> This is basically what I was thinking as well.. > > My fear is that we have lots and lots of places where the > cma_comp_exch "lock" isn't right in this same sort of way, and > everything has to be audited.. OK, glad you like the approach. Yeah, I think at least any function that has a cma_comp_exch() to unwind things in the error return path is suspect. For example, can rdma_connect() race against rdma_set_ib_path()? I can take a more comprehensive look tomorrow. What do you think about applying this patch to close this syzbot bug? Would you rather wait for a bigger patch? - R. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
