| Message ID | 20240710203031.188081-1-anthony.l.nguyen@intel.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | idpf: XDP chapter I: convert Rx to libeth | expand |
Hello: This series was applied to netdev/net-next.git (main) by Tony Nguyen <anthony.l.nguyen@intel.com>: On Wed, 10 Jul 2024 13:30:16 -0700 you wrote: > Alexander Lobakin says: > > XDP for idpf is currently 5 chapters: > * convert Rx to libeth (this); > * convert Tx and stats to libeth; > * generic XDP and XSk code changes, libeth_xdp; > * actual XDP for idpf via libeth_xdp; > * XSk for idpf (^). > > [...] Here is the summary with links: - [net-next,01/14] cache: add __cacheline_group_{begin, end}_aligned() (+ couple more) https://git.kernel.org/netdev/net-next/c/2cb13dec8c5e - [net-next,02/14] page_pool: use __cacheline_group_{begin, end}_aligned() https://git.kernel.org/netdev/net-next/c/39daa09d34ad - [net-next,03/14] libeth: add cacheline / struct layout assertion helpers https://git.kernel.org/netdev/net-next/c/62c884256ea1 - [net-next,04/14] idpf: stop using macros for accessing queue descriptors https://git.kernel.org/netdev/net-next/c/66c27e3b19d5 - [net-next,05/14] idpf: split &idpf_queue into 4 strictly-typed queue structures https://git.kernel.org/netdev/net-next/c/e4891e4687c8 - [net-next,06/14] idpf: avoid bloating &idpf_q_vector with big %NR_CPUS https://git.kernel.org/netdev/net-next/c/bf9bf7042a38 - [net-next,07/14] idpf: strictly assert cachelines of queue and queue vector structures https://git.kernel.org/netdev/net-next/c/5a816aae2d46 - [net-next,08/14] idpf: merge singleq and splitq &net_device_ops https://git.kernel.org/netdev/net-next/c/14f662b43bf8 - [net-next,09/14] idpf: compile singleq code only under default-n CONFIG_IDPF_SINGLEQ https://git.kernel.org/netdev/net-next/c/f771314d6b75 - [net-next,10/14] idpf: reuse libeth's definitions of parsed ptype structures https://git.kernel.org/netdev/net-next/c/1b1b26208515 - [net-next,11/14] idpf: remove legacy Page Pool Ethtool stats https://git.kernel.org/netdev/net-next/c/4309363f1959 - [net-next,12/14] libeth: support different types of buffers for Rx https://git.kernel.org/netdev/net-next/c/5aaac1aece4e - [net-next,13/14] idpf: convert header split mode to libeth + napi_build_skb() https://git.kernel.org/netdev/net-next/c/90912f9f4f2d - [net-next,14/14] idpf: use libeth Rx buffer management for payload buffer https://git.kernel.org/netdev/net-next/c/74d1412ac8f3 You are awesome, thank you!
Alexander Lobakin says: XDP for idpf is currently 5 chapters: * convert Rx to libeth (this); * convert Tx and stats to libeth; * generic XDP and XSk code changes, libeth_xdp; * actual XDP for idpf via libeth_xdp; * XSk for idpf (^). Part I does the following: * splits &idpf_queue into 4 (RQ, SQ, FQ, CQ) and puts them on a diet; * ensures optimal cacheline placement, strictly asserts CL sizes; * moves currently unused/dead singleq mode out of line; * reuses libeth's Rx ptype definitions and helpers; * uses libeth's Rx buffer management for both header and payload; * eliminates memcpy()s and coherent DMA uses on hotpath, uses napi_build_skb() instead of in-place short skb allocation. Most idpf patches, except for the queue split, removes more lines than adds. Expect far better memory utilization and +5-8% on Rx depending on the case (+17% on skb XDP_DROP :>). --- IWL history: From v1[0]: * *: pick Reviewed-bys from Jake; * 01: new, add generic __cacheline_group_{begin,end}_aligned() and a couple more cache macros; * 02: new, make use of new macros from 01; * 03: use macros from 01 (no more struct_group()), leave only aggressive assertions here; * 07: adjust to the changes made in 01 and 03; fix typos in the kdocs; * 13: fix typos in the commit message (Jakub); * 14: fix possible unhandled null skb (Simon, static checker). From RFC[1]: * *: add kdocs where needed and fix the existing ones to build cleanly; fix minor checkpatch and codespell warnings; add RBs from Przemek; * 01: fix kdoc script to understand new libeth_cacheline_group() macro; add an additional assert for queue struct alignment; * 02: pick RB from Mina; * 06: make idpf_chk_linearize() static as it's now used only in one file; * 07: rephrase the commitmsg: HW supports it, but never wants; * 08: fix crashes on some configurations (Mina); * 11: constify header buffer pointer in idpf_rx_hsplit_wa(). Testing hints: basic Rx regression tests (+ perf and memory usage before/after if needed). [0] https://lore.kernel.org/netdev/20240528134846.148890-1-aleksander.lobakin@intel.com [1] https://lore.kernel.org/netdev/20240510152620.2227312-1-aleksander.lobakin@intel.com The following are changes since commit ce2f84ebcd857556f5ae9fc2d4ac720df1eb49d1: Merge branch 'aquantia-phy-aqr115c' into main and are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue 200GbE Alexander Lobakin (14): cache: add __cacheline_group_{begin, end}_aligned() (+ couple more) page_pool: use __cacheline_group_{begin, end}_aligned() libeth: add cacheline / struct layout assertion helpers idpf: stop using macros for accessing queue descriptors idpf: split &idpf_queue into 4 strictly-typed queue structures idpf: avoid bloating &idpf_q_vector with big %NR_CPUS idpf: strictly assert cachelines of queue and queue vector structures idpf: merge singleq and splitq &net_device_ops idpf: compile singleq code only under default-n CONFIG_IDPF_SINGLEQ idpf: reuse libeth's definitions of parsed ptype structures idpf: remove legacy Page Pool Ethtool stats libeth: support different types of buffers for Rx idpf: convert header split mode to libeth + napi_build_skb() idpf: use libeth Rx buffer management for payload buffer drivers/net/ethernet/intel/Kconfig | 13 +- drivers/net/ethernet/intel/idpf/Kconfig | 26 + drivers/net/ethernet/intel/idpf/Makefile | 3 +- drivers/net/ethernet/intel/idpf/idpf.h | 11 +- .../net/ethernet/intel/idpf/idpf_ethtool.c | 152 +- .../net/ethernet/intel/idpf/idpf_lan_txrx.h | 2 + drivers/net/ethernet/intel/idpf/idpf_lib.c | 88 +- drivers/net/ethernet/intel/idpf/idpf_main.c | 1 + .../ethernet/intel/idpf/idpf_singleq_txrx.c | 306 ++-- drivers/net/ethernet/intel/idpf/idpf_txrx.c | 1412 +++++++++-------- drivers/net/ethernet/intel/idpf/idpf_txrx.h | 734 +++++---- .../net/ethernet/intel/idpf/idpf_virtchnl.c | 178 ++- drivers/net/ethernet/intel/libeth/rx.c | 132 +- include/linux/cache.h | 59 + include/net/libeth/cache.h | 66 + include/net/libeth/rx.h | 19 + include/net/page_pool/types.h | 22 +- net/core/page_pool.c | 3 +- 18 files changed, 1824 insertions(+), 1403 deletions(-) create mode 100644 drivers/net/ethernet/intel/idpf/Kconfig create mode 100644 include/net/libeth/cache.h