- 31 Dec, 2019 1 commit
-
-
Vladimir Oltean authored
When used in Extended SPI mode on LS1021A, the DSPI controller wants to have the least significant 16-bit word written first to the TX FIFO. In fact, the LS1021A reference manual says: 33.5.2.4.2 Draining the TX FIFO When Extended SPI Mode (DSPIx_MCR[XSPI]) is enabled, if the frame size of SPI Data to be transmitted is more than 16 bits, then it causes two Data entries to be popped from TX FIFO simultaneously which are transferred to the shift register. The first of the two popped entries forms the 16 least significant bits of the SPI frame to be transmitted. So given the following TX buffer: +-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+ | 0x0 | 0x1 | 0x2 | 0x3 | 0x4 | 0x5 | 0x6 | 0x7 | 0x8 | 0x9 | 0xa | 0xb | +-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+ | 32-bit word 1 | 32-bit word 2 | 32-bit word 3 | +-----------------------+-----------------------+-----------------------+ The correct way that a little-endian system should transmit it on the wire when bits_per_word is 32 is: 0x03020100 0x07060504 0x0b0a0908 But it is actually transmitted as following, as seen with a scope: 0x01000302 0x05040706 0x09080b0a It appears that this patch has been submitted at least once before: https://lkml.org/lkml/2018/9/21/286 but in that case Chuanhua Han did not manage to explain the problem clearly enough and the patch did not get merged, leaving XSPI mode broken. Fixes: 8fcd151d ("spi: spi-fsl-dspi: XSPI FIFO handling (in TCFQ mode)") Cc: Esben Haabendal <eha@deif.com> Cc: Chuanhua Han <chuanhua.han@nxp.com> Signed-off-by:
Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/20191228135536.14284-1-olteanv@gmail.comSigned-off-by:
Mark Brown <broonie@kernel.org> Cc: stable@vger.kernel.org
-
- 27 Dec, 2019 1 commit
-
-
Vladimir Oltean authored
The API for PTP system timestamping (associating a SPI transaction with the system time at which it was transferred) is flawed: it assumes that the xfer->tx_buf pointer will always be present. This is, of course, not always the case. So introduce a "progress" variable that denotes how many word have been transferred. Fix the Freescale DSPI driver, the only user of the API so far, in the same patch. Fixes: b42faeee ("spi: Add a PTP system timestamp to the transfer structure") Fixes: d6b71dfa ("spi: spi-fsl-dspi: Implement the PTP system timestamping for TCFQ mode") Signed-off-by:
-
- 25 Dec, 2019 1 commit
-
-
Kunihiko Hayashi authored
Rx threshold means the value to inform the receiver when the number of words in Rx FIFO is equal to or more than the value. Similarly, Tx threshold means the value to inform the sender when the number of words in Tx FIFO is equal to or less than the value. The controller triggers the driver to start the transfer. In case of Rx, the driver wants to detect that the specified number of words N are in Rx FIFO, so the value of Rx threshold should be N. In case of Tx, the driver wants to detect that the same number of spaces as Rx are in Tx FIFO, so the value of Tx threshold should be (FIFO size - N). For example, in order for the driver to receive at least 3 words from Rx FIFO, set 3 to Rx threshold. +-+-+-+-+-+-+-+-+ | | | | | |*|*|*| +-+-+-+-+-+-+-+-+ In order for the driver to send at least 3 words to Tx FIFO, because it needs at least 3 spaces, set 8(FIFO size) - 3 = 5 to Tx threshold. +-+-+-+-+-+-+-+-+ |*|*|*|*|*| | | | +-+-+-+-+-+-+-+-+ This adds new function uniphier_spi_set_fifo_threshold() to set threshold value to the register. And more, FIFO counts by 'words', so this renames 'fill_bytes' with 'fill_words', and fixes the calculation using bytes_per_words. Fixes: 37ffab81 ("spi: uniphier: introduce polling mode") Cc: Keiji Hayashibara <hayashibara.keiji@socionext.com> Signed-off-by:
Kunihiko Hayashi <hayashi.kunihiko@socionext.com> Link: https://lore.kernel.org/r/1577149107-30670-2-git-send-email-hayashi.kunihiko@socionext.comSigned-off-by:
-
- 16 Dec, 2019 2 commits
-
-
Christophe Leroy authored
Unlike irq_of_parse_and_map() which has a dummy definition on SPARC, of_irq_to_resource() hasn't. But as platform_get_irq() can be used instead and is generic, use it. Reported-by:
kbuild test robot <lkp@intel.com> Suggested-by:
Christophe Leroy <christophe.leroy@c-s.fr> Link: https://lore.kernel.org/r/091a277fd0b3356dca1e29858c1c96983fc9cb25.1576172743.git.christophe.leroy@c-s.frSigned-off-by:
-
Michael Walle authored
Make use of a core helper to ensure the desired width is respected when calling spi-mem operators. Otherwise only the SPI controller will be matched with the flash chip, which might lead to wrong widths. Also consider the width specified by the user in the device tree. Fixes: a5356aef ("spi: spi-mem: Add driver for NXP FlexSPI controller") Signed-off-by:
Michael Walle <michael@walle.cc> Link: https://lore.kernel.org/r/20191211195730.26794-1-michael@walle.ccSigned-off-by:
-
- 11 Dec, 2019 1 commit
-
-
Vignesh Raghavendra authored
When switching ChipSelect from default CS0 to any other CS, driver fails to update the bits in system control module register that control which CS is mapped for MMIO access. This causes reads to fail when driver tries to access QSPI flash on CS1/2/3. Fix this by updating appropriate bits whenever active CS changes. Reported-by:
Andreas Dannenberg <dannenberg@ti.com> Signed-off-by:
Vignesh Raghavendra <vigneshr@ti.com> Link: https://lore.kernel.org/r/20191211155216.30212-1-vigneshr@ti.comSigned-off-by:
-
- 09 Dec, 2019 2 commits
-
-
Christophe Leroy authored
With lastest kernel, the following warning is observed at startup: [ 1.500609] ------------[ cut here ]------------ [ 1.505225] remove_proc_entry: removing non-empty directory 'irq/22', leaking at least 'fsl_spi' [ 1.514234] WARNING: CPU: 0 PID: 1 at fs/proc/generic.c:682 remove_proc_entry+0x198/0x1c0 [ 1.522403] CPU: 0 PID: 1 Comm: swapper Not tainted 5.4.0-s3k-dev-02248-g93532430a4ff #2564 [ 1.530724] NIP: c0197694 LR: c0197694 CTR: c0050d80 [ 1.535762] REGS: df4a5af0 TRAP: 0700 Not tainted (5.4.0-02248-g93532430a4ff) [ 1.543818] MSR: 00029032 <EE,ME,IR,DR,RI> CR: 22028222 XER: 00000000 [ 1.550524] [ 1.550524] GPR00: c0197694 df4a5ba8 df4a0000 00000054 00000000 00000000 00004a38 00000010 [ 1.550524] GPR08: c07c5a30 00000800 00000000 00001032 22000208 00000000 c0004b14 00000000 [ 1.550524] GPR16: 00000000 00000000 00000000 00000000 00000000 00000000 c0830000 c07fc078 [ 1.550524] GPR24: c08e8ca0 df665d10 df60ea98 c07c9db8 00000001 df5d5ae3 df5d5a80 df43f8e3 [ 1.585327] NIP [c0197694] remove_proc_entry+0x198/0x1c0 [ 1.590628] LR [c0197694] remove_proc_entry+0x198/0x1c0 [ 1.595829] Call Trace: [ 1.598280] [df4a5ba8] [c0197694] remove_proc_entry+0x198/0x1c0 (unreliable) [ 1.605321] [df4a5bd8] [c0067acc] unregister_irq_proc+0x5c/0x70 [ 1.611238] [df4a5bf8] [c005fbc4] free_desc+0x3c/0x80 [ 1.616286] [df4a5c18] [c005fe2c] irq_free_descs+0x70/0xa8 [ 1.621778] [df4a5c38] [c033d3fc] of_fsl_spi_probe+0xdc/0x3cc [ 1.627525] [df4a5c88] [c02f0f64] platform_drv_probe+0x44/0xa4 [ 1.633350] [df4a5c98] [c02eee44] really_probe+0x1ac/0x418 [ 1.638829] [df4a5cc8] [c02ed3e8] bus_for_each_drv+0x64/0xb0 [ 1.644481] [df4a5cf8] [c02ef950] __device_attach+0xd4/0x128 [ 1.650132] [df4a5d28] [c02ed61c] bus_probe_device+0xa0/0xbc [ 1.655783] [df4a5d48] [c02ebbe8] device_add+0x544/0x74c [ 1.661096] [df4a5d88] [c0382b78] of_platform_device_create_pdata+0xa4/0x100 [ 1.668131] [df4a5da8] [c0382cf4] of_platform_bus_create+0x120/0x20c [ 1.674474] [df4a5df8] [c0382d50] of_platform_bus_create+0x17c/0x20c [ 1.680818] [df4a5e48] [c0382e88] of_platform_bus_probe+0x9c/0xf0 [ 1.686907] [df4a5e68] [c0751404] __machine_initcall_cmpcpro_cmpcpro_declare_of_platform_devices+0x74/0x1a4 [ 1.696629] [df4a5e98] [c072a4cc] do_one_initcall+0x8c/0x1d4 [ 1.702282] [df4a5ef8] [c072a768] kernel_init_freeable+0x154/0x204 [ 1.708455] [df4a5f28] [c0004b2c] kernel_init+0x18/0x110 [ 1.713769] [df4a5f38] [c00122ac] ret_from_kernel_thread+0x14/0x1c [ 1.719926] Instruction dump: [ 1.722889] 2c030000 4182004c 3863ffb0 3c80c05f 80e3005c 388436a0 3c60c06d 7fa6eb78 [ 1.730630] 7fe5fb78 38840280 38634178 4be8c611 <0fe00000> 4bffff6c 3c60c071 7fe4fb78 [ 1.738556] ---[ end trace 05d0720bf2e352e2 ]--- The problem comes from the error path which calls irq_dispose_mapping() while the IRQ has been requested with devm_request_irq(). IRQ doesn't need to be mapped with irq_of_parse_and_map(). The only need is to get the IRQ virtual number. For that, use of_irq_to_resource() instead of the irq_of_parse_and_map()/irq_dispose_mapping() pair. Fixes: 500a32ab ("spi: fsl: Call irq_dispose_mapping in err path") Cc: stable@vger.kernel.org Signed-off-by:
-
Chuhong Yuan authored
The driver forgets to call pci_release_regions() in probe failure and remove. Add the missed calls to fix it. Signed-off-by:
Chuhong Yuan <hslester96@gmail.com> Link: https://lore.kernel.org/r/20191206075500.18525-1-hslester96@gmail.comSigned-off-by:
-
- 04 Dec, 2019 1 commit
-
-
Huanpeng Xin authored
The original code used an incorrect SPI register to initialize the SPI controller in sprd_spi_init_hw(), thus fix it. Fixes: e7d973a3 ("spi: sprd: Add SPI driver for Spreadtrum SC9860") Signed-off-by:
Huanpeng Xin <huanpeng.xin@unisoc.com> Signed-off-by:
Baolin Wang <baolin.wang7@gmail.com> Link: https://lore.kernel.org/r/b4f7f89ec0fdc595335687bfbd9f962213bc4a1d.1575443510.git.baolin.wang7@gmail.comSigned-off-by:
-
- 03 Dec, 2019 1 commit
-
-
YueHaibing authored
Fix sparse warning: drivers/gpio/gpiolib-of.c:35:5: warning: symbol 'of_gpio_spi_cs_get_count' was not declared. Should it be static? Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Reviewed-by:
Linus Walleij <linus.walleij@linaro.org> Link: https://lore.kernel.org/r/20191130012828.14504-1-yuehaibing@huawei.comSigned-off-by:
-
- 28 Nov, 2019 4 commits
-
-
Linus Walleij authored
The Freescale MPC8xxx had a special quirk for handling a single hardwired chipselect, the case when we're using neither GPIO nor native chip select: when inspecting the device tree and finding zero "cs-gpios" on the device node the code would assume we have a single hardwired chipselect that leaves the device always selected. This quirk is not handled by the new core code, so we need to check the "cs-gpios" explicitly in the driver and set pdata->max_chipselect = 1 which will later fall through to the SPI master ->num_chipselect. Make sure not to assign the chip select handler in this case: there is no handling needed since the chip is always selected, and this is what the old code did as well. Cc: Christophe Leroy <christophe.leroy@c-s.fr> Reported-by:
-
Linus Walleij authored
We have a special quirk to handle the Freescale nonstandard SPI chipselect GPIOs in the gpiolib-of.c file, but it currently only handles the case where the GPIOs are actually requested (gpiod_*get()). We also need to handle that the SPI core attempts to count the GPIOs before use, and that needs a similar quirk in the OF part of the library. Cc: Christophe Leroy <christophe.leroy@c-s.fr> Reported-by:
-
Linus Walleij authored
This makes the driver actually support looking up GPIO descriptor. A coding mistake in the initial descriptor support patch was that it was failing to turn on the very feature it was implementing. Mea culpa. Cc: Christophe Leroy <christophe.leroy@c-s.fr> Reported-by:
-
Charles Keepax authored
This patch reverts commit 6e0a32d6 ("spi: dw: Fix default polarity of native chipselect"). The SPI framework always called the set_cs callback with the logic level it desired on the chip select line, which is what the drivers original handling supported. commit f3186dd8 ("spi: Optionally use GPIO descriptors for CS GPIOs") changed these symantics, but only in the case of drivers that also support GPIO chip selects, to true meaning apply slave select rather than logic high. This left things in an odd state where a driver that only supports hardware chip selects, the core would handle polarity but if the driver supported GPIOs as well the driver should handle polarity. At this point the reverted change was applied to change the logic in the driver to match new system. This was then broken by commit 3e5ec1db ("spi: Fix SPI_CS_HIGH setting when using native and GPIO CS") which reverted the core back to consistently calling set_cs with a logic level. This fix reverts the driver code back to its original state to match the current core code. This is probably a better fix as a) the set_cs callback is always called with consistent symantics and b) the inversion for SPI_CS_HIGH can be handled in the core and doesn't need to be coded in each driver supporting it. Fixes: 3e5ec1db ("spi: Fix SPI_CS_HIGH setting when using native and GPIO CS") Signed-off-by:
Charles Keepax <ckeepax@opensource.cirrus.com> Acked-by:
-
- 27 Nov, 2019 1 commit
-
-
Charles Keepax authored
To fix a regression on the Cadence SPI driver, this patch reverts commit 6046f540 ("spi: cadence: Fix default polarity of native chipselect"). This patch was not the correct fix for the issue. The SPI framework calls the set_cs line with the logic level it desires on the chip select line, as such the old is_high handling was correct. However, this was broken by the fact that before commit 3e5ec1db ("spi: Fix SPI_CS_HIGH setting when using native and GPIO CS") all controllers that offered the use of a GPIO chip select had SPI_CS_HIGH applied, even for hardware chip selects. This caused the value passed into the driver to be inverted. Which unfortunately makes it look like a logical enable the chip select value. Since the core was corrected to not unconditionally apply SPI_CS_HIGH, the Cadence driver, whilst using the hardware chip select, will deselect the chip select every time we attempt to communicate with the device, which results in failed communications. Fixes: 3e5ec1db ("spi: Fix SPI_CS_HIGH setting when using native and GPIO CS") Signed-off-by:
-
- 25 Nov, 2019 1 commit
-
-
Jarkko Nikula authored
LPSS SPI on Intel Jasper Lake is compatible with Intel Ice Lake which follows Intel Cannon Lake. Add PCI IDs of Jasper Lake. Signed-off-by:
Jarkko Nikula <jarkko.nikula@linux.intel.com> Link: https://lore.kernel.org/r/20191125125159.15404-1-jarkko.nikula@linux.intel.comSigned-off-by:
-
- 22 Nov, 2019 2 commits
-
-
Benjamin Gaignard authored
Convert the STM32 QSPI binding to DT schema format using json-schema Signed-off-by:
Benjamin Gaignard <benjamin.gaignard@st.com> Reviewed-by:
Rob Herring <robh@kernel.org> Link: https://lore.kernel.org/r/20191120194444.10540-1-benjamin.gaignard@st.comSigned-off-by:
-
Peter Ujfalusi authored
There is no reason to use the dma_request_slave_channel_compat() as no filter function and parameter is provided. Switch the driver to use dma_request_chan() instead and add support for deferred probing against DMA channel. Signed-off-by:
Peter Ujfalusi <peter.ujfalusi@ti.com> Reviewed-by:
Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20191121092703.30465-1-peter.ujfalusi@ti.comSigned-off-by:
-
- 20 Nov, 2019 1 commit
-
-
Krzysztof Kozlowski authored
Adjust indentation from spaces to tab (+optional two spaces) as in coding style with command like: $ sed -e 's/^ /\t/' -i */Kconfig Signed-off-by:
Krzysztof Kozlowski <krzk@kernel.org> Link: https://lore.kernel.org/r/20191120133916.13595-1-krzk@kernel.orgSigned-off-by:
-
- 18 Nov, 2019 2 commits
-
-
Luhua Xu authored
Change to use SPI_CS_HIGH to support spi CS polarity setting for chips support enhance_timing. Signed-off-by:
Luhua Xu <luhua.xu@mediatek.com> Link: https://lore.kernel.org/r/1574053037-26721-2-git-send-email-luhua.xu@mediatek.comSigned-off-by:
-
Chuhong Yuan authored
The driver forgets to call pm_runtime_disable in probe failure and remove. Add the missed calls to fix it. Signed-off-by:
-
- 15 Nov, 2019 10 commits
-
-
Chuhong Yuan authored
The driver misses calling clk_unprepare in probe failure and remove. Add the calls to fix it. Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
Jon Hunter <jonathanh@nvidia.com> Link: https://lore.kernel.org/r/20191113094256.1108-10-peter.ujfalusi@ti.comSigned-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
Andi Shyti <andi@etezian.org> Link: https://lore.kernel.org/r/20191113094256.1108-8-peter.ujfalusi@ti.comSigned-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
Nicolas Ferre <nicolas.ferre@microchip.com> Link: https://lore.kernel.org/r/20191113094256.1108-3-peter.ujfalusi@ti.comSigned-off-by:
-
Peter Ujfalusi authored
dma_request_slave_channel_reason() is: #define dma_request_slave_channel_reason(dev, name) \ dma_request_chan(dev, name) Signed-off-by:
-
- 12 Nov, 2019 1 commit
-
-
Tony Lindgren authored
Commit d948e6ca ("spi: add power control when set_cs") added generic runtime PM handling, but also changed the return value to be 1 instead of 0 that we had earlier as pm_runtime_get functions return a positve value on success. This causes SPI devices to return errors for cases where they do: ret = spi_setup(spi); if (ret) return ret; As in many cases the SPI devices do not check for if (ret < 0). Let's fix this by setting the status to 0 on succeess after the runtime PM calls. Let's not return 0 at the end of the function as this might break again later on if the function changes and starts returning status again. Fixes: d948e6ca ("spi: add power control when set_cs") Cc: Luhua Xu <luhua.xu@mediatek.com> Cc: wsd_upstream@mediatek.com Signed-off-by:
Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191111195334.44833-1-tony@atomide.comSigned-off-by:
-
- 11 Nov, 2019 3 commits
-
-
Chuhong Yuan authored
pxa2xx_spi_init_pdata misses checks for devm_clk_get and platform_get_irq. Add checks for them to fix the bugs. Since ssp->clk and ssp->irq are used in probe, they are mandatory here. So we cannot use _optional() for devm_clk_get and platform_get_irq. Signed-off-by:
-
Chuhong Yuan authored
This driver forgets to unregister controller when remove. Use devm API to unregister it automatically to fix it. Signed-off-by:
-
Vignesh Raghavendra authored
Both omap2_mcspi_tx_dma() and omap2_mcspi_rx_dma() are only called from omap2_mcspi_txrx_dma() and omap2_mcspi_txrx_dma() is always called after making sure that mcspi_dma->dma_rx and mcspi_dma->dma_tx are not NULL (see omap2_mcspi_transfer_one()). Therefore remove redundant NULL checks for omap2_mcspi->dma_tx and omap2_mcspi->dma_rx pointers in omap2_mcspi_tx_dma() and omap2_mcspi_rx_dma() respectively. Signed-off-by:
-
- 08 Nov, 2019 5 commits
-
-
Miquel Raynal authored
The Zynq QSPI controller features 2 CS. When the num-cs DT property is set to 2, the hardware will be initialized to support having two devices connected over each CS. In this case, both CS lines are driven by the state of the U_PAGE (upper page) bit. When unset, the lower page (CS0) is selected, otherwise it is the upper page (CS1). Change tested on a custom design featuring two SPI-NORs with different CS on the Zynq-7000 QSPI bus. Signed-off-by:
Miquel Raynal <miquel.raynal@bootlin.com> Link: https://lore.kernel.org/r/20191108140744.1734-8-miquel.raynal@bootlin.comSigned-off-by:
-
Miquel Raynal authored
Supporting more than one CS will need some tweaking of the linear configuration register which is (rightfully) initialized in the hardware initialization helper. The extra initialization needs the knowledge of the actual number of CS, which is retrieved by reading the value of the num-cs DT property. As the initialization helper is called pretty early and might be called much later in the probe without side effect, let's delay it a bit so that the number of CS will be available when running this helper. This way, adding support for multiple CS lines in a next patch will be eased. Signed-off-by:
-
Miquel Raynal authored
The code used to assert and de-assert a chip select line is very complicated for no reason. Simplify the logic by either setting or resetting the concerned bit, which actually only changes an electrical state. Update the comment to reflect that there is no possibility to actually choose a CS as the default (CS0) will be driven in any case. Signed-off-by:
-
Miquel Raynal authored
Using masks makes sense when manipulating fields of several bits. When only one bit is involved, it is usual to just use the BIT() macro but in this case using the term mask is abusive. Fix the #define macros and their comments. Signed-off-by:
-
Miquel Raynal authored
Most of the bits/bitfields #define'd in this driver are composed with: 1/ the driver prefix 2/ the name of the register they apply to Keep the naming consistent by applying this rule to the CONFIG register internals. These definitions will be used in a following change set. Signed-off-by:
-
