> -----Original Message----- > From: Jakub Kicinski <[email protected]> > Sent: Tuesday, July 21, 2020 11:05 AM > To: Nguyen, Anthony L <[email protected]> > Cc: [email protected]; Michael, Alice <[email protected]>; > [email protected]; [email protected]; [email protected]; > Kirsher, Jeffrey T <[email protected]>; Brady, Alan > <[email protected]>; Burra, Phani R <[email protected]>; Hay, > Joshua A <[email protected]>; Chittim, Madhu > <[email protected]>; Linga, Pavan Kumar > <[email protected]>; Skidmore, Donald C > <[email protected]>; Brandeburg, Jesse > <[email protected]>; Samudrala, Sridhar > <[email protected]> > Subject: Re: [net-next v4 06/15] iecm: Implement mailbox functionality > > On Mon, 20 Jul 2020 17:38:01 -0700 Tony Nguyen wrote: > > + (cq->next_to_use)++; > > + if (cq->next_to_use == cq->ring_size) > > + cq->next_to_use = 0; > > + } > > + > > + /* Force memory write to complete before letting hardware > > + * know that there are new descriptors to fetch. > > + */ > > + iecm_wmb(); > > dma_wmb() would probably be sufficient here? > > > + wr32(hw, cq->reg.tail, cq->next_to_use);
We will investigate and see if dma_wmb is sufficient. We do have a test for triggering the weak-ordered issue so we should be able to tell for certain. The reasoning makes sense dma_wmb is sufficient but we want to test to be sure. Alan
