From 79c913a42f0e9c0ca493d4ed8f91efa7cb92d68a Mon Sep 17 00:00:00 2001 From: Kuba Kozak Date: Thu, 27 Apr 2017 16:42:36 +0200 Subject: [PATCH] ethdev: retrieve xstats by ID Extended xstats API in ethdev library to allow grouping of stats logically so they can be retrieved per logical grouping managed by the application. Added new functions rte_eth_xstats_get_names_by_id and rte_eth_xstats_get_by_id using additional arguments (in compare to rte_eth_xstats_get_names and rte_eth_xstats_get) - array of ids and array of values. doc: add description for modified xstats API Documentation change for new extended statistics API functions. The old API only allows retrieval of *all* of the NIC statistics at once. Given this requires a MMIO read PCI transaction per statistic it is an inefficient way of retrieving just a few key statistics. Often a monitoring agent only has an interest in a few key statistics, and the old API forces wasting CPU time and PCIe bandwidth in retrieving *all* statistics; even those that the application didn't explicitly show an interest in. The new, more flexible API allow retrieval of statistics per ID. If a PMD wishes, it can be implemented to read just the required NIC registers. As a result, the monitoring application no longer wastes PCIe bandwidth and CPU time. Signed-off-by: Kuba Kozak --- doc/guides/prog_guide/poll_mode_drv.rst | 167 ++++++++++++++-- lib/librte_ether/rte_ethdev.c | 250 ++++++++++++++++++++++++ lib/librte_ether/rte_ethdev.h | 69 +++++++ lib/librte_ether/rte_ether_version.map | 2 + 4 files changed, 471 insertions(+), 17 deletions(-) diff --git a/doc/guides/prog_guide/poll_mode_drv.rst b/doc/guides/prog_guide/poll_mode_drv.rst index e48c121c00..4987f70a17 100644 --- a/doc/guides/prog_guide/poll_mode_drv.rst +++ b/doc/guides/prog_guide/poll_mode_drv.rst @@ -334,24 +334,21 @@ The Ethernet device API exported by the Ethernet PMDs is described in the *DPDK Extended Statistics API ~~~~~~~~~~~~~~~~~~~~~~~ -The extended statistics API allows each individual PMD to expose a unique set -of statistics. Accessing these from application programs is done via two -functions: - -* ``rte_eth_xstats_get``: Fills in an array of ``struct rte_eth_xstat`` - with extended statistics. -* ``rte_eth_xstats_get_names``: Fills in an array of - ``struct rte_eth_xstat_name`` with extended statistic name lookup - information. - -Each ``struct rte_eth_xstat`` contains an identifier and value pair, and -each ``struct rte_eth_xstat_name`` contains a string. Each identifier -within the ``struct rte_eth_xstat`` lookup array must have a corresponding -entry in the ``struct rte_eth_xstat_name`` lookup array. Within the latter -the index of the entry is the identifier the string is associated with. -These identifiers, as well as the number of extended statistic exposed, must -remain constant during runtime. Note that extended statistic identifiers are +The extended statistics API allows a PMD to expose all statistics that are +available to it, including statistics that are unique to the device. +Each statistic has three properties ``name``, ``id`` and ``value``: + +* ``name``: A human readable string formatted by the scheme detailed below. +* ``id``: An integer that represents only that statistic. +* ``value``: A unsigned 64-bit integer that is the value of the statistic. + +Note that extended statistic identifiers are driver-specific, and hence might not be the same for different ports. +The API consists of various ``rte_eth_xstats_*()`` functions, and allows an +application to be flexible in how it retrieves statistics. + +Scheme for Human Readable Names +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ A naming scheme exists for the strings exposed to clients of the API. This is to allow scraping of the API for statistics of interest. The naming scheme uses @@ -392,3 +389,139 @@ Some additions in the metadata scheme are as follows: An example where queue numbers are used is as follows: ``tx_q7_bytes`` which indicates this statistic applies to queue number 7, and represents the number of transmitted bytes on that queue. + +API Design +^^^^^^^^^^ + +The xstats API uses the ``name``, ``id``, and ``value`` to allow performant +lookup of specific statistics. Performant lookup means two things; + +* No string comparisons with the ``name`` of the statistic in fast-path +* Allow requesting of only the statistics of interest + +The API ensures these requirements are met by mapping the ``name`` of the +statistic to a unique ``id``, which is used as a key for lookup in the fast-path. +The API allows applications to request an array of ``id`` values, so that the +PMD only performs the required calculations. Expected usage is that the +application scans the ``name`` of each statistic, and caches the ``id`` +if it has an interest in that statistic. On the fast-path, the integer can be used +to retrieve the actual ``value`` of the statistic that the ``id`` represents. + +API Functions +^^^^^^^^^^^^^ + +The API is built out of a small number of functions, which can be used to +retrieve the number of statistics and the names, IDs and values of those +statistics. + +* ``rte_eth_xstats_get_names_by_id()``: returns the names of the statistics. When given a + ``NULL`` parameter the function returns the number of statistics that are available. + +* ``rte_eth_xstats_get_id_by_name()``: Searches for the statistic ID that matches + ``xstat_name``. If found, the ``id`` integer is set. + +* ``rte_eth_xstats_get_by_id()``: Fills in an array of ``uint64_t`` values + with matching the provided ``ids`` array. If the ``ids`` array is NULL, it + returns all statistics that are available. + + +Application Usage +^^^^^^^^^^^^^^^^^ + +Imagine an application that wants to view the dropped packet count. If no +packets are dropped, the application does not read any other metrics for +performance reasons. If packets are dropped, the application has a particular +set of statistics that it requests. This "set" of statistics allows the app to +decide what next steps to perform. The following code-snippets show how the +xstats API can be used to achieve this goal. + +First step is to get all statistics names and list them: + +.. code-block:: c + + struct rte_eth_xstat_name *xstats_names; + uint64_t *values; + int len, i; + + /* Get number of stats */ + len = rte_eth_xstats_get_names_by_id(port_id, NULL, NULL, 0); + if (len < 0) { + printf("Cannot get xstats count\n"); + goto err; + } + + xstats_names = malloc(sizeof(struct rte_eth_xstat_name) * len); + if (xstats_names == NULL) { + printf("Cannot allocate memory for xstat names\n"); + goto err; + } + + /* Retrieve xstats names, passing NULL for IDs to return all statistics */ + if (len != rte_eth_xstats_get_names_by_id(port_id, xstats_names, NULL, len)) { + printf("Cannot get xstat names\n"); + goto err; + } + + values = malloc(sizeof(values) * len); + if (values == NULL) { + printf("Cannot allocate memory for xstats\n"); + goto err; + } + + /* Getting xstats values */ + if (len != rte_eth_xstats_get_by_id(port_id, NULL, values, len)) { + printf("Cannot get xstat values\n"); + goto err; + } + + /* Print all xstats names and values */ + for (i = 0; i < len; i++) { + printf("%s: %"PRIu64"\n", xstats_names[i].name, values[i]); + } + +The application has access to the names of all of the statistics that the PMD +exposes. The application can decide which statistics are of interest, cache the +ids of those statistics by looking up the name as follows: + +.. code-block:: c + + uint64_t id; + uint64_t value; + const char *xstat_name = "rx_errors"; + + if(!rte_eth_xstats_get_id_by_name(port_id, xstat_name, &id)) { + rte_eth_xstats_get_by_id(port_id, &id, &value, 1); + printf("%s: %"PRIu64"\n", xstat_name, value); + } + else { + printf("Cannot find xstats with a given name\n"); + goto err; + } + +The API provides flexibility to the application so that it can look up multiple +statistics using an array containing multiple ``id`` numbers. This reduces the +function call overhead of retrieving statistics, and makes lookup of multiple +statistics simpler for the application. + +.. code-block:: c + + #define APP_NUM_STATS 4 + /* application cached these ids previously; see above */ + uint64_t ids_array[APP_NUM_STATS] = {3,4,7,21}; + uint64_t value_array[APP_NUM_STATS]; + + /* Getting multiple xstats values from array of IDs */ + rte_eth_xstats_get_by_id(port_id, ids_array, value_array, APP_NUM_STATS); + + uint32_t i; + for(i = 0; i < APP_NUM_STATS; i++) { + printf("%d: %"PRIu64"\n", ids_array[i], value_array[i]); + } + + +This array lookup API for xstats allows the application create multiple +"groups" of statistics, and look up the values of those IDs using a single API +call. As an end result, the application is able to achieve its goal of +monitoring a single statistic ("rx_errors" in this case), and if that shows +packets being dropped, it can easily retrieve a "set" of statistics using the +IDs array parameter to ``rte_eth_xstats_get_by_id`` function. diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c index 06b0924803..0a012ccbfc 100644 --- a/lib/librte_ether/rte_ethdev.c +++ b/lib/librte_ether/rte_ethdev.c @@ -1362,12 +1362,19 @@ get_xstats_count(uint8_t port_id) RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL); dev = &rte_eth_devices[port_id]; + if (dev->dev_ops->xstats_get_names_by_id != NULL) { + count = (*dev->dev_ops->xstats_get_names_by_id)(dev, NULL, + NULL, 0); + if (count < 0) + return count; + } if (dev->dev_ops->xstats_get_names != NULL) { count = (*dev->dev_ops->xstats_get_names)(dev, NULL, 0); if (count < 0) return count; } else count = 0; + count += RTE_NB_STATS; count += RTE_MIN(dev->data->nb_rx_queues, RTE_ETHDEV_QUEUE_STAT_CNTRS) * RTE_NB_RXQ_STATS; @@ -1376,6 +1383,123 @@ get_xstats_count(uint8_t port_id) return count; } +int +rte_eth_xstats_get_names_by_id(uint8_t port_id, + struct rte_eth_xstat_name *xstats_names, unsigned int size, + uint64_t *ids) +{ + /* Get all xstats */ + if (!ids) { + struct rte_eth_dev *dev; + int cnt_used_entries; + int cnt_expected_entries; + int cnt_driver_entries; + uint32_t idx, id_queue; + uint16_t num_q; + + cnt_expected_entries = get_xstats_count(port_id); + if (xstats_names == NULL || cnt_expected_entries < 0 || + (int)size < cnt_expected_entries) + return cnt_expected_entries; + + /* port_id checked in get_xstats_count() */ + dev = &rte_eth_devices[port_id]; + cnt_used_entries = 0; + + for (idx = 0; idx < RTE_NB_STATS; idx++) { + snprintf(xstats_names[cnt_used_entries].name, + sizeof(xstats_names[0].name), + "%s", rte_stats_strings[idx].name); + cnt_used_entries++; + } + num_q = RTE_MIN(dev->data->nb_rx_queues, + RTE_ETHDEV_QUEUE_STAT_CNTRS); + for (id_queue = 0; id_queue < num_q; id_queue++) { + for (idx = 0; idx < RTE_NB_RXQ_STATS; idx++) { + snprintf(xstats_names[cnt_used_entries].name, + sizeof(xstats_names[0].name), + "rx_q%u%s", + id_queue, + rte_rxq_stats_strings[idx].name); + cnt_used_entries++; + } + + } + num_q = RTE_MIN(dev->data->nb_tx_queues, + RTE_ETHDEV_QUEUE_STAT_CNTRS); + for (id_queue = 0; id_queue < num_q; id_queue++) { + for (idx = 0; idx < RTE_NB_TXQ_STATS; idx++) { + snprintf(xstats_names[cnt_used_entries].name, + sizeof(xstats_names[0].name), + "tx_q%u%s", + id_queue, + rte_txq_stats_strings[idx].name); + cnt_used_entries++; + } + } + + if (dev->dev_ops->xstats_get_names_by_id != NULL) { + /* If there are any driver-specific xstats, append them + * to end of list. + */ + cnt_driver_entries = + (*dev->dev_ops->xstats_get_names_by_id)( + dev, + xstats_names + cnt_used_entries, + NULL, + size - cnt_used_entries); + if (cnt_driver_entries < 0) + return cnt_driver_entries; + cnt_used_entries += cnt_driver_entries; + + } else if (dev->dev_ops->xstats_get_names != NULL) { + /* If there are any driver-specific xstats, append them + * to end of list. + */ + cnt_driver_entries = (*dev->dev_ops->xstats_get_names)( + dev, + xstats_names + cnt_used_entries, + size - cnt_used_entries); + if (cnt_driver_entries < 0) + return cnt_driver_entries; + cnt_used_entries += cnt_driver_entries; + } + + return cnt_used_entries; + } + /* Get only xstats given by IDS */ + else { + uint16_t len, i; + struct rte_eth_xstat_name *xstats_names_copy; + + len = rte_eth_xstats_get_names_by_id(port_id, NULL, 0, NULL); + + xstats_names_copy = + malloc(sizeof(struct rte_eth_xstat_name) * len); + if (!xstats_names_copy) { + RTE_PMD_DEBUG_TRACE( + "ERROR: can't allocate memory for values_copy\n"); + free(xstats_names_copy); + return -1; + } + + rte_eth_xstats_get_names_by_id(port_id, xstats_names_copy, + len, NULL); + + for (i = 0; i < size; i++) { + if (ids[i] >= len) { + RTE_PMD_DEBUG_TRACE( + "ERROR: id value isn't valid\n"); + return -1; + } + strcpy(xstats_names[i].name, + xstats_names_copy[ids[i]].name); + } + free(xstats_names_copy); + return size; + } +} + int rte_eth_xstats_get_names(uint8_t port_id, struct rte_eth_xstat_name *xstats_names, @@ -1442,6 +1566,132 @@ rte_eth_xstats_get_names(uint8_t port_id, } /* retrieve ethdev extended statistics */ +int +rte_eth_xstats_get_by_id(uint8_t port_id, const uint64_t *ids, uint64_t *values, + unsigned int n) +{ + /* If need all xstats */ + if (!ids) { + struct rte_eth_stats eth_stats; + struct rte_eth_dev *dev; + unsigned int count = 0, i, q; + signed int xcount = 0; + uint64_t val, *stats_ptr; + uint16_t nb_rxqs, nb_txqs; + + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); + dev = &rte_eth_devices[port_id]; + + nb_rxqs = RTE_MIN(dev->data->nb_rx_queues, + RTE_ETHDEV_QUEUE_STAT_CNTRS); + nb_txqs = RTE_MIN(dev->data->nb_tx_queues, + RTE_ETHDEV_QUEUE_STAT_CNTRS); + + /* Return generic statistics */ + count = RTE_NB_STATS + (nb_rxqs * RTE_NB_RXQ_STATS) + + (nb_txqs * RTE_NB_TXQ_STATS); + + + /* implemented by the driver */ + if (dev->dev_ops->xstats_get_by_id != NULL) { + /* Retrieve the xstats from the driver at the end of the + * xstats struct. Retrieve all xstats. + */ + xcount = (*dev->dev_ops->xstats_get_by_id)(dev, + NULL, + values ? values + count : NULL, + (n > count) ? n - count : 0); + + if (xcount < 0) + return xcount; + /* implemented by the driver */ + } else if (dev->dev_ops->xstats_get != NULL) { + /* Retrieve the xstats from the driver at the end of the + * xstats struct. Retrieve all xstats. + * Compatibility for PMD without xstats_get_by_ids + */ + unsigned int size = (n > count) ? n - count : 1; + struct rte_eth_xstat xstats[size]; + + xcount = (*dev->dev_ops->xstats_get)(dev, + values ? xstats : NULL, size); + + if (xcount < 0) + return xcount; + + if (values != NULL) + for (i = 0 ; i < (unsigned int)xcount; i++) + values[i + count] = xstats[i].value; + } + + if (n < count + xcount || values == NULL) + return count + xcount; + + /* now fill the xstats structure */ + count = 0; + rte_eth_stats_get(port_id, ð_stats); + + /* global stats */ + for (i = 0; i < RTE_NB_STATS; i++) { + stats_ptr = RTE_PTR_ADD(ð_stats, + rte_stats_strings[i].offset); + val = *stats_ptr; + values[count++] = val; + } + + /* per-rxq stats */ + for (q = 0; q < nb_rxqs; q++) { + for (i = 0; i < RTE_NB_RXQ_STATS; i++) { + stats_ptr = RTE_PTR_ADD(ð_stats, + rte_rxq_stats_strings[i].offset + + q * sizeof(uint64_t)); + val = *stats_ptr; + values[count++] = val; + } + } + + /* per-txq stats */ + for (q = 0; q < nb_txqs; q++) { + for (i = 0; i < RTE_NB_TXQ_STATS; i++) { + stats_ptr = RTE_PTR_ADD(ð_stats, + rte_txq_stats_strings[i].offset + + q * sizeof(uint64_t)); + val = *stats_ptr; + values[count++] = val; + } + } + + return count + xcount; + } + /* Need only xstats given by IDS array */ + else { + uint16_t i, size; + uint64_t *values_copy; + + size = rte_eth_xstats_get_by_id(port_id, NULL, NULL, 0); + + values_copy = malloc(sizeof(values_copy) * size); + if (!values_copy) { + RTE_PMD_DEBUG_TRACE( + "ERROR: can't allocate memory for values_copy\n"); + return -1; + } + + rte_eth_xstats_get_by_id(port_id, NULL, values_copy, size); + + for (i = 0; i < n; i++) { + if (ids[i] >= size) { + RTE_PMD_DEBUG_TRACE( + "ERROR: id value isn't valid\n"); + return -1; + } + values[i] = values_copy[ids[i]]; + } + free(values_copy); + return n; + } +} + int rte_eth_xstats_get(uint8_t port_id, struct rte_eth_xstat *xstats, unsigned int n) diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h index 5beb857b15..e15ae882f5 100644 --- a/lib/librte_ether/rte_ethdev.h +++ b/lib/librte_ether/rte_ethdev.h @@ -1112,6 +1112,12 @@ typedef int (*eth_xstats_get_t)(struct rte_eth_dev *dev, struct rte_eth_xstat *stats, unsigned n); /**< @internal Get extended stats of an Ethernet device. */ +typedef int (*eth_xstats_get_by_id_t)(struct rte_eth_dev *dev, + const uint64_t *ids, + uint64_t *values, + unsigned int n); +/**< @internal Get extended stats of an Ethernet device. */ + typedef void (*eth_xstats_reset_t)(struct rte_eth_dev *dev); /**< @internal Reset extended stats of an Ethernet device. */ @@ -1119,6 +1125,11 @@ typedef int (*eth_xstats_get_names_t)(struct rte_eth_dev *dev, struct rte_eth_xstat_name *xstats_names, unsigned size); /**< @internal Get names of extended stats of an Ethernet device. */ +typedef int (*eth_xstats_get_names_by_id_t)(struct rte_eth_dev *dev, + struct rte_eth_xstat_name *xstats_names, const uint64_t *ids, + unsigned int size); +/**< @internal Get names of extended stats of an Ethernet device. */ + typedef int (*eth_queue_stats_mapping_set_t)(struct rte_eth_dev *dev, uint16_t queue_id, uint8_t stat_idx, @@ -1557,6 +1568,11 @@ struct eth_dev_ops { eth_timesync_adjust_time timesync_adjust_time; /** Adjust the device clock. */ eth_timesync_read_time timesync_read_time; /** Get the device clock time. */ eth_timesync_write_time timesync_write_time; /** Set the device clock time. */ + + eth_xstats_get_by_id_t xstats_get_by_id; + /**< Get extended device statistic values by ID. */ + eth_xstats_get_names_by_id_t xstats_get_names_by_id; + /**< Get name of extended device statistics by ID. */ }; /** @@ -2309,6 +2325,59 @@ int rte_eth_xstats_get_names(uint8_t port_id, int rte_eth_xstats_get(uint8_t port_id, struct rte_eth_xstat *xstats, unsigned int n); +/** + * Retrieve names of extended statistics of an Ethernet device. + * + * @param port_id + * The port identifier of the Ethernet device. + * @param xstats_names + * An rte_eth_xstat_name array of at least *size* elements to + * be filled. If set to NULL, the function returns the required number + * of elements. + * @param ids + * IDs array given by app to retrieve specific statistics + * @param size + * The size of the xstats_names array (number of elements). + * @return + * - A positive value lower or equal to size: success. The return value + * is the number of entries filled in the stats table. + * - A positive value higher than size: error, the given statistics table + * is too small. The return value corresponds to the size that should + * be given to succeed. The entries in the table are not valid and + * shall not be used by the caller. + * - A negative value on error (invalid port id). + */ +int +rte_eth_xstats_get_names_by_id(uint8_t port_id, + struct rte_eth_xstat_name *xstats_names, unsigned int size, + uint64_t *ids); + +/** + * Retrieve extended statistics of an Ethernet device. + * + * @param port_id + * The port identifier of the Ethernet device. + * @param ids + * A pointer to an ids array passed by application. This tells wich + * statistics values function should retrieve. This parameter + * can be set to NULL if n is 0. In this case function will retrieve + * all avalible statistics. + * @param values + * A pointer to a table to be filled with device statistics values. + * @param n + * The size of the ids array (number of elements). + * @return + * - A positive value lower or equal to n: success. The return value + * is the number of entries filled in the stats table. + * - A positive value higher than n: error, the given statistics table + * is too small. The return value corresponds to the size that should + * be given to succeed. The entries in the table are not valid and + * shall not be used by the caller. + * - A negative value on error (invalid port id). + */ +int rte_eth_xstats_get_by_id(uint8_t port_id, const uint64_t *ids, + uint64_t *values, unsigned int n); + /** * Reset extended statistics of an Ethernet device. * diff --git a/lib/librte_ether/rte_ether_version.map b/lib/librte_ether/rte_ether_version.map index 238c2a15ef..1abd717368 100644 --- a/lib/librte_ether/rte_ether_version.map +++ b/lib/librte_ether/rte_ether_version.map @@ -151,5 +151,7 @@ DPDK_17.05 { rte_eth_dev_attach_secondary; rte_eth_find_next; + rte_eth_xstats_get_by_id; + rte_eth_xstats_get_names_by_id; } DPDK_17.02; -- 2.20.1