Re: [PATCH for-next v2] RDMA/rxe: Fix error paths in MR alloc routines

[Jason Gunthorpe <jgg@nvidia.com>]

On Thu, Jul 21, 2022 at 03:22:14PM -0500, Bob Pearson wrote:
> Currently the rxe driver has incorrect code in error paths for
> allocating MR objects. The PD and umem are always freed in
> rxe_mr_cleanup() but in some error paths they are already
> freed or never set. This patch makes sure that the PD is always
> set and checks to see if umem is set before freeing it in
> rxe_mr_cleanup(). umem and map sets are freed if an error
> occurs in an allocate mr call.
> 
> Reported-by: Li Zhijian <lizhijian@fujitsu.com>
> Link: https://lore.kernel.org/linux-rdma/11dafa5f-c52d-16c1-fe37-2cd45ab20474@fujitsu.com/
> Fixes: 3902b429ca14 ("Implement invalidate MW operations")
> Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
> ---
> v2:
>   Moved setting mr->umem until after checks to avoid sending
>   an ERR_PTR to ib_umem_release().
>   Cleaned up umem and map sets if errors occur in alloc mr calls.
>   Rebased to current for-next.

It seems it doesn't apply now, can you resend on the next rc1?

Thanks,
Jason