Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Sasha Levin <sashal@kernel.org>
To:     linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc:     Akhil P Oommen <quic_akhilpo@quicinc.com>,
        Rob Clark <robdclark@chromium.org>,
        Sasha Levin <sashal@kernel.org>, robdclark@gmail.com,
        quic_abhinavk@quicinc.com, dmitry.baryshkov@linaro.org,
        airlied@gmail.com, daniel@ffwll.ch, konrad.dybcio@somainline.org,
        geert@linux-m68k.org, dianders@chromium.org, olvaffe@gmail.com,
        linux-arm-msm@vger.kernel.org, dri-devel@lists.freedesktop.org,
        freedreno@lists.freedesktop.org
Subject: [PATCH AUTOSEL 6.1 16/53] drm/msm/a6xx: Avoid gx gbit halt during rpm suspend
Date:   Mon, 16 Jan 2023 09:01:16 -0500
Message-Id: <20230116140154.114951-16-sashal@kernel.org>
In-Reply-To: <20230116140154.114951-1-sashal@kernel.org>
References: <20230116140154.114951-1-sashal@kernel.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Akhil P Oommen <quic_akhilpo@quicinc.com>

[ Upstream commit f4a75b5933c998e60fd812a7680e0971eb1c7cee ]

As per the downstream driver, gx gbif halt is required only during
recovery sequence. So lets avoid it during regular rpm suspend.

Signed-off-by: Akhil P Oommen <quic_akhilpo@quicinc.com>
Patchwork: https://patchwork.freedesktop.org/patch/515279/
Link: https://lore.kernel.org/r/20221216223253.1.Ice9c47bfeb1fddb8dc377a3491a043a3ee7fca7d@changeid
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 15 +++++++++------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c |  7 +++++++
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
 3 files changed, 17 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
index e033d6a67a20..870252bef23f 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
@@ -876,7 +876,8 @@ static void a6xx_gmu_rpmh_off(struct a6xx_gmu *gmu)
 #define GBIF_CLIENT_HALT_MASK             BIT(0)
 #define GBIF_ARB_HALT_MASK                BIT(1)
 
-static void a6xx_bus_clear_pending_transactions(struct adreno_gpu *adreno_gpu)
+static void a6xx_bus_clear_pending_transactions(struct adreno_gpu *adreno_gpu,
+		bool gx_off)
 {
 	struct msm_gpu *gpu = &adreno_gpu->base;
 
@@ -889,9 +890,11 @@ static void a6xx_bus_clear_pending_transactions(struct adreno_gpu *adreno_gpu)
 		return;
 	}
 
-	/* Halt the gx side of GBIF */
-	gpu_write(gpu, REG_A6XX_RBBM_GBIF_HALT, 1);
-	spin_until(gpu_read(gpu, REG_A6XX_RBBM_GBIF_HALT_ACK) & 1);
+	if (gx_off) {
+		/* Halt the gx side of GBIF */
+		gpu_write(gpu, REG_A6XX_RBBM_GBIF_HALT, 1);
+		spin_until(gpu_read(gpu, REG_A6XX_RBBM_GBIF_HALT_ACK) & 1);
+	}
 
 	/* Halt new client requests on GBIF */
 	gpu_write(gpu, REG_A6XX_GBIF_HALT, GBIF_CLIENT_HALT_MASK);
@@ -929,7 +932,7 @@ static void a6xx_gmu_force_off(struct a6xx_gmu *gmu)
 	/* Halt the gmu cm3 core */
 	gmu_write(gmu, REG_A6XX_GMU_CM3_SYSRESET, 1);
 
-	a6xx_bus_clear_pending_transactions(adreno_gpu);
+	a6xx_bus_clear_pending_transactions(adreno_gpu, true);
 
 	/* Reset GPU core blocks */
 	gpu_write(gpu, REG_A6XX_RBBM_SW_RESET_CMD, 1);
@@ -1083,7 +1086,7 @@ static void a6xx_gmu_shutdown(struct a6xx_gmu *gmu)
 			return;
 		}
 
-		a6xx_bus_clear_pending_transactions(adreno_gpu);
+		a6xx_bus_clear_pending_transactions(adreno_gpu, a6xx_gpu->hung);
 
 		/* tell the GMU we want to slumber */
 		ret = a6xx_gmu_notify_slumber(gmu);
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
index e846e629c00d..9d7fc44c1e2a 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
@@ -1277,6 +1277,12 @@ static void a6xx_recover(struct msm_gpu *gpu)
 	if (hang_debug)
 		a6xx_dump(gpu);
 
+	/*
+	 * To handle recovery specific sequences during the rpm suspend we are
+	 * about to trigger
+	 */
+	a6xx_gpu->hung = true;
+
 	/* Halt SQE first */
 	gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 3);
 
@@ -1319,6 +1325,7 @@ static void a6xx_recover(struct msm_gpu *gpu)
 	mutex_unlock(&gpu->active_lock);
 
 	msm_gpu_hw_init(gpu);
+	a6xx_gpu->hung = false;
 }
 
 static const char *a6xx_uche_fault_block(struct msm_gpu *gpu, u32 mid)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.h b/drivers/gpu/drm/msm/adreno/a6xx_gpu.h
index ab853f61db63..eea2e60ce3b7 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.h
@@ -32,6 +32,7 @@ struct a6xx_gpu {
 	void *llc_slice;
 	void *htw_llc_slice;
 	bool have_mmu500;
+	bool hung;
 };
 
 #define to_a6xx_gpu(x) container_of(x, struct a6xx_gpu, base)
-- 
2.35.1