aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/spi/spi.c
Commit message (Collapse)AuthorAgeFilesLines
...
| * | spi: add devm_spi_optimize_message() helperMark Brown2024-06-251-0/+1
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Merge series from David Lechner <[email protected]>: In the IIO subsystem, we are finding that it is common to call spi_optimize_message() during driver probe since the SPI message doesn't change for the lifetime of the driver. This patch adds a devm_spi_optimize_message() helper to simplify this common pattern.
| | * | spi: add EXPORT_SYMBOL_GPL(devm_spi_optimize_message)David Lechner2024-06-241-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | devm_spi_optimize_message() is a public function and needs EXPORT_SYMBOL_GPL. Reported-by: Jonathan Cameron <[email protected]> Closes: https://lore.kernel.org/linux-iio/20240624204424.6a91a5e4@jic23-huawei/ Fixes: d4a0055fdc22 ("spi: add devm_spi_optimize_message() helper") Signed-off-by: David Lechner <[email protected]> Link: https://patch.msgid.link/20240624-devm_spi_optimize_message-v2-1-58155c0180c2@baylibre.com Signed-off-by: Mark Brown <[email protected]>
| * | | spi: add devm_spi_optimize_message() helperMark Brown2024-06-231-0/+27
| |\| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Merge series from David Lechner <[email protected]>: In the IIO subsystem, we are finding that it is common to call spi_optimize_message() during driver probe since the SPI message doesn't change for the lifetime of the driver. This patch adds a devm_spi_optimize_message() helper to simplify this common pattern.
| | * | spi: add devm_spi_optimize_message() helperDavid Lechner2024-06-221-0/+27
| | |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds a new helper function devm_spi_optimize_message() that automatically registers spi_unoptimize_message() to be called when the device is removed. Signed-off-by: David Lechner <[email protected]> Link: https://patch.msgid.link/20240621-devm_spi_optimize_message-v1-2-3f9dcba6e95e@baylibre.com Signed-off-by: Mark Brown <[email protected]>
| * | spi: Merge up fixesMark Brown2024-06-201-6/+10
| |\ \ | | | | | | | | | | | | We need the fixes to apply new changes to the Cirrus drivers.
| * | | spi: Rework per message DMA mapped flag to be per transferAndy Shevchenko2024-06-101-46/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The granularity of DMA mappings is transfer and moreover, the direction is also important as it can be unidirect. The current cur_msg_mapped flag doesn't fit well the DMA mapping and syncing calls and we have tons of checks around on top of it. So, instead of doing that rework the code to use per transfer per direction flag to show if it's DMA mapped or not. Signed-off-by: Andy Shevchenko <[email protected]> Tested-by: Neil Armstrong <[email protected]> # on SM8650-QRD Link: https://lore.kernel.org/r/[email protected] Reviewed-by: Serge Semin <[email protected]> Tested-by: Nícolas F. R. A. Prado <[email protected]> Signed-off-by: Mark Brown <[email protected]>
| * | | spi: Merge up fixesMark Brown2024-06-031-0/+9
| |\ \ \ | | | |/ | | |/| | | | | We need these to get the i.MX8 boards working in CI again.
| * | | spi: Refactor spi_stop_queue()Andy Shevchenko2024-05-271-15/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The refactoring makes code less verbose and easier to read. Besides that the binary size is also reduced, which sounds like a win-win case: add/remove: 0/1 grow/shrink: 2/2 up/down: 210/-226 (-16) Function old new delta spi_destroy_queue 42 156 +114 spi_controller_suspend 101 197 +96 spi_unregister_controller 346 319 -27 spi_register_controller 1834 1794 -40 spi_stop_queue 159 - -159 Total: Before=49230, After=49214, chg -0.03% Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * | | spi: Replace custom code with device_match_acpi_handle()Andy Shevchenko2024-05-271-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Since driver core provides a generic device_match_acpi_handle() we may replace the custom code with it. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | | | spi: add defer_optimize_message controller flagDavid Lechner2024-07-091-1/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Adding spi_optimize_message() broke the spi-mux driver because it calls spi_async() from it's transfer_one_message() callback. This resulted in passing an incorrectly optimized message to the controller. For example, if the underlying controller has an optimize_message() callback, this would have not been called and can cause a crash when the underlying controller driver tries to transfer the message. Also, since the spi-mux driver swaps out the controller pointer by replacing msg->spi, __spi_unoptimize_message() was being called with a different controller than the one used in __spi_optimize_message(). This could cause a crash when attempting to free the message resources when __spi_unoptimize_message() is called in spi_finalize_current_message() since it is being called with a controller that did not allocate the resources. This is fixed by adding a defer_optimize_message flag for controllers. This flag causes all of the spi_[maybe_][un]optimize_message() calls to be a no-op (other than attaching a pointer to the spi device to the message). This allows the spi-mux driver to pass an unmodified message to spi_async() in spi_mux_transfer_one_message() after the spi device has been swapped out. This causes __spi_optimize_message() and __spi_unoptimize_message() to be called only once per message and with the correct/same controller in each case. Reported-by: Oleksij Rempel <[email protected]> Closes: https://lore.kernel.org/linux-spi/[email protected]/ Reported-by: Marc Kleine-Budde <[email protected]> Closes: https://lore.kernel.org/linux-spi/[email protected]/ Fixes: 7b1d87af14d9 ("spi: add spi_optimize_message() APIs") Signed-off-by: David Lechner <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | | | spi: don't unoptimize message in spi_async()David Lechner2024-07-091-2/+0
| |_|/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Calling spi_maybe_unoptimize_message() in spi_async() is wrong because the message is likely to be in the queue and not transferred yet. This can corrupt the message while it is being used by the controller driver. spi_maybe_unoptimize_message() is already called in the correct place in spi_finalize_current_message() to balance the call to spi_maybe_optimize_message() in spi_async(). Fixes: 7b1d87af14d9 ("spi: add spi_optimize_message() APIs") Signed-off-by: David Lechner <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | | spi: Fix SPI slave probe failureAmit Kumar Mahapatra2024-06-181-4/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While adding a SPI device, the SPI core ensures that multiple logical CS doesn't map to the same physical CS. For example, spi->chip_select[0] != spi->chip_select[1] and so forth. However, unlike the SPI master, the SPI slave doesn't have the list of chip selects, this leads to probe failure when the SPI controller is configured as slave. Update the __spi_add_device() function to perform this check only if the SPI controller is configured as master. Fixes: 4d8ff6b0991d ("spi: Add multi-cs memories support in SPI core") Signed-off-by: Amit Kumar Mahapatra <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | | spi: Fix OCTAL mode supportPatrice Chotard2024-06-181-2/+4
| |/ |/| | | | | | | | | | | | | | | Add OCTAL mode support. Issue detected using "--octal" spidev_test's option. Signed-off-by: Patrice Chotard <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: Assign dummy scatterlist to unidirectional transfersAndy Shevchenko2024-05-291-0/+9
|/ | | | | | | | | | | | | | | | | | | | | | | | | | Commit 8cc3bad9d9d6 ("spi: Remove unneded check for orig_nents") introduced a regression: unmapped data could now be passed to the DMA APIs, resulting in null pointer dereferences. Commit 9f788ba457b4 ("spi: Don't mark message DMA mapped when no transfer in it is") and commit da560097c056 ("spi: Check if transfer is mapped before calling DMA sync APIs") addressed the problem, but only partially. Unidirectional transactions will still result in null pointer dereference. To prevent that from happening, assign a dummy scatterlist when no data is mapped, so that the DMA API can be called and not result in a null pointer dereference. Signed-off-by: Andy Shevchenko <[email protected]> Reported-by: Neil Armstrong <[email protected]> Closes: https://lore.kernel.org/r/[email protected] Reported-by: Nícolas F. R. A. Prado <[email protected]> Closes: https://lore.kernel.org/all/d3679496-2e4e-4a7c-97ed-f193bd53af1d@notapiano Closes: https://lore.kernel.org/all/4748499f-789c-45a8-b50a-2dd09f4bac8c@notapiano Fixes: 8cc3bad9d9d6 ("spi: Remove unneded check for orig_nents") Tested-by: Nícolas F. R. A. Prado <[email protected]> [nfraprado: wrote the commit message] Signed-off-by: Nícolas F. R. A. Prado <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* soi: Don't call DMA sync API when not neededMark Brown2024-05-231-73/+75
|\ | | | | | | | | | | | | | | | | Merge series from Andy Shevchenko <[email protected]>: A couple of fixes to avoid calling DMA sync API when it's not needed. This doesn't stop from discussing if IOMMU code is doing the right thing, i.e. dereferences SG list when orig_nents == 0, but this is a separate story.
| * spi: Check if transfer is mapped before calling DMA sync APIsAndy Shevchenko2024-05-231-5/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The resent update to remove the orig_nents checks revealed that not all DMA sync backends can cope with the unallocated SG list, while supplying orig_nents == 0 (the commit 861370f49ce4 ("iommu/dma: force bouncing if the size is not cacheline-aligned"), for example, makes that happen for the IOMMU case). It means we have to check if the buffers are DMA mapped before trying to sync them. Re-introduce that check in a form of calling ->can_dma() in the same way as it's done in the DMA mapping loop for the SPI transfers. Reported-by: Nícolas F. R. A. Prado <[email protected]> Reported-by: Neil Armstrong <[email protected]> Closes: https://lore.kernel.org/r/[email protected] Closes: https://lore.kernel.org/all/d3679496-2e4e-4a7c-97ed-f193bd53af1d@notapiano Fixes: 8cc3bad9d9d6 ("spi: Remove unneded check for orig_nents") Suggested-by: Nícolas F. R. A. Prado <[email protected]> Tested-by: Nícolas F. R. A. Prado <[email protected]> Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * spi: Don't mark message DMA mapped when no transfer in it isAndy Shevchenko2024-05-231-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | There is no need to set the DMA mapped flag of the message if it has no mapped transfers. Moreover, it may give the code a chance to take the wrong paths, i.e. to exercise DMA related APIs on unmapped data. Make __spi_map_msg() to bail earlier on the above mentioned cases. Fixes: 99adef310f68 ("spi: Provide core support for DMA mapping transfers") Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * spi: Remove unneded check for orig_nentsAndy Shevchenko2024-05-081-14/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | Both dma_unmap_sgtable() and sg_free_table() in spi_unmap_buf_attrs() have checks for orig_nents against 0. No need to duplicate this. All the same applies to other DMA mapping API calls. Also note, there is no other user in the kernel that does this kind of checks. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * spi: use spi_valid_{tx,rx}buf() in stats functionDavid Lechner2024-05-011-7/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are macros spi_valid_txbuf() and spi_valid_rxbuf() for determining if an xfer actually intended to send or receive data. These checks were hard-coded in spi_statistics_add_transfer_stats(). We can make use of the macros instead to make the code more readable and more robust against potential future changes in case the definition of what valid means changes. The macro takes the spi_message as an argument, so we need to change spi_statistics_add_transfer_stats() to take the spi_message as an argument instead of the controller. Signed-off-by: David Lechner <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * Add bridged amplifiers to cs42l43Mark Brown2024-04-171-3/+9
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Merge series from Charles Keepax <[email protected]>: In some cs42l43 systems a couple of cs35l56 amplifiers are attached to the cs42l43's SPI and I2S. On Windows the cs42l43 is controlled by a SDCA class driver and these two amplifiers are controlled by firmware running on the cs42l43. However, under Linux the decision was made to interact with the cs42l43 directly, affording the user greater control over the audio system. However, this has resulted in an issue where these two bridged cs35l56 amplifiers are not populated in ACPI and must be added manually. There is at least an SDCA extension unit DT entry we can key off. The process of adding this is handled using a software node, firstly the ability to add native chip selects to software nodes must be added. Secondly, an additional flag for naming the SPI devices is added this allows the machine driver to key to the correct amplifier. Then finally, the cs42l43 SPI driver adds the two amplifiers directly onto its SPI bus. An additional series will follow soon to add the audio machine driver parts (in the sof-sdw driver), however that is fairly orthogonal to this part of the process, getting the actual amplifiers registered.
| | * spi: Update swnode based SPI devices to use the fwnode nameCharles Keepax2024-04-161-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Update the name for software node based SPI devices to use the fwnode name as the device name. This is helpful since swnode devices are usually added within the kernel, and the kernel often then requires a predictable name such that it can refer back to the device. Reviewed-by: Andy Shevchenko <[email protected]> Signed-off-by: Charles Keepax <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| | * spi: Switch to using is_acpi_device_node() in spi_dev_set_name()Charles Keepax2024-04-161-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use is_acpi_device_node() rather than checking ACPI_COMPANION(), such that when checking for other types of firmware node, the code can consistently do checks against the fwnode. Suggested-by: Andy Shevchenko <[email protected]> Reviewed-by: Andy Shevchenko <[email protected]> Signed-off-by: Charles Keepax <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * | spi: More refacroings after multi-CS supportMark Brown2024-04-161-30/+33
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Merge series from Andy Shevchenko <[email protected]>: A couple of additional refactorings on top of the multi-CS support. One is to make sure that the comment and the code are not disrupted if additional changes come in the future and second one is f or the sake of deduplication. In both cases it also makes indentation level smaller in the affected pieces of the code. No functional changes intended.
| | * | spi: Introduce spi_for_each_valid_cs() in order of deduplicationAndy Shevchenko2024-04-151-7/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In order of deduplication and better maintenance introduce a new spi_for_each_valid_cs() helper macro. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| | * | spi: Extract spi_toggle_csgpiod() helper for better maintananceAndy Shevchenko2024-04-151-24/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The multi-CS support splits the comment and the code in the spi_set_cs(). To avoid this in the future extract spi_toggle_csgpiod() helper. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * | | spi: Consistently use BIT for cs_index_mask (part 2)Andy Shevchenko2024-04-151-7/+3
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | For some reason the commit 1209c5566f9b ("spi: Consistently use BIT for cs_index_mask") missed one place to change, do it here to finish the job. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
| * / spi: remove struct spi_message::is_dma_mappedDavid Lechner2024-03-251-7/+0
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are no more users of the deprecated is_dma_mapped in struct spi_message so it can be removed. References in documentation and comments are also removed. A few similar checks if xfer->tx_dma or xfer->rx_dma are not NULL are also removed since these are now guaranteed to be NULL because they were previously set only if is_dma_mapped was true. Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* / spi: fix null pointer dereference within spi_syncMans Rullgard2024-05-011-0/+1
|/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If spi_sync() is called with the non-empty queue and the same spi_message is then reused, the complete callback for the message remains set while the context is cleared, leading to a null pointer dereference when the callback is invoked from spi_finalize_current_message(). With function inlining disabled, the call stack might look like this: _raw_spin_lock_irqsave from complete_with_flags+0x18/0x58 complete_with_flags from spi_complete+0x8/0xc spi_complete from spi_finalize_current_message+0xec/0x184 spi_finalize_current_message from spi_transfer_one_message+0x2a8/0x474 spi_transfer_one_message from __spi_pump_transfer_message+0x104/0x230 __spi_pump_transfer_message from __spi_transfer_message_noqueue+0x30/0xc4 __spi_transfer_message_noqueue from __spi_sync+0x204/0x248 __spi_sync from spi_sync+0x24/0x3c spi_sync from mcp251xfd_regmap_crc_read+0x124/0x28c [mcp251xfd] mcp251xfd_regmap_crc_read [mcp251xfd] from _regmap_raw_read+0xf8/0x154 _regmap_raw_read from _regmap_bus_read+0x44/0x70 _regmap_bus_read from _regmap_read+0x60/0xd8 _regmap_read from regmap_read+0x3c/0x5c regmap_read from mcp251xfd_alloc_can_err_skb+0x1c/0x54 [mcp251xfd] mcp251xfd_alloc_can_err_skb [mcp251xfd] from mcp251xfd_irq+0x194/0xe70 [mcp251xfd] mcp251xfd_irq [mcp251xfd] from irq_thread_fn+0x1c/0x78 irq_thread_fn from irq_thread+0x118/0x1f4 irq_thread from kthread+0xd8/0xf4 kthread from ret_from_fork+0x14/0x28 Fix this by also setting message->complete to NULL when the transfer is complete. Fixes: ae7d2346dc89 ("spi: Don't use the message queue if possible in spi_sync") Signed-off-by: Mans Rullgard <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Restore delays for non-GPIO chip selectJanne Grunau2024-03-121-10/+14
| | | | | | | | | | | | | | SPI controller with integrated chip select handling still need to adhere to SPI device's CS setup, hold and inactive delays. For controller without set_cs_timing spi core shall handle the delays to avoid duplicated delay handling in each controller driver. Fixes a regression for the out of tree SPI controller and SPI HID transport on Apple M1/M1 Pro/Max notebooks. Fixes: 4d8ff6b0991d ("spi: Add multi-cs memories support in SPI core") Signed-off-by: Janne Grunau <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Introduce SPI_INVALID_CS and is_valid_cs()Andy Shevchenko2024-03-071-16/+23
| | | | | | | | | | The SPI core inconsistently uses the marker value for unused chip select pin. Define a constant (with appropriate type) and introduce is_valid_cs() helper function to avoid spreading this inconsistency in the future. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Consistently use BIT for cs_index_maskAndy Shevchenko2024-03-071-24/+11
| | | | | | | | | | | | Some of the parts related to the chip select are using BIT() macro the rest are using plain numbers. Unify all of them to use BIT(). While at it, make the (repetitive) comment clearer when assigning cs_index_mask during SPI target device enumeration. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Exctract spi_dev_check_cs() helperAndy Shevchenko2024-03-061-20/+27
| | | | | | | | | | It seems a few functions implement the similar for-loop to validate chip select pins for uniqueness. Let's deduplicate that code in order to have a single place of that for better maintenance. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Exctract spi_set_all_cs_unused() helperAndy Shevchenko2024-03-061-50/+24
| | | | | | | | | | It seems a few functions implement the similar for-loop to mark all chip select pins unused. Let's deduplicate that code in order to have a single place of that for better maintenance. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: move splitting transfers to spi_optimize_message()David Lechner2024-02-261-42/+68
| | | | | | | | | | | | | | | | | | | Splitting transfers is an expensive operation so we can potentially optimize it by doing it only once per optimization of the message instead of repeating each time the message is transferred. The transfer splitting functions are currently the only user of spi_res_alloc() so spi_res_release() can be safely moved at this time from spi_finalize_current_message() to spi_unoptimize_message(). The doc comments of the public functions for splitting transfers are also updated so that callers will know when it is safe to call them to ensure proper resource management. Reviewed-by: Jonathan Cameron <[email protected]> Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/20240219-mainline-spi-precook-message-v2-2-4a762c6701b9@baylibre.com Signed-off-by: Mark Brown <[email protected]>
* spi: add spi_optimize_message() APIsDavid Lechner2024-02-261-4/+147
| | | | | | | | | | | | | | | | | | | | | | | | This adds a new spi_optimize_message() function that can be used to optimize SPI messages that are used more than once. Peripheral drivers that use the same message multiple times can use this API to perform SPI message validation and controller-specific optimizations once and then reuse the message while avoiding the overhead of revalidating the message on each spi_(a)sync() call. Internally, the SPI core will also call this function for each message if the peripheral driver did not explicitly call it. This is done to so that controller drivers don't have to have multiple code paths for optimized and non-optimized messages. A hook is provided for controller drivers to perform controller-specific optimizations. Suggested-by: Martin Sperl <[email protected]> Link: https://lore.kernel.org/linux-spi/[email protected]/ Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/20240219-mainline-spi-precook-message-v2-1-4a762c6701b9@baylibre.com Reviewed-by: Jonathan Cameron <[email protected]> Signed-off-by: Mark Brown <[email protected]>
* spi: get rid of some legacy macrosMark Brown2024-02-081-1/+5
|\ | | | | | | | | | | | | Merge series from Uwe Kleine-König <[email protected]>: This series finishes off the removal of some of the legacy names for SPI controllers and devices.
| * spi: Drop compat layer from renaming "master" to "controller"Uwe Kleine-König2024-02-081-1/+1
| | | | | | | | | | | | | | | | | | | | | | Now that all in-tree users followed the rename, the compat stuff can go away. This completes the renaming started with commit 8caab75fd2c2 ("spi: Generalize SPI "master" to "controller"") Acked-by: Jonathan Cameron <[email protected]> Signed-off-by: Uwe Kleine-König <[email protected]> Link: https://lore.kernel.org/r/ad1d949325b61a4682e8d6ecf9d05da751e6a99f.1707324794.git.u.kleine-koenig@pengutronix.de Signed-off-by: Mark Brown <[email protected]>
| * spi: fix finalize message on error returnDavid Lechner2024-01-251-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | In __spi_pump_transfer_message(), the message was not finalized in the first error return as it is in the other error return paths. Not finalizing the message could cause anything waiting on the message to complete to hang forever. This adds the missing call to spi_finalize_current_message(). Fixes: ae7d2346dc89 ("spi: Don't use the message queue if possible in spi_sync") Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: drop gpf arg from __spi_split_transfer_maxsize()David Lechner2024-02-071-14/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The __spi_split_transfer_maxsize() function has a gpf argument to allow callers to specify the type of memory allocation that needs to be used. However, this function only allocates struct spi_transfer and is not intended to be used from atomic contexts so this type should always be GFP_KERNEL, so we can just drop the argument. Some callers of these functions also passed GFP_DMA, but since only struct spi_transfer is allocated and not any tx/rx buffers, this is not actually necessary and is removed in this commit. Signed-off-by: David Lechner <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: move split xfers for CS_WORD emulationDavid Lechner2024-02-051-32/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This moves splitting transfers for CS_WORD software emulation to the same place where we split transfers for controller-specific reasons. This fixes a few subtle bugs. The calculation for maxsize was wrong for bit sizes between 17 and 24. This is fixed by making use of spi_split_transfers_maxwords() which already has the correct calculation. Also, since this indirectly calls spi_res_alloc(), to avoid leaking resources, spi_finalize_current_message() would need to be called on all error paths in __spi_validate() and callers of __spi_validate() would need to do the same. This is fixed by moving the call to __spi_pump_transfer_message() where it is already splitting transfers for other reasons and correctly releases resources in the subsequent error paths. Fixes: cbaa62e0094a ("spi: add software implementation for SPI_CS_WORD") Signed-off-by: David Lechner <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: avoid double validation in __spi_sync()David Lechner2024-01-261-52/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The __spi_sync() function calls __spi_validate() early in the function. Later, it can call spi_async_locked() which calls __spi_validate() again. __spi_validate() is an expensive function, so we can improve performance measurably by avoiding calling it twice. Instead of calling spi_async_locked(), we can call __spi_async() with the spin lock held. spi_async_locked() is removed since there are no more callers. Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: consolidate setting message->spiDavid Lechner2024-01-241-7/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously, __spi_sync() and __spi_async() set message->spi to the spi device independently after calling __spi_validate(). __spi_validate() also would conditionally set this if it needed to split the message since it wasn't set yet. Since both __spi_sync() and __spi_async() call __spi_validate(), we can consolidate this into only setting message->spi once (unconditionally) in __spi_validate(). This will also save any future callers of __spi_validate() from also needing to set message->spi. Signed-off-by: David Lechner <[email protected]> Link: https://msgid.link/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: make spi_bus_type constGreg Kroah-Hartman2024-01-221-1/+1
|/ | | | | | | | | | | | Now that the driver core can properly handle constant struct bus_type, move the spi_bus_type variable to be a constant structure as well, placing it into read-only memory which can not be modified at runtime. Cc: Mark Brown <[email protected]> Cc: <[email protected]> Signed-off-by: Greg Kroah-Hartman <[email protected]> Link: https://msgid.link/r/2024010549-erasure-swoop-1cc6@gregkh Signed-off-by: Mark Brown <[email protected]>
* spi: Add support for stacked/parallel memoriesMark Brown2023-12-071-41/+216
|\ | | | | | | | | | | | | Merge series from Amit Kumar Mahapatra <[email protected]>: This patch series adds support to the SPI framework for using multiple chip selects.
| * spi: Add multi-cs memories support in SPI coreAmit Kumar Mahapatra2023-12-071-41/+216
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | AMD-Xilinx GQSPI controller has two advanced mode that allows the controller to consider two flashes as one single device. One of these two mode is the parallel mode in which each byte of data is stored in both devices, the even bits in the lower flash & the odd bits in the upper flash. The byte split is automatically handled by the QSPI controller. The other mode is the stacked mode in which both the flashes share the same SPI bus but each of the device contain half of the data. In this mode, the controller does not follow CS requests but instead internally wires the two CS levels with the value of the most significant address bit. For supporting both these modes SPI core need to be updated for providing multiple CS for a single SPI device. For adding multi CS support the SPI device need to be aware of all the CS values. So, the "chip_select" member in the spi_device structure is now an array that holds all the CS values. spi_device structure now has a "cs_index_mask" member. This acts as an index to the chip_select array. If nth bit of spi->cs_index_mask is set then the driver would assert spi->chip_select[n]. In parallel mode all the chip selects are asserted/de-asserted simultaneously and each byte of data is stored in both devices, the even bits in one, the odd bits in the other. The split is automatically handled by the GQSPI controller. The GQSPI controller supports a maximum of two flashes connected in parallel mode. A SPI_CONTROLLER_MULTI_CS flag bit is added in the spi controller flags, through ctlr->flags the spi core will make sure that the controller is capable of handling multiple chip selects at once. For supporting multiple CS via GPIO the cs_gpiod member of the spi_device structure is now an array that holds the gpio descriptor for each chipselect. CS GPIO is not tested on our hardware, but it has been tested by @Stefan https://lore.kernel.org/all/[email protected]/ Signed-off-by: Amit Kumar Mahapatra <[email protected]> Tested-by: Stefan Binding <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* | spi: introduce SPI_TRANS_FAIL_IO for error reportingNam Cao2023-11-301-0/+3
|/ | | | | | | | | | | | | | The default message transfer implementation - spi_transfer_one_message - invokes the specific device driver's transfer_one(), then waits for completion. However, there is no mechanism for the device driver to report failure in the middle of the transfer. Introduce SPI_TRANS_FAIL_IO for drivers to report transfer failure. Signed-off-by: Nam Cao <[email protected]> Acked-by: Linus Walleij <[email protected]> Link: https://lore.kernel.org/r/4b420dac528e60f122adde16851da88e4798c1ea.1701274975.git.namcao@linutronix.de Signed-off-by: Mark Brown <[email protected]>
* spi: Fix null dereference on suspendMark Hasemeyer2023-11-101-17/+39
| | | | | | | | | | | | | | | | | | | | | | | | | A race condition exists where a synchronous (noqueue) transfer can be active during a system suspend. This can cause a null pointer dereference exception to occur when the system resumes. Example order of events leading to the exception: 1. spi_sync() calls __spi_transfer_message_noqueue() which sets ctlr->cur_msg 2. Spi transfer begins via spi_transfer_one_message() 3. System is suspended interrupting the transfer context 4. System is resumed 6. spi_controller_resume() calls spi_start_queue() which resets cur_msg to NULL 7. Spi transfer context resumes and spi_finalize_current_message() is called which dereferences cur_msg (which is now NULL) Wait for synchronous transfers to complete before suspending by acquiring the bus mutex and setting/checking a suspend flag. Signed-off-by: Mark Hasemeyer <[email protected]> Link: https://lore.kernel.org/r/20231107144743.v1.1.I7987f05f61901f567f7661763646cb7d7919b528@changeid Signed-off-by: Mark Brown <[email protected]> Cc: [email protected]
* spi: Export acpi_spi_find_controller_by_adev()Hans de Goede2023-10-161-3/+2
| | | | | | | | | | | Export acpi_spi_find_controller_by_adev() so that ACPI glue code which wants to dynamically create a spi_device using acpi_spi_device_alloc() or spi_new_device() on a controller, to which the code does not already have a reference, can find the controller. Signed-off-by: Hans de Goede <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Drop warning from spi_stop_queue()Uwe Kleine-König2023-09-181-4/+0
| | | | | | | | | | | Both callers of spi_stop_queue() (i.e. spi_destroy_queue() and spi_controller_suspend()) already emit an error message if spi_stop_queue() fails. Another warning in this case isn't helpful, so drop it. Signed-off-by: Uwe Kleine-König <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
* spi: Use BITS_TO_BYTES()Andy Shevchenko2023-07-141-3/+1
| | | | | | | | | BITS_TO_BYTES() is the existing macro which takes care about full bytes that may fully hold the given amount of bits. Use it. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Mark Brown <[email protected]>
generated by cgit (git 2.51.2) at 2025-12-04 04:35:39 +0000