kernel.git/drivers/remoteproc/imx_rproc.c, branch master

remoteproc: imx_rproc: Fix unreachable platform prepare_ops

2026-03-05T17:18:23Z

Smatch reports unreachable code in imx_rproc_prepare(), where an early return inside the reserved-memory parsing loop prevents platform prepare_ops from being executed. When of_reserved_mem_region_to_resource() fails, imx_rproc_prepare() returns immediately, so the platform-specific prepare callback is never called. As a result, prepare_ops such as imx_rproc_sm_lmm_prepare() on i.MX95 have no chance to run. This is problematic when Linux controls the M7 Logical Machine and is responsible for preparing resources such as TCM. Without running the platform prepare callback, loading the M7 ELF into TCM may fail if the bootloader did not power up and initialize TCM. Fix this by breaking out of the reserved-memory loop instead of returning, allowing the platform prepare_ops to be executed as intended. Fixes: edd2a9956055 ("remoteproc: imx_rproc: Introduce prepare ops for imx_rproc_dcfg") Reported-by: Dan Carpenter Closes: https://lore.kernel.org/linux-remoteproc/aYYXAa2Fj36XG4yQ@p14s/T/#t Signed-off-by: Peng Fan Reviewed-by: Daniel Baluta Link: https://lore.kernel.org/r/20260208-imx-rproc-fix-v1-1-ad74555eb9a4@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Fix invalid loaded resource table detection

2026-02-03T16:23:12Z

imx_rproc_elf_find_loaded_rsc_table() may incorrectly report a loaded resource table even when the current firmware does not provide one. When the device tree contains a "rsc-table" entry, priv->rsc_table is non-NULL and denotes where a resource table would be located if one is present in memory. However, when the current firmware has no resource table, rproc->table_ptr is NULL. The function still returns priv->rsc_table, and the remoteproc core interprets this as a valid loaded resource table. Fix this by returning NULL from imx_rproc_elf_find_loaded_rsc_table() when there is no resource table for the current firmware (i.e. when rproc->table_ptr is NULL). This aligns the function's semantics with the remoteproc core: a loaded resource table is only reported when a valid table_ptr exists. With this change, starting firmware without a resource table no longer triggers a crash. Fixes: e954a1bd1610 ("remoteproc: imx_rproc: Use imx specific hook for find_loaded_rsc_table") Cc: stable@vger.kernel.org Signed-off-by: Peng Fan Acked-by: Daniel Baluta Link: https://lore.kernel.org/r/20260129-imx-rproc-fix-v3-1-fc4e41e6e750@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Add support for i.MX95

2026-01-13T17:01:52Z

Add imx_rproc_cfg_imx95_m7 and address(TCM and DDR) mapping. Add i.MX95 of_device_id entry. Reviewed-by: Daniel Baluta Reviewed-by: Frank Li Signed-off-by: Peng Fan Link: https://lore.kernel.org/r/20260109-imx95-rproc-2026-1-8-v6-6-d2fefb36263d@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Add support for System Manager CPU API

2026-01-13T16:59:24Z

When the System Manager configuration places the M7 core in the same Logical Machine(LM) as the A55 cores (M7 LM ID equals A55 LM ID), Linux can control M7 using the CPU protocol API. For more details, see the previous patch that adds LMM API support. Changes include: - Introduce imx_rproc_ops_sm_cpu for CPU API operations. - Reuse imx_rproc_sm_detect_mode to detect shared LM and set priv->ops to imx_rproc_ops_sm_cpu. - Implement imx_rproc_sm_cpu_{start,stop} to handle M7 start and stop. Signed-off-by: Peng Fan Reviewed-by: Daniel Baluta Link: https://lore.kernel.org/r/20260109-imx95-rproc-2026-1-8-v6-5-d2fefb36263d@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Add support for System Manager LMM API

2026-01-13T16:52:35Z

i.MX95 features a Cortex-M33 core, six Cortex-A55 cores, and one Cortex-M7 core. The System Control Management Interface(SCMI) firmware runs on the M33 core. The i.MX95 SCMI firmware named System Manager(SM) includes vendor extension protocols, Logical Machine Management(LMM) protocol and CPU protocol and etc. Depending on SM configuration, M7 can be used as follows: (1) M7 in a separate Logical Machine (LM) from A55 cores, that Linux can't control (2) M7 in a separate LM from A55 cores that Linux can control using LMM protocol. (3) M7 runs in same Logical Machine as A55 cores, so Linux can control it using CPU protocol So extend the driver to using LMM and CPU protocol to manage the M7 core. - Compare linux LM ID(got using scmi_imx_lmm_info) and M7 LM ID(the ID is fixed as 1 in SM firmware if M7 is in a separate LM), if Linux LM ID is not same as M7 LM ID(linux and M7 in same LM), use LMM protocol to start/stop. CPU protocol support will be added in the following patch. Whether using CPU or LMM protocol to start/stop, the M7 status detection could use CPU protocol to detect started or not. So in imx_rproc_detect_mode, use scmi_imx_cpu_started to check the status of M7. - For above case (1) and (2), Use SCMI_IMX_LMM_POWER_ON to detect whether the M7 LM is under control of A55 LM. - For above case , after using SCMI_IMX_LMM_POWER_ON to check permission, SCMI_IMX_LMM_SHUTDOWN API should be called to shutdown the M7 LM to save power only when M7 LM is going to be started by remoteproc framework. Otherwise bypass SCMI_IMX_LMM_SHUTDOWN API if M7 LM is started before booting Linux. Current setup relies on pre-Linux software(U-Boot) to do M7 TCM ECC initialization. In future, we could add the support in Linux to decouple U-Boot and Linux. Reviewed-by: Daniel Baluta Reviewed-by: Frank Li Signed-off-by: Peng Fan Link: https://lore.kernel.org/r/20260109-imx95-rproc-2026-1-8-v6-4-d2fefb36263d@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Introduce prepare ops for imx_rproc_dcfg

2026-01-13T16:47:59Z

Allow each platform to provide its own prepare operations, preparing for i.MX95 LMM and CPU ops support. No functional changes. Reviewed-by: Daniel Baluta Reviewed-by: Frank Li Signed-off-by: Peng Fan Link: https://lore.kernel.org/r/20260109-imx95-rproc-2026-1-8-v6-3-d2fefb36263d@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Add runtime ops copy to support dynamic behavior

2026-01-13T16:46:53Z

Structure imx_rproc_dcfg contains a const pointer to imx_rproc_plat_ops, which defines the start/stop/detect_mode operations for a remote processor. To preserve the const correctness of the static configuration while allowing runtime modification of ops behavior, introduce a new imx_rproc_plat_ops member in struct imx_rproc named `ops`. During initialization, dcfg->ops is assigned to priv->ops. Enable the driver to safely override ops at runtime without affecting the original const configuration. Improve flexibility for platforms that require dynamic operation switching (e.g. i.MX95 Logical Machine ops and CPU ops). Reviewed-by: Frank Li Reviewed-by: Daniel Baluta Signed-off-by: Peng Fan Link: https://lore.kernel.org/r/20260109-imx95-rproc-2026-1-8-v6-2-d2fefb36263d@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Use strstarts for "rsc-table" check

2025-12-16T21:50:36Z

The resource name may include an address suffix, for example: rsc-table@1fff8000. To handle such cases, use strstarts() instead of strcmp() when checking for "rsc-table". Signed-off-by: Shenwei Wang Reviewed-by: Daniel Baluta Reviewed-by: Frank Li Reviewed-by: Zhongqiu Han Reviewed-by: Peng Fan Fixes: 67a7bc7f0358 ("remoteproc: Use of_reserved_mem_region_* functions for "memory-region"") Link: https://lore.kernel.org/r/20251208233302.684139-1-shenwei.wang@nxp.com Signed-off-by: Mathieu Poirier

remoteproc: Use of_reserved_mem_region_* functions for "memory-region"

2025-11-25T15:34:31Z

Use the newly added of_reserved_mem_region_to_resource() and of_reserved_mem_region_count() functions to handle "memory-region" properties. The error handling is a bit different in some cases. Often "memory-region" is optional, so failed lookup is not an error. But then an error in of_reserved_mem_lookup() is treated as an error. However, that distinction is not really important. Either the region is available and usable or it is not. So now, it is just of_reserved_mem_region_to_resource() which is checked for an error. Acked-by: Arnaud Pouliquen Tested-by: Peng Fan # i.MX93-11x11-EVK for imx_rproc.c Reviewed-by: Geert Uytterhoeven # rcar Tested-by: Beleswar Padhi # TI Signed-off-by: Rob Herring (Arm) Link: https://lore.kernel.org/r/20251124182751.507624-1-robh@kernel.org Signed-off-by: Mathieu Poirier

remoteproc: imx_rproc: Remove the assignement to method

2025-10-24T14:48:22Z

'method' is no longer used in imx_rproc.c, so remove the assignment. But imx_dsp_rproc.c is still using 'method', so still keep the field in struct imx_rrpoc_dcfg. No functional changes. Reviewed-by: Frank Li Reviewed-by: Daniel Baluta Signed-off-by: Peng Fan Link: https://lore.kernel.org/r/20251024-imx_rproc_c4-v4-4-af83ed3fdbba@nxp.com Signed-off-by: Mathieu Poirier