Re: [PATCH v6] drm/msm/a6xx: Add support for an interconnect path

[Evan Green]

On Fri, Jan 18, 2019 at 12:24 PM Jordan Crouse <jcrouse@xxxxxxxxxxxxxx> wrote:
>
> Try to get the interconnect path for the GPU and vote for the maximum
> bandwidth to support all frequencies. This is needed for performance.
> Later we will want to scale the bandwidth based on the frequency to
> also optimize for power but that will require some device tree
> infrastructure that does not yet exist.
>
> v6: use icc_set_bw() instead of icc_set()
> v5: Remove hardcoded interconnect name and just use the default
> v4: Don't use a port string at all to skip the need for names in the DT
> v3: Use macros and change port string per Georgi Djakov
>
> Signed-off-by: Jordan Crouse <jcrouse@xxxxxxxxxxxxxx>
> ---
>
>  drivers/gpu/drm/msm/Kconfig             |  1 +
>  drivers/gpu/drm/msm/adreno/a6xx_gmu.c   | 20 ++++++++++++++++++++
>  drivers/gpu/drm/msm/adreno/adreno_gpu.c |  9 +++++++++
>  drivers/gpu/drm/msm/msm_gpu.h           |  3 +++
>  4 files changed, 33 insertions(+)
>
> diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
> index cf549f1..78c9e5a5 100644
> --- a/drivers/gpu/drm/msm/Kconfig
> +++ b/drivers/gpu/drm/msm/Kconfig
> @@ -5,6 +5,7 @@ config DRM_MSM
>         depends on ARCH_QCOM || SOC_IMX5 || (ARM && COMPILE_TEST)
>         depends on OF && COMMON_CLK
>         depends on MMU
> +       depends on INTERCONNECT || !INTERCONNECT
>         select QCOM_MDT_LOADER if ARCH_QCOM
>         select REGULATOR
>         select DRM_KMS_HELPER
> diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> index 5beb83d..c48fe46 100644
> --- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> +++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
> @@ -2,6 +2,7 @@
>  /* Copyright (c) 2017-2018 The Linux Foundation. All rights reserved. */
>
>  #include <linux/clk.h>
> +#include <linux/interconnect.h>
>  #include <linux/pm_opp.h>
>  #include <soc/qcom/cmd-db.h>
>
> @@ -84,6 +85,9 @@ bool a6xx_gmu_gx_is_on(struct a6xx_gmu *gmu)
>
>  static void __a6xx_gmu_set_freq(struct a6xx_gmu *gmu, int index)
>  {
> +       struct a6xx_gpu *a6xx_gpu = container_of(gmu, struct a6xx_gpu, gmu);
> +       struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
> +       struct msm_gpu *gpu = &adreno_gpu->base;
>         int ret;
>
>         gmu_write(gmu, REG_A6XX_GMU_DCVS_ACK_OPTION, 0);
> @@ -106,6 +110,12 @@ static void __a6xx_gmu_set_freq(struct a6xx_gmu *gmu, int index)
>                 dev_err(gmu->dev, "GMU set GPU frequency error: %d\n", ret);
>
>         gmu->freq = gmu->gpu_freqs[index];
> +
> +       /*
> +        * Eventually we will want to scale the path vote with the frequency but
> +        * for now leave it at max so that the performance is nominal.
> +        */
> +       icc_set_bw(gpu->icc_path, 0, MBps_to_icc(7216));
>  }
>
>  void a6xx_gmu_set_freq(struct msm_gpu *gpu, unsigned long freq)
> @@ -705,6 +715,8 @@ int a6xx_gmu_reset(struct a6xx_gpu *a6xx_gpu)
>
>  int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
>  {
> +       struct adreno_gpu *adreno_gpu = &a6xx_gpu->base;
> +       struct msm_gpu *gpu = &adreno_gpu->base;
>         struct a6xx_gmu *gmu = &a6xx_gpu->gmu;
>         int status, ret;
>
> @@ -720,6 +732,9 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
>         if (ret)
>                 goto out;
>
> +       /* Set the bus quota to a reasonable value for boot */
> +       icc_set_bw(gpu->icc_path, 0, MBps_to_icc(3072));

Does the comment mention boot because this resume call happens during
init? How come this number is different from the one in
__a6xx_gmu_set_freq? Also, you're setting an average bandwidth of 0, I
guess you don't know really what your average is, and the hardware is
fine with that setting. Is that true?
-Evan