devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Cristian Marussi <cristian.marussi@arm.com>
To: "Peng Fan (OSS)" <peng.fan@oss.nxp.com>
Cc: Sudeep Holla <sudeep.holla@arm.com>,
	Cristian Marussi <cristian.marussi@arm.com>,
	Shawn Guo <shawnguo@kernel.org>,
	Sascha Hauer <s.hauer@pengutronix.de>,
	Pengutronix Kernel Team <kernel@pengutronix.de>,
	Fabio Estevam <festevam@gmail.com>, Rob Herring <robh@kernel.org>,
	Krzysztof Kozlowski <krzk+dt@kernel.org>,
	Conor Dooley <conor+dt@kernel.org>,
	Dan Carpenter <dan.carpenter@linaro.org>,
	linux-kernel@vger.kernel.org, arm-scmi@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org, imx@lists.linux.dev,
	devicetree@vger.kernel.org, Peng Fan <peng.fan@nxp.com>
Subject: Re: [PATCH v2 4/7] firmware: arm_scmi: imx: Add i.MX95 CPU Protocol
Date: Mon, 24 Feb 2025 18:10:48 +0000	[thread overview]
Message-ID: <Z7y2KFYt5gIY19Qz@pluto> (raw)
In-Reply-To: <20250212-imx-lmm-cpu-v2-4-3aee005968c1@nxp.com>

On Wed, Feb 12, 2025 at 03:40:26PM +0800, Peng Fan (OSS) wrote:
> From: Peng Fan <peng.fan@nxp.com>
> 
> This protocol allows an agent to start, stop a CPU or set reset vector. It
> is used to manage auxiliary CPUs in an LM (e.g. additional cores in an AP
> cluster).
> 

Hi,

> Signed-off-by: Peng Fan <peng.fan@nxp.com>
> ---
>  drivers/firmware/arm_scmi/vendors/imx/Kconfig      |  11 +
>  drivers/firmware/arm_scmi/vendors/imx/Makefile     |   1 +
>  drivers/firmware/arm_scmi/vendors/imx/imx-sm-cpu.c | 287 +++++++++++++++++++++
>  include/linux/scmi_imx_protocol.h                  |  10 +
>  4 files changed, 309 insertions(+)
> 

[snip]

> +
> +struct scmi_imx_cpu_info_get_out {
> +#define	CPU_RUN_MODE_START	0
> +#define	CPU_RUN_MODE_HOLD	1
> +#define	CPU_RUN_MODE_STOP	2
> +#define	CPU_RUN_MODE_SLEEP	3
> +	__le32 runmode;
> +	__le32 sleepmode;
> +	__le32 resetvectorlow;
> +	__le32 resetvectorhigh;
> +};
> +
> +static int scmi_imx_cpu_validate_cpuid(const struct scmi_protocol_handle *ph,
> +				       u32 cpuid)
> +{
> +	struct scmi_imx_cpu_info *info = ph->get_priv(ph);
> +
> +	if (cpuid >= info->nr_cpu)
> +		return -EINVAL;
> +
> +	return 0;
> +}
> +
> +static int scmi_imx_cpu_start(const struct scmi_protocol_handle *ph, u32 cpuid)
> +{
> +	struct scmi_xfer *t;
> +	int ret;
> +
> +	ret = scmi_imx_cpu_validate_cpuid(ph, cpuid);
> +	if (ret)
> +		return ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, SCMI_IMX_CPU_START, sizeof(u32),
> +				      0, &t);
> +	if (ret)
> +		return ret;
> +
> +	put_unaligned_le32(cpuid, t->tx.buf);
> +	ret = ph->xops->do_xfer(ph, t);
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +
> +static int scmi_imx_cpu_stop(const struct scmi_protocol_handle *ph, u32 cpuid)
> +{
> +	struct scmi_xfer *t;
> +	int ret;
> +
> +	ret = scmi_imx_cpu_validate_cpuid(ph, cpuid);
> +	if (ret)
> +		return ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, SCMI_IMX_CPU_STOP, sizeof(u32),
> +				      0, &t);
> +	if (ret)
> +		return ret;
> +
> +	put_unaligned_le32(cpuid, t->tx.buf);
> +	ret = ph->xops->do_xfer(ph, t);
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +

Please refactor and unify this 2 seemingly identical start/stop funcs by
defining a common helper...

> +static int scmi_imx_cpu_reset_vector_set(const struct scmi_protocol_handle *ph,
> +					 u32 cpuid, u64 vector, bool start,
> +					 bool boot, bool resume)
> +{
> +	struct scmi_imx_cpu_reset_vector_set_in *in;
> +	struct scmi_xfer *t;
> +	int ret;
> +	u32 flags;
> +
> +	ret = scmi_imx_cpu_validate_cpuid(ph, cpuid);
> +	if (ret)
> +		return ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, SCMI_IMX_CPU_RESET_VECTOR_SET, sizeof(*in),
> +				      0, &t);
> +	if (ret)
> +		return ret;
> +
> +	in = t->tx.buf;
> +	in->cpuid = cpu_to_le32(cpuid);
> +	flags = start ? CPU_VEC_FLAGS_START : 0;
> +	flags |= boot ? CPU_VEC_FLAGS_BOOT : 0;
> +	flags |= resume ? CPU_VEC_FLAGS_RESUME : 0;
> +	in->flags = cpu_to_le32(flags);

...you should just avoid the endianess helper given that is a bitfield (I cannot
exclude that there are other places where we wrongly endianIZE bitfields...) and
I think that the best way to do it without cause smatch to cry would be
to use le32_encode_bits() which should just produce the desired flags
into an __le32

le32_encode_bits and friends are used throughout the code base and added

https://patchwork.ozlabs.org/project/ubuntu-kernel/patch/20190606034255.2192-2-aaron.ma@canonical.com/

which seems to be the best (and only) documentation :P

> +	in->resetvectorlow = cpu_to_le32(lower_32_bits(vector));
> +	in->resetvectorhigh = cpu_to_le32(upper_32_bits(vector));
> +	ret = ph->xops->do_xfer(ph, t);
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +
> +static int scmi_imx_cpu_started(const struct scmi_protocol_handle *ph, u32 cpuid,
> +				bool *started)
> +{
> +	struct scmi_imx_cpu_info_get_out *out;
> +	struct scmi_xfer *t;
> +	u32 mode;
> +	int ret;
> +

maybe overlay paranoid...but

	if (!started)
		return -EINVAL;

...up to you, if you feel paranoid too

> +	*started = false;
> +	ret = scmi_imx_cpu_validate_cpuid(ph, cpuid);
> +	if (ret)
> +		return ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, SCMI_IMX_CPU_INFO_GET, sizeof(u32),
> +				      0, &t);
> +	if (ret)
> +		return ret;
> +
> +	put_unaligned_le32(cpuid, t->tx.buf);
> +	ret = ph->xops->do_xfer(ph, t);
> +	if (!ret) {
> +		out = t->rx.buf;
> +		mode = le32_to_cpu(out->runmode);
> +		if ((mode == CPU_RUN_MODE_START) || (mode == CPU_RUN_MODE_SLEEP))
> +			*started = true;
> +	}
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +
> +static const struct scmi_imx_cpu_proto_ops scmi_imx_cpu_proto_ops = {
> +	.cpu_reset_vector_set = scmi_imx_cpu_reset_vector_set,
> +	.cpu_start = scmi_imx_cpu_start,
> +	.cpu_started = scmi_imx_cpu_started,
> +	.cpu_stop = scmi_imx_cpu_stop,
> +};
> +
> +static int scmi_imx_cpu_protocol_attributes_get(const struct scmi_protocol_handle *ph,
> +						struct scmi_imx_cpu_info *info)
> +{
> +	struct scmi_msg_imx_cpu_protocol_attributes *attr;
> +	struct scmi_xfer *t;
> +	int ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, PROTOCOL_ATTRIBUTES, 0,
> +				      sizeof(*attr), &t);
> +	if (ret)
> +		return ret;
> +
> +	attr = t->rx.buf;
> +
> +	ret = ph->xops->do_xfer(ph, t);
> +	if (!ret) {
> +		info->nr_cpu = SCMI_IMX_CPU_PROTO_ATTR_NUM_CPUS(attr->attributes);

	
	le32_get_bits(attr->attributes, GENMASK())

> +		dev_info(ph->dev, "i.MX SM CPU: %d cpus\n",
> +			 info->nr_cpu);
> +	}
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +
> +static int scmi_imx_cpu_attributes_get(const struct scmi_protocol_handle *ph,
> +				       u32 cpuid)
> +{
> +	struct scmi_msg_imx_cpu_attributes_out *out;
> +	char name[SCMI_SHORT_NAME_MAX_SIZE] = {'\0'};
> +	struct scmi_xfer *t;
> +	int ret;
> +
> +	ret = ph->xops->xfer_get_init(ph, SCMI_IMX_CPU_ATTRIBUTES, sizeof(u32), 0, &t);
> +	if (ret)
> +		return ret;
> +
> +	put_unaligned_le32(cpuid, t->tx.buf);
> +	ret = ph->xops->do_xfer(ph, t);
> +	if (!ret) {
> +		out = t->rx.buf;
> +		strscpy(name, out->name, SCMI_SHORT_NAME_MAX_SIZE);
> +		dev_info(ph->dev, "i.MX CPU: name: %s\n", name);
> +	} else {
> +		dev_err(ph->dev, "i.MX cpu: Failed to get info of cpu(%u)\n", cpuid);
> +	}
> +
> +	ph->xops->xfer_put(ph, t);
> +
> +	return ret;
> +}
> +
> +static int scmi_imx_cpu_protocol_init(const struct scmi_protocol_handle *ph)
> +{
> +	struct scmi_imx_cpu_info *info;
> +	u32 version;
> +	int ret, i;
> +
> +	ret = ph->xops->version_get(ph, &version);
> +	if (ret)
> +		return ret;
> +
> +	dev_info(ph->dev, "NXP SM CPU Protocol Version %d.%d\n",
> +		 PROTOCOL_REV_MAJOR(version), PROTOCOL_REV_MINOR(version));
> +
> +	info = devm_kzalloc(ph->dev, sizeof(*info), GFP_KERNEL);
> +	if (!info)
> +		return -ENOMEM;
> +
> +	ret = scmi_imx_cpu_protocol_attributes_get(ph, info);
> +	if (ret)
> +		return ret;
> +
> +	for (i = 0; i < info->nr_cpu; i++) {
> +		ret = scmi_imx_cpu_attributes_get(ph, i);
> +		if (ret)
> +			return ret;
> +	}
> +
> +	return ph->set_priv(ph, info, version);
> +}
> +
> +static const struct scmi_protocol scmi_imx_cpu = {
> +	.id = SCMI_PROTOCOL_IMX_CPU,
> +	.owner = THIS_MODULE,
> +	.instance_init = &scmi_imx_cpu_protocol_init,
> +	.ops = &scmi_imx_cpu_proto_ops,
> +	.supported_version = SCMI_PROTOCOL_SUPPORTED_VERSION,
> +	.vendor_id = SCMI_IMX_VENDOR,
> +	.sub_vendor_id = SCMI_IMX_SUBVENDOR,
> +};
> +module_scmi_protocol(scmi_imx_cpu);

similarly as LMM...

MODULE_ALIAS("scmi-protocol-" __stringify(SCMI_PROTOCOL_IMX_CPU) "-" SCMI_IMX_VENDOR);

Thanks
Cristian


  reply	other threads:[~2025-02-24 18:10 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-12  7:40 [PATCH v2 0/7] firmware: scmi/imx: Add i.MX95 LMM/CPU Protocol Peng Fan (OSS)
2025-02-12  7:40 ` [PATCH v2 1/7] firmware: arm_scmi: imx: Add LMM and CPU documentation Peng Fan (OSS)
2025-02-24 12:28   ` Cristian Marussi
2025-02-25  9:07     ` Peng Fan
2025-02-25 11:54       ` Cristian Marussi
2025-02-12  7:40 ` [PATCH v2 2/7] dt-bindings: firmware: Add i.MX95 SCMI LMM and CPU protocol Peng Fan (OSS)
2025-02-19 21:46   ` Rob Herring
2025-02-24  3:01     ` Peng Fan
2025-02-12  7:40 ` [PATCH v2 3/7] firmware: arm_scmi: imx: Add i.MX95 LMM protocol Peng Fan (OSS)
2025-02-24 15:54   ` Cristian Marussi
2025-02-25  9:25     ` Peng Fan
2025-02-12  7:40 ` [PATCH v2 4/7] firmware: arm_scmi: imx: Add i.MX95 CPU Protocol Peng Fan (OSS)
2025-02-24 18:10   ` Cristian Marussi [this message]
2025-02-25  9:29     ` Peng Fan
2025-02-12  7:40 ` [PATCH v2 5/7] firmware: imx: Add i.MX95 SCMI LMM driver Peng Fan (OSS)
2025-02-25 11:25   ` Cristian Marussi
2025-02-12  7:40 ` [PATCH v2 6/7] firmware: imx: Add i.MX95 SCMI CPU driver Peng Fan (OSS)
2025-02-25 11:28   ` Cristian Marussi
2025-02-12  7:40 ` [PATCH v2 7/7] MAINTAINERS: add entry for i.MX SCMI extensions Peng Fan (OSS)

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Z7y2KFYt5gIY19Qz@pluto \
    --to=cristian.marussi@arm.com \
    --cc=arm-scmi@vger.kernel.org \
    --cc=conor+dt@kernel.org \
    --cc=dan.carpenter@linaro.org \
    --cc=devicetree@vger.kernel.org \
    --cc=festevam@gmail.com \
    --cc=imx@lists.linux.dev \
    --cc=kernel@pengutronix.de \
    --cc=krzk+dt@kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peng.fan@nxp.com \
    --cc=peng.fan@oss.nxp.com \
    --cc=robh@kernel.org \
    --cc=s.hauer@pengutronix.de \
    --cc=shawnguo@kernel.org \
    --cc=sudeep.holla@arm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).