[4/5] mbuf: add a pktmbuf copy routine
Checks
Commit Message
This is a commonly used operation that surprisingly the
DPDK has not supported. The new rte_pktmbuf_copy does a
deep copy of packet. This is a complete copy including
meta-data.
It handles the case where the source mbuf comes from a pool
with larger data area than the destination pool. The routine
also has options for skipping data, or truncating at a fixed
length.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
---
lib/librte_mbuf/rte_mbuf.c | 70 ++++++++++++++++++++++++++++
lib/librte_mbuf/rte_mbuf.h | 26 +++++++++++
lib/librte_mbuf/rte_mbuf_version.map | 1 +
3 files changed, 97 insertions(+)
Comments
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Saturday, September 28, 2019 2:38 AM
>
> This is a commonly used operation that surprisingly the
> DPDK has not supported. The new rte_pktmbuf_copy does a
> deep copy of packet. This is a complete copy including
> meta-data.
>
> It handles the case where the source mbuf comes from a pool
> with larger data area than the destination pool. The routine
> also has options for skipping data, or truncating at a fixed
> length.
Great initiative, and the offset/length params makes it even better!
>
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
> ---
> lib/librte_mbuf/rte_mbuf.c | 70 ++++++++++++++++++++++++++++
> lib/librte_mbuf/rte_mbuf.h | 26 +++++++++++
> lib/librte_mbuf/rte_mbuf_version.map | 1 +
> 3 files changed, 97 insertions(+)
>
> diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
> index 12d0258a120d..6888d6bd5dfc 100644
> --- a/lib/librte_mbuf/rte_mbuf.c
> +++ b/lib/librte_mbuf/rte_mbuf.c
> @@ -324,6 +324,76 @@ rte_pktmbuf_linearize(struct rte_mbuf *mbuf)
> return 0;
> }
>
> +/* Create a deep copy of mbuf */
> +struct rte_mbuf *
> +rte_pktmbuf_copy(const struct rte_mbuf *m, struct rte_mempool *mp,
> + uint32_t off, uint32_t len)
> +{
> + const struct rte_mbuf *seg = m;
> + struct rte_mbuf *mc, *m_last, **prev;
> +
> + if (unlikely(off >= m->pkt_len))
> + return NULL;
> +
> + mc = rte_pktmbuf_alloc(mp);
> + if (unlikely(mc == NULL))
> + return NULL;
> +
> + if (len > m->pkt_len - off)
> + len = m->pkt_len - off;
> +
> + /* clone meta data from original */
> + mc->port = m->port;
> + mc->vlan_tci = m->vlan_tci;
> + mc->vlan_tci_outer = m->vlan_tci_outer;
> + mc->tx_offload = m->tx_offload;
> + mc->hash = m->hash;
> + mc->packet_type = m->packet_type;
> + mc->timestamp = m->timestamp;
At the dynamic mbuf presentation at DPDK Userspace, it was mentioned that the last 8 bytes of the mbuf could be used by applications. Consider if their values should be copied too, and thus if they should be explicitly define in the mbuf, e.g. dynfield[1,2] like in the dynamic mbuf patch. I don't have an opinion on this, I'm just bringing it to your attention.
Med venlig hilsen / kind regards
- Morten Brørup
@@ -324,6 +324,76 @@ rte_pktmbuf_linearize(struct rte_mbuf *mbuf)
return 0;
}
+/* Create a deep copy of mbuf */
+struct rte_mbuf *
+rte_pktmbuf_copy(const struct rte_mbuf *m, struct rte_mempool *mp,
+ uint32_t off, uint32_t len)
+{
+ const struct rte_mbuf *seg = m;
+ struct rte_mbuf *mc, *m_last, **prev;
+
+ if (unlikely(off >= m->pkt_len))
+ return NULL;
+
+ mc = rte_pktmbuf_alloc(mp);
+ if (unlikely(mc == NULL))
+ return NULL;
+
+ if (len > m->pkt_len - off)
+ len = m->pkt_len - off;
+
+ /* clone meta data from original */
+ mc->port = m->port;
+ mc->vlan_tci = m->vlan_tci;
+ mc->vlan_tci_outer = m->vlan_tci_outer;
+ mc->tx_offload = m->tx_offload;
+ mc->hash = m->hash;
+ mc->packet_type = m->packet_type;
+ mc->timestamp = m->timestamp;
+
+ prev = &mc->next;
+ m_last = mc;
+ while (len > 0) {
+ uint32_t copy_len;
+
+ while (off >= seg->data_len) {
+ off -= seg->data_len;
+ seg = seg->next;
+ }
+
+ /* current buffer is full, chain a new one */
+ if (rte_pktmbuf_tailroom(m_last) == 0) {
+ m_last = rte_pktmbuf_alloc(mp);
+ if (unlikely(m_last == NULL)) {
+ rte_pktmbuf_free(mc);
+ return NULL;
+ }
+ ++mc->nb_segs;
+ *prev = m_last;
+ prev = &m_last->next;
+ }
+
+ copy_len = RTE_MIN(seg->data_len - off, len);
+ if (copy_len > rte_pktmbuf_tailroom(m_last))
+ copy_len = rte_pktmbuf_tailroom(m_last);
+
+ /* append from seg to m_last */
+ rte_memcpy(rte_pktmbuf_mtod_offset(m_last, char *,
+ m_last->data_len),
+ rte_pktmbuf_mtod_offset(seg, char *,
+ off),
+ copy_len);
+
+ m_last->data_len += copy_len;
+ mc->pkt_len += copy_len;
+ off += copy_len;
+ len -= copy_len;
+ }
+
+ __rte_mbuf_sanity_check(mc, 1);
+ return mc;
+}
+
/* dump a mbuf on console */
void
rte_pktmbuf_dump(FILE *f, const struct rte_mbuf *m, unsigned dump_len)
@@ -1927,6 +1927,32 @@ static inline void rte_pktmbuf_free(struct rte_mbuf *m)
struct rte_mbuf *
rte_pktmbuf_clone(struct rte_mbuf *md, struct rte_mempool *mp);
+/**
+ * Creates a full copy of a given packet mbuf.
+ *
+ * Copies all the data from a given packet mbuf to a newly allocated
+ * set of mbufs.
+ *
+ * @param m
+ * The packet mbuf to be cloned.
+ * @param mp
+ * The mempool from which the "clone" mbufs are allocated.
+ * @param offset
+ * The number of bytes to skip before copying.
+ * If the mbuf does not have that many bytes, it is an error
+ * and NULL is returned.
+ * @param length
+ * The upper limit on bytes to copy. Passing UINT32_MAX
+ * means all data (after offset).
+ * @return
+ * - The pointer to the new "clone" mbuf on success.
+ * - NULL if allocation fails.
+ */
+__rte_experimental
+struct rte_mbuf *
+rte_pktmbuf_copy(const struct rte_mbuf *m, struct rte_mempool *mp,
+ uint32_t offset, uint32_t length);
+
/**
* Adds given value to the refcnt of all packet mbuf segments.
*
@@ -57,4 +57,5 @@ EXPERIMENTAL {
global:
rte_mbuf_check;
+ rte_pktmbuf_copy;
} DPDK_18.08;