kernel/linux.git/drivers/mfd/sec-common.c, branch v6.19.11

mfd: sec: Add myself as module author

2025-05-23T07:48:50+00:00

Add myself as module author, so people know whom to complain to about after the recent updates. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-23-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec-common: Convert to using MFD_CELL macros

2025-05-23T07:48:48+00:00

Use MFD_CELL macro helpers instead of open coding. This makes the code a bit shorter and more obvious. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-21-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec-common: Use sizeof(*var), not sizeof(struct type_of_var)

2025-05-23T07:48:47+00:00

Using sizeof(*var) is generally preferred over using the size of its open-coded type when allocating memory. This helps avoiding bugs when the variable type changes but the memory allocation isn't updated, and it simplifies renaming of the struct if ever necessary. No functional change. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-20-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec: Don't compare against NULL / 0 for errors, use !

2025-05-23T07:48:45+00:00

Follow general style and use if (!arg) instead of comparing against NULL. While at it, drop a useless init in sec-irq.c. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-19-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec: Change device_type to int

2025-05-23T07:48:44+00:00

Now that sec-i2c doesn't match device type by pointer casting anymore, we can switch the device type from unsigned long to int easily. This saves a few bytes in struct sec_pmic_dev due to member alignment. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-18-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec-common: Don't ignore errors from sec_irq_init()

2025-05-23T07:48:41+00:00

sec_irq_init() can fail, we shouldn't continue and ignore the error in that case, but actually error out. Reviewed-by: Krzysztof Kozlowski Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-16-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec: Use dev_err_probe() where appropriate

2025-05-23T07:48:38+00:00

dev_err_probe() exists to simplify code and harmonise error messages, there's no reason not to use it here. While at it, harmonise some error messages. Reviewed-by: Krzysztof Kozlowski Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-13-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec-common: Fix multiple trivial whitespace issues

2025-05-23T07:48:36+00:00

Rectify a couple of alignment problems reported by Checkpatch. Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-11-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones

mfd: sec: Merge separate core and irq modules

2025-05-23T07:48:34+00:00

There is no reason to have these two kernel modules separate. Having them merged into one kernel module also slightly reduces memory consumption and module load times a little. mapped size (lsmod): before: after: sec_core 20480 sec_core 24576 sec_irq 16384 ---------------- total 36864 Section sizes (size -A): before: after: sec_core 6780 sec_core 13239 sec_irq 8046 ---------------- Total 14826 Reviewed-by: Krzysztof Kozlowski Signed-off-by: André Draszik Link: https://lore.kernel.org/r/20250409-s2mpg10-v4-10-d66d5f39b6bf@linaro.org Signed-off-by: Lee Jones