Re: [PATCH rdma-next 1/2] RDMA/bnxt_re: Validate udata before destroying resources

[Jacob Moroni]

These changes look good but there is also a call to ib_respond_empty_udata
in bnxt_re_resize_cq (but that method does take input data).

Is that one a problem? I guess the resize could complete but the upper
layers would think it failed if the ib_respond_empty_udata call fails?

Reviewed-by: Jacob Moroni <jmoroni@xxxxxxxxxx>