On 2/13/20 12:38, Jason Gunthorpe wrote: > > Any reason to skip these two? > > drivers/infiniband/core/mad_priv.h: u8 mad[0]; > drivers/infiniband/core/mad_priv.h: u8 data[0]; > > And may as well touch these in the subsystem headers too: > > include/rdma/ib_fmr_pool.h: u64 page_list[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/ib_verbs.h: u8 real_sz[0]; > include/rdma/opa_vnic.h: char *dev_priv[0]; > include/rdma/rdmavt_mr.h: struct rvt_segarray *map[0]; /* the segments */ > include/rdma/rdmavt_qp.h: struct rvt_sge sg_list[0]; > > ? OK. Let me tweak my script a bit to catch all of those. Thanks for the feedback. -- Gustavo
