> -----邮件原件----- > 发件人: dev <dev-boun...@dpdk.org> 代表 Ferruh Yigit > 发送时间: Thursday, October 14, 2021 4:16 AM > 收件人: Konstantin Ananyev <konstantin.anan...@intel.com>; > dev@dpdk.org; jer...@marvell.com; Ajit Khaparde > (ajit.khapa...@broadcom.com) <ajit.khapa...@broadcom.com>; Raslan > Darawsheh <rasl...@nvidia.com>; Andrew Rybchenko > <andrew.rybche...@oktetlabs.ru>; Qi Zhang <qi.z.zh...@intel.com>; > Honnappa Nagarahalli <honnappa.nagaraha...@arm.com> > 抄送: xiaoyun...@intel.com; ano...@marvell.com; jer...@marvell.com; > ndabilpu...@marvell.com; adwiv...@marvell.com; > shepard.sie...@atomicrules.com; ed.cz...@atomicrules.com; > john.mil...@atomicrules.com; irussk...@marvell.com; Ajit Khaparde > (ajit.khapa...@broadcom.com) <ajit.khapa...@broadcom.com>; > somnath.ko...@broadcom.com; rahul.lakkire...@chelsio.com; > hemant.agra...@nxp.com; sachin.sax...@oss.nxp.com; > haiyue.w...@intel.com; johnd...@cisco.com; hyon...@cisco.com; > qi.z.zh...@intel.com; xiao.w.w...@intel.com; humi...@huawei.com; > yisen.zhu...@huawei.com; ouli...@huawei.com; beilei.x...@intel.com; > jingjing...@intel.com; qiming.y...@intel.com; ma...@nvidia.com; > viachesl...@nvidia.com; sthem...@microsoft.com; lon...@microsoft.com; > heinrich.k...@corigine.com; kirankum...@marvell.com; > andrew.rybche...@oktetlabs.ru; mcze...@marvell.com; > jiawe...@trustnetic.com; jianw...@trustnetic.com; > maxime.coque...@redhat.com; chenbo....@intel.com; > tho...@monjalon.net; m...@ashroe.eu; jay.jayatheert...@intel.com > 主题: Re: [dpdk-dev] [PATCH v6 0/6] hide eth dev related structures > > On 10/13/2021 2:36 PM, Konstantin Ananyev wrote: > > v6 changes: > > - Update comments (Andrew) > > - Move callback related variables under corresponding ifdefs (Andrew) > > - Few nits in rte_eth_macaddrs_get (Andrew) > > - Rebased on top of next-net tree > > > > v5 changes: > > - Fix spelling (Thomas/David) > > - Rename internal helper functions (David) > > - Reorder patches and update commit messages (Thomas) > > - Update comments (Thomas) > > - Changed layout in rte_eth_fp_ops, to group functions and > > related data based on their functionality: > > first 64B line for Rx, second one for Tx. > > Didn't observe any real performance difference comparing to > > original layout. Though decided to keep a new one, as it seems > > a bit more plausible. > > > > v4 changes: > > - Fix secondary process attach (Pavan) > > - Fix build failure (Ferruh) > > - Update lib/ethdev/verion.map (Ferruh) > > Note that moving newly added symbols from EXPERIMENTAL to DPDK_22 > > section makes checkpatch.sh to complain. > > > > v3 changes: > > - Changes in public struct naming (Jerin/Haiyue) > > - Split patches > > - Update docs > > - Shamelessly included Andrew's patch: > > > https://patches.dpdk.org/project/dpdk/patch/20210928154856.1015020-1- > andrew.rybche...@oktetlabs.ru/ > > into these series. > > I have to do similar thing here, so decided to avoid duplicated effort. > > > > The aim of these patch series is to make rte_ethdev core data > > structures (rte_eth_dev, rte_eth_dev_data, rte_eth_rxtx_callback, > > etc.) internal to DPDK and not visible to the user. > > That should allow future possible changes to core ethdev related > > structures to be transparent to the user and help to improve ABI/API > stability. > > Note that current ethdev API is preserved, but it is a formal ABI break. > > > > The work is based on previous discussions at: > > https://www.mail-archive.com/dev@dpdk.org/msg211405.html > > https://www.mail-archive.com/dev@dpdk.org/msg216685.html > > and consists of the following main points: > > 1. Copy public 'fast' function pointers (rx_pkt_burst(), etc.) and > > related data pointer from rte_eth_dev into a separate flat array. > > We keep it public to still be able to use inline functions for these > > 'fast' calls (like rte_eth_rx_burst(), etc.) to avoid/minimize slowdown. > > Note that apart from function pointers itself, each element of this > > flat array also contains two opaque pointers for each ethdev: > > 1) a pointer to an array of internal queue data pointers > > 2) points to array of queue callback data pointers. > > Note that exposing this extra information allows us to avoid extra > > changes inside PMD level, plus should help to avoid possible > > performance degradation. > > 2. Change implementation of 'fast' inline ethdev functions > > (rte_eth_rx_burst(), etc.) to use new public flat array. > > While it is an ABI breakage, this change is intended to be transparent > > for both users (no changes in user app is required) and PMD developers > > (no changes in PMD is required). > > One extra note - with new implementation RX/TX callback invocation > > will cost one extra function call with this changes. That might cause > > some slowdown for code-path with RX/TX callbacks heavily involved. > > Hope such trade-off is acceptable for the community. > > 3. Move rte_eth_dev, rte_eth_dev_data, rte_eth_rxtx_callback and related > > things into internal header: <ethdev_driver.h>. > > > > That approach was selected to: > > - Avoid(/minimize) possible performance losses. > > - Minimize required changes inside PMDs. > > > > Performance testing results (ICX 2.0GHz, E810 (ice)): > > - testpmd macswap fwd mode, plus > > a) no RX/TX callbacks: > > no actual slowdown observed > > b) bpf-load rx 0 0 JM ./dpdk.org/examples/bpf/t3.o: > > ~2% slowdown > > - l3fwd: no actual slowdown observed > > > > Would like to thank everyone who already reviewed and tested previous > > versions of these series. All other interested parties please don't be > > shy and provide your feedback. > > > > Konstantin Ananyev (6): > > ethdev: allocate max space for internal queue array > > ethdev: change input parameters for rx_queue_count > > ethdev: copy fast-path API into separate structure > > ethdev: make fast-path functions to use new flat array > > ethdev: add API to retrieve multiple ethernet addresses > > ethdev: hide eth dev related structures > > > > For series, > Reviewed-by: Ferruh Yigit <ferruh.yi...@intel.com> > > No performance regression detected on my testing. > > I am merging the series to next-net now which helps testing, but before > merging to main repo it will be good to get more ack and test results (I can > squash new tags later). > > @Jerin, @Ajit, @Raslan, @Andrew, @Qi, @Honnappa, Can you please test > this set for any possible regression? > > Series applied to dpdk-next-net/main, thanks. >
For series, there is no performance regression in n1sdp/thunderx2 with i40e and mlx5 40G NIC for l3fwd and testpmd. Tested-by: Feifei Wang <feifei.wa...@arm.com>
