On Mon, Oct 08, 2018 at 02:57:29PM +0530, Sravanthi Kollukuduru wrote:
> The interconnect framework is designed to provide a
> standard kernel interface to control the settings of
> the interconnects on a SoC.
>
> The interconnect API uses a consumer/provider-based model,
> where the providers are the interconnect buses and the
> consumers could be various drivers.
>
> MDSS is one of the interconnect consumers which uses the
> interconnect APIs to get the path between endpoints and
> set its bandwidth/latency/QoS requirements for the given
> interconnected path.
>
> Signed-off-by: Sravanthi Kollukuduru <skolluku@xxxxxxxxxxxxxx>
> ---
> drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c | 56 +++++++++++++++++++++++++++++---
> 1 file changed, 52 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> index 2235ef8129f4..8391e5c1e559 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c
> @@ -4,10 +4,12 @@
> */
>
> #include "dpu_kms.h"
> +#include <linux/interconnect.h>
>
> #define to_dpu_mdss(x) container_of(x, struct dpu_mdss, base)
>
> #define HW_INTR_STATUS 0x0010
> +#define MAX_AXI_PORT_COUNT 3
>
> struct dpu_mdss {
> struct msm_mdss base;
> @@ -16,8 +18,36 @@ struct dpu_mdss {
> u32 hwversion;
> struct dss_module_power mp;
> struct dpu_irq_controller irq_controller;
> + struct icc_path *path[MAX_AXI_PORT_COUNT];
> + u32 num_paths;
> };
>
> +static int dpu_mdss_parse_data_bus_icc_path(
> + struct drm_device *dev, struct dpu_mdss *dpu_mdss)
> +{
> + struct icc_path *path0 = of_icc_get(dev->dev, "port0");
> + struct icc_path *path1 = of_icc_get(dev->dev, "port1");
> + int total_num_paths = 0;
> +
> + if (IS_ERR(path0))
> + return PTR_ERR(path0);
> +
> + dpu_mdss->path[0] = path0;
> + total_num_paths = 1;
> +
> + if (!IS_ERR(path1)) {
> + dpu_mdss->path[1] = path1;
> + total_num_paths++;
> + }
> +
> + if (total_num_paths > MAX_AXI_PORT_COUNT)
> + return -EINVAL;
Since you are using fixed names you know that by design this isn't possible.
> + dpu_mdss->num_paths = total_num_paths;
> +
> + return 0;
> +}
> +
> static irqreturn_t dpu_mdss_irq(int irq, void *arg)
> {
> struct dpu_mdss *dpu_mdss = arg;
> @@ -127,7 +157,12 @@ static int dpu_mdss_enable(struct msm_mdss *mdss)
> {
> struct dpu_mdss *dpu_mdss = to_dpu_mdss(mdss);
> struct dss_module_power *mp = &dpu_mdss->mp;
> - int ret;
> + int ret, i;
> + u64 ab = (dpu_mdss->num_paths) ? 6800000000/dpu_mdss->num_paths : 0;
> + u64 ib = 6800000000;
> +
> + for (i = 0; i < dpu_mdss->num_paths; i++)
> + icc_set(dpu_mdss->path[i], ab, ib);
>
> ret = msm_dss_enable_clk(mp->clk_config, mp->num_clk, true);
> if (ret)
> @@ -140,12 +175,15 @@ static int dpu_mdss_disable(struct msm_mdss *mdss)
> {
> struct dpu_mdss *dpu_mdss = to_dpu_mdss(mdss);
> struct dss_module_power *mp = &dpu_mdss->mp;
> - int ret;
> + int ret, i;
>
> ret = msm_dss_enable_clk(mp->clk_config, mp->num_clk, false);
> if (ret)
> DPU_ERROR("clock disable failed, ret:%d\n", ret);
>
> + for (i = 0; i < dpu_mdss->num_paths; i++)
> + icc_set(dpu_mdss->path[i], 0, 0);
> +
> return ret;
> }
>
> @@ -155,6 +193,7 @@ static void dpu_mdss_destroy(struct drm_device *dev)
> struct msm_drm_private *priv = dev->dev_private;
> struct dpu_mdss *dpu_mdss = to_dpu_mdss(priv->mdss);
> struct dss_module_power *mp = &dpu_mdss->mp;
> + int i;
>
> _dpu_mdss_irq_domain_fini(dpu_mdss);
>
> @@ -163,6 +202,9 @@ static void dpu_mdss_destroy(struct drm_device *dev)
> msm_dss_put_clk(mp->clk_config, mp->num_clk);
> devm_kfree(&pdev->dev, mp->clk_config);
>
> + for (i = 0; i < dpu_mdss->num_paths; i++)
> + icc_put(dpu_mdss->path[i]);
> +
> if (dpu_mdss->mmio)
> devm_iounmap(&pdev->dev, dpu_mdss->mmio);
> dpu_mdss->mmio = NULL;
> @@ -203,6 +245,12 @@ int dpu_mdss_init(struct drm_device *dev)
> }
> dpu_mdss->mmio_len = resource_size(res);
>
> + ret = dpu_mdss_parse_data_bus_icc_path(dev, dpu_mdss);
> + if (ret) {
> + DPU_ERROR("failed to parse icc path, ret=%d\n", ret);
Depending on the ordering between DPU and interconnect, this might return
-EPROBE_DEFER for a few times before working. You should skip printing
an error message on -EPROBE_DEFER so it doesn't spam the log.
> + return ret;
> + }
> +
> mp = &dpu_mdss->mp;
> ret = msm_dss_parse_clock(pdev, mp);
> if (ret) {
> @@ -224,14 +272,14 @@ int dpu_mdss_init(struct drm_device *dev)
> goto irq_error;
> }
>
> + priv->mdss = &dpu_mdss->base;
> +
This seems like it might have snuck in from a different patch?
> pm_runtime_enable(dev->dev);
>
> pm_runtime_get_sync(dev->dev);
> dpu_mdss->hwversion = readl_relaxed(dpu_mdss->mmio);
> pm_runtime_put_sync(dev->dev);
>
> - priv->mdss = &dpu_mdss->base;
> -
> return ret;
>
> irq_error:
--
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
