From: Yangchao Zhou Date: Tue, 25 Jun 2019 15:04:14 +0000 (+0800) Subject: kni: fix kernel crash with multi-segments X-Git-Url: http://git.droids-corp.org/?a=commitdiff_plain;h=5eb1708ec1db1f2d644f44a42468139df0b0ad6c;p=dpdk.git kni: fix kernel crash with multi-segments va2pa depends on the physical address and virtual address offset of current mbuf. It may get the wrong physical address of next mbuf which allocated in another hugepage segment. In rte_mempool_populate_default(), trying to allocate whole block of contiguous memory could be failed. Then, it would reserve memory in several memzones that have different physical address and virtual address offsets. The rte_mempool_populate_default() is used by rte_pktmbuf_pool_create(). Fixes: 8451269e6d7b ("kni: remove continuous memory restriction") Cc: stable@dpdk.org Signed-off-by: Yangchao Zhou Acked-by: Ferruh Yigit --- diff --git a/doc/guides/prog_guide/kernel_nic_interface.rst b/doc/guides/prog_guide/kernel_nic_interface.rst index 5afd05c0a7..38369b30e0 100644 --- a/doc/guides/prog_guide/kernel_nic_interface.rst +++ b/doc/guides/prog_guide/kernel_nic_interface.rst @@ -268,12 +268,16 @@ Use Case: Ingress ----------------- On the DPDK RX side, the mbuf is allocated by the PMD in the RX thread context. -This thread will enqueue the mbuf in the rx_q FIFO. +This thread will enqueue the mbuf in the rx_q FIFO, +and the next pointers in mbuf-chain will convert to physical address. The KNI thread will poll all KNI active devices for the rx_q. If an mbuf is dequeued, it will be converted to a sk_buff and sent to the net stack via netif_rx(). -The dequeued mbuf must be freed, so the same pointer is sent back in the free_q FIFO. +The dequeued mbuf must be freed, so the same pointer is sent back in the free_q FIFO, +and next pointers must convert back to virtual address if exists before put in the free_q FIFO. The RX thread, in the same main loop, polls this FIFO and frees the mbuf after dequeuing it. +The address conversion of the next pointer is to prevent the chained mbuf +in different hugepage segments from causing kernel crash. Use Case: Egress ---------------- diff --git a/kernel/linux/kni/kni_net.c b/kernel/linux/kni/kni_net.c index 319ee2dcb1..a736407e66 100644 --- a/kernel/linux/kni/kni_net.c +++ b/kernel/linux/kni/kni_net.c @@ -62,18 +62,6 @@ kva2data_kva(struct rte_kni_mbuf *m) return phys_to_virt(m->buf_physaddr + m->data_off); } -/* virtual address to physical address */ -static void * -va2pa(void *va, struct rte_kni_mbuf *m) -{ - void *pa; - - pa = (void *)((unsigned long)va - - ((unsigned long)m->buf_addr - - (unsigned long)m->buf_physaddr)); - return pa; -} - /* * It can be called to process the request. */ @@ -174,7 +162,10 @@ kni_fifo_trans_pa2va(struct kni_dev *kni, struct rte_kni_fifo *src_pa, struct rte_kni_fifo *dst_va) { uint32_t ret, i, num_dst, num_rx; - void *kva; + struct rte_kni_mbuf *kva, *prev_kva; + int nb_segs; + int kva_nb_segs; + do { num_dst = kni_fifo_free_count(dst_va); if (num_dst == 0) @@ -189,6 +180,17 @@ kni_fifo_trans_pa2va(struct kni_dev *kni, for (i = 0; i < num_rx; i++) { kva = pa2kva(kni->pa[i]); kni->va[i] = pa2va(kni->pa[i], kva); + + kva_nb_segs = kva->nb_segs; + for (nb_segs = 0; nb_segs < kva_nb_segs; nb_segs++) { + if (!kva->next) + break; + + prev_kva = kva; + kva = pa2kva(kva->next); + /* Convert physical address to virtual address */ + prev_kva->next = pa2va(prev_kva->next, kva); + } } ret = kni_fifo_put(dst_va, kni->va, num_rx); @@ -314,7 +316,7 @@ kni_net_rx_normal(struct kni_dev *kni) uint32_t ret; uint32_t len; uint32_t i, num_rx, num_fq; - struct rte_kni_mbuf *kva; + struct rte_kni_mbuf *kva, *prev_kva; void *data_kva; struct sk_buff *skb; struct net_device *dev = kni->net_dev; @@ -361,8 +363,11 @@ kni_net_rx_normal(struct kni_dev *kni) if (!kva->next) break; - kva = pa2kva(va2pa(kva->next, kva)); + prev_kva = kva; + kva = pa2kva(kva->next); data_kva = kva2data_kva(kva); + /* Convert physical address to virtual address */ + prev_kva->next = pa2va(prev_kva->next, kva); } } @@ -393,7 +398,7 @@ kni_net_rx_lo_fifo(struct kni_dev *kni) uint32_t ret; uint32_t len; uint32_t i, num, num_rq, num_tq, num_aq, num_fq; - struct rte_kni_mbuf *kva; + struct rte_kni_mbuf *kva, *next_kva; void *data_kva; struct rte_kni_mbuf *alloc_kva; void *alloc_data_kva; @@ -437,6 +442,13 @@ kni_net_rx_lo_fifo(struct kni_dev *kni) data_kva = kva2data_kva(kva); kni->va[i] = pa2va(kni->pa[i], kva); + while (kva->next) { + next_kva = pa2kva(kva->next); + /* Convert physical address to virtual address */ + kva->next = pa2va(kva->next, next_kva); + kva = next_kva; + } + alloc_kva = pa2kva(kni->alloc_pa[i]); alloc_data_kva = kva2data_kva(alloc_kva); kni->alloc_va[i] = pa2va(kni->alloc_pa[i], alloc_kva); @@ -479,7 +491,7 @@ kni_net_rx_lo_fifo_skb(struct kni_dev *kni) uint32_t ret; uint32_t len; uint32_t i, num_rq, num_fq, num; - struct rte_kni_mbuf *kva; + struct rte_kni_mbuf *kva, *prev_kva; void *data_kva; struct sk_buff *skb; struct net_device *dev = kni->net_dev; @@ -537,8 +549,11 @@ kni_net_rx_lo_fifo_skb(struct kni_dev *kni) if (!kva->next) break; - kva = pa2kva(va2pa(kva->next, kva)); + prev_kva = kva; + kva = pa2kva(kva->next); data_kva = kva2data_kva(kva); + /* Convert physical address to virtual address */ + prev_kva->next = pa2va(prev_kva->next, kva); } } diff --git a/lib/librte_eal/linux/eal/include/rte_kni_common.h b/lib/librte_eal/linux/eal/include/rte_kni_common.h index 91a1c14088..37d9ee8f0f 100644 --- a/lib/librte_eal/linux/eal/include/rte_kni_common.h +++ b/lib/librte_eal/linux/eal/include/rte_kni_common.h @@ -86,7 +86,7 @@ struct rte_kni_mbuf { /* fields on second cache line */ char pad3[8] __attribute__((__aligned__(RTE_CACHE_LINE_MIN_SIZE))); void *pool; - void *next; + void *next; /**< Physical address of next mbuf in kernel. */ }; /* diff --git a/lib/librte_kni/rte_kni.c b/lib/librte_kni/rte_kni.c index 1de053deb0..4b51fb4fed 100644 --- a/lib/librte_kni/rte_kni.c +++ b/lib/librte_kni/rte_kni.c @@ -349,6 +349,19 @@ va2pa(struct rte_mbuf *m) (unsigned long)m->buf_iova)); } +static void * +va2pa_all(struct rte_mbuf *mbuf) +{ + void *phy_mbuf = va2pa(mbuf); + struct rte_mbuf *next = mbuf->next; + while (next) { + mbuf->next = va2pa(next); + mbuf = next; + next = mbuf->next; + } + return phy_mbuf; +} + static void obj_free(struct rte_mempool *mp __rte_unused, void *opaque, void *obj, unsigned obj_idx __rte_unused) @@ -541,12 +554,13 @@ rte_kni_handle_request(struct rte_kni *kni) unsigned rte_kni_tx_burst(struct rte_kni *kni, struct rte_mbuf **mbufs, unsigned int num) { + num = RTE_MIN(kni_fifo_free_count(kni->rx_q), num); void *phy_mbufs[num]; unsigned int ret; unsigned int i; for (i = 0; i < num; i++) - phy_mbufs[i] = va2pa(mbufs[i]); + phy_mbufs[i] = va2pa_all(mbufs[i]); ret = kni_fifo_put(kni->rx_q, phy_mbufs, num); diff --git a/lib/librte_kni/rte_kni_fifo.h b/lib/librte_kni/rte_kni_fifo.h index 287d7deb29..d2ec82fe87 100644 --- a/lib/librte_kni/rte_kni_fifo.h +++ b/lib/librte_kni/rte_kni_fifo.h @@ -104,3 +104,14 @@ kni_fifo_count(struct rte_kni_fifo *fifo) unsigned fifo_read = __KNI_LOAD_ACQUIRE(&fifo->read); return (fifo->len + fifo_write - fifo_read) & (fifo->len - 1); } + +/** + * Get the num of available elements in the fifo + */ +static inline uint32_t +kni_fifo_free_count(struct rte_kni_fifo *fifo) +{ + uint32_t fifo_write = __KNI_LOAD_ACQUIRE(&fifo->write); + uint32_t fifo_read = __KNI_LOAD_ACQUIRE(&fifo->read); + return (fifo_read - fifo_write - 1) & (fifo->len - 1); +}