vfio/mlx5: Enforce a single SAVE command at a time

Enforce a single SAVE command at a time.

As the SAVE command is an asynchronous one, we must enforce running only
a single command at a time.

This will preserve ordering between multiple calls and protect from
races on the migration file data structure.

This is a must for the next patches from the series where as part of
PRE_COPY we may have multiple images to be saved and multiple SAVE
commands may be issued from different flows.

Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Yishai Hadas <yishaih@nvidia.com>
Link: https://lore.kernel.org/r/20221206083438.37807-4-yishaih@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>

diff --git a/drivers/vfio/pci/mlx5/cmd.c b/drivers/vfio/pci/mlx5/cmd.c
index 0848bc9..55ee803 100644 (file)
--- a/drivers/vfio/pci/mlx5/cmd.c
+++ b/drivers/vfio/pci/mlx5/cmd.c
@@ -281,6 +281,7 @@ void mlx5vf_mig_file_cleanup_cb(struct work_struct *_work)
        dma_unmap_sgtable(mdev->device, &migf->table.sgt, DMA_FROM_DEVICE, 0);
        mlx5_core_dealloc_pd(mdev, async_data->pdn);
        kvfree(async_data->out);
+       complete(&migf->save_comp);
        fput(migf->filp);
 }
 
@@ -321,6 +322,10 @@ int mlx5vf_cmd_save_vhca_state(struct mlx5vf_pci_core_device *mvdev,
                return -ENOTCONN;
 
        mdev = mvdev->mdev;
+       err = wait_for_completion_interruptible(&migf->save_comp);
+       if (err)
+               return err;
+
        err = mlx5_core_alloc_pd(mdev, &pdn);
        if (err)
                return err;
@@ -371,6 +376,7 @@ err_create_mkey:
        dma_unmap_sgtable(mdev->device, &migf->table.sgt, DMA_FROM_DEVICE, 0);
 err_dma_map:
        mlx5_core_dealloc_pd(mdev, pdn);
+       complete(&migf->save_comp);
        return err;
 }
 

diff --git a/drivers/vfio/pci/mlx5/cmd.h b/drivers/vfio/pci/mlx5/cmd.h
index 921d572..8ffa769 100644 (file)
--- a/drivers/vfio/pci/mlx5/cmd.h
+++ b/drivers/vfio/pci/mlx5/cmd.h
@@ -37,6 +37,7 @@ struct mlx5_vf_migration_file {
        unsigned long last_offset;
        struct mlx5vf_pci_core_device *mvdev;
        wait_queue_head_t poll_wait;
+       struct completion save_comp;
        struct mlx5_async_ctx async_ctx;
        struct mlx5vf_async_data async_data;
 };

diff --git a/drivers/vfio/pci/mlx5/main.c b/drivers/vfio/pci/mlx5/main.c
index 6e9cf2a..0d71ebb 100644 (file)
--- a/drivers/vfio/pci/mlx5/main.c
+++ b/drivers/vfio/pci/mlx5/main.c
@@ -245,6 +245,13 @@ mlx5vf_pci_save_device_data(struct mlx5vf_pci_core_device *mvdev)
        stream_open(migf->filp->f_inode, migf->filp);
        mutex_init(&migf->lock);
        init_waitqueue_head(&migf->poll_wait);
+       init_completion(&migf->save_comp);
+       /*
+        * save_comp is being used as a binary semaphore built from
+        * a completion. A normal mutex cannot be used because the lock is
+        * passed between kernel threads and lockdep can't model this.
+        */
+       complete(&migf->save_comp);
        mlx5_cmd_init_async_ctx(mvdev->mdev, &migf->async_ctx);
        INIT_WORK(&migf->async_data.work, mlx5vf_mig_file_cleanup_cb);
        ret = mlx5vf_cmd_query_vhca_migration_state(mvdev,