On Thu Mar 13, 2025 at 9:31 AM CET, zhang.guanghui@xxxxxxxx wrote: > Hi, > in fact, the nvme_tcp_try_send() failure, the target may send C2HTermReq immediately. while the host receives the C2HTermReq and still starting error recovery. > so when queue->rd_enabled is false, can avoid starting error recovery agagin. Not all targets send C2HTermReq (for example, the Linux target doesn't at the moment) so you can't rely on that. In any case, calling nvme_tcp_error_recovery() twice is harmless; the first call moves the controller to the resetting state, the second call is ignored. Maurizio
