| Message ID | 20220317183629.340350-4-maciej.fijalkowski@intel.com (mailing list archive) |
|---|---|
| State | Awaiting Upstream |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | ice: xsk: various fixes | expand |
> -----Original Message----- > From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf Of > Maciej Fijalkowski > Sent: Friday, March 18, 2022 12:06 AM > To: intel-wired-lan@lists.osuosl.org > Cc: netdev@vger.kernel.org; kuba@kernel.org; bpf@vger.kernel.org; > davem@davemloft.net; Karlsson, Magnus <magnus.karlsson@intel.com> > Subject: [Intel-wired-lan] [PATCH intel-net 3/3] ice: clear > cmd_type_offset_bsz for TX rings > > Currently when XDP rings are created, each descriptor gets its DD bit set, > which turns out to be the wrong approach as it can lead to a situation where > more descriptors get cleaned than it was supposed to, e.g. when AF_XDP > busy poll is run with a large batch size. In this situation, the driver would > request for more buffers than it is able to handle. > > Fix this by not setting the DD bits in ice_xdp_alloc_setup_rings(). They should > be initialized to zero instead. > > Fixes: 9610bd988df9 ("ice: optimize XDP_TX workloads") > Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> > --- > drivers/net/ethernet/intel/ice/ice_main.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > Tested-by: Shwetha Nagaraju <Shwetha.nagaraju@intel.com>
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index c3c73a61bfd0..5332bb24001a 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -2533,7 +2533,7 @@ static int ice_xdp_alloc_setup_rings(struct ice_vsi *vsi) spin_lock_init(&xdp_ring->tx_lock); for (j = 0; j < xdp_ring->count; j++) { tx_desc = ICE_TX_DESC(xdp_ring, j); - tx_desc->cmd_type_offset_bsz = cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE); + tx_desc->cmd_type_offset_bsz = 0; } }
Currently when XDP rings are created, each descriptor gets its DD bit set, which turns out to be the wrong approach as it can lead to a situation where more descriptors get cleaned than it was supposed to, e.g. when AF_XDP busy poll is run with a large batch size. In this situation, the driver would request for more buffers than it is able to handle. Fix this by not setting the DD bits in ice_xdp_alloc_setup_rings(). They should be initialized to zero instead. Fixes: 9610bd988df9 ("ice: optimize XDP_TX workloads") Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> --- drivers/net/ethernet/intel/ice/ice_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)