Re: [PATCH V4 09/13] interconnect: mediatek: Add mt8183 interconnect provider driver

From: Henry Chen
Date: Wed Apr 01 2020 - 22:48:17 EST

Next message: Henry Chen: "Re: [PATCH V4 06/13] soc: mediatek: add MT8183 dvfsrc support"
Previous message: Henry Chen: "Re: [PATCH V4 00/13] Add driver for dvfsrc, support for active state of scpsys"
In reply to: Georgi Djakov: "Re: [PATCH V4 09/13] interconnect: mediatek: Add mt8183 interconnect provider driver"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hi Georgi,

On Wed, 2020-04-01 at 17:55 +0300, Georgi Djakov wrote:
> Hi Henry,
>
> Thanks for the patch!
>
> On 3/13/20 11:34, Henry Chen wrote:
> > Introduce Mediatek MT8183 specific provider driver using the
> > interconnect framework.
> >
> > Signed-off-by: Henry Chen <henryc.chen@xxxxxxxxxxxx>
> > ---
> > drivers/interconnect/Kconfig | 1 +
> > drivers/interconnect/Makefile | 1 +
> > drivers/interconnect/mediatek/Kconfig | 13 ++
> > drivers/interconnect/mediatek/Makefile | 3 +
> > drivers/interconnect/mediatek/mtk-emi.c | 242 ++++++++++++++++++++++++++++++++
> > 5 files changed, 260 insertions(+)
> > create mode 100644 drivers/interconnect/mediatek/Kconfig
> > create mode 100644 drivers/interconnect/mediatek/Makefile
> > create mode 100644 drivers/interconnect/mediatek/mtk-emi.c
> >
> > diff --git a/drivers/interconnect/Kconfig b/drivers/interconnect/Kconfig
> > index bfa4ca3..dd9ecb6 100644
> > --- a/drivers/interconnect/Kconfig
> > +++ b/drivers/interconnect/Kconfig
> > @@ -12,5 +12,6 @@ menuconfig INTERCONNECT
> > if INTERCONNECT
> >
> > source "drivers/interconnect/qcom/Kconfig"
> > +source "drivers/interconnect/mediatek/Kconfig"
> >
> > endif
> > diff --git a/drivers/interconnect/Makefile b/drivers/interconnect/Makefile
> > index 725029a..330a108 100644
> > --- a/drivers/interconnect/Makefile
> > +++ b/drivers/interconnect/Makefile
> > @@ -5,3 +5,4 @@ icc-core-objs := core.o
> >
> > obj-$(CONFIG_INTERCONNECT) += icc-core.o
> > obj-$(CONFIG_INTERCONNECT_QCOM) += qcom/
> > +obj-$(CONFIG_INTERCONNECT_MTK) += mediatek/
> > diff --git a/drivers/interconnect/mediatek/Kconfig b/drivers/interconnect/mediatek/Kconfig
> > new file mode 100644
> > index 0000000..972d3bb
> > --- /dev/null
> > +++ b/drivers/interconnect/mediatek/Kconfig
> > @@ -0,0 +1,13 @@
> > +config INTERCONNECT_MTK
> > + bool "Mediatek Network-on-Chip interconnect drivers"
> > + depends on ARCH_MEDIATEK
> > + help
> > + Support for Mediatek's Network-on-Chip interconnect hardware.
> > +
> > +config INTERCONNECT_MTK_EMI
> > + tristate "Mediatek EMI interconnect driver"
> > + depends on INTERCONNECT_MTK
> > + depends on (MTK_DVFSRC && OF)
> > + help
> > + This is a driver for the Mediatek Network-on-Chip on DVFSRC-based
> > + platforms.
> > diff --git a/drivers/interconnect/mediatek/Makefile b/drivers/interconnect/mediatek/Makefile
> > new file mode 100644
> > index 0000000..353842b
> > --- /dev/null
> > +++ b/drivers/interconnect/mediatek/Makefile
> > @@ -0,0 +1,3 @@
> > +# SPDX-License-Identifier: GPL-2.0
> > +
> > +obj-$(CONFIG_INTERCONNECT_MTK_EMI) += mtk-emi.o
> > \ No newline at end of file
> > diff --git a/drivers/interconnect/mediatek/mtk-emi.c b/drivers/interconnect/mediatek/mtk-emi.c
> > new file mode 100644
> > index 0000000..0a16992
> > --- /dev/null
> > +++ b/drivers/interconnect/mediatek/mtk-emi.c
> > @@ -0,0 +1,242 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Copyright (c) 2019, The Linux Foundation. All rights reserved.
>
> Nit: Maybe 2020?
Yes, will change to 2020
>
> > + *
> > + */
> > +
> > +#include <dt-bindings/interconnect/mtk,mt8183-emi.h>
>
> Please move this below the all other linux/ headers
OK
>
> > +#include <linux/device.h>
> > +#include <linux/interconnect-provider.h>
> > +#include <linux/module.h>
> > +#include <linux/of_device.h>
> > +#include <linux/of_platform.h>
>
> Is this needed?
You're right, it not necessary and I will remove this.
>
> > +#include <linux/platform_device.h>
> > +#include <soc/mediatek/mtk_dvfsrc.h>
> > +
> > +enum mtk_icc_name {
> > + SLAVE_DDR_EMI,
> > + MASTER_MCUSYS,
> > + MASTER_GPUSYS,
> > + MASTER_MMSYS,
> > + MASTER_MM_VPU,
> > + MASTER_MM_DISP,
> > + MASTER_MM_VDEC,
> > + MASTER_MM_VENC,
> > + MASTER_MM_CAM,
> > + MASTER_MM_IMG,
> > + MASTER_MM_MDP,
> > +};
> > +
> > +#define MT8183_MAX_LINKS 1
> > +
> > +/**
> > + * struct mtk_icc_node - Mediatek specific interconnect nodes
> > + * @name: the node name used in debugfs
> > + * @ep: true if the node is an end point.
> > + * @id: a unique node identifier
> > + * @links: an array of nodes where we can go next while traversing
> > + * @num_links: the total number of @links
> > + * @buswidth: width of the interconnect between a node and the bus
> > + * @sum_avg: current sum aggregate value of all avg bw kBps requests
> > + * @max_peak: current max aggregate value of all peak bw kBps requests
> > + */
> > +struct mtk_icc_node {
> > + unsigned char *name;
> > + bool ep;
> > + u16 id;
> > + u16 links[MT8183_MAX_LINKS];
> > + u16 num_links;
> > + u16 buswidth;
> > + u64 sum_avg;
> > + u64 max_peak;
> > +};
> > +
> > +struct mtk_icc_desc {
> > + struct mtk_icc_node **nodes;
> > + size_t num_nodes;
> > +};
> > +
> > +#define DEFINE_MNODE(_name, _id, _buswidth, _ep, ...) \
> > + static struct mtk_icc_node _name = { \
> > + .name = #_name, \
> > + .id = _id, \
> > + .buswidth = _buswidth, \
> > + .ep = _ep, \
> > + .num_links = ARRAY_SIZE(((int[]){ __VA_ARGS__ })), \
> > +}
> > +
> > +DEFINE_MNODE(ddr_emi, SLAVE_DDR_EMI, 1024, 1, 0);
> > +DEFINE_MNODE(mcusys, MASTER_MCUSYS, 256, 0, SLAVE_DDR_EMI);
> > +DEFINE_MNODE(gpu, MASTER_GPUSYS, 256, 0, SLAVE_DDR_EMI);
> > +DEFINE_MNODE(mmsys, MASTER_MMSYS, 256, 0, SLAVE_DDR_EMI);
> > +DEFINE_MNODE(mm_vpu, MASTER_MM_VPU, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_disp, MASTER_MM_DISP, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_vdec, MASTER_MM_VDEC, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_venc, MASTER_MM_VENC, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_cam, MASTER_MM_CAM, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_img, MASTER_MM_IMG, 128, 0, MASTER_MMSYS);
> > +DEFINE_MNODE(mm_mdp, MASTER_MM_MDP, 128, 0, MASTER_MMSYS);
> > +
> > +static struct mtk_icc_node *mt8183_icc_nodes[] = {
> > + [MT8183_SLAVE_DDR_EMI] = &ddr_emi,
> > + [MT8183_MASTER_MCUSYS] = &mcusys,
> > + [MT8183_MASTER_GPU] = &gpu,
> > + [MT8183_MASTER_MMSYS] = &mmsys,
> > + [MT8183_MASTER_MM_VPU] = &mm_vpu,
> > + [MT8183_MASTER_MM_DISP] = &mm_disp,
> > + [MT8183_MASTER_MM_VDEC] = &mm_vdec,
> > + [MT8183_MASTER_MM_VENC] = &mm_venc,
> > + [MT8183_MASTER_MM_CAM] = &mm_cam,
> > + [MT8183_MASTER_MM_IMG] = &mm_img,
> > + [MT8183_MASTER_MM_MDP] = &mm_mdp,
> > +};
> > +
> > +static struct mtk_icc_desc mt8183_icc = {
>
> const?
ok
>
> > + .nodes = mt8183_icc_nodes,
> > + .num_nodes = ARRAY_SIZE(mt8183_icc_nodes),
> > +};
> > +
> > +static int emi_icc_aggregate(struct icc_node *node, u32 tag, u32 avg_bw,
> > + u32 peak_bw, u32 *agg_avg, u32 *agg_peak)
> > +{
> > + struct mtk_icc_node *in;
> > +
> > + in = node->data;
> > +
> > + *agg_avg += avg_bw;
> > + *agg_peak += peak_bw;
> > +
> > + in->sum_avg = *agg_avg;
> > + in->max_peak = *agg_peak;
> > +
> > + return 0;
> > +}
> > +
> > +static int emi_icc_set(struct icc_node *src, struct icc_node *dst)
> > +{
> > + int ret = 0;
> > + struct mtk_icc_node *node;
> > +
> > + node = dst->data;
> > + if (node->ep) {
> > + pr_debug("sum_avg (%llu), max_peak (%llu)\n",
> > + node->sum_avg, node->max_peak);
> > + mtk_dvfsrc_send_request(src->provider->dev->parent,
> > + MTK_DVFSRC_CMD_BW_REQUEST,
> > + node->max_peak);
> > + }
> > +
> > + return ret;
> > +}
> > +
> > +static int emi_icc_remove(struct platform_device *pdev);
> > +static int emi_icc_probe(struct platform_device *pdev)
> > +{
> > + int ret;
> > + const struct mtk_icc_desc *desc;
> > + struct icc_node *node;
> > + struct icc_onecell_data *data;
> > + struct icc_provider *provider;
> > + struct mtk_icc_node **mnodes;
> > + size_t num_nodes, i, j;
> > +
> > + desc = of_device_get_match_data(&pdev->dev);
> > + if (!desc)
> > + return -EINVAL;
> > +
> > + mnodes = desc->nodes;
> > + num_nodes = desc->num_nodes;
> > +
> > + provider = devm_kzalloc(&pdev->dev, sizeof(*provider), GFP_KERNEL);
> > + if (!provider)
> > + return -ENOMEM;
> > +
> > + data = devm_kcalloc(&pdev->dev, num_nodes, sizeof(*node), GFP_KERNEL);
>
> This should be:
> devm_kzalloc(&pdev->dev, struct_size(data, nodes, num_nodes), GFP_KERNEL);
thanks, will fix this.
>
> > + if (!data)
> > + return -ENOMEM;
> > +
> > + provider->dev = &pdev->dev;
> > + provider->set = emi_icc_set;
> > + provider->aggregate = emi_icc_aggregate;
> > + provider->xlate = of_icc_xlate_onecell;
> > + INIT_LIST_HEAD(&provider->nodes);
> > + provider->data = data;
> > +
> > + ret = icc_provider_add(provider);
> > + if (ret) {
> > + dev_err(&pdev->dev, "error adding interconnect provider\n");
> > + return ret;
> > + }
> > +
> > + for (i = 0; i < num_nodes; i++) {
> > + node = icc_node_create(mnodes[i]->id);
> > + if (IS_ERR(node)) {
> > + ret = PTR_ERR(node);
> > + goto err;
> > + }
> > +
> > + node->name = mnodes[i]->name;
> > + node->data = mnodes[i];
> > + icc_node_add(node, provider);
> > +
> > + dev_dbg(&pdev->dev, "registered node %s, num link: %d\n",
> > + mnodes[i]->name, mnodes[i]->num_links);
>
> Please remove this.
ok
>
> > +
> > + /* populate links */
>
> Please remove the comment too.
ok
>
> > + for (j = 0; j < mnodes[i]->num_links; j++)
> > + icc_link_create(node, mnodes[i]->links[j]);
> > +
> > + data->nodes[i] = node;
> > + }
> > + data->num_nodes = num_nodes;
> > +
> > + platform_set_drvdata(pdev, provider);
> > +
> > + return 0;
> > +err:
> > + emi_icc_remove(pdev);
>
> This will not work because platform_set_drvdata() is called only when the
> probe function completes successfully. So platform_get_drvdata() in
> emi_icc_remove() will return NULL.
>
> > + return ret;
> > +}
> > +
> > +static int emi_icc_remove(struct platform_device *pdev)
> > +{
> > + struct icc_provider *provider = platform_get_drvdata(pdev);
> > + struct icc_node *n;
> > +
> > + list_for_each_entry(n, &provider->nodes, node_list) {
> > + icc_node_del(n);
> > + icc_node_destroy(n->id);
> > + }
>
> You can use icc_nodes_remove() instead of the above.
Got it, thanks for advise.
>
> Thanks,
> Georgi

Thanks for the review.

Henry

Next message: Henry Chen: "Re: [PATCH V4 06/13] soc: mediatek: add MT8183 dvfsrc support"
Previous message: Henry Chen: "Re: [PATCH V4 00/13] Add driver for dvfsrc, support for active state of scpsys"
In reply to: Georgi Djakov: "Re: [PATCH V4 09/13] interconnect: mediatek: Add mt8183 interconnect provider driver"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]