On Fri, Oct 31, 2025 at 03:28:07PM -0400, Frank Li wrote:
>On Fri, Oct 31, 2025 at 10:24:55AM +0800, Peng Fan wrote:
>> 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.
>>
>> There are three cases for M7:
>> (1) M7 in a separate Logical Machine(LM) that Linux can't control it.
>> (2) M7 in a separate Logical Machine that Linux can control it using
>> LMM protocol
>> (3) M7 runs in same Logical Machine as A55, 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 seprate LM),
>> if Linux LM ID equals M7 LM ID(linux and M7 in same LM), use CPU
>> protocol to start/stop. Otherwise, use LMM protocol to start/stop.
>> 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.
>>
>> Signed-off-by: Peng Fan <[email protected]>
>> ---
>> drivers/remoteproc/Kconfig | 2 +
>> drivers/remoteproc/imx_rproc.c | 192
>> +++++++++++++++++++++++++++++++++++++++++
>> drivers/remoteproc/imx_rproc.h | 3 +
>> 3 files changed, 197 insertions(+)
>>
>> diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig
>> index
>> 48a0d3a69ed08057716f1e7ea950899f60bbe0cf..ee54436fea5ad08a9c198ce74d44ce7a9aa206de
>> 100644
>> --- a/drivers/remoteproc/Kconfig
>> +++ b/drivers/remoteproc/Kconfig
>> @@ -27,6 +27,8 @@ config IMX_REMOTEPROC
>> tristate "i.MX remoteproc support"
>> depends on ARCH_MXC
>> depends on HAVE_ARM_SMCCC
>> + depends on IMX_SCMI_CPU_DRV || !IMX_SCMI_CPU_DRV
>> + depends on IMX_SCMI_LMM_DRV || !IMX_SCMI_LMM_DRV
>> select MAILBOX
>> help
>> Say y here to support iMX's remote processors via the remote
>> diff --git a/drivers/remoteproc/imx_rproc.c b/drivers/remoteproc/imx_rproc.c
>> index
>> 232eb91e0b5dc2432131b1c140d6688b073fea1d..1fb17701964ca4ee4b73d343b5ec1be8e2ee5fda
>> 100644
>> --- a/drivers/remoteproc/imx_rproc.c
>> +++ b/drivers/remoteproc/imx_rproc.c
>> @@ -8,6 +8,7 @@
>...
>> +
>> +static int imx_rproc_sm_lmm_check(struct rproc *rproc, bool started)
>> +{
>> + struct imx_rproc *priv = rproc->priv;
>> + const struct imx_rproc_dcfg *dcfg = priv->dcfg;
>> + struct device *dev = priv->dev;
>> + int ret;
>> +
>> + /*
>> + * Use power on to do permission check. If rproc is in different LM,
>> + * and linux has permission to handle the LM, set
>> IMX_RPROC_FLAGS_SM_LMM_AVAIL.
>> + */
>> + ret = scmi_imx_lmm_operation(dcfg->lmid, SCMI_IMX_LMM_POWER_ON, 0);
>> + if (ret != 0) {
>
> if (!ret)
Fix in next version.
>
>> + if (ret == -EACCES) {
>> + /* Not under Linux Control, so only do IPC between
>> rproc and Linux */
>> + dev_info(dev, "lmm(%d) not under Linux Control\n",
>> dcfg->lmid);
>> + return 0;
>> + }
>> +
>> + dev_info(dev, "power on lmm(%d) failed: %d\n", dcfg->lmid, ret);
>
>dev_err()? because it is "... failed"
Yeah.
>
>> + return ret;
>> + }
>> + is_cpu_ops = dcfg->lmid == info.lmid;
>> +
>> + if (is_cpu_ops) {
>> + priv->ops = &imx_rproc_ops_sm_cpu;
>> + dev_info(dev, "Using CPU Protocol OPS\n");
>> + return 0;
>> + }
>
>does need else branch to set priv->ops?
The platform default ops should be set to lmm ops, so there is no need
else branch, see patch 5.
Thanks,
Peng
>
>Frank
>> +
>> + dev_info(dev, "Using LMM Protocol OPS\n");
>> +
>> + return imx_rproc_sm_lmm_check(rproc, started);
>> +}
>> +
>> static int imx_rproc_detect_mode(struct imx_rproc *priv)
>> {
>> /*
>> diff --git a/drivers/remoteproc/imx_rproc.h b/drivers/remoteproc/imx_rproc.h
>> index
>> e4b9ede656506142b260aa7515823f489a168ba4..c05bfc4528f3b7518d4f2272d15fdeab1a68c2a3
>> 100644
>> --- a/drivers/remoteproc/imx_rproc.h
>> +++ b/drivers/remoteproc/imx_rproc.h
>> @@ -52,6 +52,9 @@ struct imx_rproc_dcfg {
>> enum imx_rproc_method method;
>> u32 flags;
>> const struct imx_rproc_plat_ops *ops;
>> + /* For System Manager(SM) based SoCs, the IDs are from SM firmware */
>> + u32 cpuid;
>> + u32 lmid;
>> };
>>
>> #endif /* _IMX_RPROC_H */
>>
>> --
>> 2.37.1
>>
