lwn.git/drivers/i3c/master/i3c-master-cdns.c, branch docs-mw

i3c: master: Move entdaa error suppression

2026-04-12T14:54:37+00:00

Prepare to fix improper Mx positive error propagation in later commits by handling Mx error codes where the i3c_ccc_cmd command is allocated. The CCC ENTDAA is invoked with i3c_master_entdaa_locked() and yields error I3C_ERROR_M2 if there are no devices active on the bus. Some controllers may also yield if there are no more devices need an dynamic address, since the sequence do always end in a NACK. Handle inside i3c_master_entdaa_locked(), checking cmd->err directly. Both call sites are updated, adi_i3c_master_do_daa() and cdns_i3c_master_do_daa(). Reviewed-by: Frank Li Reviewed-by: Adrian Hunter Signed-off-by: Jorge Marques Link: https://patch.msgid.link/20260323-ad4062-positive-error-fix-v3-2-30bdc68004be@analog.com Signed-off-by: Alexandre Belloni

Convert 'alloc_flex' family to use the new default GFP_KERNEL argument

2026-02-22T01:09:51+00:00

This is the exact same thing as the 'alloc_obj()' version, only much smaller because there are a lot fewer users of the *alloc_flex() interface. As with alloc_obj() version, this was done entirely with mindless brute force, using the same script, except using 'flex' in the pattern rather than 'objs*'. Signed-off-by: Linus Torvalds

Convert 'alloc_obj' family to use the new default GFP_KERNEL argument

2026-02-22T01:09:51+00:00

This was done entirely with mindless brute force, using git grep -l '\

treewide: Replace kmalloc with kmalloc_obj for non-scalar types

2026-02-21T09:02:28+00:00

This is the result of running the Coccinelle script from scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to avoid scalar types (which need careful case-by-case checking), and instead replace kmalloc-family calls that allocate struct or union object instances: Single allocations: kmalloc(sizeof(TYPE), ...) are replaced with: kmalloc_obj(TYPE, ...) Array allocations: kmalloc_array(COUNT, sizeof(TYPE), ...) are replaced with: kmalloc_objs(TYPE, COUNT, ...) Flex array allocations: kmalloc(struct_size(PTR, FAM, COUNT), ...) are replaced with: kmalloc_flex(*PTR, FAM, COUNT, ...) (where TYPE may also be *VAR) The resulting allocations no longer return "void *", instead returning "TYPE *". Signed-off-by: Kees Cook

i3c: master: switch to use new callback .i3c_xfers() from .priv_xfers()

2025-12-08T02:34:17+00:00

Switch to use the new .i3c_xfers() callback, which supports all I3C transfer modes (SDR and HDR). Also replace struct i3c_priv_xfer with the new struct i3c_xfer, as i3c_priv_xfer is now an alias of i3c_xfer. No functional changes. Signed-off-by: Frank Li Tested-by: Tommaso Merciai Link: https://patch.msgid.link/20251203-i3c_xfer_cleanup_master-v2-1-7dd94d04ee2d@nxp.com Signed-off-by: Alexandre Belloni

i3c: master: cdns: Simplify handling clocks in probe()

2025-07-30T23:24:25+00:00

The two clocks, driver is getting, are not being disabled/re-enabled during runtime of the device. Eliminate one variable in state struct, all error paths and a lot of code from probe() and remove() by using devm_clk_get_enabled(). Signed-off-by: Krzysztof Kozlowski Reviewed-by: Frank Li Link: https://lore.kernel.org/r/20250713152411.74917-2-krzysztof.kozlowski@linaro.org Signed-off-by: Alexandre Belloni

i3c: master: cdns: Use i3c_writel_fifo() and i3c_readl_fifo()

2025-07-30T23:21:10+00:00

Use common inline i3c_writel_fifo()/i3c_readl_fifo() methods to simplify code since the FIFO of controller is a 32bit width. Signed-off-by: Jorge Marques Reviewed-by: Frank Li Link: https://lore.kernel.org/r/20250624-i3c-writesl-readsl-v3-2-63ccf0870f01@analog.com Signed-off-by: Alexandre Belloni

i3c: master: cdns: replace ENOTSUPP with SUSV4-compliant EOPNOTSUPP

2025-07-07T21:14:46+00:00

Replace non-standard ENOTSUPP with the SUSV4-defined error code EOPNOTSUPP to fix below checkpatch warning: "ENOTSUPP is not a SUSV4 error code, prefer EOPNOTSUPP" Signed-off-by: Wolfram Sang Reviewed-by: Frank Li Link: https://lore.kernel.org/r/20250628192027.3932-8-wsa+renesas@sang-engineering.com Signed-off-by: Alexandre Belloni

i3c: master: cdns: use adapter timeout value for I2C transfers

2025-07-03T23:10:07+00:00

I2C adapters have their own timeout value which can be changed by userspace if desired. Use it for I2C transfers. The default is 1Hz, so the default behaviour is unchanged. Signed-off-by: Wolfram Sang Reviewed-by: Frank Li Link: https://lore.kernel.org/r/20250604101831.56585-3-wsa+renesas@sang-engineering.com Signed-off-by: Alexandre Belloni

i3c: Remove the const qualifier from i2c_msg pointer in i2c_xfers API

2025-02-20T21:54:28+00:00

The change is necessary to enable the use of the `i2c_get_dma_safe_msg_buf()` API, which requires a non-const `struct i2c_msg *` to operate. The `i2c_get_dma_safe_msg_buf()` function ensures safe handling of I2C messages when using DMA, making it essential for scenarios where DMA transfers are involved. By removing the `const` qualifier, this patch allows drivers to prepare and manage DMA-safe buffers directly. Signed-off-by: Billy Tsai Reviewed-by: Frank Li Reviewed-by: Wolfram Sang Acked-by: Mukesh Kumar Savaliya Link: https://lore.kernel.org/r/20250204091702.4014466-1-billy_tsai@aspeedtech.com Signed-off-by: Alexandre Belloni