2006-06-26 23:41:07 +00:00
|
|
|
/* $OpenBSD: if_nfereg.h,v 1.16 2006/02/22 19:23:44 damien Exp $ */
|
2006-06-26 17:25:22 +00:00
|
|
|
|
|
|
|
|
/*-
|
|
|
|
|
* Copyright (c) 2005 Jonathan Gray <jsg@openbsd.org>
|
|
|
|
|
*
|
|
|
|
|
* Permission to use, copy, modify, and distribute this software for any
|
|
|
|
|
* purpose with or without fee is hereby granted, provided that the above
|
|
|
|
|
* copyright notice and this permission notice appear in all copies.
|
|
|
|
|
*
|
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
|
|
|
|
|
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
|
|
|
|
|
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
|
|
|
|
|
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
|
|
|
|
|
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
|
|
|
|
|
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
|
|
|
|
|
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
|
2006-06-26 23:41:07 +00:00
|
|
|
*
|
|
|
|
|
* $FreeBSD$
|
2006-06-26 17:25:22 +00:00
|
|
|
*/
|
|
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_RX_RING_COUNT 256
|
|
|
|
|
#define NFE_JUMBO_RX_RING_COUNT NFE_RX_RING_COUNT
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_TX_RING_COUNT 256
|
|
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_PROC_DEFAULT ((NFE_RX_RING_COUNT * 3) / 4)
|
|
|
|
|
#define NFE_PROC_MIN 50
|
|
|
|
|
#define NFE_PROC_MAX (NFE_RX_RING_COUNT - 1)
|
|
|
|
|
|
|
|
|
|
#define NFE_INC(x, y) (x) = ((x) + 1) % y
|
|
|
|
|
|
2006-11-27 00:16:47 +00:00
|
|
|
/* RX/TX MAC addr + type + VLAN + align + slack */
|
|
|
|
|
#define NFE_RX_HEADERS 64
|
|
|
|
|
|
|
|
|
|
/* Maximum MTU size. */
|
|
|
|
|
#define NV_PKTLIMIT_1 ETH_DATA_LEN /* Hard limit not known. */
|
|
|
|
|
#define NV_PKTLIMIT_2 9100 /* Actual limit according to NVidia:9202 */
|
2006-10-19 10:04:46 +00:00
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_JUMBO_FRAMELEN NV_PKTLIMIT_2
|
|
|
|
|
#define NFE_JUMBO_MTU \
|
|
|
|
|
(NFE_JUMBO_FRAMELEN - NFE_RX_HEADERS)
|
|
|
|
|
#define NFE_MIN_FRAMELEN (ETHER_MIN_LEN - ETHER_CRC_LEN)
|
|
|
|
|
#define NFE_JSLOTS ((NFE_JUMBO_RX_RING_COUNT * 3) / 2)
|
|
|
|
|
|
|
|
|
|
#define NFE_JRAWLEN (NFE_JUMBO_FRAMELEN + ETHER_ALIGN)
|
|
|
|
|
#define NFE_JLEN \
|
|
|
|
|
(NFE_JRAWLEN + (sizeof(uint64_t) - (NFE_JRAWLEN % sizeof(uint64_t))))
|
|
|
|
|
#define NFE_JPAGESZ PAGE_SIZE
|
|
|
|
|
#define NFE_RESID \
|
|
|
|
|
(NFE_JPAGESZ - (NFE_JLEN * NFE_JSLOTS) % NFE_JPAGESZ)
|
|
|
|
|
#define NFE_JMEM ((NFE_JLEN * NFE_JSLOTS) + NFE_RESID)
|
|
|
|
|
|
|
|
|
|
#define NFE_MAX_SCATTER 32
|
2007-06-12 02:35:01 +00:00
|
|
|
#define NFE_TSO_MAXSGSIZE 4096
|
2007-06-12 10:51:47 +00:00
|
|
|
#define NFE_TSO_MAXSIZE (65535 + sizeof(struct ether_vlan_header))
|
2006-10-19 10:04:46 +00:00
|
|
|
|
|
|
|
|
#define NFE_IRQ_STATUS 0x000
|
|
|
|
|
#define NFE_IRQ_MASK 0x004
|
|
|
|
|
#define NFE_SETUP_R6 0x008
|
|
|
|
|
#define NFE_IMTIMER 0x00c
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_MSI_MAP0 0x020
|
|
|
|
|
#define NFE_MSI_MAP1 0x024
|
|
|
|
|
#define NFE_MSI_IRQ_MASK 0x030
|
|
|
|
|
#define NFE_MAC_RESET 0x03c
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_MISC1 0x080
|
|
|
|
|
#define NFE_TX_CTL 0x084
|
|
|
|
|
#define NFE_TX_STATUS 0x088
|
|
|
|
|
#define NFE_RXFILTER 0x08c
|
|
|
|
|
#define NFE_RXBUFSZ 0x090
|
|
|
|
|
#define NFE_RX_CTL 0x094
|
|
|
|
|
#define NFE_RX_STATUS 0x098
|
|
|
|
|
#define NFE_RNDSEED 0x09c
|
|
|
|
|
#define NFE_SETUP_R1 0x0a0
|
|
|
|
|
#define NFE_SETUP_R2 0x0a4
|
|
|
|
|
#define NFE_MACADDR_HI 0x0a8
|
|
|
|
|
#define NFE_MACADDR_LO 0x0ac
|
|
|
|
|
#define NFE_MULTIADDR_HI 0x0b0
|
|
|
|
|
#define NFE_MULTIADDR_LO 0x0b4
|
|
|
|
|
#define NFE_MULTIMASK_HI 0x0b8
|
|
|
|
|
#define NFE_MULTIMASK_LO 0x0bc
|
|
|
|
|
#define NFE_PHY_IFACE 0x0c0
|
|
|
|
|
#define NFE_TX_RING_ADDR_LO 0x100
|
|
|
|
|
#define NFE_RX_RING_ADDR_LO 0x104
|
|
|
|
|
#define NFE_RING_SIZE 0x108
|
|
|
|
|
#define NFE_TX_UNK 0x10c
|
|
|
|
|
#define NFE_LINKSPEED 0x110
|
|
|
|
|
#define NFE_SETUP_R5 0x130
|
|
|
|
|
#define NFE_SETUP_R3 0x13C
|
|
|
|
|
#define NFE_SETUP_R7 0x140
|
|
|
|
|
#define NFE_RXTX_CTL 0x144
|
|
|
|
|
#define NFE_TX_RING_ADDR_HI 0x148
|
|
|
|
|
#define NFE_RX_RING_ADDR_HI 0x14c
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_TX_PAUSE_FRAME 0x170
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_PHY_STATUS 0x180
|
|
|
|
|
#define NFE_SETUP_R4 0x184
|
|
|
|
|
#define NFE_STATUS 0x188
|
|
|
|
|
#define NFE_PHY_SPEED 0x18c
|
|
|
|
|
#define NFE_PHY_CTL 0x190
|
|
|
|
|
#define NFE_PHY_DATA 0x194
|
|
|
|
|
#define NFE_WOL_CTL 0x200
|
|
|
|
|
#define NFE_PATTERN_CRC 0x204
|
|
|
|
|
#define NFE_PATTERN_MASK 0x208
|
|
|
|
|
#define NFE_PWR_CAP 0x268
|
|
|
|
|
#define NFE_PWR_STATE 0x26c
|
|
|
|
|
#define NFE_VTAG_CTL 0x300
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_MSIX_MAP0 0x3e0
|
|
|
|
|
#define NFE_MSIX_MAP1 0x3e4
|
|
|
|
|
#define NFE_MSIX_IRQ_STATUS 0x3f0
|
|
|
|
|
#define NFE_PWR2_CTL 0x600
|
|
|
|
|
|
|
|
|
|
#define NFE_MAC_RESET_MAGIC 0x00f3
|
|
|
|
|
|
|
|
|
|
#define NFE_MAC_ADDR_INORDER 0x8000
|
2006-10-19 10:04:46 +00:00
|
|
|
|
|
|
|
|
#define NFE_PHY_ERROR 0x00001
|
|
|
|
|
#define NFE_PHY_WRITE 0x00400
|
|
|
|
|
#define NFE_PHY_BUSY 0x08000
|
|
|
|
|
#define NFE_PHYADD_SHIFT 5
|
|
|
|
|
|
|
|
|
|
#define NFE_STATUS_MAGIC 0x140000
|
|
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_R1_MAGIC_1000 0x14050f
|
|
|
|
|
#define NFE_R1_MAGIC_10_100 0x16070f
|
|
|
|
|
#define NFE_R1_MAGIC_DEFAULT 0x15050f
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_R2_MAGIC 0x16
|
|
|
|
|
#define NFE_R4_MAGIC 0x08
|
|
|
|
|
#define NFE_R6_MAGIC 0x03
|
|
|
|
|
#define NFE_WOL_MAGIC 0x1111
|
|
|
|
|
#define NFE_RX_START 0x01
|
|
|
|
|
#define NFE_TX_START 0x01
|
|
|
|
|
|
|
|
|
|
#define NFE_IRQ_RXERR 0x0001
|
|
|
|
|
#define NFE_IRQ_RX 0x0002
|
|
|
|
|
#define NFE_IRQ_RX_NOBUF 0x0004
|
|
|
|
|
#define NFE_IRQ_TXERR 0x0008
|
|
|
|
|
#define NFE_IRQ_TX_DONE 0x0010
|
|
|
|
|
#define NFE_IRQ_TIMER 0x0020
|
|
|
|
|
#define NFE_IRQ_LINK 0x0040
|
|
|
|
|
#define NFE_IRQ_TXERR2 0x0080
|
|
|
|
|
#define NFE_IRQ_TX1 0x0100
|
|
|
|
|
|
|
|
|
|
#define NFE_IRQ_WANTED \
|
2006-06-26 17:25:22 +00:00
|
|
|
(NFE_IRQ_RXERR | NFE_IRQ_RX_NOBUF | NFE_IRQ_RX | \
|
|
|
|
|
NFE_IRQ_TXERR | NFE_IRQ_TXERR2 | NFE_IRQ_TX_DONE | \
|
|
|
|
|
NFE_IRQ_LINK)
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_RXTX_KICKTX 0x0001
|
|
|
|
|
#define NFE_RXTX_BIT1 0x0002
|
|
|
|
|
#define NFE_RXTX_BIT2 0x0004
|
|
|
|
|
#define NFE_RXTX_RESET 0x0010
|
|
|
|
|
#define NFE_RXTX_VTAG_STRIP 0x0040
|
|
|
|
|
#define NFE_RXTX_VTAG_INSERT 0x0080
|
|
|
|
|
#define NFE_RXTX_RXCSUM 0x0400
|
|
|
|
|
#define NFE_RXTX_V2MAGIC 0x2100
|
|
|
|
|
#define NFE_RXTX_V3MAGIC 0x2200
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_RXFILTER_MAGIC 0x007f0000
|
|
|
|
|
#define NFE_PFF_RX_PAUSE (1 << 3)
|
|
|
|
|
#define NFE_PFF_LOOPBACK (1 << 4)
|
|
|
|
|
#define NFE_PFF_U2M (1 << 5)
|
|
|
|
|
#define NFE_PFF_PROMISC (1 << 7)
|
2006-11-27 04:47:27 +00:00
|
|
|
#define NFE_CSUM_FEATURES (CSUM_IP | CSUM_TCP | CSUM_UDP)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
|
|
|
|
/* default interrupt moderation timer of 128us */
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_IM_DEFAULT ((128 * 100) / 1024)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_VTAG_ENABLE (1 << 13)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_PWR_VALID (1 << 8)
|
|
|
|
|
#define NFE_PWR_WAKEUP (1 << 15)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_PWR2_WAKEUP_MASK 0x0f11
|
|
|
|
|
#define NFE_PWR2_REVA3 (1 << 0)
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_MEDIA_SET 0x10000
|
2006-06-26 17:25:22 +00:00
|
|
|
#define NFE_MEDIA_1000T 0x00032
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_MEDIA_100TX 0x00064
|
|
|
|
|
#define NFE_MEDIA_10T 0x003e8
|
2006-06-26 17:25:22 +00:00
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_PHY_100TX (1 << 0)
|
|
|
|
|
#define NFE_PHY_1000T (1 << 1)
|
|
|
|
|
#define NFE_PHY_HDX (1 << 8)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_MISC1_MAGIC 0x003b0f3c
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_MISC1_TX_PAUSE (1 << 0)
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_MISC1_HDX (1 << 1)
|
2006-06-26 17:25:22 +00:00
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_TX_PAUSE_FRAME_DISABLE 0x1ff0080
|
|
|
|
|
#define NFE_TX_PAUSE_FRAME_ENABLE 0x0c00030
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_SEED_MASK 0x0003ff00
|
|
|
|
|
#define NFE_SEED_10T 0x00007f00
|
|
|
|
|
#define NFE_SEED_100TX 0x00002d00
|
|
|
|
|
#define NFE_SEED_1000T 0x00007400
|
2006-06-26 17:25:22 +00:00
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_MSI_MESSAGES 8
|
|
|
|
|
#define NFE_MSI_VECTOR_0_ENABLED 0x01
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* It seems that nForce supports only the lower 40 bits of a DMA address.
|
|
|
|
|
*/
|
|
|
|
|
#if (BUS_SPACE_MAXADDR < 0xFFFFFFFFFF)
|
|
|
|
|
#define NFE_DMA_MAXADDR BUS_SPACE_MAXADDR
|
|
|
|
|
#else
|
|
|
|
|
#define NFE_DMA_MAXADDR 0xFFFFFFFFFF
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#define NFE_ADDR_LO(x) ((u_int64_t) (x) & 0xffffffff)
|
|
|
|
|
#define NFE_ADDR_HI(x) ((u_int64_t) (x) >> 32)
|
|
|
|
|
|
2006-06-26 17:25:22 +00:00
|
|
|
/* Rx/Tx descriptor */
|
|
|
|
|
struct nfe_desc32 {
|
|
|
|
|
uint32_t physaddr;
|
|
|
|
|
uint16_t length;
|
|
|
|
|
uint16_t flags;
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_RX_FIXME_V1 0x6004
|
|
|
|
|
#define NFE_RX_VALID_V1 (1 << 0)
|
|
|
|
|
#define NFE_TX_ERROR_V1 0x7808
|
|
|
|
|
#define NFE_TX_LASTFRAG_V1 (1 << 0)
|
|
|
|
|
#define NFE_RX_ERROR1_V1 (1<<7)
|
|
|
|
|
#define NFE_RX_ERROR2_V1 (1<<8)
|
|
|
|
|
#define NFE_RX_ERROR3_V1 (1<<9)
|
|
|
|
|
#define NFE_RX_ERROR4_V1 (1<<10)
|
2006-06-26 17:25:22 +00:00
|
|
|
} __packed;
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_V1_TXERR "\020" \
|
2006-06-26 17:25:22 +00:00
|
|
|
"\14TXERROR\13UNDERFLOW\12LATECOLLISION\11LOSTCARRIER\10DEFERRED" \
|
|
|
|
|
"\08FORCEDINT\03RETRY\00LASTPACKET"
|
|
|
|
|
|
|
|
|
|
/* V2 Rx/Tx descriptor */
|
|
|
|
|
struct nfe_desc64 {
|
|
|
|
|
uint32_t physaddr[2];
|
|
|
|
|
uint32_t vtag;
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_RX_VTAG (1 << 16)
|
|
|
|
|
#define NFE_TX_VTAG (1 << 18)
|
2006-06-26 17:25:22 +00:00
|
|
|
uint16_t length;
|
|
|
|
|
uint16_t flags;
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_RX_FIXME_V2 0x4300
|
|
|
|
|
#define NFE_RX_VALID_V2 (1 << 13)
|
|
|
|
|
#define NFE_TX_ERROR_V2 0x5c04
|
|
|
|
|
#define NFE_TX_LASTFRAG_V2 (1 << 13)
|
|
|
|
|
#define NFE_RX_ERROR1_V2 (1<<2)
|
|
|
|
|
#define NFE_RX_ERROR2_V2 (1<<3)
|
|
|
|
|
#define NFE_RX_ERROR3_V2 (1<<4)
|
|
|
|
|
#define NFE_RX_ERROR4_V2 (1<<5)
|
2006-06-26 17:25:22 +00:00
|
|
|
} __packed;
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_V2_TXERR "\020" \
|
2006-06-26 17:25:22 +00:00
|
|
|
"\14FORCEDINT\13LASTPACKET\12UNDERFLOW\10LOSTCARRIER\09DEFERRED\02RETRY"
|
|
|
|
|
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_RING_ALIGN (sizeof(struct nfe_desc64))
|
|
|
|
|
|
2006-06-26 17:25:22 +00:00
|
|
|
/* flags common to V1/V2 descriptors */
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_RX_UDP_CSUMOK (1 << 10)
|
|
|
|
|
#define NFE_RX_TCP_CSUMOK (1 << 11)
|
|
|
|
|
#define NFE_RX_IP_CSUMOK (1 << 12)
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_RX_ERROR (1 << 14)
|
|
|
|
|
#define NFE_RX_READY (1 << 15)
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_RX_LEN_MASK 0x3fff
|
|
|
|
|
#define NFE_TX_TCP_UDP_CSUM (1 << 10)
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_TX_IP_CSUM (1 << 11)
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define NFE_TX_TSO (1 << 12)
|
|
|
|
|
#define NFE_TX_TSO_SHIFT 14
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_TX_VALID (1 << 15)
|
|
|
|
|
|
|
|
|
|
#define NFE_READ(sc, reg) \
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
bus_read_4((sc)->nfe_res[0], (reg))
|
2006-06-26 17:25:22 +00:00
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_WRITE(sc, reg, val) \
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
bus_write_4((sc)->nfe_res[0], (reg), (val))
|
|
|
|
|
|
|
|
|
|
#define NFE_TIMEOUT 1000
|
2006-06-26 23:41:07 +00:00
|
|
|
|
|
|
|
|
#ifndef PCI_VENDOR_NVIDIA
|
|
|
|
|
#define PCI_VENDOR_NVIDIA 0x10DE
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE_LAN 0x01C3
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE2_LAN 0x0066
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_LAN1 0x00D6
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE2_400_LAN1 0x0086
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE2_400_LAN2 0x008C
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_250_LAN 0x00E6
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_LAN4 0x00DF
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE4_LAN1 0x0056
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE4_LAN2 0x0057
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP04_LAN1 0x0037
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP04_LAN2 0x0038
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE430_LAN1 0x0268
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE430_LAN2 0x0269
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP55_LAN1 0x0372
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP55_LAN2 0x0373
|
2006-09-11 07:31:55 +00:00
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP61_LAN1 0x03e5
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP61_LAN2 0x03e6
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP61_LAN3 0x03ee
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP61_LAN4 0x03ef
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP65_LAN1 0x0450
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP65_LAN2 0x0451
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP65_LAN3 0x0452
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP65_LAN4 0x0453
|
Bring overhauled nfe(4) into tree.
o s/printf/device_printf/g
o Nuke OpenBSDism.
o Nuke NetBSD/OpenBSD specific DMA sync operations.(we don't have a way
to sync a single descriptor within a DMA map.)
o Remove recursive mutex.
o bus_dma(9) clean up.
o 40bit DMA address support.
o Add protection for Rx map load failure.
o Fix a long standing bug for watchdog timeout. [1]
o Add additional protections, missing Tx completion interrupt, losing
start Tx command, for watchdog timeout.
o Switch to taskqueue(9) API to handle interrupts.
o Use our own timer for watchdog instead of if_watchdog/if_timer
interface.
o Advertise VLAN header length/capability correctly to upper layer.
o Remove excessive kernel stack consumption in nfe_encap().
o Handle highly fragmented mbuf chains correctly.
o Enable etherenet address reprogramming with ifconfig(8).
o Add ALTQ/TSO, MSI/MSIX support.
o Increased Rx ring to 256 descriptors from 128.
o Align Tx/Rx descriptor ring on sizeof(struct nfe_desc64) boundary.
o Remove alignment restrictions on Tx/Rx buffers.
o Rewritten jumbo frame support code.
o Add support for hardware assistend VLAN tag insertion/stripping.
o Add support for Tx/Rx flow control based on patches from Peer Chen. [2]
o Add a routine that detects whether ethernet address swap routines is
required. [3]
o Add a workaround that take MAC/PHY out of power down mode.
o Add suspend/resume support.
o style(9) and code clean up.
Special thanks to Shigeaki Tagashira, the original porter of nfe(4),
who submitted lots of patches, performed uncountable number of
regression tests and maintained nfe(4) for a long time. Without his
enthusiastic help and support I could never have completed this
overhauling task.
The only weak point of nfe(4) compared to nve(4) is instability of
manual half-duplex media selection on certain hardwares(auto sensing
media type should work for all cases, though). This was a long
standing bug of nfe(4) and I still have no idea why it doesn't work
on some hardwares.
Obtained from: OpenBSD [1]
Submitted by: Peer Chen < pchen at nvidia dot com > [2], [3]
Reviewed by: Shigeaki Tagashira < shigeaki AT se DOT hiroshima-u DOT ac DOT jp >
Tested by: Shigeaki Tagashira, current
Discussed with: current
Silence from: obrien
2007-06-12 02:16:02 +00:00
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP67_LAN1 0x054c
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP67_LAN2 0x054d
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP67_LAN3 0x054e
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP67_LAN4 0x054f
|
2006-06-26 23:41:07 +00:00
|
|
|
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_LAN2 PCI_PRODUCT_NVIDIA_NFORCE2_400_LAN1
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_LAN3 PCI_PRODUCT_NVIDIA_NFORCE2_400_LAN2
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_NFORCE3_LAN5 PCI_PRODUCT_NVIDIA_NFORCE3_250_LAN
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_CK804_LAN1 PCI_PRODUCT_NVIDIA_NFORCE4_LAN1
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_CK804_LAN2 PCI_PRODUCT_NVIDIA_NFORCE4_LAN2
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP51_LAN1 PCI_PRODUCT_NVIDIA_NFORCE430_LAN1
|
|
|
|
|
#define PCI_PRODUCT_NVIDIA_MCP51_LAN2 PCI_PRODUCT_NVIDIA_NFORCE430_LAN2
|
|
|
|
|
|
2006-10-19 10:04:46 +00:00
|
|
|
#define NFE_DEBUG 0x0000
|
|
|
|
|
#define NFE_DEBUG_INIT 0x0001
|
|
|
|
|
#define NFE_DEBUG_RUNNING 0x0002
|
|
|
|
|
#define NFE_DEBUG_DEINIT 0x0004
|
|
|
|
|
#define NFE_DEBUG_IOCTL 0x0008
|
|
|
|
|
#define NFE_DEBUG_INTERRUPT 0x0010
|
|
|
|
|
#define NFE_DEBUG_API 0x0020
|
|
|
|
|
#define NFE_DEBUG_LOCK 0x0040
|
|
|
|
|
#define NFE_DEBUG_BROKEN 0x0080
|
|
|
|
|
#define NFE_DEBUG_MII 0x0100
|
|
|
|
|
#define NFE_DEBUG_ALL 0xFFFF
|
