Commit f0ae4afe authored Nov 22, 2021 by Alaa Hleihel Committed by Jason Gunthorpe Nov 25, 2021

RDMA/mlx5: Fix releasing unallocated memory in dereg MR flow

For the case of IB_MR_TYPE_DM the mr does doesn't have a umem, even though
it is a user MR. This causes function mlx5_free_priv_descs() to think that
it is a kernel MR, leading to wrongly accessing mr->descs that will get
wrong values in the union which leads to attempt to release resources that
were not allocated in the first place.

For example:
 DMA-API: mlx5_core 0000:08:00.1: device driver tries to free DMA memory it has not allocated [device address=0x0000000000000000] [size=0 bytes]
 WARNING: CPU: 8 PID: 1021 at kernel/dma/debug.c:961 check_unmap+0x54f/0x8b0
 RIP: 0010:check_unmap+0x54f/0x8b0
 Call Trace:
  debug_dma_unmap_page+0x57/0x60
  mlx5_free_priv_descs+0x57/0x70 [mlx5_ib]
  mlx5_ib_dereg_mr+0x1fb/0x3d0 [mlx5_ib]
  ib_dereg_mr_user+0x60/0x140 [ib_core]
  uverbs_destroy_uobject+0x59/0x210 [ib_uverbs]
  uobj_destroy+0x3f/0x80 [ib_uverbs]
  ib_uverbs_cmd_verbs+0x435/0xd10 [ib_uverbs]
  ? uverbs_finalize_object+0x50/0x50 [ib_uverbs]
  ? lock_acquire+0xc4/0x2e0
  ? lock_acquired+0x12/0x380
  ? lock_acquire+0xc4/0x2e0
  ? lock_acquire+0xc4/0x2e0
  ? ib_uverbs_ioctl+0x7c/0x140 [ib_uverbs]
  ? lock_release+0x28a/0x400
  ib_uverbs_ioctl+0xc0/0x140 [ib_uverbs]
  ? ib_uverbs_ioctl+0x7c/0x140 [ib_uverbs]
  __x64_sys_ioctl+0x7f/0xb0
  do_syscall_64+0x38/0x90

Fix it by reorganizing the dereg flow and mlx5_ib_mr structure:
 - Move the ib_umem field into the user MRs structure in the union as it's
   applicable only there.
 - Function mlx5_ib_dereg_mr() will now call mlx5_free_priv_descs() only
   in case there isn't udata, which indicates that this isn't a user MR.

Fixes: f18ec422 ("RDMA/mlx5: Use a union inside mlx5_ib_mr")
Link: https://lore.kernel.org/r/66bb1dd253c1fd7ceaa9fc411061eefa457b86fb.1637581144.git.leonro@nvidia.com


Signed-off-by: Alaa Hleihel <alaa@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

parent 84b01721

drivers/infiniband/hw/mlx5/mlx5_ib.h

+3 −3

Original line number	Diff line number	Diff line
		@@ -664,7 +664,6 @@ struct mlx5_ib_mr {

		/* User MR data */
		struct mlx5_cache_ent *cache_ent;
		struct ib_umem *umem;

		/* This is zero'd when the MR is allocated */
		union {
		@@ -676,7 +675,7 @@ struct mlx5_ib_mr {
		struct list_head list;
		};

		/* Used only by kernel MRs (umem == NULL) */
		/* Used only by kernel MRs */
		struct {
		void *descs;
		void *descs_alloc;
		@@ -697,8 +696,9 @@ struct mlx5_ib_mr {
		int data_length;
		};

		/* Used only by User MRs (umem != NULL) */
		/* Used only by User MRs */
		struct {
		struct ib_umem *umem;
		unsigned int page_shift;
		/* Current access_flags */
		int access_flags;

drivers/infiniband/hw/mlx5/mr.c

+12 −14

Original line number	Diff line number	Diff line
		@@ -1904,20 +1904,19 @@ mlx5_alloc_priv_descs(struct ib_device *device,
		return ret;
		}

		static void
		mlx5_free_priv_descs(struct mlx5_ib_mr *mr)
		static void mlx5_free_priv_descs(struct mlx5_ib_mr *mr)
		{
		if (!mr->umem && mr->descs) {
		struct ib_device *device = mr->ibmr.device;
		struct mlx5_ib_dev *dev = to_mdev(mr->ibmr.device);
		int size = mr->max_descs * mr->desc_size;
		struct mlx5_ib_dev *dev = to_mdev(device);

		if (!mr->descs)
		return;

		dma_unmap_single(&dev->mdev->pdev->dev, mr->desc_map, size,
		DMA_TO_DEVICE);
		kfree(mr->descs_alloc);
		mr->descs = NULL;
		}
		}

		int mlx5_ib_dereg_mr(struct ib_mr ibmr, struct ib_udata udata)
		{
		@@ -1992,6 +1991,7 @@ int mlx5_ib_dereg_mr(struct ib_mr ibmr, struct ib_udata udata)
		if (mr->cache_ent) {
		mlx5_mr_cache_free(dev, mr);
		} else {
		if (!udata)
		mlx5_free_priv_descs(mr);
		kfree(mr);
		}
		@@ -2079,7 +2079,6 @@ static struct mlx5_ib_mr mlx5_ib_alloc_pi_mr(struct ib_pd pd,
		if (err)
		goto err_free_in;

		mr->umem = NULL;
		kfree(in);

		return mr;
		@@ -2206,7 +2205,6 @@ static struct ib_mr __mlx5_ib_alloc_mr(struct ib_pd pd,
		}

		mr->ibmr.device = pd->device;
		mr->umem = NULL;

		switch (mr_type) {
		case IB_MR_TYPE_MEM_REG: