[PATCH] accel/ivpu: Fix sporadic VPU boot failure

[PATCH] accel/ivpu: Fix sporadic VPU boot failure

[Stanislaw Gruszka]

From: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com>

Wait for AON bit in HOST_SS_CPR_RST_CLR to return 0 before
starting VPUIP power up sequence, otherwise the VPU device
may sporadically fail to boot.

An error in power up sequence is propagated to the runtime
power management - the device will be in an error state
until the VPU driver is reloaded.

Fixes: 35b137630f08 ("accel/ivpu: Introduce a new DRM driver for Intel VPU")
Cc: stable@vger.kernel.org # 6.3.x
Signed-off-by: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com>
Reviewed-by: Krystian Pradzynski <krystian.pradzynski@linux.intel.com>
Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
---
 drivers/accel/ivpu/ivpu_hw_mtl.c     | 13 ++++++++++++-
 drivers/accel/ivpu/ivpu_hw_mtl_reg.h |  1 +
 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/drivers/accel/ivpu/ivpu_hw_mtl.c b/drivers/accel/ivpu/ivpu_hw_mtl.c
index 382ec127be8e..efae679d7b7a 100644
--- a/drivers/accel/ivpu/ivpu_hw_mtl.c
+++ b/drivers/accel/ivpu/ivpu_hw_mtl.c
@@ -197,6 +197,11 @@ static void ivpu_pll_init_frequency_ratios(struct ivpu_device *vdev)
 	hw->pll.pn_ratio = clamp_t(u8, fuse_pn_ratio, hw->pll.min_ratio, hw->pll.max_ratio);
 }
 
+static int ivpu_hw_mtl_wait_for_vpuip_bar(struct ivpu_device *vdev)
+{
+	return REGV_POLL_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, AON, 0, 100);
+}
+
 static int ivpu_pll_drive(struct ivpu_device *vdev, bool enable)
 {
 	struct ivpu_hw_info *hw = vdev->hw;
@@ -239,6 +244,12 @@ static int ivpu_pll_drive(struct ivpu_device *vdev, bool enable)
 			ivpu_err(vdev, "Timed out waiting for PLL ready status\n");
 			return ret;
 		}
+
+		ret = ivpu_hw_mtl_wait_for_vpuip_bar(vdev);
+		if (ret) {
+			ivpu_err(vdev, "Timed out waiting for VPUIP bar\n");
+			return ret;
+		}
 	}
 
 	return 0;
@@ -256,7 +267,7 @@ static int ivpu_pll_disable(struct ivpu_device *vdev)
 
 static void ivpu_boot_host_ss_rst_clr_assert(struct ivpu_device *vdev)
 {
-	u32 val = REGV_RD32(MTL_VPU_HOST_SS_CPR_RST_CLR);
+	u32 val = 0;
 
 	val = REG_SET_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, TOP_NOC, val);
 	val = REG_SET_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, DSS_MAS, val);
diff --git a/drivers/accel/ivpu/ivpu_hw_mtl_reg.h b/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
index d83ccfd9a871..593b8ff07417 100644
--- a/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
+++ b/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
@@ -91,6 +91,7 @@
 #define MTL_VPU_HOST_SS_CPR_RST_SET_MSS_MAS_MASK			BIT_MASK(11)
 
 #define MTL_VPU_HOST_SS_CPR_RST_CLR					0x00000098u
+#define MTL_VPU_HOST_SS_CPR_RST_CLR_AON_MASK				BIT_MASK(0)
 #define MTL_VPU_HOST_SS_CPR_RST_CLR_TOP_NOC_MASK			BIT_MASK(1)
 #define MTL_VPU_HOST_SS_CPR_RST_CLR_DSS_MAS_MASK			BIT_MASK(10)
 #define MTL_VPU_HOST_SS_CPR_RST_CLR_MSS_MAS_MASK			BIT_MASK(11)
-- 
2.25.1

Re: [PATCH] accel/ivpu: Fix sporadic VPU boot failure

[Stanislaw Gruszka]

On Wed, Jun 07, 2023 at 11:45:02AM +0200, Stanislaw Gruszka wrote:
> From: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com>
> 
> Wait for AON bit in HOST_SS_CPR_RST_CLR to return 0 before
> starting VPUIP power up sequence, otherwise the VPU device
> may sporadically fail to boot.
> 
> An error in power up sequence is propagated to the runtime
> power management - the device will be in an error state
> until the VPU driver is reloaded.
> 
> Fixes: 35b137630f08 ("accel/ivpu: Introduce a new DRM driver for Intel VPU")
> Cc: stable@vger.kernel.org # 6.3.x
> Signed-off-by: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com>
> Reviewed-by: Krystian Pradzynski <krystian.pradzynski@linux.intel.com>
> Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Applied to drm-misc-fixes

Thanks
Stanislaw

> ---
>  drivers/accel/ivpu/ivpu_hw_mtl.c     | 13 ++++++++++++-
>  drivers/accel/ivpu/ivpu_hw_mtl_reg.h |  1 +
>  2 files changed, 13 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/accel/ivpu/ivpu_hw_mtl.c b/drivers/accel/ivpu/ivpu_hw_mtl.c
> index 382ec127be8e..efae679d7b7a 100644
> --- a/drivers/accel/ivpu/ivpu_hw_mtl.c
> +++ b/drivers/accel/ivpu/ivpu_hw_mtl.c
> @@ -197,6 +197,11 @@ static void ivpu_pll_init_frequency_ratios(struct ivpu_device *vdev)
>  	hw->pll.pn_ratio = clamp_t(u8, fuse_pn_ratio, hw->pll.min_ratio, hw->pll.max_ratio);
>  }
>  
> +static int ivpu_hw_mtl_wait_for_vpuip_bar(struct ivpu_device *vdev)
> +{
> +	return REGV_POLL_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, AON, 0, 100);
> +}
> +
>  static int ivpu_pll_drive(struct ivpu_device *vdev, bool enable)
>  {
>  	struct ivpu_hw_info *hw = vdev->hw;
> @@ -239,6 +244,12 @@ static int ivpu_pll_drive(struct ivpu_device *vdev, bool enable)
>  			ivpu_err(vdev, "Timed out waiting for PLL ready status\n");
>  			return ret;
>  		}
> +
> +		ret = ivpu_hw_mtl_wait_for_vpuip_bar(vdev);
> +		if (ret) {
> +			ivpu_err(vdev, "Timed out waiting for VPUIP bar\n");
> +			return ret;
> +		}
>  	}
>  
>  	return 0;
> @@ -256,7 +267,7 @@ static int ivpu_pll_disable(struct ivpu_device *vdev)
>  
>  static void ivpu_boot_host_ss_rst_clr_assert(struct ivpu_device *vdev)
>  {
> -	u32 val = REGV_RD32(MTL_VPU_HOST_SS_CPR_RST_CLR);
> +	u32 val = 0;
>  
>  	val = REG_SET_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, TOP_NOC, val);
>  	val = REG_SET_FLD(MTL_VPU_HOST_SS_CPR_RST_CLR, DSS_MAS, val);
> diff --git a/drivers/accel/ivpu/ivpu_hw_mtl_reg.h b/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
> index d83ccfd9a871..593b8ff07417 100644
> --- a/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
> +++ b/drivers/accel/ivpu/ivpu_hw_mtl_reg.h
> @@ -91,6 +91,7 @@
>  #define MTL_VPU_HOST_SS_CPR_RST_SET_MSS_MAS_MASK			BIT_MASK(11)
>  
>  #define MTL_VPU_HOST_SS_CPR_RST_CLR					0x00000098u
> +#define MTL_VPU_HOST_SS_CPR_RST_CLR_AON_MASK				BIT_MASK(0)
>  #define MTL_VPU_HOST_SS_CPR_RST_CLR_TOP_NOC_MASK			BIT_MASK(1)
>  #define MTL_VPU_HOST_SS_CPR_RST_CLR_DSS_MAS_MASK			BIT_MASK(10)
>  #define MTL_VPU_HOST_SS_CPR_RST_CLR_MSS_MAS_MASK			BIT_MASK(11)
> -- 
> 2.25.1
>