From: "TingHan Shen (沈廷翰)" <TingHan.Shen@mediatek.com>
To: "mathieu.poirier@linaro.org" <mathieu.poirier@linaro.org>,
"matthias.bgg@gmail.com" <matthias.bgg@gmail.com>,
"angelogioacchino.delregno@collabora.com"
<angelogioacchino.delregno@collabora.com>,
"andersson@kernel.org" <andersson@kernel.org>,
"robh+dt@kernel.org" <robh+dt@kernel.org>,
"krzysztof.kozlowski+dt@linaro.org"
<krzysztof.kozlowski+dt@linaro.org>
Cc: "linux-remoteproc@vger.kernel.org"
<linux-remoteproc@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"linux-mediatek@lists.infradead.org"
<linux-mediatek@lists.infradead.org>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
Project_Global_Chrome_Upstream_Group
<Project_Global_Chrome_Upstream_Group@mediatek.com>
Subject: Re: [PATCH v7 07/12] remoteproc: mediatek: Probe multi-core SCP
Date: Thu, 16 Feb 2023 02:40:18 +0000 [thread overview]
Message-ID: <2534c41188d6f740da9930d1de3daf16cbc99614.camel@mediatek.com> (raw)
In-Reply-To: <1fb45a25-38e9-d9cf-6c69-0ba0dac83824@collabora.com>
On Wed, 2023-02-15 at 13:29 +0100, AngeloGioacchino Del Regno wrote:
> Il 15/02/23 05:12, Tinghan Shen ha scritto:
> > The difference of single-core SCP and multi-core SCP device tree is
> > the presence of child device nodes described SCP cores. The SCP
> > driver populates the platform device and checks the child nodes
> > to identify whether it's a single-core SCP or a multi-core SCP.
> >
> > The resource structure of the multi-core SCP is a list of remoteproc
> > instances which is different to the single-core SCP. The corresponding
> > resource releasing action is based on the type of SCP.
> >
> > Signed-off-by: Tinghan Shen <tinghan.shen@mediatek.com>
> > ---
> > drivers/remoteproc/mtk_common.h | 4 +
> > drivers/remoteproc/mtk_scp.c | 179 +++++++++++++++++++++++++++++---
> > 2 files changed, 168 insertions(+), 15 deletions(-)
> >
> > diff --git a/drivers/remoteproc/mtk_common.h b/drivers/remoteproc/mtk_common.h
> > index 3778894c96f3..7821cb15d6fd 100644
> > --- a/drivers/remoteproc/mtk_common.h
> > +++ b/drivers/remoteproc/mtk_common.h
> > @@ -140,6 +140,10 @@ struct mtk_scp {
> > size_t dram_size;
> >
> > struct rproc_subdev *rpmsg_subdev;
> > +
> > + struct list_head elem;
> > + struct list_head cluster_cores;
> > + struct list_head *cluster;
>
> I don't understand why you need both `cluster_cores` and `cluster`... check below.
>
> > };
> >
> > /**
> > diff --git a/drivers/remoteproc/mtk_scp.c b/drivers/remoteproc/mtk_scp.c
> > index 6270c388727a..6588e5acb159 100644
> > --- a/drivers/remoteproc/mtk_scp.c
> > +++ b/drivers/remoteproc/mtk_scp.c
> > @@ -862,7 +862,9 @@ static void scp_remove_rpmsg_subdev(struct mtk_scp *scp)
> > }
> >
> > static int scp_rproc_init(struct platform_device *pdev,
> > - struct mtk_scp_of_regs *of_regs)
> > + struct mtk_scp_of_regs *of_regs,
> > + const struct mtk_scp_of_data *of_data,
> > + bool is_multi_core)
> > {
> > struct device *dev = &pdev->dev;
> > struct device_node *np = dev->of_node;
> > @@ -883,7 +885,7 @@ static int scp_rproc_init(struct platform_device *pdev,
> > scp = (struct mtk_scp *)rproc->priv;
> > scp->rproc = rproc;
> > scp->dev = dev;
> > - scp->data = of_device_get_match_data(dev);
> > + scp->data = of_data;
> > platform_set_drvdata(pdev, scp);
> >
> > scp->reg_base = of_regs->reg_base;
> > @@ -932,9 +934,11 @@ static int scp_rproc_init(struct platform_device *pdev,
> > goto remove_subdev;
> > }
> >
> > - ret = rproc_add(rproc);
> > - if (ret)
> > - goto remove_subdev;
> > + if (!is_multi_core) {
> > + ret = rproc_add(rproc);
> > + if (ret)
> > + goto remove_subdev;
> > + }
> >
> > return 0;
> >
> > @@ -950,9 +954,125 @@ static int scp_rproc_init(struct platform_device *pdev,
> > return ret;
> > }
> >
> > +static void scp_rproc_free(struct mtk_scp *scp)
> > +{
> > + int i;
> > +
> > + scp_remove_rpmsg_subdev(scp);
> > + scp_ipi_unregister(scp, SCP_IPI_INIT);
> > + scp_unmap_memory_region(scp);
> > + for (i = 0; i < SCP_IPI_MAX; i++)
> > + mutex_destroy(&scp->ipi_desc[i].lock);
> > + mutex_destroy(&scp->send_lock);
> > +}
> > +
> > +static void scp_rproc_exit(struct platform_device *pdev)
> > +{
> > + struct mtk_scp *scp = platform_get_drvdata(pdev);
> > +
> > + rproc_del(scp->rproc);
> > + scp_rproc_free(scp);
> > +}
> > +
> > +static int scp_cluster_init(struct platform_device *pdev,
> > + struct mtk_scp_of_regs *of_regs)
> > +{
> > + struct device *dev = &pdev->dev;
> > + struct device_node *np = dev_of_node(dev);
> > + struct platform_device *cpdev;
> > + struct device_node *child;
> > + const struct mtk_scp_of_data **cluster_of_data;
> > + struct list_head *cluster;
>
> Here you have a local `cluster` pointer.....
>
> > + struct mtk_scp *scp, *temp;
> > + int core_id, num_cores;
> > + int ret;
> > +
> > + cluster_of_data = (const struct mtk_scp_of_data **)of_device_get_match_data(dev);
> > +
> > + for (num_cores = 0; cluster_of_data[num_cores]; num_cores++)
> > + ;
> > +
> > + cluster = NULL;
>
> (unrelated: please initialize cluster and core_id on the declaration, not here)
Ok. I'll update this in the next version.
>
> > + core_id = 0;
> > + for_each_available_child_of_node(np, child) {
> > + if (core_id >= num_cores) {
> > + ret = -EINVAL;
> > + dev_err(dev, "Not support core %d\n", core_id);
> > + of_node_put(child);
> > + goto init_fail;
> > + }
> > +
> > + cpdev = of_find_device_by_node(child);
> > + if (!cpdev) {
> > + ret = -ENODEV;
> > + dev_err(dev, "Not found platform device for core %d\n", core_id);
> > + of_node_put(child);
> > + goto init_fail;
> > + }
> > +
> > + ret = scp_rproc_init(cpdev, of_regs, cluster_of_data[core_id], true);
> > + if (ret) {
> > + dev_err(dev, "Failed to initialize core %d rproc\n", core_id);
> > + put_device(&cpdev->dev);
> > + of_node_put(child);
> > + goto init_fail;
> > + }
> > +
> > + scp = platform_get_drvdata(cpdev);
> > + put_device(&cpdev->dev);
> > + if (core_id == 0) {
> > + INIT_LIST_HEAD(&scp->cluster_cores);
>
> ...then here you initialize cluster_cores,
>
> > + cluster = &scp->cluster_cores;
>
> and assign to your local `cluster` pointer and even set it
>
> > + platform_set_drvdata(pdev, cluster);
>
> as platform drvdata.
>
> > + }
> > +
> > + list_add_tail(&scp->elem, cluster);
>
> Still using the local `cluster` pointer, which is `scp->cluster_cores`
>
> > + scp->cluster = cluster;
>
> ...and here, `scp->cluster` effectively points to local `cluster`, which is
> anyway `scp->cluster_cores`.
>
> As far as I understand, `scp->cluster` is EITHER `scp->cluster_cores` or NULL,
> hence the question: why do you need both scp->cluster and scp->cluster_cores,
> if they're always pointing to the same list?
>
> P.S.: I won't cut the rest of the patch so that the readers of this reply won't
> have to click around to see the whole code; comments end here!
>
> Regards,
> Angelo
It's because each core is represented by a mtk_scp structure and they are linked in the list of core 0.
After initialized, the data structure is depicted as below,
mtk_scp for core 0 mtk_scp for core 1
elem <----------------------elem (list)
|
V
cluster_cores <----┐ cluster_cores (not used)
Λ |
| |
*cluster └--------*cluster (pointer)
For the 2nd core scp, its cluster_cores is useless.
So, I add the *cluster to point back to the cores list in the 1st core's data structure.
Best regards,
TingHan
>
> > +
> > + of_node_put(child);
> > + core_id++;
> > + }
> > +
> > + list_for_each_entry_safe_reverse(scp, temp, cluster, elem) {
> > + ret = rproc_add(scp->rproc);
> > + if (ret)
> > + goto add_fail;
> > + }
> > +
> > + return 0;
> > +
> > +add_fail:
> > + list_for_each_entry_continue(scp, cluster, elem) {
> > + rproc_del(scp->rproc);
> > + }
> > +init_fail:
> > + if (cluster) {
> > + list_for_each_entry_safe_reverse(scp, temp, cluster, elem) {
> > + list_del(&scp->elem);
> > + scp_rproc_free(scp);
> > + }
> > + }
> > + return ret;
> > +}
> > +
> > +static void scp_cluster_exit(struct platform_device *pdev)
> > +{
> > + struct list_head *cluster = platform_get_drvdata(pdev);
> > + struct platform_device *cpdev;
> > + struct mtk_scp *scp, *temp;
> > +
> > + list_for_each_entry_safe_reverse(scp, temp, cluster, elem) {
> > + list_del(&scp->elem);
> > + cpdev = to_platform_device(scp->dev);
> > + scp_rproc_exit(cpdev);
> > + }
> > +}
> > +
> > static int scp_probe(struct platform_device *pdev)
> > {
> > struct device *dev = &pdev->dev;
> > + struct device_node *np = dev->of_node;
> > + struct device_node *core_node;
> > struct resource *res;
> > struct mtk_scp_of_regs scp_regs;
> > int ret;
> > @@ -975,21 +1095,43 @@ static int scp_probe(struct platform_device *pdev)
> > scp_regs.l1tcm_phys = res->start;
> > }
> >
> > - return scp_rproc_init(pdev, &scp_regs);
> > + ret = devm_of_platform_populate(dev);
> > + if (ret)
> > + return dev_err_probe(dev, ret, "Failed to populate platform devices\n");
> > +
> > + core_node = of_get_child_by_name(np, "scp");
> > + of_node_put(core_node);
> > +
> > + if (!core_node) {
> > + dev_info(dev, "single-core scp\n");
> > +
> > + ret = scp_rproc_init(pdev, &scp_regs, of_device_get_match_data(dev), false);
> > + if (ret)
> > + return dev_err_probe(dev, ret, "Failed to initialize single-core scp\n");
> > + } else {
> > + dev_info(dev, "multi-core scp\n");
> > +
> > + ret = scp_cluster_init(pdev, &scp_regs);
> > + if (ret)
> > + return dev_err_probe(dev, ret, "Failed to initialize scp cluster\n");
> > + }
> > +
> > + return 0;
> > }
> >
> > static int scp_remove(struct platform_device *pdev)
> > {
> > - struct mtk_scp *scp = platform_get_drvdata(pdev);
> > - int i;
> > + struct device *dev = &pdev->dev;
> > + struct device_node *np = dev->of_node;
> > + struct device_node *core_node;
> >
> > - rproc_del(scp->rproc);
> > - scp_remove_rpmsg_subdev(scp);
> > - scp_ipi_unregister(scp, SCP_IPI_INIT);
> > - scp_unmap_memory_region(scp);
> > - for (i = 0; i < SCP_IPI_MAX; i++)
> > - mutex_destroy(&scp->ipi_desc[i].lock);
> > - mutex_destroy(&scp->send_lock);
> > + core_node = of_get_child_by_name(np, "scp");
> > + of_node_put(core_node);
> > +
> > + if (!core_node)
> > + scp_rproc_exit(pdev);
> > + else
> > + scp_cluster_exit(pdev);
> >
> > return 0;
> > }
> > @@ -1068,12 +1210,19 @@ static const struct mtk_scp_of_data mt8195_of_data_c1 = {
> > .host_to_scp_int_bit = MT8195_CORE1_HOST_IPC_INT_BIT,
> > };
> >
> > +static const struct mtk_scp_of_data *mt8195_of_data_cores[] = {
> > + &mt8195_of_data,
> > + &mt8195_of_data_c1,
> > + NULL
> > +};
> > +
> > static const struct of_device_id mtk_scp_of_match[] = {
> > { .compatible = "mediatek,mt8183-scp", .data = &mt8183_of_data },
> > { .compatible = "mediatek,mt8186-scp", .data = &mt8186_of_data },
> > { .compatible = "mediatek,mt8188-scp", .data = &mt8188_of_data },
> > { .compatible = "mediatek,mt8192-scp", .data = &mt8192_of_data },
> > { .compatible = "mediatek,mt8195-scp", .data = &mt8195_of_data },
> > + { .compatible = "mediatek,mt8195-scp-dual", .data = &mt8195_of_data_cores },
> > {},
> > };
> > MODULE_DEVICE_TABLE(of, mtk_scp_of_match);
>
>
--
Best regards,
TingHan
next prev parent reply other threads:[~2023-02-16 2:40 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-15 4:12 [PATCH v7 00/12] Add support for MT8195 SCP 2nd core Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 01/12] dt-bindings: remoteproc: mediatek: Improve the rpmsg subnode definition Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 02/12] arm64: dts: mediatek: mt8183-kukui: Update the node name of SCP rpmsg subnode Tinghan Shen
2023-02-16 11:47 ` Matthias Brugger
2023-02-15 4:12 ` [PATCH v7 03/12] dt-bindings: remoteproc: mediatek: Support MT8195 dual-core SCP Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 04/12] remoteproc: mediatek: Add SCP core 1 register definitions Tinghan Shen
2023-02-16 11:42 ` Matthias Brugger
2023-02-15 4:12 ` [PATCH v7 05/12] remoteproc: mediatek: Add MT8195 SCP core 1 operations Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 06/12] remoteproc: mediatek: Extract remoteproc initialization flow Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 07/12] remoteproc: mediatek: Probe multi-core SCP Tinghan Shen
2023-02-15 12:29 ` AngeloGioacchino Del Regno
2023-02-16 2:40 ` TingHan Shen (沈廷翰) [this message]
2023-02-15 4:12 ` [PATCH v7 08/12] remoteproc: mediatek: Control SCP core 1 by rproc subdevice Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 09/12] remoteproc: mediatek: Setup MT8195 SCP core 1 SRAM offset Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 10/12] remoteproc: mediatek: Handle MT8195 SCP core 1 watchdog timeout Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 11/12] remoteproc: mediatek: Refine ipi handler error message Tinghan Shen
2023-02-15 4:12 ` [PATCH v7 12/12] arm64: dts: mediatek: mt8195: Add SCP 2nd core Tinghan Shen
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=2534c41188d6f740da9930d1de3daf16cbc99614.camel@mediatek.com \
--to=tinghan.shen@mediatek.com \
--cc=Project_Global_Chrome_Upstream_Group@mediatek.com \
--cc=andersson@kernel.org \
--cc=angelogioacchino.delregno@collabora.com \
--cc=devicetree@vger.kernel.org \
--cc=krzysztof.kozlowski+dt@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux-remoteproc@vger.kernel.org \
--cc=mathieu.poirier@linaro.org \
--cc=matthias.bgg@gmail.com \
--cc=robh+dt@kernel.org \
/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).