regmap: mmio: Add missing MODULE_DESCRIPTION()

2025-08-10T20:09:27Z

There were already several commits to add module descriptions to regmap modules. But this one was still missing: WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/base/regmap/regmap-mmio.o Signed-off-by: Luis Henriques Link: https://patch.msgid.link/20250728150829.11890-1-luis@igalia.com Signed-off-by: Mark Brown

regmap: mmio: Remove unused 64-bit support code

2023-07-09T21:48:46Z

regmap API does not support 64-bit data size, so there is no point to have it in regmap MMIO. Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20230622183613.58762-4-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown

regmap: mmio: Allow passing an empty config->reg_stride

2023-05-12T01:28:19Z

Regmap's stride is used for MMIO regmaps to check the correctness of reg_width. However, it's acceptable to pass an empty config->reg_stride, in that case the actual stride used is 1. There are valid cases now to pass an empty stride, when using down/upshifting of register address. In this case, the stride value loses its sense, so ignore the reg_width when the stride isn't set. Signed-off-by: Maxime Chevallier

regmap: mmio: replace return 0 with break in switch statement

2022-09-22T11:33:18Z

Variable min_stride is assigned a value that is never read, fix this by replacing the return 0 with a break statement. This also makes the case statement consistent with the other cases in the switch statement. Signed-off-by: Colin Ian King Link: https://lore.kernel.org/r/20220922080445.818020-1-colin.i.king@gmail.com Signed-off-by: Mark Brown

regmap: mmio: Use swabXX_array() helpers

2022-09-07T11:42:26Z

Since we have a few helpers to swab elements of a given size in an array use them instead of open coded variants. Signed-off-by: Andy Shevchenko Reviewed-by: Linus Walleij Link: https://lore.kernel.org/r/20220831212744.56435-2-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown

regmap: mmio: Fix rebase error

2022-09-01T10:56:26Z

A dangling pointless "ret 0" was left in and some unneeded whitespace can go too. Fixes: 81c0386c1376 ("regmap: mmio: Support accelerared noinc operations") Reported-by: Andy Shevchenko Signed-off-by: Linus Walleij Link: https://lore.kernel.org/r/20220831141303.501548-1-linus.walleij@linaro.org Signed-off-by: Mark Brown

regmap: mmio: Support accelerared noinc operations

2022-08-17T12:00:32Z

Use the newly added callback for accelerated noinc MMIO to provide writesb, writesw, writesl, writesq, readsb, readsw, readsl and readsq. A special quirk is needed to deal with big endian regmaps: there are no accelerated operations defined for big endian, so fall back to calling the big endian operations itereatively for this case. The Hexagon architecture turns out to have an incomplete : writesb() is not implemented. Fix this by doing what other architectures do: include into the file. Cc: Brian Cain Cc: linux-hexagon@vger.kernel.org Cc: Arnd Bergmann Signed-off-by: Linus Walleij Link: https://lore.kernel.org/r/20220816204832.265837-2-linus.walleij@linaro.org Signed-off-by: Mark Brown

regmap: mmio: Fix MMIO accessors to avoid talking to IO port

2022-08-15T00:20:11Z

Currently regmap MMIO is inconsistent with IO accessors. I.e. the Big Endian counterparts are using ioreadXXbe() / iowriteXXbe() which are not clean implementations of readXXbe(). That said, reimplement current Big Endian MMIO accessors by replacing ioread()/iowrite() with respective read()/write() and swab() calls. Note, there are no current in-kernel users that may utilize the functionality of the IO ports on Big Endian hardware. All drivers that use regmap MMIO either Little Endian, or they don't map IO ports in a way that ioreadXX()/iowriteXX() may be utilized. Signed-off-by: Andy Shevchenko Acked-by: William Breathitt Gray Link: https://lore.kernel.org/r/20220808203401.35153-5-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown

regmap: mmio: Introduce IO accessors that can talk to IO port

2022-08-15T00:20:10Z

Some users may use regmap MMIO for IO ports, and this can be done by assigning ioreadXX()/iowriteXX() and their Big Endian counterparts to the regmap context. Add IO port support with a corresponding flag added. While doing that, make sure that user won't select relaxed MMIO access along with IO port because the latter have no relaxed variants. Signed-off-by: Andy Shevchenko Acked-by: William Breathitt Gray Link: https://lore.kernel.org/r/20220808203401.35153-4-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown

regmap: mmio: Get rid of broken 64-bit IO

2022-08-15T00:20:09Z

The current implementation, besides having no active users, is broken by design of regmap. For 64-bit IO we need to supply 64-bit value, otherwise there is no way to handle upper 32 bits in 64-bit register. Hence, remove the broken IO accessors for good and wait for real user that can fix entire regmap API for that. Signed-off-by: Andy Shevchenko Acked-by: William Breathitt Gray Link: https://lore.kernel.org/r/20220808203401.35153-3-andriy.shevchenko@linux.intel.com Signed-off-by: Mark Brown

kernel/drivers/base/regmap/regmap-mmio.c, branch linux-rolling-stable