Re: [PATCH v3 16/28] drm/xe/pf: Handle GuC migration data as part of PF control

[Michal Wajdeczko <michal.wajdeczko@intel.com>]

On 10/30/2025 9:31 PM, Michał Winiarski wrote:
> Connect the helpers to allow save and restore of GuC migration data in
> stop_copy / resume device state.
> 
> Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c   | 25 ++++++-
>  drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c | 66 +++++++++++++++----
>  drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h |  8 ++-
>  .../drm/xe/xe_gt_sriov_pf_migration_types.h   |  5 ++
>  4 files changed, 87 insertions(+), 17 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c
> index 203ba1fbeefcd..cb45e97f4c4d9 100644
> --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c
> +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_control.c
> @@ -848,6 +848,18 @@ static void pf_enter_vf_save_failed(struct xe_gt *gt, unsigned int vfid)
>  
>  static int pf_handle_vf_save_data(struct xe_gt *gt, unsigned int vfid)
>  {
> +	int ret;
> +
> +	if (xe_gt_sriov_pf_migration_save_test(gt, vfid, XE_SRIOV_MIGRATION_DATA_TYPE_GUC)) {
> +		ret = xe_gt_sriov_pf_migration_guc_save(gt, vfid);
> +		if (ret)
> +			return ret;
> +
> +		xe_gt_sriov_pf_migration_save_clear(gt, vfid, XE_SRIOV_MIGRATION_DATA_TYPE_GUC);
> +
> +		return -EAGAIN;
> +	}
> +
>  	return 0;
>  }
>  
> @@ -887,6 +899,7 @@ static void pf_exit_vf_save_wait_data(struct xe_gt *gt, unsigned int vfid)
>  static bool pf_enter_vf_save_wip(struct xe_gt *gt, unsigned int vfid)
>  {
>  	if (pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_WIP)) {
> +		xe_gt_sriov_pf_migration_save_init(gt, vfid);
>  		pf_enter_vf_state(gt, vfid, XE_GT_SRIOV_STATE_SAVE_PROCESS_DATA);
>  		pf_enter_vf_wip(gt, vfid);
>  		pf_queue_vf(gt, vfid);
> @@ -1048,14 +1061,22 @@ static void pf_enter_vf_restore_failed(struct xe_gt *gt, unsigned int vfid)
>  static int pf_handle_vf_restore_data(struct xe_gt *gt, unsigned int vfid)
>  {
>  	struct xe_sriov_migration_data *data = xe_gt_sriov_pf_migration_restore_consume(gt, vfid);
> +	int ret = 0;
>  
>  	xe_gt_assert(gt, data);
>  
> -	xe_gt_sriov_notice(gt, "Skipping VF%u unknown data type: %d\n", vfid, data->type);
> +	switch (data->type) {
> +	case XE_SRIOV_MIGRATION_DATA_TYPE_GUC:
> +		ret = xe_gt_sriov_pf_migration_guc_restore(gt, vfid, data);
> +		break;
> +	default:
> +		xe_gt_sriov_notice(gt, "Skipping VF%u unknown data type: %d\n", vfid, data->type);
> +		break;
> +	}
>  
>  	xe_sriov_migration_data_free(data);
>  
> -	return 0;
> +	return ret;
>  }
>  
>  static bool pf_handle_vf_restore(struct xe_gt *gt, unsigned int vfid)
> diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c
> index 4a716e0a29fe4..fbb451767712a 100644
> --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c
> +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.c
> @@ -174,22 +174,10 @@ static int pf_save_vf_guc_mig_data(struct xe_gt *gt, unsigned int vfid)
>  	return ret;
>  }
>  
> -/**
> - * xe_gt_sriov_pf_migration_guc_size() - Get the size of VF GuC migration data.
> - * @gt: the &xe_gt
> - * @vfid: the VF identifier
> - *
> - * This function is for PF only.
> - *
> - * Return: size in bytes or a negative error code on failure.
> - */
> -ssize_t xe_gt_sriov_pf_migration_guc_size(struct xe_gt *gt, unsigned int vfid)
> +static ssize_t pf_migration_guc_size(struct xe_gt *gt, unsigned int vfid)
>  {
>  	ssize_t size;
>  
> -	xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
> -	xe_gt_assert(gt, vfid != PFID);
> -	xe_gt_assert(gt, vfid <= xe_sriov_pf_get_totalvfs(gt_to_xe(gt)));
>  
>  	if (!pf_migration_supported(gt))
>  		return -ENOPKG;
> @@ -278,12 +266,19 @@ int xe_gt_sriov_pf_migration_guc_restore(struct xe_gt *gt, unsigned int vfid,
>  ssize_t xe_gt_sriov_pf_migration_size(struct xe_gt *gt, unsigned int vfid)
>  {
>  	ssize_t total = 0;
> +	ssize_t size;
>  
>  	xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
>  	xe_gt_assert(gt, vfid != PFID);
>  	xe_gt_assert(gt, vfid <= xe_sriov_pf_get_totalvfs(gt_to_xe(gt)));
>  
> -	/* Nothing to query yet - will be updated once per-GT migration data types are added */
> +	size = pf_migration_guc_size(gt, vfid);
> +	if (size < 0)
> +		return size;
> +	if (size > 0)
> +		size += sizeof(struct xe_sriov_pf_migration_hdr);
> +	total += size;
> +
>  	return total;
>  }
>  
> @@ -330,6 +325,49 @@ void xe_gt_sriov_pf_migration_ring_free(struct xe_gt *gt, unsigned int vfid)
>  		xe_sriov_migration_data_free(data);
>  }
>  
> +/**
> + * xe_gt_sriov_pf_migration_save_init() - Initialize per-GT migration related data.
> + * @gt: the &xe_gt
> + * @vfid: the VF identifier (can't be 0)
> + *
> + * Return: 0 on success or a negative error code on failure.

but it's void

> + */
> +void xe_gt_sriov_pf_migration_save_init(struct xe_gt *gt, unsigned int vfid)
> +{
> +	struct xe_gt_sriov_migration_data *migration = pf_pick_gt_migration(gt, vfid);
> +
> +	migration->save.data_remaining = 0;
> +
> +	xe_gt_assert(gt, pf_migration_guc_size(gt, vfid) > 0);
> +	set_bit(XE_SRIOV_MIGRATION_DATA_TYPE_GUC, &migration->save.data_remaining);
> +}
> +
> +/**
> + * xe_gt_sriov_pf_migration_save_test() - Test if migration data type needs to be saved.
> + * @gt: the &xe_gt
> + * @vfid: the VF identifier (can't be 0)
> + * @type: the &xe_sriov_migration_data_type of data to be saved

.. to check

> + *
> + * Return: true if the data needs saving, otherwise false.
> + */
> +bool xe_gt_sriov_pf_migration_save_test(struct xe_gt *gt, unsigned int vfid,
> +					enum xe_sriov_migration_data_type type)
> +{
> +	return test_bit(type, &pf_pick_gt_migration(gt, vfid)->save.data_remaining);
> +}
> +
> +/**
> + * xe_gt_sriov_pf_migration_save_clear() - Clear migration data type from saving.
> + * @gt: the &xe_gt
> + * @vfid: the VF identifier (can't be 0)
> + * @type: the &xe_sriov_migration_data_type of data to be cleared

.. to mark as done/saved

> + */
> +void xe_gt_sriov_pf_migration_save_clear(struct xe_gt *gt, unsigned int vfid,
> +					 enum xe_sriov_migration_data_type type)
> +{
> +	clear_bit(type, &pf_pick_gt_migration(gt, vfid)->save.data_remaining);
> +}
> +
>  /**
>   * xe_gt_sriov_pf_migration_save_produce() - Add VF save data packet to migration ring.
>   * @gt: the &xe_gt
> diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h
> index b3c18e369df79..66a45d6234245 100644
> --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h
> +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration.h
> @@ -10,12 +10,12 @@
>  
>  struct xe_gt;
>  struct xe_sriov_migration_data;
> +enum xe_sriov_migration_data_type;
>  
>  /* TODO: get this information by querying GuC in the future */
>  #define XE_GT_SRIOV_PF_MIGRATION_GUC_DATA_MAX_SIZE SZ_8M
>  
>  int xe_gt_sriov_pf_migration_init(struct xe_gt *gt);
> -ssize_t xe_gt_sriov_pf_migration_guc_size(struct xe_gt *gt, unsigned int vfid);
>  int xe_gt_sriov_pf_migration_guc_save(struct xe_gt *gt, unsigned int vfid);
>  int xe_gt_sriov_pf_migration_guc_restore(struct xe_gt *gt, unsigned int vfid,
>  					 struct xe_sriov_migration_data *data);
> @@ -26,6 +26,12 @@ bool xe_gt_sriov_pf_migration_ring_empty(struct xe_gt *gt, unsigned int vfid);
>  bool xe_gt_sriov_pf_migration_ring_full(struct xe_gt *gt, unsigned int vfid);
>  void xe_gt_sriov_pf_migration_ring_free(struct xe_gt *gt, unsigned int vfid);
>  
> +void xe_gt_sriov_pf_migration_save_init(struct xe_gt *gt, unsigned int vfid);
> +bool xe_gt_sriov_pf_migration_save_test(struct xe_gt *gt, unsigned int vfid,
> +					enum xe_sriov_migration_data_type type);
> +void xe_gt_sriov_pf_migration_save_clear(struct xe_gt *gt, unsigned int vfid,
> +					 enum xe_sriov_migration_data_type type);

nit: "clear" is little confusing, maybe

	xe_gt_sriov_pf_migration_save_data_done(gt, vfid, type);	// instead of save_clear
	xe_gt_sriov_pf_migration_save_data_pending(gt, vfid, type);	// instead of save_test
and 
	pf_migration_save_data_todo(gt, vfid, type);	// instead of set_bit


> +
>  int xe_gt_sriov_pf_migration_save_produce(struct xe_gt *gt, unsigned int vfid,
>  					  struct xe_sriov_migration_data *data);
>  struct xe_sriov_migration_data *
> diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration_types.h b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration_types.h
> index 75d8b94cbbefb..9f24878690d9c 100644
> --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration_types.h
> +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_migration_types.h
> @@ -16,6 +16,11 @@
>  struct xe_gt_sriov_migration_data {
>  	/** @ring: queue containing VF save / restore migration data */
>  	struct ptr_ring ring;
> +	/** @save: structure for currently processed save migration data */
> +	struct {
> +		/** @save.data_remaining: bitmap of migration types that need to be saved */
> +		unsigned long data_remaining;
> +	} save;
>  };
>  
>  #endif

otherwise LGTM, so with kernel-doc fixed, and (hopefully) with better names,

Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>