* [dpdk-dev] [PATCH 0/2] pdump: cleanups
@ 2019-11-08 4:38 Stephen Hemminger
2019-11-08 4:38 ` [dpdk-dev] [PATCH 1/2] pdump: use new pktmbuf copy function Stephen Hemminger
` (2 more replies)
0 siblings, 3 replies; 12+ messages in thread
From: Stephen Hemminger @ 2019-11-08 4:38 UTC (permalink / raw)
To: dev; +Cc: Stephen Hemminger
These are a couple of small cleanups for 19.10 which
came out of work on pcapng support. Full pcapng support
and BPF are planned for DPDK 20.02.
Stephen Hemminger (2):
pdump: use new pktmbuf copy function
pdump: use dynamic logtype
lib/librte_pdump/rte_pdump.c | 137 ++++++++++-------------------------
1 file changed, 39 insertions(+), 98 deletions(-)
--
2.20.1
^ permalink raw reply [flat|nested] 12+ messages in thread* [dpdk-dev] [PATCH 1/2] pdump: use new pktmbuf copy function 2019-11-08 4:38 [dpdk-dev] [PATCH 0/2] pdump: cleanups Stephen Hemminger @ 2019-11-08 4:38 ` Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger 2 siblings, 0 replies; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 4:38 UTC (permalink / raw) To: dev; +Cc: Stephen Hemminger The rte_pktmbuf_copy handles varying size mbuf pools correctly. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> --- lib/librte_pdump/rte_pdump.c | 69 +----------------------------------- 1 file changed, 1 insertion(+), 68 deletions(-) diff --git a/lib/librte_pdump/rte_pdump.c b/lib/librte_pdump/rte_pdump.c index ac94fea93656..7816c7254615 100644 --- a/lib/librte_pdump/rte_pdump.c +++ b/lib/librte_pdump/rte_pdump.c @@ -64,73 +64,6 @@ static struct pdump_rxtx_cbs { } rx_cbs[RTE_MAX_ETHPORTS][RTE_MAX_QUEUES_PER_PORT], tx_cbs[RTE_MAX_ETHPORTS][RTE_MAX_QUEUES_PER_PORT]; -static inline int -pdump_pktmbuf_copy_data(struct rte_mbuf *seg, const struct rte_mbuf *m) -{ - if (rte_pktmbuf_tailroom(seg) < m->data_len) { - RTE_LOG(ERR, PDUMP, - "User mempool: insufficient data_len of mbuf\n"); - return -EINVAL; - } - - seg->port = m->port; - seg->vlan_tci = m->vlan_tci; - seg->hash = m->hash; - seg->tx_offload = m->tx_offload; - seg->ol_flags = m->ol_flags; - seg->packet_type = m->packet_type; - seg->vlan_tci_outer = m->vlan_tci_outer; - seg->data_len = m->data_len; - seg->pkt_len = seg->data_len; - rte_memcpy(rte_pktmbuf_mtod(seg, void *), - rte_pktmbuf_mtod(m, void *), - rte_pktmbuf_data_len(seg)); - - return 0; -} - -static inline struct rte_mbuf * -pdump_pktmbuf_copy(struct rte_mbuf *m, struct rte_mempool *mp) -{ - struct rte_mbuf *m_dup, *seg, **prev; - uint32_t pktlen; - uint16_t nseg; - - m_dup = rte_pktmbuf_alloc(mp); - if (unlikely(m_dup == NULL)) - return NULL; - - seg = m_dup; - prev = &seg->next; - pktlen = m->pkt_len; - nseg = 0; - - do { - nseg++; - if (pdump_pktmbuf_copy_data(seg, m) < 0) { - if (seg != m_dup) - rte_pktmbuf_free_seg(seg); - rte_pktmbuf_free(m_dup); - return NULL; - } - *prev = seg; - prev = &seg->next; - } while ((m = m->next) != NULL && - (seg = rte_pktmbuf_alloc(mp)) != NULL); - - *prev = NULL; - m_dup->nb_segs = nseg; - m_dup->pkt_len = pktlen; - - /* Allocation of new indirect segment failed */ - if (unlikely(seg == NULL)) { - rte_pktmbuf_free(m_dup); - return NULL; - } - - __rte_mbuf_sanity_check(m_dup, 1); - return m_dup; -} static inline void pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) @@ -148,7 +81,7 @@ pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) ring = cbs->ring; mp = cbs->mp; for (i = 0; i < nb_pkts; i++) { - p = pdump_pktmbuf_copy(pkts[i], mp); + p = rte_pktmbuf_copy(pkts[i], mp, 0, UINT32_MAX); if (p) dup_bufs[d_pkts++] = p; } -- 2.20.1 ^ permalink raw reply [flat|nested] 12+ messages in thread
* [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype 2019-11-08 4:38 [dpdk-dev] [PATCH 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 1/2] pdump: use new pktmbuf copy function Stephen Hemminger @ 2019-11-08 4:38 ` Stephen Hemminger 2019-11-08 8:13 ` David Marchand 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger 2 siblings, 1 reply; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 4:38 UTC (permalink / raw) To: dev; +Cc: Stephen Hemminger The logtype USER1 should not be overloaded for library function. Instead use a dynamic log type. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> --- lib/librte_pdump/rte_pdump.c | 68 ++++++++++++++++++++---------------- 1 file changed, 38 insertions(+), 30 deletions(-) diff --git a/lib/librte_pdump/rte_pdump.c b/lib/librte_pdump/rte_pdump.c index 7816c7254615..ae25e2a064a4 100644 --- a/lib/librte_pdump/rte_pdump.c +++ b/lib/librte_pdump/rte_pdump.c @@ -13,8 +13,12 @@ #include "rte_pdump.h" #define DEVICE_ID_SIZE 64 -/* Macros for printing using RTE_LOG */ -#define RTE_LOGTYPE_PDUMP RTE_LOGTYPE_USER1 + +/* Macro for printing using RTE_LOG */ +static int pdump_logtype; +#define PDUMP_LOG(level, fmt, args...) \ + rte_log(RTE_LOG_ ## level, pdump_logtype, "%s(): " fmt, \ + __func__, ## args) /* Used for the multi-process communication */ #define PDUMP_MP "mp_pdump" @@ -88,7 +92,7 @@ pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) ring_enq = rte_ring_enqueue_burst(ring, (void *)dup_bufs, d_pkts, NULL); if (unlikely(ring_enq < d_pkts)) { - RTE_LOG(DEBUG, PDUMP, + PDUMP_LOG(DEBUG, "only %d of packets enqueued to ring\n", ring_enq); do { rte_pktmbuf_free(dup_bufs[ring_enq]); @@ -127,7 +131,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs = &rx_cbs[port][qid]; if (cbs && operation == ENABLE) { if (cbs->cb) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add rx callback for port=%d " "and queue=%d, callback already exists\n", port, qid); @@ -138,7 +142,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs->cb = rte_eth_add_first_rx_callback(port, qid, pdump_rx, cbs); if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add rx callback, errno=%d\n", rte_errno); return rte_errno; @@ -148,7 +152,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, int ret; if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to delete non existing rx " "callback for port=%d and queue=%d\n", port, qid); @@ -156,7 +160,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, } ret = rte_eth_remove_rx_callback(port, qid, cbs->cb); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to remove rx callback, errno=%d\n", -ret); return ret; @@ -182,7 +186,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs = &tx_cbs[port][qid]; if (cbs && operation == ENABLE) { if (cbs->cb) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add tx callback for port=%d " "and queue=%d, callback already exists\n", port, qid); @@ -193,7 +197,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs->cb = rte_eth_add_tx_callback(port, qid, pdump_tx, cbs); if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add tx callback, errno=%d\n", rte_errno); return rte_errno; @@ -203,7 +207,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, int ret; if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to delete non existing tx " "callback for port=%d and queue=%d\n", port, qid); @@ -211,7 +215,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, } ret = rte_eth_remove_tx_callback(port, qid, cbs->cb); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to remove tx callback, errno=%d\n", -ret); return ret; @@ -240,7 +244,7 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) ret = rte_eth_dev_get_port_by_name(p->data.en_v1.device, &port); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to get port id for device id=%s\n", p->data.en_v1.device); return -EINVAL; @@ -252,7 +256,7 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) ret = rte_eth_dev_get_port_by_name(p->data.dis_v1.device, &port); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to get port id for device id=%s\n", p->data.dis_v1.device); return -EINVAL; @@ -277,18 +281,18 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) nb_rx_q = dev_info.nb_rx_queues; nb_tx_q = dev_info.nb_tx_queues; if (nb_rx_q == 0 && flags & RTE_PDUMP_FLAG_RX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "number of rx queues cannot be 0\n"); return -EINVAL; } if (nb_tx_q == 0 && flags & RTE_PDUMP_FLAG_TX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "number of tx queues cannot be 0\n"); return -EINVAL; } if ((nb_tx_q == 0 || nb_rx_q == 0) && flags == RTE_PDUMP_FLAG_RXTX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "both tx&rx queues must be non zero\n"); return -EINVAL; } @@ -324,7 +328,7 @@ pdump_server(const struct rte_mp_msg *mp_msg, const void *peer) /* recv client requests */ if (mp_msg->len_param != sizeof(*cli_req)) { - RTE_LOG(ERR, PDUMP, "failed to recv from client\n"); + PDUMP_LOG(ERR, "failed to recv from client\n"); resp->err_value = -EINVAL; } else { cli_req = (const struct pdump_request *)mp_msg->param; @@ -337,8 +341,8 @@ pdump_server(const struct rte_mp_msg *mp_msg, const void *peer) mp_resp.len_param = sizeof(*resp); mp_resp.num_fds = 0; if (rte_mp_reply(&mp_resp, peer) < 0) { - RTE_LOG(ERR, PDUMP, "failed to send to client:%s, %s:%d\n", - strerror(rte_errno), __func__, __LINE__); + PDUMP_LOG(ERR, "failed to send to client:%s\n", + strerror(rte_errno)); return -1; } @@ -366,19 +370,18 @@ static int pdump_validate_ring_mp(struct rte_ring *ring, struct rte_mempool *mp) { if (ring == NULL || mp == NULL) { - RTE_LOG(ERR, PDUMP, "NULL ring or mempool are passed %s:%d\n", - __func__, __LINE__); + PDUMP_LOG(ERR, "NULL ring or mempool\n"); rte_errno = EINVAL; return -1; } if (mp->flags & MEMPOOL_F_SP_PUT || mp->flags & MEMPOOL_F_SC_GET) { - RTE_LOG(ERR, PDUMP, "mempool with either SP or SC settings" + PDUMP_LOG(ERR, "mempool with either SP or SC settings" " is not valid for pdump, should have MP and MC settings\n"); rte_errno = EINVAL; return -1; } if (ring->prod.single || ring->cons.single) { - RTE_LOG(ERR, PDUMP, "ring with either SP or SC settings" + PDUMP_LOG(ERR, "ring with either SP or SC settings" " is not valid for pdump, should have MP and MC settings\n"); rte_errno = EINVAL; return -1; @@ -392,7 +395,7 @@ pdump_validate_flags(uint32_t flags) { if (flags != RTE_PDUMP_FLAG_RX && flags != RTE_PDUMP_FLAG_TX && flags != RTE_PDUMP_FLAG_RXTX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "invalid flags, should be either rx/tx/rxtx\n"); rte_errno = EINVAL; return -1; @@ -407,17 +410,15 @@ pdump_validate_port(uint16_t port, char *name) int ret = 0; if (port >= RTE_MAX_ETHPORTS) { - RTE_LOG(ERR, PDUMP, "Invalid port id %u, %s:%d\n", port, - __func__, __LINE__); + PDUMP_LOG(ERR, "Invalid port id %u\n", port); rte_errno = EINVAL; return -1; } ret = rte_eth_dev_get_name_by_port(port, name); if (ret < 0) { - RTE_LOG(ERR, PDUMP, - "port id to name mapping failed for port id=%u, %s:%d\n", - port, __func__, __LINE__); + PDUMP_LOG(ERR, "port %u to name mapping failed\n", + port); rte_errno = EINVAL; return -1; } @@ -472,7 +473,7 @@ pdump_prepare_client_request(char *device, uint16_t queue, } if (ret < 0) - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "client request for pdump enable/disable failed\n"); return ret; } @@ -559,3 +560,10 @@ rte_pdump_disable_by_deviceid(char *device_id, uint16_t queue, return ret; } + +RTE_INIT(pdump_log) +{ + pdump_logtype = rte_log_register("lib.pdump"); + if (pdump_logtype >= 0) + rte_log_set_level(pdump_logtype, RTE_LOG_NOTICE); +} -- 2.20.1 ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype 2019-11-08 4:38 ` [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype Stephen Hemminger @ 2019-11-08 8:13 ` David Marchand 2019-11-08 16:19 ` Stephen Hemminger 0 siblings, 1 reply; 12+ messages in thread From: David Marchand @ 2019-11-08 8:13 UTC (permalink / raw) To: Stephen Hemminger; +Cc: dev On Fri, Nov 8, 2019 at 5:39 AM Stephen Hemminger <stephen@networkplumber.org> wrote: > > The logtype USER1 should not be overloaded for library function. > Instead use a dynamic log type. > > Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> This patch is incomplete. $ git grep -w RTE_LOG lib/librte_pdump/ lib/librte_pdump/rte_pdump.c:/* Macro for printing using RTE_LOG */ lib/librte_pdump/rte_pdump.c: RTE_LOG(ERR, PDUMP, == Build lib/librte_pdump CC rte_pdump.o In file included from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_bus.h:24:0, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_eal.h:21, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_lcore.h:16, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/generic/rte_spinlock.h:21, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_spinlock.h:12, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_mempool.h:44, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_mbuf.h:38, from /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/lib/librte_pdump/rte_pdump.c:6: /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/lib/librte_pdump/rte_pdump.c: In function ‘set_pdump_rxtx_cbs’: /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_log.h:337:4: error: ‘RTE_LOGTYPE_PDUMP’ undeclared (first use in this function); did you mean ‘RTE_LOGTYPE_PMD’? RTE_LOGTYPE_ ## t, # t ": " __VA_ARGS__) ^ /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/lib/librte_pdump/rte_pdump.c:275:4: note: in expansion of macro ‘RTE_LOG’ RTE_LOG(ERR, PDUMP, ^~~~~~~ /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/x86_64-native-linuxapp-gcc/include/rte_log.h:337:4: note: each undeclared identifier is reported only once for each function it appears in RTE_LOGTYPE_ ## t, # t ": " __VA_ARGS__) ^ /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/lib/librte_pdump/rte_pdump.c:275:4: note: in expansion of macro ‘RTE_LOG’ RTE_LOG(ERR, PDUMP, ^~~~~~~ /home-local/jenkins-local/jenkins-agent/workspace/Apply-Custom-Patch-Set/dpdk/lib/librte_pdump/rte_pdump.c: At top level: -- David Marchand ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype 2019-11-08 8:13 ` David Marchand @ 2019-11-08 16:19 ` Stephen Hemminger 0 siblings, 0 replies; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 16:19 UTC (permalink / raw) To: David Marchand; +Cc: dev On Fri, 8 Nov 2019 09:13:56 +0100 David Marchand <david.marchand@redhat.com> wrote: > On Fri, Nov 8, 2019 at 5:39 AM Stephen Hemminger > <stephen@networkplumber.org> wrote: > > > > The logtype USER1 should not be overloaded for library function. > > Instead use a dynamic log type. > > > > Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> > > This patch is incomplete. > $ git grep -w RTE_LOG lib/librte_pdump/ > lib/librte_pdump/rte_pdump.c:/* Macro for printing using RTE_LOG */ > lib/librte_pdump/rte_pdump.c: RTE_LOG(ERR, PDUMP, Thanks missed one that got added by later code. > ^ permalink raw reply [flat|nested] 12+ messages in thread
* [dpdk-dev] [PATCH v2 0/2] pdump: cleanups 2019-11-08 4:38 [dpdk-dev] [PATCH 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 1/2] pdump: use new pktmbuf copy function Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype Stephen Hemminger @ 2019-11-08 16:47 ` Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 1/2] pdump: use new pktmbuf copy function Stephen Hemminger ` (4 more replies) 2 siblings, 5 replies; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 16:47 UTC (permalink / raw) To: dev; +Cc: Stephen Hemminger These are a couple of small cleanups for 19.10 which came out of work on pcapng support. Full pcapng support and BPF are planned for DPDK 20.02. Stephen Hemminger (2): pdump: use new pktmbuf copy function pdump: use dynamic logtype lib/librte_pdump/rte_pdump.c | 139 ++++++++++------------------------- 1 file changed, 40 insertions(+), 99 deletions(-) v2 - replace new log message from info_get as well -- 2.20.1 ^ permalink raw reply [flat|nested] 12+ messages in thread
* [dpdk-dev] [PATCH v2 1/2] pdump: use new pktmbuf copy function 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger @ 2019-11-08 16:47 ` Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 2/2] pdump: use dynamic logtype Stephen Hemminger ` (3 subsequent siblings) 4 siblings, 0 replies; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 16:47 UTC (permalink / raw) To: dev; +Cc: Stephen Hemminger The rte_pktmbuf_copy handles varying size mbuf pools correctly. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> --- lib/librte_pdump/rte_pdump.c | 69 +----------------------------------- 1 file changed, 1 insertion(+), 68 deletions(-) diff --git a/lib/librte_pdump/rte_pdump.c b/lib/librte_pdump/rte_pdump.c index ac94fea93656..7816c7254615 100644 --- a/lib/librte_pdump/rte_pdump.c +++ b/lib/librte_pdump/rte_pdump.c @@ -64,73 +64,6 @@ static struct pdump_rxtx_cbs { } rx_cbs[RTE_MAX_ETHPORTS][RTE_MAX_QUEUES_PER_PORT], tx_cbs[RTE_MAX_ETHPORTS][RTE_MAX_QUEUES_PER_PORT]; -static inline int -pdump_pktmbuf_copy_data(struct rte_mbuf *seg, const struct rte_mbuf *m) -{ - if (rte_pktmbuf_tailroom(seg) < m->data_len) { - RTE_LOG(ERR, PDUMP, - "User mempool: insufficient data_len of mbuf\n"); - return -EINVAL; - } - - seg->port = m->port; - seg->vlan_tci = m->vlan_tci; - seg->hash = m->hash; - seg->tx_offload = m->tx_offload; - seg->ol_flags = m->ol_flags; - seg->packet_type = m->packet_type; - seg->vlan_tci_outer = m->vlan_tci_outer; - seg->data_len = m->data_len; - seg->pkt_len = seg->data_len; - rte_memcpy(rte_pktmbuf_mtod(seg, void *), - rte_pktmbuf_mtod(m, void *), - rte_pktmbuf_data_len(seg)); - - return 0; -} - -static inline struct rte_mbuf * -pdump_pktmbuf_copy(struct rte_mbuf *m, struct rte_mempool *mp) -{ - struct rte_mbuf *m_dup, *seg, **prev; - uint32_t pktlen; - uint16_t nseg; - - m_dup = rte_pktmbuf_alloc(mp); - if (unlikely(m_dup == NULL)) - return NULL; - - seg = m_dup; - prev = &seg->next; - pktlen = m->pkt_len; - nseg = 0; - - do { - nseg++; - if (pdump_pktmbuf_copy_data(seg, m) < 0) { - if (seg != m_dup) - rte_pktmbuf_free_seg(seg); - rte_pktmbuf_free(m_dup); - return NULL; - } - *prev = seg; - prev = &seg->next; - } while ((m = m->next) != NULL && - (seg = rte_pktmbuf_alloc(mp)) != NULL); - - *prev = NULL; - m_dup->nb_segs = nseg; - m_dup->pkt_len = pktlen; - - /* Allocation of new indirect segment failed */ - if (unlikely(seg == NULL)) { - rte_pktmbuf_free(m_dup); - return NULL; - } - - __rte_mbuf_sanity_check(m_dup, 1); - return m_dup; -} static inline void pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) @@ -148,7 +81,7 @@ pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) ring = cbs->ring; mp = cbs->mp; for (i = 0; i < nb_pkts; i++) { - p = pdump_pktmbuf_copy(pkts[i], mp); + p = rte_pktmbuf_copy(pkts[i], mp, 0, UINT32_MAX); if (p) dup_bufs[d_pkts++] = p; } -- 2.20.1 ^ permalink raw reply [flat|nested] 12+ messages in thread
* [dpdk-dev] [PATCH v2 2/2] pdump: use dynamic logtype 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 1/2] pdump: use new pktmbuf copy function Stephen Hemminger @ 2019-11-08 16:47 ` Stephen Hemminger 2019-11-12 20:37 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups David Marchand ` (2 subsequent siblings) 4 siblings, 0 replies; 12+ messages in thread From: Stephen Hemminger @ 2019-11-08 16:47 UTC (permalink / raw) To: dev; +Cc: Stephen Hemminger The logtype USER1 should not be overloaded for library function. Instead use a dynamic log type. Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> --- lib/librte_pdump/rte_pdump.c | 70 ++++++++++++++++++++---------------- 1 file changed, 39 insertions(+), 31 deletions(-) diff --git a/lib/librte_pdump/rte_pdump.c b/lib/librte_pdump/rte_pdump.c index 7816c7254615..8a01ac51071a 100644 --- a/lib/librte_pdump/rte_pdump.c +++ b/lib/librte_pdump/rte_pdump.c @@ -13,8 +13,12 @@ #include "rte_pdump.h" #define DEVICE_ID_SIZE 64 -/* Macros for printing using RTE_LOG */ -#define RTE_LOGTYPE_PDUMP RTE_LOGTYPE_USER1 + +/* Macro for printing using RTE_LOG */ +static int pdump_logtype; +#define PDUMP_LOG(level, fmt, args...) \ + rte_log(RTE_LOG_ ## level, pdump_logtype, "%s(): " fmt, \ + __func__, ## args) /* Used for the multi-process communication */ #define PDUMP_MP "mp_pdump" @@ -88,7 +92,7 @@ pdump_copy(struct rte_mbuf **pkts, uint16_t nb_pkts, void *user_params) ring_enq = rte_ring_enqueue_burst(ring, (void *)dup_bufs, d_pkts, NULL); if (unlikely(ring_enq < d_pkts)) { - RTE_LOG(DEBUG, PDUMP, + PDUMP_LOG(DEBUG, "only %d of packets enqueued to ring\n", ring_enq); do { rte_pktmbuf_free(dup_bufs[ring_enq]); @@ -127,7 +131,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs = &rx_cbs[port][qid]; if (cbs && operation == ENABLE) { if (cbs->cb) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add rx callback for port=%d " "and queue=%d, callback already exists\n", port, qid); @@ -138,7 +142,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs->cb = rte_eth_add_first_rx_callback(port, qid, pdump_rx, cbs); if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add rx callback, errno=%d\n", rte_errno); return rte_errno; @@ -148,7 +152,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, int ret; if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to delete non existing rx " "callback for port=%d and queue=%d\n", port, qid); @@ -156,7 +160,7 @@ pdump_register_rx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, } ret = rte_eth_remove_rx_callback(port, qid, cbs->cb); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to remove rx callback, errno=%d\n", -ret); return ret; @@ -182,7 +186,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs = &tx_cbs[port][qid]; if (cbs && operation == ENABLE) { if (cbs->cb) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add tx callback for port=%d " "and queue=%d, callback already exists\n", port, qid); @@ -193,7 +197,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, cbs->cb = rte_eth_add_tx_callback(port, qid, pdump_tx, cbs); if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to add tx callback, errno=%d\n", rte_errno); return rte_errno; @@ -203,7 +207,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, int ret; if (cbs->cb == NULL) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to delete non existing tx " "callback for port=%d and queue=%d\n", port, qid); @@ -211,7 +215,7 @@ pdump_register_tx_callbacks(uint16_t end_q, uint16_t port, uint16_t queue, } ret = rte_eth_remove_tx_callback(port, qid, cbs->cb); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to remove tx callback, errno=%d\n", -ret); return ret; @@ -240,7 +244,7 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) ret = rte_eth_dev_get_port_by_name(p->data.en_v1.device, &port); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to get port id for device id=%s\n", p->data.en_v1.device); return -EINVAL; @@ -252,7 +256,7 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) ret = rte_eth_dev_get_port_by_name(p->data.dis_v1.device, &port); if (ret < 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "failed to get port id for device id=%s\n", p->data.dis_v1.device); return -EINVAL; @@ -268,7 +272,7 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) ret = rte_eth_dev_info_get(port, &dev_info); if (ret != 0) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "Error during getting device (port %u) info: %s\n", port, strerror(-ret)); return ret; @@ -277,18 +281,18 @@ set_pdump_rxtx_cbs(const struct pdump_request *p) nb_rx_q = dev_info.nb_rx_queues; nb_tx_q = dev_info.nb_tx_queues; if (nb_rx_q == 0 && flags & RTE_PDUMP_FLAG_RX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "number of rx queues cannot be 0\n"); return -EINVAL; } if (nb_tx_q == 0 && flags & RTE_PDUMP_FLAG_TX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "number of tx queues cannot be 0\n"); return -EINVAL; } if ((nb_tx_q == 0 || nb_rx_q == 0) && flags == RTE_PDUMP_FLAG_RXTX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "both tx&rx queues must be non zero\n"); return -EINVAL; } @@ -324,7 +328,7 @@ pdump_server(const struct rte_mp_msg *mp_msg, const void *peer) /* recv client requests */ if (mp_msg->len_param != sizeof(*cli_req)) { - RTE_LOG(ERR, PDUMP, "failed to recv from client\n"); + PDUMP_LOG(ERR, "failed to recv from client\n"); resp->err_value = -EINVAL; } else { cli_req = (const struct pdump_request *)mp_msg->param; @@ -337,8 +341,8 @@ pdump_server(const struct rte_mp_msg *mp_msg, const void *peer) mp_resp.len_param = sizeof(*resp); mp_resp.num_fds = 0; if (rte_mp_reply(&mp_resp, peer) < 0) { - RTE_LOG(ERR, PDUMP, "failed to send to client:%s, %s:%d\n", - strerror(rte_errno), __func__, __LINE__); + PDUMP_LOG(ERR, "failed to send to client:%s\n", + strerror(rte_errno)); return -1; } @@ -366,19 +370,18 @@ static int pdump_validate_ring_mp(struct rte_ring *ring, struct rte_mempool *mp) { if (ring == NULL || mp == NULL) { - RTE_LOG(ERR, PDUMP, "NULL ring or mempool are passed %s:%d\n", - __func__, __LINE__); + PDUMP_LOG(ERR, "NULL ring or mempool\n"); rte_errno = EINVAL; return -1; } if (mp->flags & MEMPOOL_F_SP_PUT || mp->flags & MEMPOOL_F_SC_GET) { - RTE_LOG(ERR, PDUMP, "mempool with either SP or SC settings" + PDUMP_LOG(ERR, "mempool with either SP or SC settings" " is not valid for pdump, should have MP and MC settings\n"); rte_errno = EINVAL; return -1; } if (ring->prod.single || ring->cons.single) { - RTE_LOG(ERR, PDUMP, "ring with either SP or SC settings" + PDUMP_LOG(ERR, "ring with either SP or SC settings" " is not valid for pdump, should have MP and MC settings\n"); rte_errno = EINVAL; return -1; @@ -392,7 +395,7 @@ pdump_validate_flags(uint32_t flags) { if (flags != RTE_PDUMP_FLAG_RX && flags != RTE_PDUMP_FLAG_TX && flags != RTE_PDUMP_FLAG_RXTX) { - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "invalid flags, should be either rx/tx/rxtx\n"); rte_errno = EINVAL; return -1; @@ -407,17 +410,15 @@ pdump_validate_port(uint16_t port, char *name) int ret = 0; if (port >= RTE_MAX_ETHPORTS) { - RTE_LOG(ERR, PDUMP, "Invalid port id %u, %s:%d\n", port, - __func__, __LINE__); + PDUMP_LOG(ERR, "Invalid port id %u\n", port); rte_errno = EINVAL; return -1; } ret = rte_eth_dev_get_name_by_port(port, name); if (ret < 0) { - RTE_LOG(ERR, PDUMP, - "port id to name mapping failed for port id=%u, %s:%d\n", - port, __func__, __LINE__); + PDUMP_LOG(ERR, "port %u to name mapping failed\n", + port); rte_errno = EINVAL; return -1; } @@ -472,7 +473,7 @@ pdump_prepare_client_request(char *device, uint16_t queue, } if (ret < 0) - RTE_LOG(ERR, PDUMP, + PDUMP_LOG(ERR, "client request for pdump enable/disable failed\n"); return ret; } @@ -559,3 +560,10 @@ rte_pdump_disable_by_deviceid(char *device_id, uint16_t queue, return ret; } + +RTE_INIT(pdump_log) +{ + pdump_logtype = rte_log_register("lib.pdump"); + if (pdump_logtype >= 0) + rte_log_set_level(pdump_logtype, RTE_LOG_NOTICE); +} -- 2.20.1 ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 1/2] pdump: use new pktmbuf copy function Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 2/2] pdump: use dynamic logtype Stephen Hemminger @ 2019-11-12 20:37 ` David Marchand 2019-11-15 12:50 ` Pattan, Reshma 2020-06-11 8:46 ` Dong Zhou 4 siblings, 0 replies; 12+ messages in thread From: David Marchand @ 2019-11-12 20:37 UTC (permalink / raw) To: Pattan, Reshma, Stephen Hemminger; +Cc: dev On Fri, Nov 8, 2019 at 5:47 PM Stephen Hemminger <stephen@networkplumber.org> wrote: > > These are a couple of small cleanups for 19.10 which Stephen, we might be a bit late for this release, but it is still 19.11 :-) > came out of work on pcapng support. Full pcapng support > and BPF are planned for DPDK 20.02. > > Stephen Hemminger (2): > pdump: use new pktmbuf copy function > pdump: use dynamic logtype > > lib/librte_pdump/rte_pdump.c | 139 ++++++++++------------------------- > 1 file changed, 40 insertions(+), 99 deletions(-) > > v2 - replace new log message from info_get as well Overall looks good. Reshma, review please. Thanks. -- David Marchand ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger ` (2 preceding siblings ...) 2019-11-12 20:37 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups David Marchand @ 2019-11-15 12:50 ` Pattan, Reshma 2020-02-05 20:20 ` David Marchand 2020-06-11 8:46 ` Dong Zhou 4 siblings, 1 reply; 12+ messages in thread From: Pattan, Reshma @ 2019-11-15 12:50 UTC (permalink / raw) To: Stephen Hemminger, dev > -----Original Message----- > From: dev <dev-bounces@dpdk.org> On Behalf Of Stephen Hemminger > Sent: Friday, November 8, 2019 4:47 PM > To: dev@dpdk.org > Cc: Stephen Hemminger <stephen@networkplumber.org> > Subject: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups > > These are a couple of small cleanups for 19.10 which came out of work on > pcapng support. Full pcapng support and BPF are planned for DPDK 20.02. > > Stephen Hemminger (2): > pdump: use new pktmbuf copy function > pdump: use dynamic logtype > > lib/librte_pdump/rte_pdump.c | 139 ++++++++++------------------------- > 1 file changed, 40 insertions(+), 99 deletions(-) > > v2 - replace new log message from info_get as well > > -- > 2.20.1 Looks good to me. Acked-by: Reshma Pattan <reshma.pattan@intel.com> ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups 2019-11-15 12:50 ` Pattan, Reshma @ 2020-02-05 20:20 ` David Marchand 0 siblings, 0 replies; 12+ messages in thread From: David Marchand @ 2020-02-05 20:20 UTC (permalink / raw) To: Stephen Hemminger; +Cc: dev, Pattan, Reshma On Fri, Nov 15, 2019 at 1:51 PM Pattan, Reshma <reshma.pattan@intel.com> wrote: > > -----Original Message----- > > From: dev <dev-bounces@dpdk.org> On Behalf Of Stephen Hemminger > > Sent: Friday, November 8, 2019 4:47 PM > > To: dev@dpdk.org > > Cc: Stephen Hemminger <stephen@networkplumber.org> > > Subject: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups > > > > These are a couple of small cleanups for 19.10 which came out of work on > > pcapng support. Full pcapng support and BPF are planned for DPDK 20.02. > > > > Stephen Hemminger (2): > > pdump: use new pktmbuf copy function > > pdump: use dynamic logtype > > > > lib/librte_pdump/rte_pdump.c | 139 ++++++++++------------------------- > > 1 file changed, 40 insertions(+), 99 deletions(-) > > > > v2 - replace new log message from info_get as well > > > > -- > > 2.20.1 > Looks good to me. > Acked-by: Reshma Pattan <reshma.pattan@intel.com> Series applied, thanks. -- David Marchand ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger ` (3 preceding siblings ...) 2019-11-15 12:50 ` Pattan, Reshma @ 2020-06-11 8:46 ` Dong Zhou 4 siblings, 0 replies; 12+ messages in thread From: Dong Zhou @ 2020-06-11 8:46 UTC (permalink / raw) To: luca.boccassi, stephen; +Cc: dev, stable > -----Original Message----- > From: dev <dev-bounces@dpdk.org> On Behalf Of Stephen Hemminger > Sent: Friday, November 8, 2019 4:47 PM > To: dev@dpdk.org > Cc: Stephen Hemminger <stephen@networkplumber.org> > Subject: [dpdk-dev] [PATCH v2 0/2] pdump: cleanups > > These are a couple of small cleanups for 19.10 which came out of work on > pcapng support. Full pcapng support and BPF are planned for DPDK 20.02. > > Stephen Hemminger (2): > pdump: use new pktmbuf copy function > pdump: use dynamic logtype Hi luca, will this patch be in 19.11.3 stable release? ^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2020-06-11 8:47 UTC | newest] Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2019-11-08 4:38 [dpdk-dev] [PATCH 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 1/2] pdump: use new pktmbuf copy function Stephen Hemminger 2019-11-08 4:38 ` [dpdk-dev] [PATCH 2/2] pdump: use dynamic logtype Stephen Hemminger 2019-11-08 8:13 ` David Marchand 2019-11-08 16:19 ` Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 1/2] pdump: use new pktmbuf copy function Stephen Hemminger 2019-11-08 16:47 ` [dpdk-dev] [PATCH v2 2/2] pdump: use dynamic logtype Stephen Hemminger 2019-11-12 20:37 ` [dpdk-dev] [PATCH v2 0/2] pdump: cleanups David Marchand 2019-11-15 12:50 ` Pattan, Reshma 2020-02-05 20:20 ` David Marchand 2020-06-11 8:46 ` Dong Zhou
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).