@kernel.org>

The R-Car Gen2 Hardware User Manual Rev. 2.00 states:

    If the master/slave mode select bit (MSTR) is modified while the SPI
    function enable bit (SPE) is set to 1 (that is, this module is
    enabled), the subsequent operation cannot be guaranteed.

Hence do not set SPCR_SPE when setting SPCR_MSTR, just like the
.set_config_register() implementations for other RSPI variants do.

Note that when booted from QSPI, the boot loader will have set SPCR_MSTR
already, hence usually the bit is never modified by the Linux driver.
Reported-by: NYoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

Use the of_device_get_match_data() helper instead of open coding.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

This patch replaced "n" by "len" bytes of data in qspi_transfer_in() and
qspi_transfer_out() function. This will make improving readability.
Signed-off-by: NDongCV <cv-dong@jinso.co.jp>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

In qspi_transfer_in(), when receiving the last n (or len) bytes of data,
one bogus byte was written in the receive buffer.
This code leads to a buffer overflow.

"jffs2: jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found
at 0x03b40000: 0x1900 instead
jffs2: jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found
at 0x03b40004: 0x000c instead"

The error message above happens when trying to mount, unmount,
and remount a jffs2-formatted device.
This patch removed the bogus write to fixes: 3be09bec
"spi: rspi: supports 32bytes buffer for DUAL and QUAD"

And here is Geert's comment:

"spi: rspi: Fix bogus received byte in qspi_transfer_in()
When there are less than QSPI_BUFFER_SIZE remaining bytes to be received,
qspi_transfer_in() writes one bogus byte in the receive buffer, possibly
leading to a buffer overflow.
This can be reproduced by mounting, unmounting, and remounting a
jffs2-formatted device, causing lots of warnings like:

"jffs2: jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found
at 0x03b40000: 0x1900 instead"

Remove the bogus write to fix this. "
Signed-off-by: NDongCV <cv-dong@jinso.co.jp>
Reviewed-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

Printing an error on memory allocation failure is unnecessary.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

The newly introduced rspi_pio_transfer_in_or_our() function must
take either a valid 'rx' or 'tx' pointer, and has undefined behavior
if both are NULL, as found by 'gcc -Wmaybe-unintialized':

drivers/spi/spi-rspi.c: In function 'rspi_pio_transfer_in_or_our':
drivers/spi/spi-rspi.c:558:5: error: 'len' may be used uninitialized in this function [-Werror=maybe-uninitialized]

The analysis of the function is correct in principle, but the code
is currently safe because both callers always pass exactly one
of the two pointers.

Looking closer at this function shows that having a combined
method for rx and tx here actually increases the complexity
and the size of the file. This simplifies it again by keeping
the two separate, which then ends up avoiding that warning.

Fixes: 3be09bec ("spi: rspi: supports 32bytes buffer for DUAL and QUAD")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NMark Brown <broonie@kernel.org>

This patch supports 32bytes of buffer for DUAL and QUAD in QSPI by
Using Transmit/Receive Buffer Data Triggering Number.
In order to improve the DUAL and QUAD's performance of SPI
while transferring data in PIO mode, it sends/receives each 32bytes
data instead of each byte data as current situation.
Signed-off-by: NHiep Cao Minh <cm-hiep@jinso.co.jp>
Signed-off-by: NMark Brown <broonie@kernel.org>

When you leave the clock divider at 0, 130kHz is the lowest you can go.
Also, by adjusting the clock divider you can get more accurate resolutions
for clock speeds lower than 16MHz. This patch uses the clock divider as
part of the bit rate setup.
Signed-off-by: NChris Brandt <chris.brandt@renesas.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

qspi_set_send_trigger() returns an unsigned value, so make it return
"unsigned int".
Update the loop variables qspi_trigger_transfer_out_int() to match the
above.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

Just use "ret" instead, for consistency with other similar functions.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

To reduce complexity of code, drop "ret" then qspi_transfer_out_in function
should return the value of "qspi_trigger_transfer_out_in" directly.
Signed-off-by: NHiep Cao Minh <cm-hiep@jinso.co.jp>
Acked-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

The name of "qspi_trigger_transfer_out_int" function should be
"qspi_trigger_transfer_out_in" without "t".
Signed-off-by: NHiep Cao Minh <cm-hiep@jinso.co.jp>
Acked-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@kernel.org>

To reduce indentation and complexity of code, insteeds of returning zero
the function rspi_dma_check_then_transfer should return rspi_dma_transfer
directly after checking error.
Signed-off-by: NHiep Cao Minh <cm-hiep@jinso.co.jp>
Signed-off-by: NMark Brown <broonie@kernel.org>

The platform_device_id is not modified by the driver and core uses it as
const.
Signed-off-by: NKrzysztof Kozlowski <k.kozlowski.k@gmail.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

We need "ret" to be unsigned for the error handling to work.  The
signedness of "i" and "n" don't matter but qspi_set_send_trigger()
returns an int so I've changed them to int as well.

Fixes: 4b6fe3ed ('spi: Using Trigger number to transmit/receive data')
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

In order to transmit and receive data when have 32 bytes of data that
ready has prepared on Transmit/Receive Buffer to transmit or receive.
Instead transmits/receives a byte data using Transmit/Receive Buffer
Data Triggering Number will improve the speed of transfer data.
Signed-off-by: NHiep Cao Minh <cm-hiep@jinso.co.jp>
Signed-off-by: NMark Brown <broonie@kernel.org>

Current rspi sets dma_slave_config :: slave_id field for DMAEngine,
but it is no longer needed. Let's remove it.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Acked-by: NMark Brown <broonie@kernel.org>
Acked-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NVinod Koul <vinod.koul@intel.com>

Signed-off-by: NJarkko Nikula <jarkko.nikula@linux.intel.com>
Signed-off-by: NMark Brown <broonie@kernel.org>

A platform_driver does not need to set an owner, it will be populated by the
driver core.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>

Add support for QSPI in:
  - r8a7792 (R-Car V2H)
  - r8a7793 (R-Car M2-N)
  - r8a7794 (R-Car E2)

r8a7791 is now called "R-Car M2-W".
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

The new Renesas R-Car Gen2 DMA Controller driver (rcar-dmac) requires
explicit configuration of the DMA slave bus width.
Hardcode the DMA transfer size to 1 byte for both directions, as that's
the only supported configuration (16-bit DMA support was removed in
commit 9c5de2c1 ("spi: rspi: Remove
unused 16-bit DMA support")).
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use the devm_kasprintf() helper function instead of open coding
error-prone buffer handling and string formatting.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Introduced by commit 426ef76d ("spi:
rspi: Add DT support").
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

If dmaengine_prep_slave_sg() or dmaengine_submit() fail, we may leak
unused DMA descriptors.

As per Documentation/dmaengine.txt, once a DMA descriptor has been
obtained, it must be submitted. Hence:
  - First prepare and submit all DMA descriptors,
  - Prepare the SPI controller for DMA,
  - Start DMA by calling dma_async_issue_pending(),
  - Make sure to call dmaengine_terminate_all() on all descriptors that
    haven't completed.
Reported-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

As typically a shmobile SoC has less DMA channels than devices that can use
DMA, we may want to prioritize access to the DMA channels in the future.
This means that dmaengine_prep_slave_sg() may start failing arbitrarily.

Handle dmaengine_prep_slave_sg() failures gracefully by falling back to
PIO.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Acked-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NMark Brown <broonie@linaro.org>

rspi_release_dma() doesn't need access to any fields in the driver private
data, except for the pointer to the SPI master object. Hence just pass the
needed pointer.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Introduced by commit 8b983e90 ("spi: rspi:
Extract rspi_common_transfer()"), which removed its users.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Extract the common parts of rspi_transfer_one(), rspi_rz_transfer_one(),
and qspi_transfer_out_in() into the new function rspi_common_transfer().
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Enable DMA support for RSPI on r7s72100 (RZ/A1H).
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Enable DMA support for QSPI on R-Car Gen2, for Single, Dual, and Quad SPI
Transfers.

Performance figures for reading from a QSPI FLASH driven at 24.375 MHz
on r8a7791/koelsch:
  - Single:  1.1 Mbps PIO, 23 Mbps DMA
  - Dual  : 12.7 Mbps PIO, 48 Mbps DMA
  - Quad  : 13   Mbps PIO, 70 Mbps DMA
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

rspi_send_dma() and rspi_send_receive_dma() are very similar. Consolidate
into a single function rspi_dma_transfer(), and add missing checks for
dmaengine_submit() failures.

Both sg_table pointer parameters can be NULL, as RSPI supports TX-only
mode, and unidirectional DMA transfers will also be needed later for
Dual/Quad DMA support.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

The DMA routines only need access to the scatter-gather tables inside the
spi_transfer structures, hence just pass those.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Refactor RSPI (on SH) DMA handling to make it reusable for other RSPI
implementations:
  - Call the DMA routines after configuring the TX Mode bit and after
    calling rspi_receive_init(), so these RSPI-specific operations can be
    removed from the DMA routines,
  - Absorb rspi_transfer_out_in() into rspi_transfer_one().
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

Use the SPI core DMA mapping framework instead of our own.
If available, DMA is used for transfers larger than the FIFO size
(8 or 32 bytes).
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>

The SPI DMA core framework needs both RX and TX DMA to function.  As a
preparation for converting the driver to use this framework, fall back to
PIO if no DMA channel or only one DMA channel is available.

This affects only RSPI, which could do DMA transfers for TX-only before.
RSPI-RZ and QSPI (at least for Single SPI Transfers) will need both RX and
TX DMA anyway.
Signed-off-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NMark Brown <broonie@linaro.org>