diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c
index 3d07a5a1f338..b371f1c04a47 100644
--- a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c
@@ -200,6 +200,11 @@ static int a4xx_hw_init(struct msm_gpu *gpu)
 	/* Turn on performance counters: */
 	gpu_write(gpu, REG_A4XX_RBBM_PERFCTR_CTL, 0x01);
 
+	/* use the first CP counter for timestamp queries.. userspace may set
+	 * this as well but it selects the same counter/countable:
+	 */
+	gpu_write(gpu, REG_A4XX_CP_PERFCTR_CP_SEL_0, CP_ALWAYS_COUNT);
+
 	if (adreno_is_a430(adreno_gpu))
 		gpu_write(gpu, REG_A4XX_UCHE_CACHE_WAYS_VFD, 0x07);
 
@@ -294,6 +299,7 @@ static int a4xx_hw_init(struct msm_gpu *gpu)
 	gpu_write(gpu, REG_A4XX_CP_ME_CNTL, 0);
 
 	a4xx_me_init(gpu);
+
 	return 0;
 }
 
@@ -585,6 +591,22 @@ static int a4xx_pm_suspend(struct msm_gpu *gpu) {
 	return 0;
 }
 
+static int a4xx_get_timestamp(struct msm_gpu *gpu, uint64_t *value)
+{
+	uint32_t hi, lo, tmp;
+
+	tmp = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_HI);
+	do {
+		hi = tmp;
+		lo = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_LO);
+		tmp = gpu_read(gpu, REG_A4XX_RBBM_PERFCTR_CP_0_HI);
+	} while (tmp != hi);
+
+	*value = (((uint64_t)hi) << 32) | lo;
+
+	return 0;
+}
+
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
 		.get_param = adreno_get_param,
@@ -602,6 +624,7 @@ static const struct adreno_gpu_funcs funcs = {
 		.show = a4xx_show,
 #endif
 	},
+	.get_timestamp = a4xx_get_timestamp,
 };
 
 struct msm_gpu *a4xx_gpu_init(struct drm_device *dev)
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index 2859b1991259..4951172ede06 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -44,6 +44,10 @@ int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value)
 	case MSM_PARAM_MAX_FREQ:
 		*value = adreno_gpu->base.fast_rate;
 		return 0;
+	case MSM_PARAM_TIMESTAMP:
+		if (adreno_gpu->funcs->get_timestamp)
+			return adreno_gpu->funcs->get_timestamp(gpu, value);
+		return -EINVAL;
 	default:
 		DBG("%s: invalid param: %u", gpu->name, param);
 		return -EINVAL;
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index c26aea13df6d..1d07511f4d22 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -114,6 +114,7 @@ struct adreno_rev {
 
 struct adreno_gpu_funcs {
 	struct msm_gpu_funcs base;
+	int (*get_timestamp)(struct msm_gpu *gpu, uint64_t *value);
 };
 
 struct adreno_info {
diff --git a/include/uapi/drm/msm_drm.h b/include/uapi/drm/msm_drm.h
index 0e3c4c48b326..254d3e92d18e 100644
--- a/include/uapi/drm/msm_drm.h
+++ b/include/uapi/drm/msm_drm.h
@@ -51,6 +51,7 @@ struct drm_msm_timespec {
 #define MSM_PARAM_GMEM_SIZE  0x02
 #define MSM_PARAM_CHIP_ID    0x03
 #define MSM_PARAM_MAX_FREQ   0x04
+#define MSM_PARAM_TIMESTAMP  0x05
 
 struct drm_msm_param {
 	__u32 pipe;           /* in, MSM_PIPE_x */