i40e/ice AF_XDP ZC fixes
Subsystems (4)
cover i40e/ice AF_XDP ZC fixes
This series address two crashes in the AF_XDP zero-copy mode for ice and i40e. More details in each individual the commit message.
Thanks,
Bj?rn
Bj?rn T?pel (2):
ice, xsk: clear the status bits for the next_to_use descriptor i40e, xsk: clear the status bits for the next_to_use descriptor
drivers/net/ethernet/intel/i40e/i40e_xsk.c | 5 ++++- drivers/net/ethernet/intel/ice/ice_xsk.c | 5 ++++- 2 files changed, 8 insertions(+), 2 deletions(-)
base-commit: d9838b1d39283c1200c13f9076474c7624b8ec34
Signature · 1 line
2.27.0
PATCH 1/2 ice, xsk: clear the status bits for the next_to_use descriptor
From: Bj?rn T?pel <[email protected]>
On the Rx side, the next_to_use index points to the next item in the HW ring to be refilled/allocated, and next_to_clean points to the next item to potentially be processed.
When the HW Rx ring is fully refilled, i.e. no packets has been processed, the next_to_use will be next_to_clean - 1. When the ring is fully processed next_to_clean will be equal to next_to_use. The latter case is where a bug is triggered.
If the next_to_use bits are not cleared, and the "fully processed" state is entered, a stale descriptor can be processed.
The skb-path correctly clear the status bit for the next_to_use descriptor, but the AF_XDP zero-copy path did not do that.
This change adds the status bits clearing of the next_to_use descriptor.
--- drivers/net/ethernet/intel/ice/ice_xsk.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/ice/ice_xsk.c +4 -1
@@ -446,8 +446,11 @@ bool ice_alloc_rx_bufs_zc(struct ice_ring *rx_ring, u16 count) } }while (--count); -if (rx_ring->next_to_use != ntu) +if (rx_ring->next_to_use != ntu) { +/* clear the status bits for the next_to_use descriptor */ + rx_desc->wb.status_error0 =0 ;ice_release_rx_desc (rx_ring, ntu); + }return ret; }
Signature · 1 line
2.27.0
PATCH 2/2 i40e, xsk: clear the status bits for the next_to_use descriptor
In reply to Björn TöpelFrom: Bj?rn T?pel <[email protected]>
On the Rx side, the next_to_use index points to the next item in the HW ring to be refilled/allocated, and next_to_clean points to the next item to potentially be processed.
When the HW Rx ring is fully refilled, i.e. no packets has been processed, the next_to_use will be next_to_clean - 1. When the ring is fully processed next_to_clean will be equal to next_to_use. The latter case is where a bug is triggered.
If the next_to_use bits are not cleared, and the "fully processed" state is entered, a stale descriptor can be processed.
The skb-path correctly clear the status bit for the next_to_use descriptor, but the AF_XDP zero-copy path did not do that.
This change adds the status bits clearing of the next_to_use descriptor.
--- drivers/net/ethernet/intel/i40e/i40e_xsk.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/i40e/i40e_xsk.c +4 -1
@@ -219,8 +219,11 @@ bool i40e_alloc_rx_buffers_zc(struct i40e_ring *rx_ring, u16 count) }while (count); no_buffers: -if (rx_ring->next_to_use != ntu) +if (rx_ring->next_to_use != ntu) { +/* clear the status bits for the next_to_use descriptor */ + rx_desc->wb.qword1.status_error_len =0 ;i40e_release_rx_desc (rx_ring, ntu); + }return ok; }
Signature · 1 line
2.27.0
Quoted 2 lines · On Fri, 11 Dec 2020 15:57:10 +0100 Bj?rn T?pel wrote:
This series address two crashes in the
AF_XDPzero-copy mode for ice and i40e. More details in each individual the commit message.
Applied, queued. Thanks!