Date: Tue, 10 Jan 2017 03:23:22 +0000 (UTC) From: Sean Bruno <sbruno@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r311849 - in head: . sys/amd64/conf sys/arm64/conf sys/conf sys/dev/e1000 sys/i386/conf sys/mips/conf sys/modules sys/modules/em sys/modules/igb sys/powerpc/conf Message-ID: <201701100323.v0A3NMAI045868@repo.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: sbruno Date: Tue Jan 10 03:23:22 2017 New Revision: 311849 URL: https://svnweb.freebsd.org/changeset/base/311849 Log: Migrate e1000 to the IFLIB framework: - em(4) igb(4) and lem(4) - deprecate the igb device from kernel configurations - create a symbolic link in /boot/kernel from if_em.ko to if_igb.ko Devices tested: - 82574L - I218-LM - 82546GB - 82579LM - I350 - I217 Please report problems to freebsd-net@freebsd.org Partial review from jhb and suggestions on how to *not* brick folks who originally would have lost their igbX device. Submitted by: mmacy@nextbsd.org MFC after: 2 weeks Relnotes: yes Sponsored by: Limelight Networks and Dell EMC Isilon Differential Revision: https://reviews.freebsd.org/D8299 Added: head/sys/dev/e1000/em_txrx.c (contents, props changed) head/sys/dev/e1000/igb_txrx.c (contents, props changed) Deleted: head/sys/dev/e1000/if_igb.c head/sys/dev/e1000/if_igb.h head/sys/dev/e1000/if_lem.c head/sys/dev/e1000/if_lem.h head/sys/modules/igb/ Modified: head/UPDATING head/sys/amd64/conf/GENERIC head/sys/arm64/conf/GENERIC head/sys/conf/NOTES head/sys/conf/files head/sys/conf/makeLINT.mk head/sys/dev/e1000/if_em.c head/sys/dev/e1000/if_em.h head/sys/i386/conf/GENERIC head/sys/mips/conf/OCTEON1 head/sys/modules/Makefile head/sys/modules/em/Makefile head/sys/powerpc/conf/GENERIC64 Modified: head/UPDATING ============================================================================== --- head/UPDATING Tue Jan 10 01:36:50 2017 (r311848) +++ head/UPDATING Tue Jan 10 03:23:22 2017 (r311849) @@ -51,6 +51,11 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 12 ****************************** SPECIAL WARNING: ****************************** +20170109: + The igb(4), em(4) and lem(4) ethernet drivers are now implemented via + IFLIB. If you have a custom kernel configuration that excludes em(4) + but you use igb(4), you need to re-add em(4) to your custom configuration. + 20161217: Clang, llvm, lldb, compiler-rt and libc++ have been upgraded to 3.9.1. Please see the 20141231 entry below for information about prerequisites Modified: head/sys/amd64/conf/GENERIC ============================================================================== --- head/sys/amd64/conf/GENERIC Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/amd64/conf/GENERIC Tue Jan 10 03:23:22 2017 (r311849) @@ -230,7 +230,6 @@ device puc # Multi I/O cards and mult device bxe # Broadcom NetXtreme II BCM5771X/BCM578XX 10GbE device de # DEC/Intel DC21x4x (``Tulip'') device em # Intel PRO/1000 Gigabit Ethernet Family -device igb # Intel PRO/1000 PCIE Server Gigabit Family device ix # Intel PRO/10GbE PCIE PF Ethernet device ixv # Intel PRO/10GbE PCIE VF Ethernet device ixl # Intel XL710 40Gbe PCIE Ethernet Modified: head/sys/arm64/conf/GENERIC ============================================================================== --- head/sys/arm64/conf/GENERIC Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/arm64/conf/GENERIC Tue Jan 10 03:23:22 2017 (r311849) @@ -120,7 +120,6 @@ device mii device miibus # MII bus support device awg # Allwinner EMAC Gigabit Ethernet device em # Intel PRO/1000 Gigabit Ethernet Family -device igb # Intel PRO/1000 PCIE Server Gigabit Family device ix # Intel 10Gb Ethernet Family device msk # Marvell/SysKonnect Yukon II Gigabit Ethernet device smc # SMSC LAN91C111 Modified: head/sys/conf/NOTES ============================================================================== --- head/sys/conf/NOTES Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/conf/NOTES Tue Jan 10 03:23:22 2017 (r311849) @@ -1972,7 +1972,6 @@ device xmphy # XaQti XMAC II # KNE110TX. # de: Digital Equipment DC21040 # em: Intel Pro/1000 Gigabit Ethernet 82542, 82543, 82544 based adapters. -# igb: Intel Pro/1000 PCI Express Gigabit Ethernet: 82575 and later adapters. # ep: 3Com 3C509, 3C529, 3C556, 3C562D, 3C563D, 3C572, 3C574X, 3C579, 3C589 # and PC Card devices using these chipsets. # ex: Intel EtherExpress Pro/10 and other i82595-based adapters, @@ -2145,7 +2144,6 @@ device cxgbe # Chelsio T4-T6 1/10/25/4 device cxgbev # Chelsio T4-T6 Virtual Functions device de # DEC/Intel DC21x4x (``Tulip'') device em # Intel Pro/1000 Gigabit Ethernet -device igb # Intel Pro/1000 PCIE Gigabit Ethernet device ixgb # Intel Pro/10Gbe PCI-X Ethernet device ix # Intel Pro/10Gbe PCIE Ethernet device ixv # Intel Pro/10Gbe PCIE Ethernet VF Modified: head/sys/conf/files ============================================================================== --- head/sys/conf/files Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/conf/files Tue Jan 10 03:23:22 2017 (r311849) @@ -1572,43 +1572,43 @@ dev/eisa/eisa_if.m standard dev/eisa/eisaconf.c optional eisa dev/e1000/if_em.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/if_lem.c optional em \ +dev/e1000/em_txrx.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/if_igb.c optional igb \ +dev/e1000/igb_txrx.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_80003es2lan.c optional em | igb \ +dev/e1000/e1000_80003es2lan.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82540.c optional em | igb \ +dev/e1000/e1000_82540.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82541.c optional em | igb \ +dev/e1000/e1000_82541.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82542.c optional em | igb \ +dev/e1000/e1000_82542.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82543.c optional em | igb \ +dev/e1000/e1000_82543.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82571.c optional em | igb \ +dev/e1000/e1000_82571.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_82575.c optional em | igb \ +dev/e1000/e1000_82575.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_ich8lan.c optional em | igb \ +dev/e1000/e1000_ich8lan.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_i210.c optional em | igb \ +dev/e1000/e1000_i210.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_api.c optional em | igb \ +dev/e1000/e1000_api.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_mac.c optional em | igb \ +dev/e1000/e1000_mac.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_manage.c optional em | igb \ +dev/e1000/e1000_manage.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_nvm.c optional em | igb \ +dev/e1000/e1000_nvm.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_phy.c optional em | igb \ +dev/e1000/e1000_phy.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_vf.c optional em | igb \ +dev/e1000/e1000_vf.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_mbx.c optional em | igb \ +dev/e1000/e1000_mbx.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" -dev/e1000/e1000_osdep.c optional em | igb \ +dev/e1000/e1000_osdep.c optional em \ compile-with "${NORMAL_C} -I$S/dev/e1000" dev/et/if_et.c optional et dev/en/if_en_pci.c optional en pci Modified: head/sys/conf/makeLINT.mk ============================================================================== --- head/sys/conf/makeLINT.mk Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/conf/makeLINT.mk Tue Jan 10 03:23:22 2017 (r311849) @@ -38,7 +38,6 @@ LINT: ${NOTES} ../../conf/makeLINT.sed echo "nodevice bxe" >> ${.TARGET}-NOIP echo "nodevice em" >> ${.TARGET}-NOIP echo "nodevice fxp" >> ${.TARGET}-NOIP - echo "nodevice igb" >> ${.TARGET}-NOIP echo "nodevice jme" >> ${.TARGET}-NOIP echo "nodevice msk" >> ${.TARGET}-NOIP echo "nodevice mxge" >> ${.TARGET}-NOIP Added: head/sys/dev/e1000/em_txrx.c ============================================================================== --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/sys/dev/e1000/em_txrx.c Tue Jan 10 03:23:22 2017 (r311849) @@ -0,0 +1,720 @@ +/* $FreeBSD$ */ +#include "if_em.h" + +#ifdef RSS +#include <net/rss_config.h> +#include <netinet/in_rss.h> +#endif + +#ifdef VERBOSE_DEBUG +#define DPRINTF device_printf +#else +#define DPRINTF(...) +#endif + +/********************************************************************* + * Local Function prototypes + *********************************************************************/ +static int em_tso_setup(struct adapter *adapter, if_pkt_info_t pi, u32 *txd_upper, u32 *txd_lower); +static int em_transmit_checksum_setup(struct adapter *adapter, if_pkt_info_t pi, u32 *txd_upper, u32 *txd_lower); +static int em_isc_txd_encap(void *arg, if_pkt_info_t pi); +static void em_isc_txd_flush(void *arg, uint16_t txqid, uint32_t pidx); +static int em_isc_txd_credits_update(void *arg, uint16_t txqid, uint32_t cidx_init, bool clear); +static void em_isc_rxd_refill(void *arg, uint16_t rxqid, uint8_t flid __unused, + uint32_t pidx, uint64_t *paddrs, caddr_t *vaddrs __unused, uint16_t count, uint16_t buflen __unused); +static void em_isc_rxd_flush(void *arg, uint16_t rxqid, uint8_t flid __unused, uint32_t pidx); +static int em_isc_rxd_available(void *arg, uint16_t rxqid, uint32_t idx, + int budget); +static int em_isc_rxd_pkt_get(void *arg, if_rxd_info_t ri); + +static void lem_isc_rxd_refill(void *arg, uint16_t rxqid, uint8_t flid __unused, + uint32_t pidx, uint64_t *paddrs, caddr_t *vaddrs __unused, uint16_t count, uint16_t buflen __unused); + +static int lem_isc_rxd_available(void *arg, uint16_t rxqid, uint32_t idx, + int budget); +static int lem_isc_rxd_pkt_get(void *arg, if_rxd_info_t ri); + +static void lem_receive_checksum(int status, int errors, if_rxd_info_t ri); +static void em_receive_checksum(uint32_t status, if_rxd_info_t ri); +extern int em_intr(void *arg); + +struct if_txrx em_txrx = { + em_isc_txd_encap, + em_isc_txd_flush, + em_isc_txd_credits_update, + em_isc_rxd_available, + em_isc_rxd_pkt_get, + em_isc_rxd_refill, + em_isc_rxd_flush, + em_intr +}; + +struct if_txrx lem_txrx = { + em_isc_txd_encap, + em_isc_txd_flush, + em_isc_txd_credits_update, + lem_isc_rxd_available, + lem_isc_rxd_pkt_get, + lem_isc_rxd_refill, + em_isc_rxd_flush, + em_intr +}; + +extern if_shared_ctx_t em_sctx; + +/********************************************************************** + * + * Setup work for hardware segmentation offload (TSO) on + * adapters using advanced tx descriptors + * + **********************************************************************/ +static int +em_tso_setup(struct adapter *adapter, if_pkt_info_t pi, u32 *txd_upper, u32 *txd_lower) +{ + if_softc_ctx_t scctx = adapter->shared; + struct em_tx_queue *que = &adapter->tx_queues[pi->ipi_qsidx]; + struct tx_ring *txr = &que->txr; + struct e1000_context_desc *TXD; + struct em_txbuffer *tx_buffer; + int cur, hdr_len; + + hdr_len = pi->ipi_ehdrlen + pi->ipi_ip_hlen + pi->ipi_tcp_hlen; + *txd_lower = (E1000_TXD_CMD_DEXT | /* Extended descr type */ + E1000_TXD_DTYP_D | /* Data descr type */ + E1000_TXD_CMD_TSE); /* Do TSE on this packet */ + + /* IP and/or TCP header checksum calculation and insertion. */ + *txd_upper = (E1000_TXD_POPTS_IXSM | E1000_TXD_POPTS_TXSM) << 8; + + cur = pi->ipi_pidx; + TXD = (struct e1000_context_desc *)&txr->tx_base[cur]; + tx_buffer = &txr->tx_buffers[cur]; + + /* + * Start offset for header checksum calculation. + * End offset for header checksum calculation. + * Offset of place put the checksum. + */ + TXD->lower_setup.ip_fields.ipcss = pi->ipi_ehdrlen; + TXD->lower_setup.ip_fields.ipcse = + htole16(pi->ipi_ehdrlen + pi->ipi_ip_hlen - 1); + TXD->lower_setup.ip_fields.ipcso = pi->ipi_ehdrlen + offsetof(struct ip, ip_sum); + + /* + * Start offset for payload checksum calculation. + * End offset for payload checksum calculation. + * Offset of place to put the checksum. + */ + TXD->upper_setup.tcp_fields.tucss = pi->ipi_ehdrlen + pi->ipi_ip_hlen; + TXD->upper_setup.tcp_fields.tucse = 0; + TXD->upper_setup.tcp_fields.tucso = + pi->ipi_ehdrlen + pi->ipi_ip_hlen + offsetof(struct tcphdr, th_sum); + + /* + * Payload size per packet w/o any headers. + * Length of all headers up to payload. + */ + TXD->tcp_seg_setup.fields.mss = htole16(pi->ipi_tso_segsz); + TXD->tcp_seg_setup.fields.hdr_len = hdr_len; + + TXD->cmd_and_length = htole32(adapter->txd_cmd | + E1000_TXD_CMD_DEXT | /* Extended descr */ + E1000_TXD_CMD_TSE | /* TSE context */ + E1000_TXD_CMD_IP | /* Do IP csum */ + E1000_TXD_CMD_TCP | /* Do TCP checksum */ + (pi->ipi_len - hdr_len)); /* Total len */ + tx_buffer->eop = -1; + txr->tx_tso = TRUE; + + if (++cur == scctx->isc_ntxd[0]) { + cur = 0; + } + DPRINTF(iflib_get_dev(adapter->ctx), "%s: pidx: %d cur: %d\n", __FUNCTION__, pi->ipi_pidx, cur); + return (cur); +} + +#define TSO_WORKAROUND 4 +#define DONT_FORCE_CTX 1 + + +/********************************************************************* + * The offload context is protocol specific (TCP/UDP) and thus + * only needs to be set when the protocol changes. The occasion + * of a context change can be a performance detriment, and + * might be better just disabled. The reason arises in the way + * in which the controller supports pipelined requests from the + * Tx data DMA. Up to four requests can be pipelined, and they may + * belong to the same packet or to multiple packets. However all + * requests for one packet are issued before a request is issued + * for a subsequent packet and if a request for the next packet + * requires a context change, that request will be stalled + * until the previous request completes. This means setting up + * a new context effectively disables pipelined Tx data DMA which + * in turn greatly slow down performance to send small sized + * frames. + **********************************************************************/ + +static int +em_transmit_checksum_setup(struct adapter *adapter, if_pkt_info_t pi, u32 *txd_upper, u32 *txd_lower) +{ + struct e1000_context_desc *TXD = NULL; + if_softc_ctx_t scctx = adapter->shared; + struct em_tx_queue *que = &adapter->tx_queues[pi->ipi_qsidx]; + struct tx_ring *txr = &que->txr; + struct em_txbuffer *tx_buffer; + int csum_flags = pi->ipi_csum_flags; + int cur, hdr_len; + u32 cmd; + + cur = pi->ipi_pidx; + hdr_len = pi->ipi_ehdrlen + pi->ipi_ip_hlen; + cmd = adapter->txd_cmd; + + /* + * The 82574L can only remember the *last* context used + * regardless of queue that it was use for. We cannot reuse + * contexts on this hardware platform and must generate a new + * context every time. 82574L hardware spec, section 7.2.6, + * second note. + */ + if (DONT_FORCE_CTX && + adapter->tx_num_queues == 1 && + txr->csum_lhlen == pi->ipi_ehdrlen && + txr->csum_iphlen == pi->ipi_ip_hlen && + txr->csum_flags == csum_flags) { + /* + * Same csum offload context as the previous packets; + * just return. + */ + *txd_upper = txr->csum_txd_upper; + *txd_lower = txr->csum_txd_lower; + return (cur); + } + + TXD = (struct e1000_context_desc *)&txr->tx_base[cur]; + if (csum_flags & CSUM_IP) { + *txd_upper |= E1000_TXD_POPTS_IXSM << 8; + /* + * Start offset for header checksum calculation. + * End offset for header checksum calculation. + * Offset of place to put the checksum. + */ + TXD->lower_setup.ip_fields.ipcss = pi->ipi_ehdrlen; + TXD->lower_setup.ip_fields.ipcse = htole16(hdr_len); + TXD->lower_setup.ip_fields.ipcso = pi->ipi_ehdrlen + offsetof(struct ip, ip_sum); + cmd |= E1000_TXD_CMD_IP; + } + + if (csum_flags & (CSUM_TCP|CSUM_UDP)) { + uint8_t tucso; + + *txd_upper |= E1000_TXD_POPTS_TXSM << 8; + *txd_lower = E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D; + + if (csum_flags & CSUM_TCP) { + tucso = hdr_len + offsetof(struct tcphdr, th_sum); + cmd |= E1000_TXD_CMD_TCP; + } else + tucso = hdr_len + offsetof(struct udphdr, uh_sum); + TXD->upper_setup.tcp_fields.tucss = hdr_len; + TXD->upper_setup.tcp_fields.tucse = htole16(0); + TXD->upper_setup.tcp_fields.tucso = tucso; + } + + txr->csum_lhlen = pi->ipi_ehdrlen; + txr->csum_iphlen = pi->ipi_ip_hlen; + txr->csum_flags = csum_flags; + txr->csum_txd_upper = *txd_upper; + txr->csum_txd_lower = *txd_lower; + + TXD->tcp_seg_setup.data = htole32(0); + TXD->cmd_and_length = + htole32(E1000_TXD_CMD_IFCS | E1000_TXD_CMD_DEXT | cmd); + + tx_buffer = &txr->tx_buffers[cur]; + tx_buffer->eop = -1; + + if (++cur == scctx->isc_ntxd[0]) { + cur = 0; + } + DPRINTF(iflib_get_dev(adapter->ctx), "checksum_setup csum_flags=%x txd_upper=%x txd_lower=%x hdr_len=%d cmd=%x\n", + csum_flags, *txd_upper, *txd_lower, hdr_len, cmd); + return (cur); +} + +static int +em_isc_txd_encap(void *arg, if_pkt_info_t pi) +{ + struct adapter *sc = arg; + if_softc_ctx_t scctx = sc->shared; + struct em_tx_queue *que = &sc->tx_queues[pi->ipi_qsidx]; + struct tx_ring *txr = &que->txr; + bus_dma_segment_t *segs = pi->ipi_segs; + int nsegs = pi->ipi_nsegs; + int csum_flags = pi->ipi_csum_flags; + int i, j, first, pidx_last; + u32 txd_upper = 0, txd_lower = 0; + + struct em_txbuffer *tx_buffer; + struct e1000_tx_desc *ctxd = NULL; + bool do_tso, tso_desc; + + i = first = pi->ipi_pidx; + do_tso = (csum_flags & CSUM_TSO); + tso_desc = FALSE; + /* + * TSO Hardware workaround, if this packet is not + * TSO, and is only a single descriptor long, and + * it follows a TSO burst, then we need to add a + * sentinel descriptor to prevent premature writeback. + */ + if ((!do_tso) && (txr->tx_tso == TRUE)) { + if (nsegs == 1) + tso_desc = TRUE; + txr->tx_tso = FALSE; + } + + /* Do hardware assists */ + if (do_tso) { + i = em_tso_setup(sc, pi, &txd_upper, &txd_lower); + tso_desc = TRUE; + } else if (csum_flags & CSUM_OFFLOAD) { + i = em_transmit_checksum_setup(sc, pi, &txd_upper, &txd_lower); + } + + if (pi->ipi_mflags & M_VLANTAG) { + /* Set the vlan id. */ + txd_upper |= htole16(pi->ipi_vtag) << 16; + /* Tell hardware to add tag */ + txd_lower |= htole32(E1000_TXD_CMD_VLE); + } + + DPRINTF(iflib_get_dev(sc->ctx), "encap: set up tx: nsegs=%d first=%d i=%d\n", nsegs, first, i); + /* XXX adapter->pcix_82544 -- lem_fill_descriptors */ + + /* Set up our transmit descriptors */ + for (j = 0; j < nsegs; j++) { + bus_size_t seg_len; + bus_addr_t seg_addr; + uint32_t cmd; + + ctxd = &txr->tx_base[i]; + tx_buffer = &txr->tx_buffers[i]; + seg_addr = segs[j].ds_addr; + seg_len = segs[j].ds_len; + cmd = E1000_TXD_CMD_IFCS | sc->txd_cmd; + + /* + ** TSO Workaround: + ** If this is the last descriptor, we want to + ** split it so we have a small final sentinel + */ + if (tso_desc && (j == (nsegs - 1)) && (seg_len > 8)) { + seg_len -= TSO_WORKAROUND; + ctxd->buffer_addr = htole64(seg_addr); + ctxd->lower.data = htole32(cmd | txd_lower | seg_len); + ctxd->upper.data = htole32(txd_upper); + + if (++i == scctx->isc_ntxd[0]) + i = 0; + + /* Now make the sentinel */ + ctxd = &txr->tx_base[i]; + tx_buffer = &txr->tx_buffers[i]; + ctxd->buffer_addr = htole64(seg_addr + seg_len); + ctxd->lower.data = htole32(cmd | txd_lower | TSO_WORKAROUND); + ctxd->upper.data = htole32(txd_upper); + pidx_last = i; + if (++i == scctx->isc_ntxd[0]) + i = 0; + DPRINTF(iflib_get_dev(sc->ctx), "TSO path pidx_last=%d i=%d ntxd[0]=%d\n", pidx_last, i, scctx->isc_ntxd[0]); + } else { + ctxd->buffer_addr = htole64(seg_addr); + ctxd->lower.data = htole32(cmd | txd_lower | seg_len); + ctxd->upper.data = htole32(txd_upper); + pidx_last = i; + if (++i == scctx->isc_ntxd[0]) + i = 0; + DPRINTF(iflib_get_dev(sc->ctx), "pidx_last=%d i=%d ntxd[0]=%d\n", pidx_last, i, scctx->isc_ntxd[0]); + } + tx_buffer->eop = -1; + } + + /* + * Last Descriptor of Packet + * needs End Of Packet (EOP) + * and Report Status (RS) + */ + ctxd->lower.data |= + htole32(E1000_TXD_CMD_EOP | E1000_TXD_CMD_RS); + + tx_buffer = &txr->tx_buffers[first]; + tx_buffer->eop = pidx_last; + DPRINTF(iflib_get_dev(sc->ctx), "tx_buffers[%d]->eop = %d ipi_new_pidx=%d\n", first, pidx_last, i); + pi->ipi_new_pidx = i; + + return (0); +} + +static void +em_isc_txd_flush(void *arg, uint16_t txqid, uint32_t pidx) +{ + struct adapter *adapter = arg; + struct em_tx_queue *que = &adapter->tx_queues[txqid]; + struct tx_ring *txr = &que->txr; + + E1000_WRITE_REG(&adapter->hw, E1000_TDT(txr->me), pidx); +} + +static int +em_isc_txd_credits_update(void *arg, uint16_t txqid, uint32_t cidx_init, bool clear) +{ + struct adapter *adapter = arg; + if_softc_ctx_t scctx = adapter->shared; + struct em_tx_queue *que = &adapter->tx_queues[txqid]; + struct tx_ring *txr = &que->txr; + + u32 cidx, processed = 0; + int last, done; + struct em_txbuffer *buf; + struct e1000_tx_desc *tx_desc, *eop_desc; + + cidx = cidx_init; + buf = &txr->tx_buffers[cidx]; + tx_desc = &txr->tx_base[cidx]; + last = buf->eop; + eop_desc = &txr->tx_base[last]; + + DPRINTF(iflib_get_dev(adapter->ctx), "credits_update: cidx_init=%d clear=%d last=%d\n", + cidx_init, clear, last); + /* + * What this does is get the index of the + * first descriptor AFTER the EOP of the + * first packet, that way we can do the + * simple comparison on the inner while loop. + */ + if (++last == scctx->isc_ntxd[0]) + last = 0; + done = last; + + + while (eop_desc->upper.fields.status & E1000_TXD_STAT_DD) { + /* We clean the range of the packet */ + while (cidx != done) { + if (clear) { + tx_desc->upper.data = 0; + tx_desc->lower.data = 0; + tx_desc->buffer_addr = 0; + buf->eop = -1; + } + tx_desc++; + buf++; + processed++; + + /* wrap the ring ? */ + if (++cidx == scctx->isc_ntxd[0]) { + cidx = 0; + } + buf = &txr->tx_buffers[cidx]; + tx_desc = &txr->tx_base[cidx]; + } + /* See if we can continue to the next packet */ + last = buf->eop; + if (last == -1) + break; + eop_desc = &txr->tx_base[last]; + /* Get new done point */ + if (++last == scctx->isc_ntxd[0]) + last = 0; + done = last; + } + + DPRINTF(iflib_get_dev(adapter->ctx), "Processed %d credits update\n", processed); + return(processed); +} + +static void +lem_isc_rxd_refill(void *arg, uint16_t rxqid, uint8_t flid __unused, + uint32_t pidx, uint64_t *paddrs, caddr_t *vaddrs __unused, uint16_t count, uint16_t buflen __unused) +{ + struct adapter *sc = arg; + if_softc_ctx_t scctx = sc->shared; + struct em_rx_queue *que = &sc->rx_queues[rxqid]; + struct rx_ring *rxr = &que->rxr; + struct e1000_rx_desc *rxd; + int i; + uint32_t next_pidx; + + for (i = 0, next_pidx = pidx; i < count; i++) { + rxd = (struct e1000_rx_desc *)&rxr->rx_base[next_pidx]; + rxd->buffer_addr = htole64(paddrs[i]); + /* status bits must be cleared */ + rxd->status = 0; + + if (++next_pidx == scctx->isc_nrxd[0]) + next_pidx = 0; + } +} + +static void +em_isc_rxd_refill(void *arg, uint16_t rxqid, uint8_t flid __unused, + uint32_t pidx, uint64_t *paddrs, caddr_t *vaddrs __unused, uint16_t count, uint16_t buflen __unused) +{ + struct adapter *sc = arg; + if_softc_ctx_t scctx = sc->shared; + struct em_rx_queue *que = &sc->rx_queues[rxqid]; + struct rx_ring *rxr = &que->rxr; + union e1000_rx_desc_extended *rxd; + int i; + uint32_t next_pidx; + + for (i = 0, next_pidx = pidx; i < count; i++) { + rxd = &rxr->rx_base[next_pidx]; + rxd->read.buffer_addr = htole64(paddrs[i]); + /* DD bits must be cleared */ + rxd->wb.upper.status_error = 0; + + if (++next_pidx == scctx->isc_nrxd[0]) + next_pidx = 0; + } +} + +static void +em_isc_rxd_flush(void *arg, uint16_t rxqid, uint8_t flid __unused, uint32_t pidx) +{ + struct adapter *sc = arg; + struct em_rx_queue *que = &sc->rx_queues[rxqid]; + struct rx_ring *rxr = &que->rxr; + + E1000_WRITE_REG(&sc->hw, E1000_RDT(rxr->me), pidx); +} + +static int +lem_isc_rxd_available(void *arg, uint16_t rxqid, uint32_t idx, int budget) +{ + struct adapter *sc = arg; + if_softc_ctx_t scctx = sc->shared; + struct em_rx_queue *que = &sc->rx_queues[rxqid]; + struct rx_ring *rxr = &que->rxr; + struct e1000_rx_desc *rxd; + u32 staterr = 0; + int cnt, i; + + for (cnt = 0, i = idx; cnt < scctx->isc_nrxd[0] && cnt <= budget;) { + rxd = (struct e1000_rx_desc *)&rxr->rx_base[i]; + staterr = rxd->status; + + if ((staterr & E1000_RXD_STAT_DD) == 0) + break; + + if (++i == scctx->isc_nrxd[0]) + i = 0; + + if (staterr & E1000_RXD_STAT_EOP) + cnt++; + } + return (cnt); +} + +static int +em_isc_rxd_available(void *arg, uint16_t rxqid, uint32_t idx, int budget) +{ + struct adapter *sc = arg; + if_softc_ctx_t scctx = sc->shared; + struct em_rx_queue *que = &sc->rx_queues[rxqid]; + struct rx_ring *rxr = &que->rxr; + union e1000_rx_desc_extended *rxd; + u32 staterr = 0; + int cnt, i; + + for (cnt = 0, i = idx; cnt < scctx->isc_nrxd[0] && cnt <= budget;) { + rxd = &rxr->rx_base[i]; + staterr = le32toh(rxd->wb.upper.status_error); + + if ((staterr & E1000_RXD_STAT_DD) == 0) + break; + + if (++i == scctx->isc_nrxd[0]) { + i = 0; + } + + if (staterr & E1000_RXD_STAT_EOP) + cnt++; + + } + return (cnt); +} + +static int +lem_isc_rxd_pkt_get(void *arg, if_rxd_info_t ri) +{ + struct adapter *adapter = arg; + if_softc_ctx_t scctx = adapter->shared; + struct em_rx_queue *que = &adapter->rx_queues[ri->iri_qsidx]; + struct rx_ring *rxr = &que->rxr; + struct e1000_rx_desc *rxd; + u16 len; + u32 status, errors; + bool eop; + int i, cidx; + + status = errors = i = 0; + cidx = ri->iri_cidx; + + do { + rxd = (struct e1000_rx_desc *)&rxr->rx_base[cidx]; + status = rxd->status; + errors = rxd->errors; + + /* Error Checking then decrement count */ + MPASS ((status & E1000_RXD_STAT_DD) != 0); + + len = le16toh(rxd->length); + ri->iri_len += len; + + eop = (status & E1000_RXD_STAT_EOP) != 0; + + /* Make sure bad packets are discarded */ + if (errors & E1000_RXD_ERR_FRAME_ERR_MASK) { + adapter->dropped_pkts++; + /* XXX fixup if common */ + return (EBADMSG); + } + + ri->iri_frags[i].irf_flid = 0; + ri->iri_frags[i].irf_idx = cidx; + ri->iri_frags[i].irf_len = len; + /* Zero out the receive descriptors status. */ + rxd->status = 0; + + if (++cidx == scctx->isc_nrxd[0]) + cidx = 0; + i++; + } while (!eop); + + /* XXX add a faster way to look this up */ + if (adapter->hw.mac.type >= e1000_82543 && !(status & E1000_RXD_STAT_IXSM)) + lem_receive_checksum(status, errors, ri); + + if (status & E1000_RXD_STAT_VP) { + ri->iri_vtag = le16toh(rxd->special); + ri->iri_flags |= M_VLANTAG; + } + + ri->iri_nfrags = i; + + return (0); +} + +static int +em_isc_rxd_pkt_get(void *arg, if_rxd_info_t ri) +{ + struct adapter *adapter = arg; + if_softc_ctx_t scctx = adapter->shared; + struct em_rx_queue *que = &adapter->rx_queues[ri->iri_qsidx]; + struct rx_ring *rxr = &que->rxr; + union e1000_rx_desc_extended *rxd; + + u16 len; + u32 staterr = 0; + bool eop; + int i, cidx, vtag; + + i = vtag = 0; + cidx = ri->iri_cidx; + + do { + rxd = &rxr->rx_base[cidx]; + staterr = le32toh(rxd->wb.upper.status_error); + + /* Error Checking then decrement count */ + MPASS ((staterr & E1000_RXD_STAT_DD) != 0); + + len = le16toh(rxd->wb.upper.length); + ri->iri_len += len; + + eop = (staterr & E1000_RXD_STAT_EOP) != 0; + + /* Make sure bad packets are discarded */ + if (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) { + adapter->dropped_pkts++; + return EBADMSG; + } + + ri->iri_frags[i].irf_flid = 0; + ri->iri_frags[i].irf_idx = cidx; + ri->iri_frags[i].irf_len = len; + /* Zero out the receive descriptors status. */ + rxd->wb.upper.status_error &= htole32(~0xFF); + + if (++cidx == scctx->isc_nrxd[0]) + cidx = 0; + i++; + } while (!eop); + + /* XXX add a faster way to look this up */ + if (adapter->hw.mac.type >= e1000_82543) + em_receive_checksum(staterr, ri); + + if (staterr & E1000_RXD_STAT_VP) { + vtag = le16toh(rxd->wb.upper.vlan); + } + + ri->iri_vtag = vtag; + ri->iri_nfrags = i; + if (vtag) + ri->iri_flags |= M_VLANTAG; + + return (0); +} + +/********************************************************************* + * + * Verify that the hardware indicated that the checksum is valid. + * Inform the stack about the status of checksum so that stack + * doesn't spend time verifying the checksum. + * + *********************************************************************/ +static void +lem_receive_checksum(int status, int errors, if_rxd_info_t ri) +{ + /* Did it pass? */ + if (status & E1000_RXD_STAT_IPCS && !(errors & E1000_RXD_ERR_IPE)) + ri->iri_csum_flags = (CSUM_IP_CHECKED|CSUM_IP_VALID); + + if (status & E1000_RXD_STAT_TCPCS) { + /* Did it pass? */ + if (!(errors & E1000_RXD_ERR_TCPE)) { + ri->iri_csum_flags |= + (CSUM_DATA_VALID | CSUM_PSEUDO_HDR); + ri->iri_csum_data = htons(0xffff); + } + } +} + +static void +em_receive_checksum(uint32_t status, if_rxd_info_t ri) +{ + ri->iri_csum_flags = 0; + + /* Ignore Checksum bit is set */ + if (status & E1000_RXD_STAT_IXSM) + return; + + /* If the IP checksum exists and there is no IP Checksum error */ + if ((status & (E1000_RXD_STAT_IPCS | E1000_RXDEXT_STATERR_IPE)) == + E1000_RXD_STAT_IPCS) { + ri->iri_csum_flags = (CSUM_IP_CHECKED | CSUM_IP_VALID); + } + + /* TCP or UDP checksum */ + if ((status & (E1000_RXD_STAT_TCPCS | E1000_RXDEXT_STATERR_TCPE)) == + E1000_RXD_STAT_TCPCS) { + ri->iri_csum_flags |= (CSUM_DATA_VALID | CSUM_PSEUDO_HDR); + ri->iri_csum_data = htons(0xffff); + } + if (status & E1000_RXD_STAT_UDPCS) { + ri->iri_csum_flags |= (CSUM_DATA_VALID | CSUM_PSEUDO_HDR); + ri->iri_csum_data = htons(0xffff); + } +} Modified: head/sys/dev/e1000/if_em.c ============================================================================== --- head/sys/dev/e1000/if_em.c Tue Jan 10 01:36:50 2017 (r311848) +++ head/sys/dev/e1000/if_em.c Tue Jan 10 03:23:22 2017 (r311849) @@ -1,99 +1,10 @@ -/****************************************************************************** - - Copyright (c) 2001-2015, Intel Corporation - All rights reserved. - - Redistribution and use in source and binary forms, with or without - modification, are permitted provided that the following conditions are met: - - 1. Redistributions of source code must retain the above copyright notice, - this list of conditions and the following disclaimer. - - 2. Redistributions in binary form must reproduce the above copyright - notice, this list of conditions and the following disclaimer in the - documentation and/or other materials provided with the distribution. - - 3. Neither the name of the Intel Corporation nor the names of its - contributors may be used to endorse or promote products derived from - this software without specific prior written permission. - - THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" - AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE - IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE - ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE - LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS - INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN - CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) - ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE - POSSIBILITY OF SUCH DAMAGE. - -******************************************************************************/ -/*$FreeBSD$*/ - -#include "opt_em.h" -#include "opt_ddb.h" -#include "opt_inet.h" -#include "opt_inet6.h" - -#ifdef HAVE_KERNEL_OPTION_HEADERS -#include "opt_device_polling.h" -#endif - -#include <sys/param.h> -#include <sys/systm.h> -#ifdef DDB -#include <sys/types.h> -#include <ddb/ddb.h> -#endif -#if __FreeBSD_version >= 800000 -#include <sys/buf_ring.h> -#endif -#include <sys/bus.h> -#include <sys/endian.h> -#include <sys/kernel.h> -#include <sys/kthread.h> -#include <sys/malloc.h> -#include <sys/mbuf.h> -#include <sys/module.h> -#include <sys/rman.h> -#include <sys/smp.h> -#include <sys/socket.h> -#include <sys/sockio.h> -#include <sys/sysctl.h> -#include <sys/taskqueue.h> -#include <sys/eventhandler.h> -#include <machine/bus.h> -#include <machine/resource.h> - -#include <net/bpf.h> -#include <net/ethernet.h> -#include <net/if.h> -#include <net/if_var.h> -#include <net/if_arp.h> -#include <net/if_dl.h> -#include <net/if_media.h> - -#include <net/if_types.h> -#include <net/if_vlan_var.h> - -#include <netinet/in_systm.h> -#include <netinet/in.h> -#include <netinet/if_ether.h> -#include <netinet/ip.h> -#include <netinet/ip6.h> -#include <netinet/tcp.h> -#include <netinet/udp.h> - -#include <machine/in_cksum.h> -#include <dev/led/led.h> -#include <dev/pci/pcivar.h> -#include <dev/pci/pcireg.h> - -#include "e1000_api.h" -#include "e1000_82571.h" +/* $FreeBSD$ */ #include "if_em.h" +#include <sys/sbuf.h> +#include <machine/_inttypes.h> + +#define em_mac_min e1000_82547 +#define igb_mac_min e1000_82575 /********************************************************************* * Driver version: @@ -110,184 +21,213 @@ char em_driver_version[] = "7.6.1-k"; * { Vendor ID, Device ID, SubVendor ID, SubDevice ID, String Index } *********************************************************************/ -static em_vendor_info_t em_vendor_info_array[] = +static pci_vendor_info_t em_vendor_info_array[] = { - /* Intel(R) PRO/1000 Network Connection */ - { 0x8086, E1000_DEV_ID_82571EB_COPPER, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_FIBER, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_SERDES, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_SERDES_DUAL, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_SERDES_QUAD, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_QUAD_COPPER, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_QUAD_COPPER_LP, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571EB_QUAD_FIBER, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82571PT_QUAD_COPPER, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82572EI_COPPER, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82572EI_FIBER, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82572EI_SERDES, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82572EI, PCI_ANY_ID, PCI_ANY_ID, 0}, - - { 0x8086, E1000_DEV_ID_82573E, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82573E_IAMT, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82573L, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_82583V, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_80003ES2LAN_COPPER_SPT, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_80003ES2LAN_SERDES_SPT, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_80003ES2LAN_COPPER_DPT, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_80003ES2LAN_SERDES_DPT, - PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_ICH8_IGP_M_AMT, PCI_ANY_ID, PCI_ANY_ID, 0}, - { 0x8086, E1000_DEV_ID_ICH8_IGP_AMT, PCI_ANY_ID, PCI_ANY_ID, 0}, *** DIFF OUTPUT TRUNCATED AT 1000 LINES ***
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201701100323.v0A3NMAI045868>