[PATCH 1/2] drm/amdgpu: change default behavior of bad_page_threshold parameter

[PATCH 1/2] drm/amdgpu: change default behavior of bad_page_threshold parameter

[Tao Zhou]

Ignore ras umc bad page threshold by default, GPU initialization won't
be stopped in this mode.

v2: refine the description of bad_page_threshold.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c        | 2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c        | 7 ++++---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c | 4 ++--
 3 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 6c2fe50b528e..8a375394db0c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -921,7 +921,7 @@ module_param_named(reset_method, amdgpu_reset_method, int, 0444);
  * result in the GPU entering bad status when the number of total
  * faulty pages by ECC exceeds the threshold value.
  */
-MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = auto(default value), 0 = disable bad page retirement, -2 = ignore bad page threshold)");
+MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = ignore threshold (default value), 0 = disable bad page retirement, -2 = driver sets threshold)");
 module_param_named(bad_page_threshold, amdgpu_bad_page_threshold, int, 0444);
 
 MODULE_PARM_DESC(num_kcq, "number of kernel compute queue user want to setup (8 if set to greater than 8 or less than 0, only affect gfx 8+)");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index 5c02c6c9f773..63dfcc98152d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -2196,11 +2196,12 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev,
 	/*
 	 * Justification of value bad_page_cnt_threshold in ras structure
 	 *
-	 * Generally, -1 <= amdgpu_bad_page_threshold <= max record length
-	 * in eeprom, and introduce two scenarios accordingly.
+	 * Generally, 0 <= amdgpu_bad_page_threshold <= max record length
+	 * in eeprom or amdgpu_bad_page_threshold == -2, introduce two
+	 * scenarios accordingly.
 	 *
 	 * Bad page retirement enablement:
-	 *    - If amdgpu_bad_page_threshold = -1,
+	 *    - If amdgpu_bad_page_threshold = -2,
 	 *      bad_page_cnt_threshold = typical value by formula.
 	 *
 	 *    - When the value from user is 0 < amdgpu_bad_page_threshold <
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 2d9f3f4cd79e..9d370465b08d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -1191,8 +1191,8 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control,
 		} else {
 			dev_err(adev->dev, "RAS records:%d exceed threshold:%d",
 				control->ras_num_recs, ras->bad_page_cnt_threshold);
-			if (amdgpu_bad_page_threshold == -2) {
-				dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -2.");
+			if (amdgpu_bad_page_threshold == -1) {
+				dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -1.");
 				res = 0;
 			} else {
 				*exceed_err_limit = true;
-- 
2.35.1

[PATCH 2/2] drm/amdgpu: add bad_page_threshold check in ras_eeprom_check_err

[Tao Zhou]

bad_page_threshold controls page retirement behavior and it should be
also checked.

v2: simplify the condition of bad page handling path.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
---
 .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 9d370465b08d..2e08fce87521 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -417,7 +417,8 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 
-	if (!__is_ras_eeprom_supported(adev))
+	if (!__is_ras_eeprom_supported(adev) ||
+	    !amdgpu_bad_page_threshold)
 		return false;
 
 	/* skip check eeprom table for VEGA20 Gaming */
@@ -428,10 +429,18 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 			return false;
 
 	if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
-		dev_warn(adev->dev, "This GPU is in BAD status.");
-		dev_warn(adev->dev, "Please retire it or set a larger "
-			 "threshold value when reloading driver.\n");
-		return true;
+		if (amdgpu_bad_page_threshold == -1) {
+			dev_warn(adev->dev, "RAS records:%d exceed threshold:%d",
+				con->eeprom_control.ras_num_recs, con->bad_page_cnt_threshold);
+			dev_warn(adev->dev,
+				"But GPU can be operated due to bad_page_threshold = -1.\n");
+			return false;
+		} else {
+			dev_warn(adev->dev, "This GPU is in BAD status.");
+			dev_warn(adev->dev, "Please retire it or set a larger "
+				 "threshold value when reloading driver.\n");
+			return true;
+		}
 	}
 
 	return false;
-- 
2.35.1

RE: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in ras_eeprom_check_err

[Yang, Stanley]

[AMD Official Use Only - General]

The series is Reviewed-by: Stanley.Yang <Stanley.Yang@amd.com>

Regards,
Stanley
> -----Original Message-----
> From: Zhou1, Tao <Tao.Zhou1@amd.com>
> Sent: Wednesday, February 22, 2023 10:52 AM
> To: amd-gfx@lists.freedesktop.org; Zhang, Hawking
> <Hawking.Zhang@amd.com>; Yang, Stanley <Stanley.Yang@amd.com>; Chai,
> Thomas <YiPeng.Chai@amd.com>; Li, Candice <Candice.Li@amd.com>; Lazar,
> Lijo <Lijo.Lazar@amd.com>
> Cc: Zhou1, Tao <Tao.Zhou1@amd.com>
> Subject: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in
> ras_eeprom_check_err
> 
> bad_page_threshold controls page retirement behavior and it should be also
> checked.
> 
> v2: simplify the condition of bad page handling path.
> 
> Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
> ---
>  .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 19 ++++++++++++++-
> ----
>  1 file changed, 14 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> index 9d370465b08d..2e08fce87521 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> @@ -417,7 +417,8 @@ bool
> amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)  {
>  	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
> 
> -	if (!__is_ras_eeprom_supported(adev))
> +	if (!__is_ras_eeprom_supported(adev) ||
> +	    !amdgpu_bad_page_threshold)
>  		return false;
> 
>  	/* skip check eeprom table for VEGA20 Gaming */ @@ -428,10
> +429,18 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct
> amdgpu_device *adev)
>  			return false;
> 
>  	if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
> -		dev_warn(adev->dev, "This GPU is in BAD status.");
> -		dev_warn(adev->dev, "Please retire it or set a larger "
> -			 "threshold value when reloading driver.\n");
> -		return true;
> +		if (amdgpu_bad_page_threshold == -1) {
> +			dev_warn(adev->dev, "RAS records:%d exceed
> threshold:%d",
> +				con->eeprom_control.ras_num_recs, con-
> >bad_page_cnt_threshold);
> +			dev_warn(adev->dev,
> +				"But GPU can be operated due to
> bad_page_threshold = -1.\n");
> +			return false;
> +		} else {
> +			dev_warn(adev->dev, "This GPU is in BAD status.");
> +			dev_warn(adev->dev, "Please retire it or set a larger
> "
> +				 "threshold value when reloading driver.\n");
> +			return true;
> +		}
>  	}
> 
>  	return false;
> --
> 2.35.1

RE: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in ras_eeprom_check_err

[Zhou1, Tao]

[AMD Official Use Only - General]

> -----Original Message-----
> From: Yang, Stanley <Stanley.Yang@amd.com>
> Sent: Tuesday, February 21, 2023 5:34 PM
> To: Zhou1, Tao <Tao.Zhou1@amd.com>; amd-gfx@lists.freedesktop.org; Zhang,
> Hawking <Hawking.Zhang@amd.com>; Chai, Thomas <YiPeng.Chai@amd.com>;
> Li, Candice <Candice.Li@amd.com>
> Subject: RE: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in
> ras_eeprom_check_err
>
> [AMD Official Use Only - General]
>
>
>
> > -----Original Message-----
> > From: Zhou1, Tao <Tao.Zhou1@amd.com>
> > Sent: Tuesday, February 21, 2023 4:29 PM
> > To: amd-gfx@lists.freedesktop.org; Zhang, Hawking
> > <Hawking.Zhang@amd.com>; Yang, Stanley <Stanley.Yang@amd.com>; Chai,
> > Thomas <YiPeng.Chai@amd.com>; Li, Candice <Candice.Li@amd.com>
> > Cc: Zhou1, Tao <Tao.Zhou1@amd.com>
> > Subject: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in
> > ras_eeprom_check_err
> >
> > bad_page_threshold controls page retirement behavior and it should be
> > also checked.
> >
> > Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
> > ---
> >  .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 20 ++++++++++++++-
> > ----
> >  1 file changed, 15 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> > b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> > index 9d370465b08d..c88123896fe8 100644
> > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> > @@ -417,7 +417,8 @@ bool
> > amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)  {
> >     struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
> >
> > -   if (!__is_ras_eeprom_supported(adev))
> > +   if (!__is_ras_eeprom_supported(adev) ||
> > +       !amdgpu_bad_page_threshold)
> >             return false;
> >
> >     /* skip check eeprom table for VEGA20 Gaming */ @@ -428,10
> > +429,19 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct
> > amdgpu_device *adev)
> >                     return false;
> >
> >     if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
> > -           dev_warn(adev->dev, "This GPU is in BAD status.");
> > -           dev_warn(adev->dev, "Please retire it or set a larger "
> > -                    "threshold value when reloading driver.\n");
> > -           return true;
> > +           if (amdgpu_bad_page_threshold == -1) {
> > +                   dev_warn(adev->dev, "RAS records:%d exceed
> > threshold:%d",
> > +                           con->eeprom_control.ras_num_recs, con-
> > >bad_page_cnt_threshold);
> > +                   dev_warn(adev->dev,
> > +                           "But GPU can be operated due to
> > bad_page_threshold = -1.\n");
> > +                   return false;
> > +           } else if (amdgpu_bad_page_threshold > 0 ||
> > +               amdgpu_bad_page_threshold == -2) {
>
> Stanley: it can't guarantee use to set amdgpu_bad_page_threshold value as
> expected for example -3, how about set this if condition as below

[Tao] Since "<= -2" and "> 0" can be treated as same thing here, will update the condition to "else".
The "-2" isn't retired, it indicates threshold number is calculated by driver.

>       else if (amdgpu_bad_page_threshold) {
>               ...
>       }
>       And in patch#1 the value -2 isn't need anymore.
>
> Regards,
> Stanley
> > +                   dev_warn(adev->dev, "This GPU is in BAD status.");
> > +                   dev_warn(adev->dev, "Please retire it or set a larger
> > "
> > +                            "threshold value when reloading driver.\n");
> > +                   return true;
> > +           }
> >     }
> >
> >     return false;
> > --
> > 2.35.1

RE: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in ras_eeprom_check_err

[Yang, Stanley]

[AMD Official Use Only - General]



> -----Original Message-----
> From: Zhou1, Tao <Tao.Zhou1@amd.com>
> Sent: Tuesday, February 21, 2023 4:29 PM
> To: amd-gfx@lists.freedesktop.org; Zhang, Hawking
> <Hawking.Zhang@amd.com>; Yang, Stanley <Stanley.Yang@amd.com>; Chai,
> Thomas <YiPeng.Chai@amd.com>; Li, Candice <Candice.Li@amd.com>
> Cc: Zhou1, Tao <Tao.Zhou1@amd.com>
> Subject: [PATCH 2/2] drm/amdgpu: add bad_page_threshold check in
> ras_eeprom_check_err
> 
> bad_page_threshold controls page retirement behavior and it should be also
> checked.
> 
> Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
> ---
>  .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 20 ++++++++++++++-
> ----
>  1 file changed, 15 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> index 9d370465b08d..c88123896fe8 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
> @@ -417,7 +417,8 @@ bool
> amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)  {
>  	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
> 
> -	if (!__is_ras_eeprom_supported(adev))
> +	if (!__is_ras_eeprom_supported(adev) ||
> +	    !amdgpu_bad_page_threshold)
>  		return false;
> 
>  	/* skip check eeprom table for VEGA20 Gaming */ @@ -428,10
> +429,19 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct
> amdgpu_device *adev)
>  			return false;
> 
>  	if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
> -		dev_warn(adev->dev, "This GPU is in BAD status.");
> -		dev_warn(adev->dev, "Please retire it or set a larger "
> -			 "threshold value when reloading driver.\n");
> -		return true;
> +		if (amdgpu_bad_page_threshold == -1) {
> +			dev_warn(adev->dev, "RAS records:%d exceed
> threshold:%d",
> +				con->eeprom_control.ras_num_recs, con-
> >bad_page_cnt_threshold);
> +			dev_warn(adev->dev,
> +				"But GPU can be operated due to
> bad_page_threshold = -1.\n");
> +			return false;
> +		} else if (amdgpu_bad_page_threshold > 0 ||
> +		    amdgpu_bad_page_threshold == -2) {

Stanley: it can't guarantee use to set amdgpu_bad_page_threshold value as expected for example -3, how about set this if condition as below
	else if (amdgpu_bad_page_threshold) {
		...
	}
	And in patch#1 the value -2 isn't need anymore.

Regards,
Stanley
> +			dev_warn(adev->dev, "This GPU is in BAD status.");
> +			dev_warn(adev->dev, "Please retire it or set a larger
> "
> +				 "threshold value when reloading driver.\n");
> +			return true;
> +		}
>  	}
> 
>  	return false;
> --
> 2.35.1

[PATCH 2/2] drm/amdgpu: add bad_page_threshold check in ras_eeprom_check_err

[Tao Zhou]

bad_page_threshold controls page retirement behavior and it should be
also checked.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
---
 .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 20 ++++++++++++++-----
 1 file changed, 15 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 9d370465b08d..c88123896fe8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -417,7 +417,8 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 
-	if (!__is_ras_eeprom_supported(adev))
+	if (!__is_ras_eeprom_supported(adev) ||
+	    !amdgpu_bad_page_threshold)
 		return false;
 
 	/* skip check eeprom table for VEGA20 Gaming */
@@ -428,10 +429,19 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev)
 			return false;
 
 	if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) {
-		dev_warn(adev->dev, "This GPU is in BAD status.");
-		dev_warn(adev->dev, "Please retire it or set a larger "
-			 "threshold value when reloading driver.\n");
-		return true;
+		if (amdgpu_bad_page_threshold == -1) {
+			dev_warn(adev->dev, "RAS records:%d exceed threshold:%d",
+				con->eeprom_control.ras_num_recs, con->bad_page_cnt_threshold);
+			dev_warn(adev->dev,
+				"But GPU can be operated due to bad_page_threshold = -1.\n");
+			return false;
+		} else if (amdgpu_bad_page_threshold > 0 ||
+		    amdgpu_bad_page_threshold == -2) {
+			dev_warn(adev->dev, "This GPU is in BAD status.");
+			dev_warn(adev->dev, "Please retire it or set a larger "
+				 "threshold value when reloading driver.\n");
+			return true;
+		}
 	}
 
 	return false;
-- 
2.35.1