On Monday 12 October 2015 23:20:25 John Garry wrote: > @@ -804,6 +818,16 @@ static irqreturn_t int_phyup_v1_hw(int irq_no, void *p) > phy->identify.target_port_protocols = > SAS_PROTOCOL_SMP; > > + wq = kmalloc(sizeof(*wq), GFP_ATOMIC); > + if (!wq) > + goto end; > + > + wq->event = PHYUP; > + wq->hisi_hba = hisi_hba; > + wq->phy_no = phy_no; > + > + INIT_WORK(&wq->work_struct, hisi_sas_wq_process); > + queue_work(hisi_hba->wq, &wq->work_struct); > > end: > hisi_sas_phy_write32(hisi_hba, phy_no, CHL_INT2, > While rereading some other parts of the code, I stumbled over this piece. You should generally not allocate work structs dynamically. Why not embed the work struct inside of the phy structure and then just queue that? Arnd -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html