AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v3] drm/amdgpu: add badpages sysfs interafce
@ 2019-05-09 10:31 Pan, Xinhui
       [not found] ` <SN6PR12MB280055871BC84DB8396BDAFB87330-kxOKjb6HO/EqkY47FTA1ogdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
  0 siblings, 1 reply; 4+ messages in thread
From: Pan, Xinhui @ 2019-05-09 10:31 UTC (permalink / raw)
  To: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
  Cc: Deucher, Alexander

add badpages node.
it will output badpages list in format
gpu pfn : gpu page size : flags

example
0x00000000 : 0x00001000 : R
0x00000001 : 0x00001000 : R
0x00000002 : 0x00001000 : R
0x00000003 : 0x00001000 : R
0x00000004 : 0x00001000 : R
0x00000005 : 0x00001000 : R
0x00000006 : 0x00001000 : R
0x00000007 : 0x00001000 : P
0x00000008 : 0x00001000 : P
0x00000009 : 0x00001000 : P

flags can be one of below characters
R: reserved.
P: pending for reserve.
F: failed to reserve for some reasons.

Signed-off-by: xinhui pan <xinhui.pan@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 146 ++++++++++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h |   1 +
 2 files changed, 147 insertions(+)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index c60d5f813801..c9e24f60938e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -90,6 +90,12 @@ struct ras_manager {
 	struct ras_err_data err_data;
 };
 
+struct ras_badpage {
+	unsigned int bp;
+	unsigned int size;
+	unsigned int flags;
+};
+
 const char *ras_error_string[] = {
 	"none",
 	"parity",
@@ -710,6 +716,77 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
 
 /* sysfs begin */
 
+static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
+		struct ras_badpage **bps, unsigned int *count);
+
+static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
+{
+	switch (flags) {
+	case 0:
+		return "R";
+	case 1:
+		return "P";
+	case 2:
+	default:
+		return "F";
+	};
+}
+
+/*
+ * DOC: ras sysfs gpu_vram_bad_pages interface
+ *
+ * It allows user to read the bad pages of vram on the gpu through
+ * /sys/class/drm/card[0/1/2...]/device/ras/gpu_vram_bad_pages
+ *
+ * It outputs multiple lines, and each line stands for one gpu page.
+ *
+ * The format of one line is below,
+ * gpu pfn : gpu page size : flags
+ *
+ * gpu pfn and gpu page size are printed in hex format.
+ * flags can be one of below character,
+ * R: reserved, this gpu page is reserved and not able to use.
+ * P: pending for reserve, this gpu page is marked as bad, will be reserved
+ *    in next window of page_reserve.
+ * F: unable to reserve. this gpu page can't be reserved due to some reasons.
+ *
+ * examples:
+ * 0x00000001 : 0x00001000 : R
+ * 0x00000002 : 0x00001000 : P
+ */
+
+static ssize_t amdgpu_ras_sysfs_badpages_read(struct file *f,
+		struct kobject *kobj, struct bin_attribute *attr,
+		char *buf, loff_t ppos, size_t count)
+{
+	struct amdgpu_ras *con =
+		container_of(attr, struct amdgpu_ras, badpages_attr);
+	struct amdgpu_device *adev = con->adev;
+	const unsigned int element_size =
+		sizeof("0xabcdabcd : 0x12345678 : R\n") - 1;
+	unsigned int start = (ppos + element_size - 1) / element_size;
+	unsigned int end = (ppos + count - 1) / element_size;
+	ssize_t s = 0;
+	struct ras_badpage *bps = NULL;
+	unsigned int bps_count = 0;
+
+	memset(buf, 0, count);
+
+	if (amdgpu_ras_badpages_read(adev, &bps, &bps_count))
+		return 0;
+
+	for (; start < end && start < bps_count; start++)
+		s += scnprintf(&buf[s], element_size + 1,
+				"0x%08x : 0x%08x : %1s\n",
+				bps[start].bp,
+				bps[start].size,
+				amdgpu_ras_badpage_flags_str(bps[start].flags));
+
+	kfree(bps);
+
+	return s;
+}
+
 static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
@@ -750,9 +827,14 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
 		&con->features_attr.attr,
 		NULL
 	};
+	struct bin_attribute *bin_attrs[] = {
+		&con->badpages_attr,
+		NULL
+	};
 	struct attribute_group group = {
 		.name = "ras",
 		.attrs = attrs,
+		.bin_attrs = bin_attrs,
 	};
 
 	con->features_attr = (struct device_attribute) {
@@ -762,7 +844,19 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
 		},
 			.show = amdgpu_ras_sysfs_features_read,
 	};
+
+	con->badpages_attr = (struct bin_attribute) {
+		.attr = {
+			.name = "gpu_vram_bad_pages",
+			.mode = S_IRUGO,
+		},
+		.size = 0,
+		.private = NULL,
+		.read = amdgpu_ras_sysfs_badpages_read,
+	};
+
 	sysfs_attr_init(attrs[0]);
+	sysfs_bin_attr_init(bin_attrs[0]);
 
 	return sysfs_create_group(&adev->dev->kobj, &group);
 }
@@ -774,9 +868,14 @@ static int amdgpu_ras_sysfs_remove_feature_node(struct amdgpu_device *adev)
 		&con->features_attr.attr,
 		NULL
 	};
+	struct bin_attribute *bin_attrs[] = {
+		&con->badpages_attr,
+		NULL
+	};
 	struct attribute_group group = {
 		.name = "ras",
 		.attrs = attrs,
+		.bin_attrs = bin_attrs,
 	};
 
 	sysfs_remove_group(&adev->dev->kobj, &group);
@@ -1108,6 +1207,53 @@ static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev)
 /* ih end */
 
 /* recovery begin */
+
+/* return 0 on success.
+ * caller need free bps.
+ */
+static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
+		struct ras_badpage **bps, unsigned int *count)
+{
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	struct ras_err_handler_data *data;
+	int i = 0;
+	int ret = 0;
+
+	if (!con || !con->eh_data || !bps || !count)
+		return -EINVAL;
+
+	mutex_lock(&con->recovery_lock);
+	data = con->eh_data;
+	if (!data || data->count == 0) {
+		*bps = NULL;
+		goto out;
+	}
+
+	*bps = kmalloc(sizeof(struct ras_badpage) * data->count, GFP_KERNEL);
+	if (!*bps) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	for (; i < data->count; i++) {
+		(*bps)[i] = (struct ras_badpage){
+			.bp = data->bps[i].bp,
+			.size = AMDGPU_GPU_PAGE_SIZE,
+			.flags = 0,
+		};
+
+		if (data->last_reserved <= i)
+			(*bps)[i].flags = 1;
+		else if (data->bps[i].bo == NULL)
+			(*bps)[i].flags = 2;
+	}
+
+	*count = data->count;
+out:
+	mutex_unlock(&con->recovery_lock);
+	return ret;
+}
+
 static void amdgpu_ras_do_recovery(struct work_struct *work)
 {
 	struct amdgpu_ras *ras =
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index 065c66baf947..e2dff00b8d1c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -93,6 +93,7 @@ struct amdgpu_ras {
 	struct dentry *ent;
 	/* sysfs */
 	struct device_attribute features_attr;
+	struct bin_attribute badpages_attr;
 	/* block array */
 	struct ras_manager *objs;
 
-- 
2.17.1

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH v3] drm/amdgpu: add badpages sysfs interafce
       [not found] ` <SN6PR12MB280055871BC84DB8396BDAFB87330-kxOKjb6HO/EqkY47FTA1ogdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
@ 2019-05-09 15:24   ` Alex Deucher
       [not found]     ` <CADnq5_MF_MQw=4aA2hW_agDhfwcsZX-BrW0TEO2Q7Bh9N2Z+7Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
  2019-05-17  7:19   ` [v3] " Nathan Chancellor
  1 sibling, 1 reply; 4+ messages in thread
From: Alex Deucher @ 2019-05-09 15:24 UTC (permalink / raw)
  To: Pan, Xinhui
  Cc: Deucher, Alexander,
	amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org

On Thu, May 9, 2019 at 6:31 AM Pan, Xinhui <Xinhui.Pan@amd.com> wrote:
>
> add badpages node.
> it will output badpages list in format
> gpu pfn : gpu page size : flags
>
> example
> 0x00000000 : 0x00001000 : R
> 0x00000001 : 0x00001000 : R
> 0x00000002 : 0x00001000 : R
> 0x00000003 : 0x00001000 : R
> 0x00000004 : 0x00001000 : R
> 0x00000005 : 0x00001000 : R
> 0x00000006 : 0x00001000 : R
> 0x00000007 : 0x00001000 : P
> 0x00000008 : 0x00001000 : P
> 0x00000009 : 0x00001000 : P
>
> flags can be one of below characters
> R: reserved.
> P: pending for reserve.
> F: failed to reserve for some reasons.
>
> Signed-off-by: xinhui pan <xinhui.pan@amd.com>

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>

> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 146 ++++++++++++++++++++++++
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h |   1 +
>  2 files changed, 147 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> index c60d5f813801..c9e24f60938e 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> @@ -90,6 +90,12 @@ struct ras_manager {
>         struct ras_err_data err_data;
>  };
>
> +struct ras_badpage {
> +       unsigned int bp;
> +       unsigned int size;
> +       unsigned int flags;
> +};
> +
>  const char *ras_error_string[] = {
>         "none",
>         "parity",
> @@ -710,6 +716,77 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
>
>  /* sysfs begin */
>
> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
> +               struct ras_badpage **bps, unsigned int *count);
> +
> +static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
> +{
> +       switch (flags) {
> +       case 0:
> +               return "R";
> +       case 1:
> +               return "P";
> +       case 2:
> +       default:
> +               return "F";
> +       };
> +}
> +
> +/*
> + * DOC: ras sysfs gpu_vram_bad_pages interface
> + *
> + * It allows user to read the bad pages of vram on the gpu through
> + * /sys/class/drm/card[0/1/2...]/device/ras/gpu_vram_bad_pages
> + *
> + * It outputs multiple lines, and each line stands for one gpu page.
> + *
> + * The format of one line is below,
> + * gpu pfn : gpu page size : flags
> + *
> + * gpu pfn and gpu page size are printed in hex format.
> + * flags can be one of below character,
> + * R: reserved, this gpu page is reserved and not able to use.
> + * P: pending for reserve, this gpu page is marked as bad, will be reserved
> + *    in next window of page_reserve.
> + * F: unable to reserve. this gpu page can't be reserved due to some reasons.
> + *
> + * examples:
> + * 0x00000001 : 0x00001000 : R
> + * 0x00000002 : 0x00001000 : P
> + */
> +
> +static ssize_t amdgpu_ras_sysfs_badpages_read(struct file *f,
> +               struct kobject *kobj, struct bin_attribute *attr,
> +               char *buf, loff_t ppos, size_t count)
> +{
> +       struct amdgpu_ras *con =
> +               container_of(attr, struct amdgpu_ras, badpages_attr);
> +       struct amdgpu_device *adev = con->adev;
> +       const unsigned int element_size =
> +               sizeof("0xabcdabcd : 0x12345678 : R\n") - 1;
> +       unsigned int start = (ppos + element_size - 1) / element_size;
> +       unsigned int end = (ppos + count - 1) / element_size;
> +       ssize_t s = 0;
> +       struct ras_badpage *bps = NULL;
> +       unsigned int bps_count = 0;
> +
> +       memset(buf, 0, count);
> +
> +       if (amdgpu_ras_badpages_read(adev, &bps, &bps_count))
> +               return 0;
> +
> +       for (; start < end && start < bps_count; start++)
> +               s += scnprintf(&buf[s], element_size + 1,
> +                               "0x%08x : 0x%08x : %1s\n",
> +                               bps[start].bp,
> +                               bps[start].size,
> +                               amdgpu_ras_badpage_flags_str(bps[start].flags));
> +
> +       kfree(bps);
> +
> +       return s;
> +}
> +
>  static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev,
>                 struct device_attribute *attr, char *buf)
>  {
> @@ -750,9 +827,14 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>                 &con->features_attr.attr,
>                 NULL
>         };
> +       struct bin_attribute *bin_attrs[] = {
> +               &con->badpages_attr,
> +               NULL
> +       };
>         struct attribute_group group = {
>                 .name = "ras",
>                 .attrs = attrs,
> +               .bin_attrs = bin_attrs,
>         };
>
>         con->features_attr = (struct device_attribute) {
> @@ -762,7 +844,19 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>                 },
>                         .show = amdgpu_ras_sysfs_features_read,
>         };
> +
> +       con->badpages_attr = (struct bin_attribute) {
> +               .attr = {
> +                       .name = "gpu_vram_bad_pages",
> +                       .mode = S_IRUGO,
> +               },
> +               .size = 0,
> +               .private = NULL,
> +               .read = amdgpu_ras_sysfs_badpages_read,
> +       };
> +
>         sysfs_attr_init(attrs[0]);
> +       sysfs_bin_attr_init(bin_attrs[0]);
>
>         return sysfs_create_group(&adev->dev->kobj, &group);
>  }
> @@ -774,9 +868,14 @@ static int amdgpu_ras_sysfs_remove_feature_node(struct amdgpu_device *adev)
>                 &con->features_attr.attr,
>                 NULL
>         };
> +       struct bin_attribute *bin_attrs[] = {
> +               &con->badpages_attr,
> +               NULL
> +       };
>         struct attribute_group group = {
>                 .name = "ras",
>                 .attrs = attrs,
> +               .bin_attrs = bin_attrs,
>         };
>
>         sysfs_remove_group(&adev->dev->kobj, &group);
> @@ -1108,6 +1207,53 @@ static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev)
>  /* ih end */
>
>  /* recovery begin */
> +
> +/* return 0 on success.
> + * caller need free bps.
> + */
> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
> +               struct ras_badpage **bps, unsigned int *count)
> +{
> +       struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
> +       struct ras_err_handler_data *data;
> +       int i = 0;
> +       int ret = 0;
> +
> +       if (!con || !con->eh_data || !bps || !count)
> +               return -EINVAL;
> +
> +       mutex_lock(&con->recovery_lock);
> +       data = con->eh_data;
> +       if (!data || data->count == 0) {
> +               *bps = NULL;
> +               goto out;
> +       }
> +
> +       *bps = kmalloc(sizeof(struct ras_badpage) * data->count, GFP_KERNEL);
> +       if (!*bps) {
> +               ret = -ENOMEM;
> +               goto out;
> +       }
> +
> +       for (; i < data->count; i++) {
> +               (*bps)[i] = (struct ras_badpage){
> +                       .bp = data->bps[i].bp,
> +                       .size = AMDGPU_GPU_PAGE_SIZE,
> +                       .flags = 0,
> +               };
> +
> +               if (data->last_reserved <= i)
> +                       (*bps)[i].flags = 1;
> +               else if (data->bps[i].bo == NULL)
> +                       (*bps)[i].flags = 2;
> +       }
> +
> +       *count = data->count;
> +out:
> +       mutex_unlock(&con->recovery_lock);
> +       return ret;
> +}
> +
>  static void amdgpu_ras_do_recovery(struct work_struct *work)
>  {
>         struct amdgpu_ras *ras =
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> index 065c66baf947..e2dff00b8d1c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> @@ -93,6 +93,7 @@ struct amdgpu_ras {
>         struct dentry *ent;
>         /* sysfs */
>         struct device_attribute features_attr;
> +       struct bin_attribute badpages_attr;
>         /* block array */
>         struct ras_manager *objs;
>
> --
> 2.17.1
>
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v3] drm/amdgpu: add badpages sysfs interafce
       [not found]     ` <CADnq5_MF_MQw=4aA2hW_agDhfwcsZX-BrW0TEO2Q7Bh9N2Z+7Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
@ 2019-05-09 15:45       ` William Lewis
  0 siblings, 0 replies; 4+ messages in thread
From: William Lewis @ 2019-05-09 15:45 UTC (permalink / raw)
  To: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org

Typo in the patch description.  s/interafce/interface/

On 5/9/19 10:24 AM, Alex Deucher wrote:
> On Thu, May 9, 2019 at 6:31 AM Pan, Xinhui <Xinhui.Pan@amd.com> wrote:
>> add badpages node.
>> it will output badpages list in format
>> gpu pfn : gpu page size : flags
>>
>> example
>> 0x00000000 : 0x00001000 : R
>> 0x00000001 : 0x00001000 : R
>> 0x00000002 : 0x00001000 : R
>> 0x00000003 : 0x00001000 : R
>> 0x00000004 : 0x00001000 : R
>> 0x00000005 : 0x00001000 : R
>> 0x00000006 : 0x00001000 : R
>> 0x00000007 : 0x00001000 : P
>> 0x00000008 : 0x00001000 : P
>> 0x00000009 : 0x00001000 : P
>>
>> flags can be one of below characters
>> R: reserved.
>> P: pending for reserve.
>> F: failed to reserve for some reasons.
>>
>> Signed-off-by: xinhui pan <xinhui.pan@amd.com>
> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
>
>> ---
>>   drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 146 ++++++++++++++++++++++++
>>   drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h |   1 +
>>   2 files changed, 147 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
>> index c60d5f813801..c9e24f60938e 100644
>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
>> @@ -90,6 +90,12 @@ struct ras_manager {
>>          struct ras_err_data err_data;
>>   };
>>
>> +struct ras_badpage {
>> +       unsigned int bp;
>> +       unsigned int size;
>> +       unsigned int flags;
>> +};
>> +
>>   const char *ras_error_string[] = {
>>          "none",
>>          "parity",
>> @@ -710,6 +716,77 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
>>
>>   /* sysfs begin */
>>
>> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
>> +               struct ras_badpage **bps, unsigned int *count);
>> +
>> +static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
>> +{
>> +       switch (flags) {
>> +       case 0:
>> +               return "R";
>> +       case 1:
>> +               return "P";
>> +       case 2:
>> +       default:
>> +               return "F";
>> +       };
>> +}
>> +
>> +/*
>> + * DOC: ras sysfs gpu_vram_bad_pages interface
>> + *
>> + * It allows user to read the bad pages of vram on the gpu through
>> + * /sys/class/drm/card[0/1/2...]/device/ras/gpu_vram_bad_pages
>> + *
>> + * It outputs multiple lines, and each line stands for one gpu page.
>> + *
>> + * The format of one line is below,
>> + * gpu pfn : gpu page size : flags
>> + *
>> + * gpu pfn and gpu page size are printed in hex format.
>> + * flags can be one of below character,
>> + * R: reserved, this gpu page is reserved and not able to use.
>> + * P: pending for reserve, this gpu page is marked as bad, will be reserved
>> + *    in next window of page_reserve.
>> + * F: unable to reserve. this gpu page can't be reserved due to some reasons.
>> + *
>> + * examples:
>> + * 0x00000001 : 0x00001000 : R
>> + * 0x00000002 : 0x00001000 : P
>> + */
>> +
>> +static ssize_t amdgpu_ras_sysfs_badpages_read(struct file *f,
>> +               struct kobject *kobj, struct bin_attribute *attr,
>> +               char *buf, loff_t ppos, size_t count)
>> +{
>> +       struct amdgpu_ras *con =
>> +               container_of(attr, struct amdgpu_ras, badpages_attr);
>> +       struct amdgpu_device *adev = con->adev;
>> +       const unsigned int element_size =
>> +               sizeof("0xabcdabcd : 0x12345678 : R\n") - 1;
>> +       unsigned int start = (ppos + element_size - 1) / element_size;
>> +       unsigned int end = (ppos + count - 1) / element_size;
>> +       ssize_t s = 0;
>> +       struct ras_badpage *bps = NULL;
>> +       unsigned int bps_count = 0;
>> +
>> +       memset(buf, 0, count);
>> +
>> +       if (amdgpu_ras_badpages_read(adev, &bps, &bps_count))
>> +               return 0;
>> +
>> +       for (; start < end && start < bps_count; start++)
>> +               s += scnprintf(&buf[s], element_size + 1,
>> +                               "0x%08x : 0x%08x : %1s\n",
>> +                               bps[start].bp,
>> +                               bps[start].size,
>> +                               amdgpu_ras_badpage_flags_str(bps[start].flags));
>> +
>> +       kfree(bps);
>> +
>> +       return s;
>> +}
>> +
>>   static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev,
>>                  struct device_attribute *attr, char *buf)
>>   {
>> @@ -750,9 +827,14 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>>                  &con->features_attr.attr,
>>                  NULL
>>          };
>> +       struct bin_attribute *bin_attrs[] = {
>> +               &con->badpages_attr,
>> +               NULL
>> +       };
>>          struct attribute_group group = {
>>                  .name = "ras",
>>                  .attrs = attrs,
>> +               .bin_attrs = bin_attrs,
>>          };
>>
>>          con->features_attr = (struct device_attribute) {
>> @@ -762,7 +844,19 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>>                  },
>>                          .show = amdgpu_ras_sysfs_features_read,
>>          };
>> +
>> +       con->badpages_attr = (struct bin_attribute) {
>> +               .attr = {
>> +                       .name = "gpu_vram_bad_pages",
>> +                       .mode = S_IRUGO,
>> +               },
>> +               .size = 0,
>> +               .private = NULL,
>> +               .read = amdgpu_ras_sysfs_badpages_read,
>> +       };
>> +
>>          sysfs_attr_init(attrs[0]);
>> +       sysfs_bin_attr_init(bin_attrs[0]);
>>
>>          return sysfs_create_group(&adev->dev->kobj, &group);
>>   }
>> @@ -774,9 +868,14 @@ static int amdgpu_ras_sysfs_remove_feature_node(struct amdgpu_device *adev)
>>                  &con->features_attr.attr,
>>                  NULL
>>          };
>> +       struct bin_attribute *bin_attrs[] = {
>> +               &con->badpages_attr,
>> +               NULL
>> +       };
>>          struct attribute_group group = {
>>                  .name = "ras",
>>                  .attrs = attrs,
>> +               .bin_attrs = bin_attrs,
>>          };
>>
>>          sysfs_remove_group(&adev->dev->kobj, &group);
>> @@ -1108,6 +1207,53 @@ static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev)
>>   /* ih end */
>>
>>   /* recovery begin */
>> +
>> +/* return 0 on success.
>> + * caller need free bps.
>> + */
>> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
>> +               struct ras_badpage **bps, unsigned int *count)
>> +{
>> +       struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
>> +       struct ras_err_handler_data *data;
>> +       int i = 0;
>> +       int ret = 0;
>> +
>> +       if (!con || !con->eh_data || !bps || !count)
>> +               return -EINVAL;
>> +
>> +       mutex_lock(&con->recovery_lock);
>> +       data = con->eh_data;
>> +       if (!data || data->count == 0) {
>> +               *bps = NULL;
>> +               goto out;
>> +       }
>> +
>> +       *bps = kmalloc(sizeof(struct ras_badpage) * data->count, GFP_KERNEL);
>> +       if (!*bps) {
>> +               ret = -ENOMEM;
>> +               goto out;
>> +       }
>> +
>> +       for (; i < data->count; i++) {
>> +               (*bps)[i] = (struct ras_badpage){
>> +                       .bp = data->bps[i].bp,
>> +                       .size = AMDGPU_GPU_PAGE_SIZE,
>> +                       .flags = 0,
>> +               };
>> +
>> +               if (data->last_reserved <= i)
>> +                       (*bps)[i].flags = 1;
>> +               else if (data->bps[i].bo == NULL)
>> +                       (*bps)[i].flags = 2;
>> +       }
>> +
>> +       *count = data->count;
>> +out:
>> +       mutex_unlock(&con->recovery_lock);
>> +       return ret;
>> +}
>> +
>>   static void amdgpu_ras_do_recovery(struct work_struct *work)
>>   {
>>          struct amdgpu_ras *ras =
>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
>> index 065c66baf947..e2dff00b8d1c 100644
>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
>> @@ -93,6 +93,7 @@ struct amdgpu_ras {
>>          struct dentry *ent;
>>          /* sysfs */
>>          struct device_attribute features_attr;
>> +       struct bin_attribute badpages_attr;
>>          /* block array */
>>          struct ras_manager *objs;
>>
>> --
>> 2.17.1
>>
>> _______________________________________________
>> amd-gfx mailing list
>> amd-gfx@lists.freedesktop.org
>> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Famd-gfx&amp;data=02%7C01%7C%7Cae69f0d4b0df49f576e008d6d4926c9c%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636930122677737006&amp;sdata=2kIXvUHQfN4G%2BJ0b7FqX4A1x7qXyTIlv3WMErXUtpgY%3D&amp;reserved=0
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@lists.freedesktop.org
> https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Famd-gfx&amp;data=02%7C01%7C%7Cae69f0d4b0df49f576e008d6d4926c9c%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636930122677747017&amp;sdata=Evfrt6pS%2B3mEGF7jH6PCzkc0Y8UArseY4GJuhr702Rs%3D&amp;reserved=0
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [v3] drm/amdgpu: add badpages sysfs interafce
       [not found] ` <SN6PR12MB280055871BC84DB8396BDAFB87330-kxOKjb6HO/EqkY47FTA1ogdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
  2019-05-09 15:24   ` Alex Deucher
@ 2019-05-17  7:19   ` Nathan Chancellor
  1 sibling, 0 replies; 4+ messages in thread
From: Nathan Chancellor @ 2019-05-17  7:19 UTC (permalink / raw)
  To: Pan, Xinhui
  Cc: Deucher, Alexander,
	amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org

On Thu, May 09, 2019 at 10:31:05AM +0000, Pan, Xinhui wrote:
> add badpages node.
> it will output badpages list in format
> gpu pfn : gpu page size : flags
> 
> example
> 0x00000000 : 0x00001000 : R
> 0x00000001 : 0x00001000 : R
> 0x00000002 : 0x00001000 : R
> 0x00000003 : 0x00001000 : R
> 0x00000004 : 0x00001000 : R
> 0x00000005 : 0x00001000 : R
> 0x00000006 : 0x00001000 : R
> 0x00000007 : 0x00001000 : P
> 0x00000008 : 0x00001000 : P
> 0x00000009 : 0x00001000 : P
> 
> flags can be one of below characters
> R: reserved.
> P: pending for reserve.
> F: failed to reserve for some reasons.
> 
> Signed-off-by: xinhui pan <xinhui.pan@amd.com>
> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 146 ++++++++++++++++++++++++
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h |   1 +
>  2 files changed, 147 insertions(+)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> index c60d5f813801..c9e24f60938e 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
> @@ -90,6 +90,12 @@ struct ras_manager {
>  	struct ras_err_data err_data;
>  };
>  
> +struct ras_badpage {
> +	unsigned int bp;
> +	unsigned int size;
> +	unsigned int flags;
> +};
> +
>  const char *ras_error_string[] = {
>  	"none",
>  	"parity",
> @@ -710,6 +716,77 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev,
>  
>  /* sysfs begin */
>  
> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
> +		struct ras_badpage **bps, unsigned int *count);
> +
> +static char *amdgpu_ras_badpage_flags_str(unsigned int flags)
> +{
> +	switch (flags) {
> +	case 0:
> +		return "R";
> +	case 1:
> +		return "P";
> +	case 2:
> +	default:
> +		return "F";
> +	};
> +}
> +
> +/*
> + * DOC: ras sysfs gpu_vram_bad_pages interface
> + *
> + * It allows user to read the bad pages of vram on the gpu through
> + * /sys/class/drm/card[0/1/2...]/device/ras/gpu_vram_bad_pages
> + *
> + * It outputs multiple lines, and each line stands for one gpu page.
> + *
> + * The format of one line is below,
> + * gpu pfn : gpu page size : flags
> + *
> + * gpu pfn and gpu page size are printed in hex format.
> + * flags can be one of below character,
> + * R: reserved, this gpu page is reserved and not able to use.
> + * P: pending for reserve, this gpu page is marked as bad, will be reserved
> + *    in next window of page_reserve.
> + * F: unable to reserve. this gpu page can't be reserved due to some reasons.
> + *
> + * examples:
> + * 0x00000001 : 0x00001000 : R
> + * 0x00000002 : 0x00001000 : P
> + */
> +
> +static ssize_t amdgpu_ras_sysfs_badpages_read(struct file *f,
> +		struct kobject *kobj, struct bin_attribute *attr,
> +		char *buf, loff_t ppos, size_t count)
> +{
> +	struct amdgpu_ras *con =
> +		container_of(attr, struct amdgpu_ras, badpages_attr);
> +	struct amdgpu_device *adev = con->adev;
> +	const unsigned int element_size =
> +		sizeof("0xabcdabcd : 0x12345678 : R\n") - 1;
> +	unsigned int start = (ppos + element_size - 1) / element_size;
> +	unsigned int end = (ppos + count - 1) / element_size;

I believe these two lines cause a link time error with arm32 defconfig +
CONFIG_DRM_AMDGPU (filtered down from allyesconfig):

arm-linux-gnueabi-ld: drivers/gpu/drm/amd/amdgpu/amdgpu_ras.o: in function `amdgpu_ras_sysfs_badpages_read':
amdgpu_ras.c:(.text+0x804): undefined reference to `__aeabi_ldivmod'
arm-linux-gnueabi-ld: amdgpu_ras.c:(.text+0x830): undefined reference to `__aeabi_ldivmod'

The assignments of start and end involve a 64-bit dividend because loff_t
is defined as long long, meaning one of the 64-bit division functions
from include/linux/math64.h should be used. I am not sure of which one
otherwise I would have sent a patch :)

Cheers,
Nathan

> +	ssize_t s = 0;
> +	struct ras_badpage *bps = NULL;
> +	unsigned int bps_count = 0;
> +
> +	memset(buf, 0, count);
> +
> +	if (amdgpu_ras_badpages_read(adev, &bps, &bps_count))
> +		return 0;
> +
> +	for (; start < end && start < bps_count; start++)
> +		s += scnprintf(&buf[s], element_size + 1,
> +				"0x%08x : 0x%08x : %1s\n",
> +				bps[start].bp,
> +				bps[start].size,
> +				amdgpu_ras_badpage_flags_str(bps[start].flags));
> +
> +	kfree(bps);
> +
> +	return s;
> +}
> +
>  static ssize_t amdgpu_ras_sysfs_features_read(struct device *dev,
>  		struct device_attribute *attr, char *buf)
>  {
> @@ -750,9 +827,14 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>  		&con->features_attr.attr,
>  		NULL
>  	};
> +	struct bin_attribute *bin_attrs[] = {
> +		&con->badpages_attr,
> +		NULL
> +	};
>  	struct attribute_group group = {
>  		.name = "ras",
>  		.attrs = attrs,
> +		.bin_attrs = bin_attrs,
>  	};
>  
>  	con->features_attr = (struct device_attribute) {
> @@ -762,7 +844,19 @@ static int amdgpu_ras_sysfs_create_feature_node(struct amdgpu_device *adev)
>  		},
>  			.show = amdgpu_ras_sysfs_features_read,
>  	};
> +
> +	con->badpages_attr = (struct bin_attribute) {
> +		.attr = {
> +			.name = "gpu_vram_bad_pages",
> +			.mode = S_IRUGO,
> +		},
> +		.size = 0,
> +		.private = NULL,
> +		.read = amdgpu_ras_sysfs_badpages_read,
> +	};
> +
>  	sysfs_attr_init(attrs[0]);
> +	sysfs_bin_attr_init(bin_attrs[0]);
>  
>  	return sysfs_create_group(&adev->dev->kobj, &group);
>  }
> @@ -774,9 +868,14 @@ static int amdgpu_ras_sysfs_remove_feature_node(struct amdgpu_device *adev)
>  		&con->features_attr.attr,
>  		NULL
>  	};
> +	struct bin_attribute *bin_attrs[] = {
> +		&con->badpages_attr,
> +		NULL
> +	};
>  	struct attribute_group group = {
>  		.name = "ras",
>  		.attrs = attrs,
> +		.bin_attrs = bin_attrs,
>  	};
>  
>  	sysfs_remove_group(&adev->dev->kobj, &group);
> @@ -1108,6 +1207,53 @@ static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev)
>  /* ih end */
>  
>  /* recovery begin */
> +
> +/* return 0 on success.
> + * caller need free bps.
> + */
> +static int amdgpu_ras_badpages_read(struct amdgpu_device *adev,
> +		struct ras_badpage **bps, unsigned int *count)
> +{
> +	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
> +	struct ras_err_handler_data *data;
> +	int i = 0;
> +	int ret = 0;
> +
> +	if (!con || !con->eh_data || !bps || !count)
> +		return -EINVAL;
> +
> +	mutex_lock(&con->recovery_lock);
> +	data = con->eh_data;
> +	if (!data || data->count == 0) {
> +		*bps = NULL;
> +		goto out;
> +	}
> +
> +	*bps = kmalloc(sizeof(struct ras_badpage) * data->count, GFP_KERNEL);
> +	if (!*bps) {
> +		ret = -ENOMEM;
> +		goto out;
> +	}
> +
> +	for (; i < data->count; i++) {
> +		(*bps)[i] = (struct ras_badpage){
> +			.bp = data->bps[i].bp,
> +			.size = AMDGPU_GPU_PAGE_SIZE,
> +			.flags = 0,
> +		};
> +
> +		if (data->last_reserved <= i)
> +			(*bps)[i].flags = 1;
> +		else if (data->bps[i].bo == NULL)
> +			(*bps)[i].flags = 2;
> +	}
> +
> +	*count = data->count;
> +out:
> +	mutex_unlock(&con->recovery_lock);
> +	return ret;
> +}
> +
>  static void amdgpu_ras_do_recovery(struct work_struct *work)
>  {
>  	struct amdgpu_ras *ras =
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> index 065c66baf947..e2dff00b8d1c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
> @@ -93,6 +93,7 @@ struct amdgpu_ras {
>  	struct dentry *ent;
>  	/* sysfs */
>  	struct device_attribute features_attr;
> +	struct bin_attribute badpages_attr;
>  	/* block array */
>  	struct ras_manager *objs;
>  
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2019-05-17  7:19 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-05-09 10:31 [PATCH v3] drm/amdgpu: add badpages sysfs interafce Pan, Xinhui
     [not found] ` <SN6PR12MB280055871BC84DB8396BDAFB87330-kxOKjb6HO/EqkY47FTA1ogdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
2019-05-09 15:24   ` Alex Deucher
     [not found]     ` <CADnq5_MF_MQw=4aA2hW_agDhfwcsZX-BrW0TEO2Q7Bh9N2Z+7Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2019-05-09 15:45       ` William Lewis
2019-05-17  7:19   ` [v3] " Nathan Chancellor

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox