From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 41CB9A04C0; Fri, 25 Sep 2020 10:47:27 +0200 (CEST) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id 08BE81D55A; Fri, 25 Sep 2020 10:47:27 +0200 (CEST) Received: from huawei.com (szxga06-in.huawei.com [45.249.212.32]) by dpdk.org (Postfix) with ESMTP id 894291D17B for ; Fri, 25 Sep 2020 10:47:24 +0200 (CEST) Received: from DGGEMS413-HUB.china.huawei.com (unknown [172.30.72.59]) by Forcepoint Email with ESMTP id 7BBF2AE82809C22A51F5; Fri, 25 Sep 2020 16:47:23 +0800 (CST) Received: from [10.67.103.128] (10.67.103.128) by DGGEMS413-HUB.china.huawei.com (10.3.19.213) with Microsoft SMTP Server id 14.3.487.0; Fri, 25 Sep 2020 16:47:15 +0800 To: Andrew Rybchenko , CC: , , , References: <1599534347-20430-1-git-send-email-humin29@huawei.com> <1600952745-55993-1-git-send-email-humin29@huawei.com> <1600952745-55993-2-git-send-email-humin29@huawei.com> From: "Min Hu (Connor)" Message-ID: Date: Fri, 25 Sep 2020 16:47:14 +0800 User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.3.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset="utf-8"; format=flowed Content-Transfer-Encoding: 8bit X-Originating-IP: [10.67.103.128] X-CFilter-Loop: Reflected Subject: Re: [dpdk-dev] [PATCH V12 1/4] ethdev: introduce FEC API X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" HI，Andrew， I fix it in V13 according to your advice. Thanks for your patient review. Please check it out again. By the way, there is always a coding check warning, like this: WARNING:PREFER_FALLTHROUGH: Prefer 'fallthrough;' over fallthrough comment #264: FILE: drivers/net/hns3/hns3_ethdev.c:5601: + /* fallthrough */ total: 0 errors, 1 warnings, 0 checks, 439 lines checked. I have tried some ways, but it does not help. Could you giver some advice? thanks. 在 2020/9/24 22:46, Andrew Rybchenko 写道: > On 9/24/20 4:05 PM, Min Hu (Connor) wrote: >> This patch adds Forward error correction(FEC) support for ethdev. >> Introduce APIs which support query and config FEC information in >> hardware. > > Almost good now. See my notes below. > Many thanks for hard work and patience. > >> >> Signed-off-by: Min Hu (Connor) >> Reviewed-by: Wei Hu (Xavier) >> Reviewed-by: Chengwen Feng >> Reviewed-by: Chengchang Tang >> Reviewed-by: Ajit Khaparde >> Acked-by: Konstantin Ananyev > > I think that tags for Ajit and Konstantin should be > dropped, since API changes significantly after their review. > >> --- >> v10->v11: >> allow to report capabilities per link speed. >> specify what should be reported if link is down >> when get FEC. >> change mode to capa bitmask. >> >> --- >> v9->v10: >> add macro RTE_ETH_FEC_MODE_CAPA_MASK(x) to indicate >> different FEC mode capa. >> >> --- >> v8->v9: >> added reviewed-by and acked-by. >> >> --- >> v7->v8: >> put AUTO just after NOFEC in rte_fec_mode definition. >> >> --- >> v6->v7: >> deleted RTE_ETH_FEC_NUM to prevent ABI breakage. >> add new macro to indicate translation from fec mode >> to capa. >> >> --- >> v5->v6: >> modified release notes. >> deleted check duplicated for FEC API >> fixed code styles according to DPDK coding style. >> added _eth prefix. >> >> --- >> v4->v5: >> Modifies FEC capa definitions using macros. >> Add RTE_ prefix for public FEC mode enum. >> add release notes about FEC for dpdk20_11. >> >> --- >> v2->v3: >> add function return value "-ENOTSUP" for API. >> >> --- >> lib/librte_ethdev/rte_ethdev.c | 37 +++++++++++++ >> lib/librte_ethdev/rte_ethdev.h | 91 ++++++++++++++++++++++++++++++++ >> lib/librte_ethdev/rte_ethdev_driver.h | 82 ++++++++++++++++++++++++++++ >> lib/librte_ethdev/rte_ethdev_version.map | 3 ++ >> 4 files changed, 213 insertions(+) >> >> diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c >> index dfe5c1b..b614bfc 100644 >> --- a/lib/librte_ethdev/rte_ethdev.c >> +++ b/lib/librte_ethdev/rte_ethdev.c >> @@ -3679,6 +3679,43 @@ rte_eth_led_off(uint16_t port_id) >> return eth_err(port_id, (*dev->dev_ops->dev_led_off)(dev)); >> } >> >> +int >> +rte_eth_fec_get_capability(uint16_t port_id, uint32_t *num, >> + struct rte_eth_fec_capa *speed_fec_capa) >> +{ >> + struct rte_eth_dev *dev; >> + >> + if (num == NULL || speed_fec_capa == NULL) >> + return -EINVAL; > > I think it is OK to have speed_fec_cap==NULL if *num is 0. > I.e. a request to get number of required array entries. > >> + >> + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); >> + dev = &rte_eth_devices[port_id]; >> + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->fec_get_capability, -ENOTSUP); >> + return eth_err(port_id, (*dev->dev_ops->fec_get_capability)(dev, num, >> + speed_fec_capa)); >> +} >> + >> +int >> +rte_eth_fec_get(uint16_t port_id, uint32_t *mode) >> +{ >> + struct rte_eth_dev *dev; > > I think it would be good to check that mode is not NULL here. > >> + >> + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); >> + dev = &rte_eth_devices[port_id]; >> + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->fec_get, -ENOTSUP); >> + return eth_err(port_id, (*dev->dev_ops->fec_get)(dev, mode)); >> +} >> + >> +int >> +rte_eth_fec_set(uint16_t port_id, uint32_t mode) >> +{ >> + struct rte_eth_dev *dev; >> + >> + dev = &rte_eth_devices[port_id]; >> + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->fec_set, -ENOTSUP); >> + return eth_err(port_id, (*dev->dev_ops->fec_set)(dev, mode)); >> +} >> + >> /* >> * Returns index into MAC address array of addr. Use 00:00:00:00:00:00 to find >> * an empty spot. >> diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h >> index 645a186..104181d 100644 >> --- a/lib/librte_ethdev/rte_ethdev.h >> +++ b/lib/librte_ethdev/rte_ethdev.h >> @@ -1544,6 +1544,29 @@ struct rte_eth_dcb_info { >> struct rte_eth_dcb_tc_queue_mapping tc_queue; >> }; >> >> +/** >> + * This enum indicates the possible (forward error correction)FEC modes > > (forward error correction)FEC -> Forward Error Correction (FEC) > >> + * of an ethdev port. >> + */ >> +enum rte_eth_fec_mode { >> + RTE_ETH_FEC_NOFEC = 0, /**< FEC is off */ >> + RTE_ETH_FEC_AUTO, /**< FEC autonegotiation modes */ >> + RTE_ETH_FEC_BASER, /**< FEC using common algorithm */ >> + RTE_ETH_FEC_RS, /**< FEC using RS algorithm */ >> +}; >> + >> +/* Translate from FEC mode to FEC capa */ >> +#define RTE_ETH_FEC_MODE_TO_CAPA(x) (1U << (x)) >> + >> +/* This macro indicates FEC capa mask*/ > > Add missing space before */ > >> +#define RTE_ETH_FEC_MODE_CAPA_MASK(x) (1U << (RTE_ETH_FEC_ ## x)) >> + >> +/* A structure used to get capabilities per link speed */ >> +struct rte_eth_fec_capa { >> + uint32_t speed; /**< Link speed (see ETH_SPEED_NUM_*) */ >> + uint32_t capa; /**< FEC capabilities bitmask (see RTE_FEC_CAPA_*) */ >> +}; >> + >> #define RTE_ETH_ALL RTE_MAX_ETHPORTS >> >> /* Macros to check for valid port */ >> @@ -3397,6 +3420,74 @@ int rte_eth_led_on(uint16_t port_id); >> int rte_eth_led_off(uint16_t port_id); >> >> /** >> + * @warning >> + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice >> + * >> + * Get Forward Error Correction(FEC) capability. >> + * >> + * @param port_id >> + * The port identifier of the Ethernet device. >> + * @param num >> + * the num is in/out with a number of elements in an array. > > Please, see below my notes on callback description. > >> + * @param speed_fec_capa >> + * speed_fec_capa is out only with per-speed capabilities. >> + * >> + * @return >> + * - (0) if successful. >> + * - (-ENOTSUP) if underlying hardware OR driver doesn't support. >> + * that operation. >> + * - (-EIO) if device is removed. >> + * - (-ENODEV) if *port_id* invalid. >> + * - (-EINVAL) if *num* or *speed_fec_capa* invalid >> + */ >> +__rte_experimental >> +int rte_eth_fec_get_capability(uint16_t port_id, uint32_t *num, >> + struct rte_eth_fec_capa *speed_fec_capa); >> + >> +/** >> + * @warning >> + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice >> + * >> + * Get current Forward Error Correction(FEC) mode. >> + * If link is down and AUTO is enabled, AUTO is returned, otherwise, >> + * configured FEC mode is returned. >> + * If link is up, current FEC mode is returned. >> + * >> + * @param port_id >> + * The port identifier of the Ethernet device. >> + * @param mode >> + * returns the FEC mode from the device. >> + * @return >> + * - (0) if successful. >> + * - (-ENOTSUP) if underlying hardware OR driver doesn't support. >> + * that operation. >> + * - (-EIO) if device is removed. >> + * - (-ENODEV) if *port_id* invalid. >> + */ >> +__rte_experimental >> +int rte_eth_fec_get(uint16_t port_id, uint32_t *mode); >> + >> +/** >> + * @warning >> + * @b EXPERIMENTAL: this API may change, or be removed, without prior notice >> + * >> + * Set Forward Error Correction(FEC) mode. >> + * >> + * @param port_id >> + * The port identifier of the Ethernet device. >> + * @param mode >> + * the FEC mode. > > See below. > >> + * @return >> + * - (0) if successful. >> + * - (-EINVAL) if the FEC mode is not valid. >> + * - (-ENOTSUP) if underlying hardware OR driver doesn't support. >> + * - (-EIO) if device is removed. >> + * - (-ENODEV) if *port_id* invalid. >> + */ >> +__rte_experimental >> +int rte_eth_fec_set(uint16_t port_id, uint32_t mode); >> + >> +/** >> * Get current status of the Ethernet link flow control for Ethernet device >> * >> * @param port_id >> diff --git a/lib/librte_ethdev/rte_ethdev_driver.h b/lib/librte_ethdev/rte_ethdev_driver.h >> index 23cc1e0..625b8c5 100644 >> --- a/lib/librte_ethdev/rte_ethdev_driver.h >> +++ b/lib/librte_ethdev/rte_ethdev_driver.h >> @@ -575,6 +575,81 @@ typedef int (*eth_tx_hairpin_queue_setup_t) >> const struct rte_eth_hairpin_conf *hairpin_conf); >> >> /** >> + * @internal >> + * Get Forward Error Correction(FEC) capability. >> + * >> + * @param dev >> + * ethdev handle of port. >> + * @param num >> + * the num is in/out with a number of elements in an array. > > I'm sorry, I should do it before my previous suggestion, but: > Looking at rte_eth_xstats_get_names() and trying to be > consistent I'd like to suggest to put the argument after > speed_fec_capa and make it input only with a number of > array elements. > Positive return values should be used to provide number of > filled in array elements. If the returned value is greater > than 'num', just provided elements are filled in, but > it is indication as well, that num is too small. > >> + * @param speed_fec_capa >> + * speed_fec_capa is out only with per-speed capabilities. >> + * >> + * @return >> + * Negative errno value on error, 0 on success. >> + * >> + * @retval 0 >> + * Success, get FEC success. > > See above. > >> + * @retval -ENOTSUP >> + * operation is not supported. > > Should start from upper case letter > >> + * @retval -EIO >> + * device is removed. > > Should start from upper case letter > >> + * @retval -ENODEV >> + * Device is gone. > > What's the difference between "device is remove" and > "Device is gone" > >> + * @retval -EINVAL >> + * *num* or *speed_fec_capa* invalid. >> + */ >> +typedef int (*eth_fec_get_capability_t)(struct rte_eth_dev *dev, uint32_t *num, >> + struct rte_eth_fec_capa *speed_fec_capa); >> + >> +/** >> + * @internal >> + * Get Forward Error Correction(FEC) mode. >> + * >> + * @param dev >> + * ethdev handle of port. >> + * @param mode >> + * returns the FEC mode from the device. >> + * >> + * @return >> + * Negative errno value on error, 0 on success. >> + * >> + * @retval 0 >> + * Success, get FEC success. >> + * @retval -ENOTSUP >> + * operation is not supported. > > Should start from upper case letter > >> + * @retval -EIO >> + * device is removed. > > Should start from upper case letter > >> + * @retval -ENODEV >> + * Device is gone. >> + */ >> +typedef int (*eth_fec_get_t)(struct rte_eth_dev *dev, >> + uint32_t *mode); >> + >> +/** >> + * @internal >> + * Set Forward Error Correction(FEC) mode. > > Remove extra spaces before "Set" > >> + * >> + * @param dev >> + * ethdev handle of port. >> + * @param mode >> + * the FEC mode. > > The description is insufficient and misleading. > It should be fec_capa and described as: > Bitmask of allowed FEC modes. If must be only one > if AUTO is disabled. If AUTO is enabled, other > bits specify FEC modes which may be negotiated. > >> + * >> + * @return >> + * Negative errno value on error, 0 on success. >> + * >> + * @retval 0 >> + * Success, set FEC success. >> + * @retval -ENOTSUP >> + * operation is not supported. > > Should start from upper case letter > > What about -EINVAL in the case of unsupported FEC mode > requested? It is listed above in API function but > missing here. > >> + * @retval -EIO >> + * device is removed. > > Should start from upper case letter > >> + * @retval -ENODEV >> + * Device is gone. > > What's the difference between "device is remove" and > "Device is gone" > >> + */ >> +typedef int (*eth_fec_set_t)(struct rte_eth_dev *dev, uint32_t mode); >> + >> +/** >> * @internal A structure containing the functions exported by an Ethernet driver. >> */ >> struct eth_dev_ops { >> @@ -713,6 +788,13 @@ struct eth_dev_ops { >> /**< Set up device RX hairpin queue. */ >> eth_tx_hairpin_queue_setup_t tx_hairpin_queue_setup; >> /**< Set up device TX hairpin queue. */ >> + >> + eth_fec_get_capability_t fec_get_capability; >> + /**< Get Forward Error Correction(FEC) capability; */ > > It should be a dot (.) at the end, not semicolon (;). > >> + eth_fec_get_t fec_get; >> + /**< Get Forward Error Correction(FEC) mode; */ > > same > >> + eth_fec_set_t fec_set; >> + /**< Set Forward Error Correction(FEC) mode; */ > > same > >> }; >> >> /** >> diff --git a/lib/librte_ethdev/rte_ethdev_version.map b/lib/librte_ethdev/rte_ethdev_version.map >> index c95ef51..b9ace3a 100644 >> --- a/lib/librte_ethdev/rte_ethdev_version.map >> +++ b/lib/librte_ethdev/rte_ethdev_version.map >> @@ -229,6 +229,9 @@ EXPERIMENTAL { >> # added in 20.11 >> rte_eth_link_speed_to_str; >> rte_eth_link_to_str; >> + rte_eth_fec_get_capability; >> + rte_eth_fec_get; >> + rte_eth_fec_set; >> }; >> >> INTERNAL { >> > > . >