This series collects fixes for the fail-safe PMD found while testing live migration of a VF-backed sub-device (iavf) paired with a virtio-net standby sub-device.
- double release of the sub-device port on hot-unplug; - NULL queue-release op on sub-devices that do not implement it; - MAC address pool size off-by-one and replay of zeroed MAC slots; - unsupported link-status control (-ENOTSUP) treated as fatal; - Rx/Tx burst still polling a removed sub-device: the fast-path ops are copied into rte_eth_fp_ops at start, so swapping the burst functions at runtime had no effect; - sub-device removal not detected without the RMV interrupt (e.g. VFIO device release), now handled through a bus device-event callback. v3: - Split the removal-detection change (v2 patch 5/5) into a fix for the Rx/Tx burst path (5/6) and the new device-event based detection (6/6). - Patch 1/6: reword and drop the redundant braces; the port is already released by rte_eth_dev_close() or rte_dev_remove(). - Patch 2/6: correct the Fixes tag to the fail-safe PMD commit and drop the unneeded ETH(sdev) != NULL check. - Patch 5/6: new, the actual crash fix for the burst path. - Patch 6/6: simplify the device-event callback and note that the bus delivers the event automatically for VFIO. v2: - Added an ETH(sdev) != NULL check (removed again in v3). - Clarified secondary-process detection in a comment. Sandeep Penigalapati (6): net/failsafe: fix double release of sub-device port net/failsafe: fix queue release without sub-device op net/failsafe: fix MAC address pool size and replay net/failsafe: ignore unsupported link status update net/failsafe: fix Rx/Tx burst after sub-device removal net/failsafe: detect sub-device removal via device event drivers/net/failsafe/failsafe_ether.c | 42 ++++++++++--- drivers/net/failsafe/failsafe_ops.c | 27 +++++++- drivers/net/failsafe/failsafe_private.h | 13 ++-- drivers/net/failsafe/failsafe_rxtx.c | 83 +------------------------ 4 files changed, 65 insertions(+), 100 deletions(-) -- 2.34.1
