Implement gpu_busy based on the downstream msm-3.4 code [1]. This
allows us to use devfreq on this old old old hardware!

[1] https://github.com/LineageOS/android_kernel_sony_apq8064/blob/lineage-16.0/drivers/gpu/msm/adreno_a2xx.c#L1975

Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
---
 drivers/gpu/drm/msm/adreno/a2xx_gpu.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)

diff --git a/drivers/gpu/drm/msm/adreno/a2xx_gpu.c b/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
index c67089a7ebc1..6258c98e5a88 100644
--- a/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
@@ -481,6 +481,33 @@ a2xx_create_address_space(struct msm_gpu *gpu, struct platform_device *pdev)
 	return aspace;
 }
 
+/* While the precise size of this field is unknown, it holds at least these three values.. */
+#define PERF_MODE_CNT		GENMASK(2, 0)
+ #define PERF_STATE_RESET	0x0
+ #define PERF_STATE_ENABLE	0x1
+ #define PERF_STATE_FREEZE	0x2
+static u64 a2xx_gpu_busy(struct msm_gpu *gpu, unsigned long *out_sample_rate)
+{
+	u64 busy_cycles;
+
+	/* Freeze the counter */
+	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_FREEZE));
+
+	busy_cycles = gpu_read64(gpu, REG_A2XX_RBBM_PERFCOUNTER1_LO);
+
+	/* Reset the counter */
+	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_RESET));
+
+	/* Re-enable the performance monitors */
+	gpu_rmw(gpu, REG_A2XX_RBBM_PM_OVERRIDE2, BIT(6), BIT(6));
+	gpu_write(gpu, REG_A2XX_RBBM_PERFCOUNTER1_SELECT, 1);
+	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_ENABLE));
+
+	*out_sample_rate = clk_get_rate(gpu->core_clk);
+
+	return busy_cycles;
+}
+
 static u32 a2xx_get_rptr(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
 {
 	ring->memptrs->rptr = gpu_read(gpu, REG_AXXX_CP_RB_RPTR);
@@ -502,6 +529,7 @@ static const struct adreno_gpu_funcs funcs = {
 #if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP)
 		.show = adreno_show,
 #endif
+		.gpu_busy = a2xx_gpu_busy,
 		.gpu_state_get = a2xx_gpu_state_get,
 		.gpu_state_put = adreno_gpu_state_put,
 		.create_address_space = a2xx_create_address_space,

-- 
2.39.2

On 22/02/2023 23:47, Konrad Dybcio wrote:
> Implement gpu_busy based on the downstream msm-3.4 code [1]. This
> allows us to use devfreq on this old old old hardware!
> 
> [1] https://github.com/LineageOS/android_kernel_sony_apq8064/blob/lineage-16.0/drivers/gpu/msm/adreno_a2xx.c#L1975
> 
> Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
> ---
>   drivers/gpu/drm/msm/adreno/a2xx_gpu.c | 28 ++++++++++++++++++++++++++++
>   1 file changed, 28 insertions(+)
> 
> diff --git a/drivers/gpu/drm/msm/adreno/a2xx_gpu.c b/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
> index c67089a7ebc1..6258c98e5a88 100644
> --- a/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
> +++ b/drivers/gpu/drm/msm/adreno/a2xx_gpu.c
> @@ -481,6 +481,33 @@ a2xx_create_address_space(struct msm_gpu *gpu, struct platform_device *pdev)
>   	return aspace;
>   }
>   
> +/* While the precise size of this field is unknown, it holds at least these three values.. */
> +#define PERF_MODE_CNT		GENMASK(2, 0)
> + #define PERF_STATE_RESET	0x0
> + #define PERF_STATE_ENABLE	0x1
> + #define PERF_STATE_FREEZE	0x2

These should go into a2xx.xml.h

LGTM otherwise.

> +static u64 a2xx_gpu_busy(struct msm_gpu *gpu, unsigned long *out_sample_rate)
> +{
> +	u64 busy_cycles;
> +
> +	/* Freeze the counter */
> +	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_FREEZE));
> +
> +	busy_cycles = gpu_read64(gpu, REG_A2XX_RBBM_PERFCOUNTER1_LO);
> +
> +	/* Reset the counter */
> +	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_RESET));
> +
> +	/* Re-enable the performance monitors */
> +	gpu_rmw(gpu, REG_A2XX_RBBM_PM_OVERRIDE2, BIT(6), BIT(6));
> +	gpu_write(gpu, REG_A2XX_RBBM_PERFCOUNTER1_SELECT, 1);
> +	gpu_write(gpu, REG_A2XX_CP_PERFMON_CNTL, FIELD_PREP(PERF_MODE_CNT, PERF_STATE_ENABLE));
> +
> +	*out_sample_rate = clk_get_rate(gpu->core_clk);
> +
> +	return busy_cycles;
> +}
> +
>   static u32 a2xx_get_rptr(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
>   {
>   	ring->memptrs->rptr = gpu_read(gpu, REG_AXXX_CP_RB_RPTR);
> @@ -502,6 +529,7 @@ static const struct adreno_gpu_funcs funcs = {
>   #if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP)
>   		.show = adreno_show,
>   #endif
> +		.gpu_busy = a2xx_gpu_busy,
>   		.gpu_state_get = a2xx_gpu_state_get,
>   		.gpu_state_put = adreno_gpu_state_put,
>   		.create_address_space = a2xx_create_address_space,
> 

-- 
With best wishes
Dmitry