https://bugs.dpdk.org/show_bug.cgi?id=643
Bug ID: 643
Summary: af_xdp驱动造成rte_rx_queue_setup的coredump问题
Product: DPDK
Version: 20.05
Hardware: x86
OS: Linux
Status: UNCONFIRMED
Severity: normal
Priority: Normal
Component: other
Assignee: dev@dpdk.org
Reporter: huangyin...@360.cn
Target Milestone: ---
driver/net/af_xdp/rte_eth_af_xdp.c
xsk_configure中调用了rte_pktmbuf_alloc_bulk,当rte_pktmbuf_alloc_bulk调用失败后xsk_configure仍然返回0，造成调用rte_eth_rx_queue_setup时不会返回失败，但是在xsk_configure中已经调用了xdp_umem_destroy回收了内存，当再次调用rte_eth_rx_queue_setup时造成coredump，代码如下：
static int
xsk_configure(struct pmd_internals *internals, struct pkt_rx_queue *rxq,
int ring_size)
{
struct xsk_socket_config cfg;
struct pkt_tx_queue *txq = rxq->pair;
int ret = 0;
...........
#if defined(XDP_UMEM_UNALIGNED_CHUNK_FLAG)
if (rte_pktmbuf_alloc_bulk(rxq->umem->mb_pool, fq_bufs, reserve_size)) {
AF_XDP_LOG(DEBUG, "Failed to get enough buffers for fq.\n");
goto err;
}
#endif
err:
xdp_umem_destroy(rxq->umem);
return ret;
}
这让使用这在使用rte_eth_rx_queue_setup时无法根据返回的错误进行相应的异常处理
--
You are receiving this mail because:
You are the assignee for the bug.
