freebsd-nq/sys/dev/sk/if_skreg.h

1530 lines
50 KiB
C
Raw Normal View History

/* $OpenBSD: if_skreg.h,v 1.10 2003/08/12 05:23:06 nate Exp $ */
/*-
* Copyright (c) 1997, 1998, 1999, 2000
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
* Bill Paul <wpaul@ctr.columbia.edu>. 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. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul 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 Bill Paul OR THE VOICES IN HIS HEAD
* 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.
*
1999-08-28 01:08:13 +00:00
* $FreeBSD$
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
*/
/*-
* Copyright (c) 2003 Nathan L. Binkert <binkertn@umich.edu>
*
* 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.
*/
/* Values to keep the different chip revisions apart (SK_CHIPVER). */
#define SK_GENESIS 0x0A
#define SK_YUKON 0xB0
#define SK_YUKON_LITE 0xB1
#define SK_YUKON_LP 0xB2
#define SK_YUKON_FAMILY(x) ((x) & 0xB0)
/* Known revisions in SK_CONFIG. */
#define SK_YUKON_LITE_REV_A0 0x0 /* invented, see test in skc_attach. */
#define SK_YUKON_LITE_REV_A1 0x3
#define SK_YUKON_LITE_REV_A3 0x7
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* SysKonnect PCI vendor ID
*/
#define VENDORID_SK 0x1148
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* Marvell PCI vendor ID
*/
#define VENDORID_MARVELL 0x11AB
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* SK-NET gigabit ethernet device IDs
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
*/
#define DEVICEID_SK_V1 0x4300
#define DEVICEID_SK_V2 0x4320
/*
* Belkin F5D5005
*/
#define DEVICEID_BELKIN_5005 0x5005
/*
* 3Com PCI vendor ID
*/
#define VENDORID_3COM 0x10b7
/*
* 3Com gigabit ethernet device ID
*/
#define DEVICEID_3COM_3C940 0x1700
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* Linksys PCI vendor ID
*/
#define VENDORID_LINKSYS 0x1737
/*
* Linksys gigabit ethernet device ID
*/
#define DEVICEID_LINKSYS_EG1032 0x1032
/*
* Linksys gigabit ethernet rev 2 sub-device ID
*/
#define SUBDEVICEID_LINKSYS_EG1032_REV2 0x0015
/*
* D-Link PCI vendor ID
*/
#define VENDORID_DLINK 0x1186
/*
* D-Link gigabit ethernet device ID
*/
#define DEVICEID_DLINK_DGE530T_A1 0x4c00
#define DEVICEID_DLINK_DGE530T_B1 0x4b01
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* GEnesis registers. The GEnesis chip has a 256-byte I/O window
* but internally it has a 16K register space. This 16K space is
* divided into 128-byte blocks. The first 128 bytes of the I/O
* window represent the first block, which is permanently mapped
* at the start of the window. The other 127 blocks can be mapped
* to the second 128 bytes of the I/O window by setting the desired
* block value in the RAP register in block 0. Not all of the 127
* blocks are actually used. Most registers are 32 bits wide, but
* there are a few 16-bit and 8-bit ones as well.
*/
/* Start of remappable register window. */
#define SK_WIN_BASE 0x0080
/* Size of a window */
#define SK_WIN_LEN 0x80
#define SK_WIN_MASK 0x3F80
#define SK_REG_MASK 0x7F
/* Compute the window of a given register (for the RAP register) */
#define SK_WIN(reg) (((reg) & SK_WIN_MASK) / SK_WIN_LEN)
/* Compute the relative offset of a register within the window */
#define SK_REG(reg) ((reg) & SK_REG_MASK)
#define SK_PORT_A 0
#define SK_PORT_B 1
/*
* Compute offset of port-specific register. Since there are two
* ports, there are two of some GEnesis modules (e.g. two sets of
* DMA queues, two sets of FIFO control registers, etc...). Normally,
* the block for port 0 is at offset 0x0 and the block for port 1 is
* at offset 0x80 (i.e. the next page over). However for the transmit
* BMUs and RAMbuffers, there are two blocks for each port: one for
* the sync transmit queue and one for the async queue (which we don't
* use). However instead of ordering them like this:
* TX sync 1 / TX sync 2 / TX async 1 / TX async 2
* SysKonnect has instead ordered them like this:
* TX sync 1 / TX async 1 / TX sync 2 / TX async 2
* This means that when referencing the TX BMU and RAMbuffer registers,
* we have to double the block offset (0x80 * 2) in order to reach the
* second queue. This prevents us from using the same formula
* (sk_port * 0x80) to compute the offsets for all of the port-specific
* blocks: we need an extra offset for the BMU and RAMbuffer registers.
* The simplest thing is to provide an extra argument to these macros:
* the 'skip' parameter. The 'skip' value is the number of extra pages
* for skip when computing the port0/port1 offsets. For most registers,
* the skip value is 0; for the BMU and RAMbuffer registers, it's 1.
*/
#define SK_IF_READ_4(sc_if, skip, reg) \
sk_win_read_4(sc_if->sk_softc, reg + \
((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN))
#define SK_IF_READ_2(sc_if, skip, reg) \
sk_win_read_2(sc_if->sk_softc, reg + \
((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN))
#define SK_IF_READ_1(sc_if, skip, reg) \
sk_win_read_1(sc_if->sk_softc, reg + \
((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN))
#define SK_IF_WRITE_4(sc_if, skip, reg, val) \
sk_win_write_4(sc_if->sk_softc, \
reg + ((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN), val)
#define SK_IF_WRITE_2(sc_if, skip, reg, val) \
sk_win_write_2(sc_if->sk_softc, \
reg + ((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN), val)
#define SK_IF_WRITE_1(sc_if, skip, reg, val) \
sk_win_write_1(sc_if->sk_softc, \
reg + ((sc_if->sk_port * (skip + 1)) * SK_WIN_LEN), val)
/* Block 0 registers, permanently mapped at iobase. */
#define SK_RAP 0x0000
#define SK_CSR 0x0004
#define SK_LED 0x0006
#define SK_ISR 0x0008 /* interrupt source */
#define SK_IMR 0x000C /* interrupt mask */
#define SK_IESR 0x0010 /* interrupt hardware error source */
#define SK_IEMR 0x0014 /* interrupt hardware error mask */
#define SK_ISSR 0x0018 /* special interrupt source */
#define SK_XM_IMR0 0x0020
#define SK_XM_ISR0 0x0028
#define SK_XM_PHYADDR0 0x0030
#define SK_XM_PHYDATA0 0x0034
#define SK_XM_IMR1 0x0040
#define SK_XM_ISR1 0x0048
#define SK_XM_PHYADDR1 0x0050
#define SK_XM_PHYDATA1 0x0054
#define SK_BMU_RX_CSR0 0x0060
#define SK_BMU_RX_CSR1 0x0064
#define SK_BMU_TXS_CSR0 0x0068
#define SK_BMU_TXA_CSR0 0x006C
#define SK_BMU_TXS_CSR1 0x0070
#define SK_BMU_TXA_CSR1 0x0074
/* SK_CSR register */
#define SK_CSR_SW_RESET 0x0001
#define SK_CSR_SW_UNRESET 0x0002
#define SK_CSR_MASTER_RESET 0x0004
#define SK_CSR_MASTER_UNRESET 0x0008
#define SK_CSR_MASTER_STOP 0x0010
#define SK_CSR_MASTER_DONE 0x0020
#define SK_CSR_SW_IRQ_CLEAR 0x0040
#define SK_CSR_SW_IRQ_SET 0x0080
#define SK_CSR_SLOTSIZE 0x0100 /* 1 == 64 bits, 0 == 32 */
#define SK_CSR_BUSCLOCK 0x0200 /* 1 == 33/66 Mhz, = 33 */
/* SK_LED register */
#define SK_LED_GREEN_OFF 0x01
#define SK_LED_GREEN_ON 0x02
/* SK_ISR register */
#define SK_ISR_TX2_AS_CHECK 0x00000001
#define SK_ISR_TX2_AS_EOF 0x00000002
#define SK_ISR_TX2_AS_EOB 0x00000004
#define SK_ISR_TX2_S_CHECK 0x00000008
#define SK_ISR_TX2_S_EOF 0x00000010
#define SK_ISR_TX2_S_EOB 0x00000020
#define SK_ISR_TX1_AS_CHECK 0x00000040
#define SK_ISR_TX1_AS_EOF 0x00000080
#define SK_ISR_TX1_AS_EOB 0x00000100
#define SK_ISR_TX1_S_CHECK 0x00000200
#define SK_ISR_TX1_S_EOF 0x00000400
#define SK_ISR_TX1_S_EOB 0x00000800
#define SK_ISR_RX2_CHECK 0x00001000
#define SK_ISR_RX2_EOF 0x00002000
#define SK_ISR_RX2_EOB 0x00004000
#define SK_ISR_RX1_CHECK 0x00008000
#define SK_ISR_RX1_EOF 0x00010000
#define SK_ISR_RX1_EOB 0x00020000
#define SK_ISR_LINK2_OFLOW 0x00040000
#define SK_ISR_MAC2 0x00080000
#define SK_ISR_LINK1_OFLOW 0x00100000
#define SK_ISR_MAC1 0x00200000
#define SK_ISR_TIMER 0x00400000
#define SK_ISR_EXTERNAL_REG 0x00800000
#define SK_ISR_SW 0x01000000
#define SK_ISR_I2C_RDY 0x02000000
#define SK_ISR_TX2_TIMEO 0x04000000
#define SK_ISR_TX1_TIMEO 0x08000000
#define SK_ISR_RX2_TIMEO 0x10000000
#define SK_ISR_RX1_TIMEO 0x20000000
#define SK_ISR_RSVD 0x40000000
#define SK_ISR_HWERR 0x80000000
/* SK_IMR register */
#define SK_IMR_TX2_AS_CHECK 0x00000001
#define SK_IMR_TX2_AS_EOF 0x00000002
#define SK_IMR_TX2_AS_EOB 0x00000004
#define SK_IMR_TX2_S_CHECK 0x00000008
#define SK_IMR_TX2_S_EOF 0x00000010
#define SK_IMR_TX2_S_EOB 0x00000020
#define SK_IMR_TX1_AS_CHECK 0x00000040
#define SK_IMR_TX1_AS_EOF 0x00000080
#define SK_IMR_TX1_AS_EOB 0x00000100
#define SK_IMR_TX1_S_CHECK 0x00000200
#define SK_IMR_TX1_S_EOF 0x00000400
#define SK_IMR_TX1_S_EOB 0x00000800
#define SK_IMR_RX2_CHECK 0x00001000
#define SK_IMR_RX2_EOF 0x00002000
#define SK_IMR_RX2_EOB 0x00004000
#define SK_IMR_RX1_CHECK 0x00008000
#define SK_IMR_RX1_EOF 0x00010000
#define SK_IMR_RX1_EOB 0x00020000
#define SK_IMR_LINK2_OFLOW 0x00040000
#define SK_IMR_MAC2 0x00080000
#define SK_IMR_LINK1_OFLOW 0x00100000
#define SK_IMR_MAC1 0x00200000
#define SK_IMR_TIMER 0x00400000
#define SK_IMR_EXTERNAL_REG 0x00800000
#define SK_IMR_SW 0x01000000
#define SK_IMR_I2C_RDY 0x02000000
#define SK_IMR_TX2_TIMEO 0x04000000
#define SK_IMR_TX1_TIMEO 0x08000000
#define SK_IMR_RX2_TIMEO 0x10000000
#define SK_IMR_RX1_TIMEO 0x20000000
#define SK_IMR_RSVD 0x40000000
#define SK_IMR_HWERR 0x80000000
#define SK_INTRS1 \
(SK_IMR_RX1_EOF|SK_IMR_TX1_S_EOF|SK_IMR_MAC1)
#define SK_INTRS2 \
(SK_IMR_RX2_EOF|SK_IMR_TX2_S_EOF|SK_IMR_MAC2)
/* SK_IESR register */
#define SK_IESR_PAR_RX2 0x00000001
#define SK_IESR_PAR_RX1 0x00000002
#define SK_IESR_PAR_MAC2 0x00000004
#define SK_IESR_PAR_MAC1 0x00000008
#define SK_IESR_PAR_WR_RAM 0x00000010
#define SK_IESR_PAR_RD_RAM 0x00000020
#define SK_IESR_NO_TSTAMP_MAC2 0x00000040
#define SK_IESR_NO_TSTAMO_MAC1 0x00000080
#define SK_IESR_NO_STS_MAC2 0x00000100
#define SK_IESR_NO_STS_MAC1 0x00000200
#define SK_IESR_IRQ_STS 0x00000400
#define SK_IESR_MASTERERR 0x00000800
/* SK_IEMR register */
#define SK_IEMR_PAR_RX2 0x00000001
#define SK_IEMR_PAR_RX1 0x00000002
#define SK_IEMR_PAR_MAC2 0x00000004
#define SK_IEMR_PAR_MAC1 0x00000008
#define SK_IEMR_PAR_WR_RAM 0x00000010
#define SK_IEMR_PAR_RD_RAM 0x00000020
#define SK_IEMR_NO_TSTAMP_MAC2 0x00000040
#define SK_IEMR_NO_TSTAMO_MAC1 0x00000080
#define SK_IEMR_NO_STS_MAC2 0x00000100
#define SK_IEMR_NO_STS_MAC1 0x00000200
#define SK_IEMR_IRQ_STS 0x00000400
#define SK_IEMR_MASTERERR 0x00000800
/* Block 2 */
#define SK_MAC0_0 0x0100
#define SK_MAC0_1 0x0104
#define SK_MAC1_0 0x0108
#define SK_MAC1_1 0x010C
#define SK_MAC2_0 0x0110
#define SK_MAC2_1 0x0114
#define SK_CONNTYPE 0x0118
#define SK_PMDTYPE 0x0119
#define SK_CONFIG 0x011A
#define SK_CHIPVER 0x011B
#define SK_EPROM0 0x011C
#define SK_EPROM1 0x011D /* yukon/genesis */
#define SK_EPROM2 0x011E /* yukon/genesis */
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_EPROM3 0x011F
#define SK_EP_ADDR 0x0120
#define SK_EP_DATA 0x0124
#define SK_EP_LOADCTL 0x0128
#define SK_EP_LOADTST 0x0129
#define SK_TIMERINIT 0x0130
#define SK_TIMER 0x0134
#define SK_TIMERCTL 0x0138
#define SK_TIMERTST 0x0139
#define SK_IMTIMERINIT 0x0140
#define SK_IMTIMER 0x0144
#define SK_IMTIMERCTL 0x0148
#define SK_IMTIMERTST 0x0149
#define SK_IMMR 0x014C
#define SK_IHWEMR 0x0150
#define SK_TESTCTL1 0x0158
#define SK_TESTCTL2 0x0159
#define SK_GPIO 0x015C
#define SK_I2CHWCTL 0x0160
#define SK_I2CHWDATA 0x0164
#define SK_I2CHWIRQ 0x0168
#define SK_I2CSW 0x016C
#define SK_BLNKINIT 0x0170
#define SK_BLNKCOUNT 0x0174
#define SK_BLNKCTL 0x0178
#define SK_BLNKSTS 0x0179
#define SK_BLNKTST 0x017A
#define SK_IMCTL_STOP 0x02
#define SK_IMCTL_START 0x04
#define SK_IMTIMER_TICKS_GENESIS 53
#define SK_IMTIMER_TICKS_YUKON 78
#define SK_IM_USECS(x, t) ((x) * (t))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_IM_MIN 10
#define SK_IM_DEFAULT 100
#define SK_IM_MAX 10000
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* The SK_EPROM0 register contains a byte that describes the
* amount of SRAM mounted on the NIC. The value also tells if
* the chips are 64K or 128K. This affects the RAMbuffer address
* offset that we need to use.
*/
#define SK_RAMSIZE_512K_64 0x1
#define SK_RAMSIZE_1024K_128 0x2
#define SK_RAMSIZE_1024K_64 0x3
#define SK_RAMSIZE_2048K_128 0x4
#define SK_RBOFF_0 0x0
#define SK_RBOFF_80000 0x80000
/*
* SK_EEPROM1 contains the PHY type, which may be XMAC for
* fiber-based cards or BCOM for 1000baseT cards with a Broadcom
* PHY.
*/
#define SK_PHYTYPE_XMAC 0 /* integeated XMAC II PHY */
#define SK_PHYTYPE_BCOM 1 /* Broadcom BCM5400 */
#define SK_PHYTYPE_LONE 2 /* Level One LXT1000 */
#define SK_PHYTYPE_NAT 3 /* National DP83891 */
#define SK_PHYTYPE_MARV_COPPER 4 /* Marvell 88E1011S */
#define SK_PHYTYPE_MARV_FIBER 5 /* Marvell 88E1011S (fiber) */
/*
* PHY addresses.
*/
#define SK_PHYADDR_XMAC 0x0
#define SK_PHYADDR_BCOM 0x1
#define SK_PHYADDR_LONE 0x3
#define SK_PHYADDR_NAT 0x0
#define SK_PHYADDR_MARV 0x0
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_CONFIG_SINGLEMAC 0x01
#define SK_CONFIG_DIS_DSL_CLK 0x02
#define SK_PMD_1000BASELX 0x4C
#define SK_PMD_1000BASESX 0x53
#define SK_PMD_1000BASECX 0x43
#define SK_PMD_1000BASETX 0x54
/* GPIO bits */
#define SK_GPIO_DAT0 0x00000001
#define SK_GPIO_DAT1 0x00000002
#define SK_GPIO_DAT2 0x00000004
#define SK_GPIO_DAT3 0x00000008
#define SK_GPIO_DAT4 0x00000010
#define SK_GPIO_DAT5 0x00000020
#define SK_GPIO_DAT6 0x00000040
#define SK_GPIO_DAT7 0x00000080
#define SK_GPIO_DAT8 0x00000100
#define SK_GPIO_DAT9 0x00000200
#define SK_GPIO_DIR0 0x00010000
#define SK_GPIO_DIR1 0x00020000
#define SK_GPIO_DIR2 0x00040000
#define SK_GPIO_DIR3 0x00080000
#define SK_GPIO_DIR4 0x00100000
#define SK_GPIO_DIR5 0x00200000
#define SK_GPIO_DIR6 0x00400000
#define SK_GPIO_DIR7 0x00800000
#define SK_GPIO_DIR8 0x01000000
#define SK_GPIO_DIR9 0x02000000
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 3 Ram interface and MAC arbiter registers */
#define SK_RAMADDR 0x0180
#define SK_RAMDATA0 0x0184
#define SK_RAMDATA1 0x0188
#define SK_TO0 0x0190
#define SK_TO1 0x0191
#define SK_TO2 0x0192
#define SK_TO3 0x0193
#define SK_TO4 0x0194
#define SK_TO5 0x0195
#define SK_TO6 0x0196
#define SK_TO7 0x0197
#define SK_TO8 0x0198
#define SK_TO9 0x0199
#define SK_TO10 0x019A
#define SK_TO11 0x019B
#define SK_RITIMEO_TMR 0x019C
#define SK_RAMCTL 0x01A0
#define SK_RITIMER_TST 0x01A2
#define SK_RAMCTL_RESET 0x0001
#define SK_RAMCTL_UNRESET 0x0002
#define SK_RAMCTL_CLR_IRQ_WPAR 0x0100
#define SK_RAMCTL_CLR_IRQ_RPAR 0x0200
/* Mac arbiter registers */
#define SK_MINIT_RX1 0x01B0
#define SK_MINIT_RX2 0x01B1
#define SK_MINIT_TX1 0x01B2
#define SK_MINIT_TX2 0x01B3
#define SK_MTIMEO_RX1 0x01B4
#define SK_MTIMEO_RX2 0x01B5
#define SK_MTIMEO_TX1 0x01B6
#define SK_MTIEMO_TX2 0x01B7
#define SK_MACARB_CTL 0x01B8
#define SK_MTIMER_TST 0x01BA
#define SK_RCINIT_RX1 0x01C0
#define SK_RCINIT_RX2 0x01C1
#define SK_RCINIT_TX1 0x01C2
#define SK_RCINIT_TX2 0x01C3
#define SK_RCTIMEO_RX1 0x01C4
#define SK_RCTIMEO_RX2 0x01C5
#define SK_RCTIMEO_TX1 0x01C6
#define SK_RCTIMEO_TX2 0x01C7
#define SK_RECOVERY_CTL 0x01C8
#define SK_RCTIMER_TST 0x01CA
/* Packet arbiter registers */
#define SK_RXPA1_TINIT 0x01D0
#define SK_RXPA2_TINIT 0x01D4
#define SK_TXPA1_TINIT 0x01D8
#define SK_TXPA2_TINIT 0x01DC
#define SK_RXPA1_TIMEO 0x01E0
#define SK_RXPA2_TIMEO 0x01E4
#define SK_TXPA1_TIMEO 0x01E8
#define SK_TXPA2_TIMEO 0x01EC
#define SK_PKTARB_CTL 0x01F0
#define SK_PKTATB_TST 0x01F2
#define SK_PKTARB_TIMEOUT 0x2000
#define SK_PKTARBCTL_RESET 0x0001
#define SK_PKTARBCTL_UNRESET 0x0002
#define SK_PKTARBCTL_RXTO1_OFF 0x0004
#define SK_PKTARBCTL_RXTO1_ON 0x0008
#define SK_PKTARBCTL_RXTO2_OFF 0x0010
#define SK_PKTARBCTL_RXTO2_ON 0x0020
#define SK_PKTARBCTL_TXTO1_OFF 0x0040
#define SK_PKTARBCTL_TXTO1_ON 0x0080
#define SK_PKTARBCTL_TXTO2_OFF 0x0100
#define SK_PKTARBCTL_TXTO2_ON 0x0200
#define SK_PKTARBCTL_CLR_IRQ_RXTO1 0x0400
#define SK_PKTARBCTL_CLR_IRQ_RXTO2 0x0800
#define SK_PKTARBCTL_CLR_IRQ_TXTO1 0x1000
#define SK_PKTARBCTL_CLR_IRQ_TXTO2 0x2000
#define SK_MINIT_XMAC_B2 54
#define SK_MINIT_XMAC_C1 63
#define SK_MACARBCTL_RESET 0x0001
#define SK_MACARBCTL_UNRESET 0x0002
#define SK_MACARBCTL_FASTOE_OFF 0x0004
#define SK_MACARBCRL_FASTOE_ON 0x0008
#define SK_RCINIT_XMAC_B2 54
#define SK_RCINIT_XMAC_C1 0
#define SK_RECOVERYCTL_RX1_OFF 0x0001
#define SK_RECOVERYCTL_RX1_ON 0x0002
#define SK_RECOVERYCTL_RX2_OFF 0x0004
#define SK_RECOVERYCTL_RX2_ON 0x0008
#define SK_RECOVERYCTL_TX1_OFF 0x0010
#define SK_RECOVERYCTL_TX1_ON 0x0020
#define SK_RECOVERYCTL_TX2_OFF 0x0040
#define SK_RECOVERYCTL_TX2_ON 0x0080
#define SK_RECOVERY_XMAC_B2 \
(SK_RECOVERYCTL_RX1_ON|SK_RECOVERYCTL_RX2_ON| \
SK_RECOVERYCTL_TX1_ON|SK_RECOVERYCTL_TX2_ON)
#define SK_RECOVERY_XMAC_C1 \
(SK_RECOVERYCTL_RX1_OFF|SK_RECOVERYCTL_RX2_OFF| \
SK_RECOVERYCTL_TX1_OFF|SK_RECOVERYCTL_TX2_OFF)
/* Block 4 -- TX Arbiter MAC 1 */
#define SK_TXAR1_TIMERINIT 0x0200
#define SK_TXAR1_TIMERVAL 0x0204
#define SK_TXAR1_LIMITINIT 0x0208
#define SK_TXAR1_LIMITCNT 0x020C
#define SK_TXAR1_COUNTERCTL 0x0210
#define SK_TXAR1_COUNTERTST 0x0212
#define SK_TXAR1_COUNTERSTS 0x0212
/* Block 5 -- TX Arbiter MAC 2 */
#define SK_TXAR2_TIMERINIT 0x0280
#define SK_TXAR2_TIMERVAL 0x0284
#define SK_TXAR2_LIMITINIT 0x0288
#define SK_TXAR2_LIMITCNT 0x028C
#define SK_TXAR2_COUNTERCTL 0x0290
#define SK_TXAR2_COUNTERTST 0x0291
#define SK_TXAR2_COUNTERSTS 0x0292
#define SK_TXARCTL_OFF 0x01
#define SK_TXARCTL_ON 0x02
#define SK_TXARCTL_RATECTL_OFF 0x04
#define SK_TXARCTL_RATECTL_ON 0x08
#define SK_TXARCTL_ALLOC_OFF 0x10
#define SK_TXARCTL_ALLOC_ON 0x20
#define SK_TXARCTL_FSYNC_OFF 0x40
#define SK_TXARCTL_FSYNC_ON 0x80
/* Block 6 -- External registers */
#define SK_EXTREG_BASE 0x300
#define SK_EXTREG_END 0x37C
/* Block 7 -- PCI config registers */
#define SK_PCI_BASE 0x0380
#define SK_PCI_END 0x03FC
/* Compute offset of mirrored PCI register */
#define SK_PCI_REG(reg) ((reg) + SK_PCI_BASE)
/* Block 8 -- RX queue 1 */
#define SK_RXQ1_BUFCNT 0x0400
#define SK_RXQ1_BUFCTL 0x0402
#define SK_RXQ1_NEXTDESC 0x0404
#define SK_RXQ1_RXBUF_LO 0x0408
#define SK_RXQ1_RXBUF_HI 0x040C
#define SK_RXQ1_RXSTAT 0x0410
#define SK_RXQ1_TIMESTAMP 0x0414
#define SK_RXQ1_CSUM1 0x0418
#define SK_RXQ1_CSUM2 0x041A
#define SK_RXQ1_CSUM1_START 0x041C
#define SK_RXQ1_CSUM2_START 0x041E
#define SK_RXQ1_CURADDR_LO 0x0420
#define SK_RXQ1_CURADDR_HI 0x0424
#define SK_RXQ1_CURCNT_LO 0x0428
#define SK_RXQ1_CURCNT_HI 0x042C
#define SK_RXQ1_CURBYTES 0x0430
#define SK_RXQ1_BMU_CSR 0x0434
#define SK_RXQ1_WATERMARK 0x0438
#define SK_RXQ1_FLAG 0x043A
#define SK_RXQ1_TEST1 0x043C
#define SK_RXQ1_TEST2 0x0440
#define SK_RXQ1_TEST3 0x0444
/* Block 9 -- RX queue 2 */
#define SK_RXQ2_BUFCNT 0x0480
#define SK_RXQ2_BUFCTL 0x0482
#define SK_RXQ2_NEXTDESC 0x0484
#define SK_RXQ2_RXBUF_LO 0x0488
#define SK_RXQ2_RXBUF_HI 0x048C
#define SK_RXQ2_RXSTAT 0x0490
#define SK_RXQ2_TIMESTAMP 0x0494
#define SK_RXQ2_CSUM1 0x0498
#define SK_RXQ2_CSUM2 0x049A
#define SK_RXQ2_CSUM1_START 0x049C
#define SK_RXQ2_CSUM2_START 0x049E
#define SK_RXQ2_CURADDR_LO 0x04A0
#define SK_RXQ2_CURADDR_HI 0x04A4
#define SK_RXQ2_CURCNT_LO 0x04A8
#define SK_RXQ2_CURCNT_HI 0x04AC
#define SK_RXQ2_CURBYTES 0x04B0
#define SK_RXQ2_BMU_CSR 0x04B4
#define SK_RXQ2_WATERMARK 0x04B8
#define SK_RXQ2_FLAG 0x04BA
#define SK_RXQ2_TEST1 0x04BC
#define SK_RXQ2_TEST2 0x04C0
#define SK_RXQ2_TEST3 0x04C4
#define SK_RXBMU_CLR_IRQ_ERR 0x00000001
#define SK_RXBMU_CLR_IRQ_EOF 0x00000002
#define SK_RXBMU_CLR_IRQ_EOB 0x00000004
#define SK_RXBMU_CLR_IRQ_PAR 0x00000008
#define SK_RXBMU_RX_START 0x00000010
#define SK_RXBMU_RX_STOP 0x00000020
#define SK_RXBMU_POLL_OFF 0x00000040
#define SK_RXBMU_POLL_ON 0x00000080
#define SK_RXBMU_TRANSFER_SM_RESET 0x00000100
#define SK_RXBMU_TRANSFER_SM_UNRESET 0x00000200
#define SK_RXBMU_DESCWR_SM_RESET 0x00000400
#define SK_RXBMU_DESCWR_SM_UNRESET 0x00000800
#define SK_RXBMU_DESCRD_SM_RESET 0x00001000
#define SK_RXBMU_DESCRD_SM_UNRESET 0x00002000
#define SK_RXBMU_SUPERVISOR_SM_RESET 0x00004000
#define SK_RXBMU_SUPERVISOR_SM_UNRESET 0x00008000
#define SK_RXBMU_PFI_SM_RESET 0x00010000
#define SK_RXBMU_PFI_SM_UNRESET 0x00020000
#define SK_RXBMU_FIFO_RESET 0x00040000
#define SK_RXBMU_FIFO_UNRESET 0x00080000
#define SK_RXBMU_DESC_RESET 0x00100000
#define SK_RXBMU_DESC_UNRESET 0x00200000
#define SK_RXBMU_SUPERVISOR_IDLE 0x01000000
#define SK_RXBMU_ONLINE \
(SK_RXBMU_TRANSFER_SM_UNRESET|SK_RXBMU_DESCWR_SM_UNRESET| \
SK_RXBMU_DESCRD_SM_UNRESET|SK_RXBMU_SUPERVISOR_SM_UNRESET| \
SK_RXBMU_PFI_SM_UNRESET|SK_RXBMU_FIFO_UNRESET| \
SK_RXBMU_DESC_UNRESET)
#define SK_RXBMU_OFFLINE \
(SK_RXBMU_TRANSFER_SM_RESET|SK_RXBMU_DESCWR_SM_RESET| \
SK_RXBMU_DESCRD_SM_RESET|SK_RXBMU_SUPERVISOR_SM_RESET| \
SK_RXBMU_PFI_SM_RESET|SK_RXBMU_FIFO_RESET| \
SK_RXBMU_DESC_RESET)
/* Block 12 -- TX sync queue 1 */
#define SK_TXQS1_BUFCNT 0x0600
#define SK_TXQS1_BUFCTL 0x0602
#define SK_TXQS1_NEXTDESC 0x0604
#define SK_TXQS1_RXBUF_LO 0x0608
#define SK_TXQS1_RXBUF_HI 0x060C
#define SK_TXQS1_RXSTAT 0x0610
#define SK_TXQS1_CSUM_STARTVAL 0x0614
#define SK_TXQS1_CSUM_STARTPOS 0x0618
#define SK_TXQS1_CSUM_WRITEPOS 0x061A
#define SK_TXQS1_CURADDR_LO 0x0620
#define SK_TXQS1_CURADDR_HI 0x0624
#define SK_TXQS1_CURCNT_LO 0x0628
#define SK_TXQS1_CURCNT_HI 0x062C
#define SK_TXQS1_CURBYTES 0x0630
#define SK_TXQS1_BMU_CSR 0x0634
#define SK_TXQS1_WATERMARK 0x0638
#define SK_TXQS1_FLAG 0x063A
#define SK_TXQS1_TEST1 0x063C
#define SK_TXQS1_TEST2 0x0640
#define SK_TXQS1_TEST3 0x0644
/* Block 13 -- TX async queue 1 */
#define SK_TXQA1_BUFCNT 0x0680
#define SK_TXQA1_BUFCTL 0x0682
#define SK_TXQA1_NEXTDESC 0x0684
#define SK_TXQA1_RXBUF_LO 0x0688
#define SK_TXQA1_RXBUF_HI 0x068C
#define SK_TXQA1_RXSTAT 0x0690
#define SK_TXQA1_CSUM_STARTVAL 0x0694
#define SK_TXQA1_CSUM_STARTPOS 0x0698
#define SK_TXQA1_CSUM_WRITEPOS 0x069A
#define SK_TXQA1_CURADDR_LO 0x06A0
#define SK_TXQA1_CURADDR_HI 0x06A4
#define SK_TXQA1_CURCNT_LO 0x06A8
#define SK_TXQA1_CURCNT_HI 0x06AC
#define SK_TXQA1_CURBYTES 0x06B0
#define SK_TXQA1_BMU_CSR 0x06B4
#define SK_TXQA1_WATERMARK 0x06B8
#define SK_TXQA1_FLAG 0x06BA
#define SK_TXQA1_TEST1 0x06BC
#define SK_TXQA1_TEST2 0x06C0
#define SK_TXQA1_TEST3 0x06C4
/* Block 14 -- TX sync queue 2 */
#define SK_TXQS2_BUFCNT 0x0700
#define SK_TXQS2_BUFCTL 0x0702
#define SK_TXQS2_NEXTDESC 0x0704
#define SK_TXQS2_RXBUF_LO 0x0708
#define SK_TXQS2_RXBUF_HI 0x070C
#define SK_TXQS2_RXSTAT 0x0710
#define SK_TXQS2_CSUM_STARTVAL 0x0714
#define SK_TXQS2_CSUM_STARTPOS 0x0718
#define SK_TXQS2_CSUM_WRITEPOS 0x071A
#define SK_TXQS2_CURADDR_LO 0x0720
#define SK_TXQS2_CURADDR_HI 0x0724
#define SK_TXQS2_CURCNT_LO 0x0728
#define SK_TXQS2_CURCNT_HI 0x072C
#define SK_TXQS2_CURBYTES 0x0730
#define SK_TXQS2_BMU_CSR 0x0734
#define SK_TXQS2_WATERMARK 0x0738
#define SK_TXQS2_FLAG 0x073A
#define SK_TXQS2_TEST1 0x073C
#define SK_TXQS2_TEST2 0x0740
#define SK_TXQS2_TEST3 0x0744
/* Block 15 -- TX async queue 2 */
#define SK_TXQA2_BUFCNT 0x0780
#define SK_TXQA2_BUFCTL 0x0782
#define SK_TXQA2_NEXTDESC 0x0784
#define SK_TXQA2_RXBUF_LO 0x0788
#define SK_TXQA2_RXBUF_HI 0x078C
#define SK_TXQA2_RXSTAT 0x0790
#define SK_TXQA2_CSUM_STARTVAL 0x0794
#define SK_TXQA2_CSUM_STARTPOS 0x0798
#define SK_TXQA2_CSUM_WRITEPOS 0x079A
#define SK_TXQA2_CURADDR_LO 0x07A0
#define SK_TXQA2_CURADDR_HI 0x07A4
#define SK_TXQA2_CURCNT_LO 0x07A8
#define SK_TXQA2_CURCNT_HI 0x07AC
#define SK_TXQA2_CURBYTES 0x07B0
#define SK_TXQA2_BMU_CSR 0x07B4
#define SK_TXQA2_WATERMARK 0x07B8
#define SK_TXQA2_FLAG 0x07BA
#define SK_TXQA2_TEST1 0x07BC
#define SK_TXQA2_TEST2 0x07C0
#define SK_TXQA2_TEST3 0x07C4
#define SK_TXBMU_CLR_IRQ_ERR 0x00000001
#define SK_TXBMU_CLR_IRQ_EOF 0x00000002
#define SK_TXBMU_CLR_IRQ_EOB 0x00000004
#define SK_TXBMU_TX_START 0x00000010
#define SK_TXBMU_TX_STOP 0x00000020
#define SK_TXBMU_POLL_OFF 0x00000040
#define SK_TXBMU_POLL_ON 0x00000080
#define SK_TXBMU_TRANSFER_SM_RESET 0x00000100
#define SK_TXBMU_TRANSFER_SM_UNRESET 0x00000200
#define SK_TXBMU_DESCWR_SM_RESET 0x00000400
#define SK_TXBMU_DESCWR_SM_UNRESET 0x00000800
#define SK_TXBMU_DESCRD_SM_RESET 0x00001000
#define SK_TXBMU_DESCRD_SM_UNRESET 0x00002000
#define SK_TXBMU_SUPERVISOR_SM_RESET 0x00004000
#define SK_TXBMU_SUPERVISOR_SM_UNRESET 0x00008000
#define SK_TXBMU_PFI_SM_RESET 0x00010000
#define SK_TXBMU_PFI_SM_UNRESET 0x00020000
#define SK_TXBMU_FIFO_RESET 0x00040000
#define SK_TXBMU_FIFO_UNRESET 0x00080000
#define SK_TXBMU_DESC_RESET 0x00100000
#define SK_TXBMU_DESC_UNRESET 0x00200000
#define SK_TXBMU_SUPERVISOR_IDLE 0x01000000
#define SK_TXBMU_ONLINE \
(SK_TXBMU_TRANSFER_SM_UNRESET|SK_TXBMU_DESCWR_SM_UNRESET| \
SK_TXBMU_DESCRD_SM_UNRESET|SK_TXBMU_SUPERVISOR_SM_UNRESET| \
SK_TXBMU_PFI_SM_UNRESET|SK_TXBMU_FIFO_UNRESET| \
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
SK_TXBMU_DESC_UNRESET|SK_TXBMU_POLL_ON)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_TXBMU_OFFLINE \
(SK_TXBMU_TRANSFER_SM_RESET|SK_TXBMU_DESCWR_SM_RESET| \
SK_TXBMU_DESCRD_SM_RESET|SK_TXBMU_SUPERVISOR_SM_RESET| \
SK_TXBMU_PFI_SM_RESET|SK_TXBMU_FIFO_RESET| \
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
SK_TXBMU_DESC_RESET|SK_TXBMU_POLL_OFF)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 16 -- Receive RAMbuffer 1 */
#define SK_RXRB1_START 0x0800
#define SK_RXRB1_END 0x0804
#define SK_RXRB1_WR_PTR 0x0808
#define SK_RXRB1_RD_PTR 0x080C
#define SK_RXRB1_UTHR_PAUSE 0x0810
#define SK_RXRB1_LTHR_PAUSE 0x0814
#define SK_RXRB1_UTHR_HIPRIO 0x0818
#define SK_RXRB1_UTHR_LOPRIO 0x081C
#define SK_RXRB1_PKTCNT 0x0820
#define SK_RXRB1_LVL 0x0824
#define SK_RXRB1_CTLTST 0x0828
/* Block 17 -- Receive RAMbuffer 2 */
#define SK_RXRB2_START 0x0880
#define SK_RXRB2_END 0x0884
#define SK_RXRB2_WR_PTR 0x0888
#define SK_RXRB2_RD_PTR 0x088C
#define SK_RXRB2_UTHR_PAUSE 0x0890
#define SK_RXRB2_LTHR_PAUSE 0x0894
#define SK_RXRB2_UTHR_HIPRIO 0x0898
#define SK_RXRB2_UTHR_LOPRIO 0x089C
#define SK_RXRB2_PKTCNT 0x08A0
#define SK_RXRB2_LVL 0x08A4
#define SK_RXRB2_CTLTST 0x08A8
/* Block 20 -- Sync. Transmit RAMbuffer 1 */
#define SK_TXRBS1_START 0x0A00
#define SK_TXRBS1_END 0x0A04
#define SK_TXRBS1_WR_PTR 0x0A08
#define SK_TXRBS1_RD_PTR 0x0A0C
#define SK_TXRBS1_PKTCNT 0x0A20
#define SK_TXRBS1_LVL 0x0A24
#define SK_TXRBS1_CTLTST 0x0A28
/* Block 21 -- Async. Transmit RAMbuffer 1 */
#define SK_TXRBA1_START 0x0A80
#define SK_TXRBA1_END 0x0A84
#define SK_TXRBA1_WR_PTR 0x0A88
#define SK_TXRBA1_RD_PTR 0x0A8C
#define SK_TXRBA1_PKTCNT 0x0AA0
#define SK_TXRBA1_LVL 0x0AA4
#define SK_TXRBA1_CTLTST 0x0AA8
/* Block 22 -- Sync. Transmit RAMbuffer 2 */
#define SK_TXRBS2_START 0x0B00
#define SK_TXRBS2_END 0x0B04
#define SK_TXRBS2_WR_PTR 0x0B08
#define SK_TXRBS2_RD_PTR 0x0B0C
#define SK_TXRBS2_PKTCNT 0x0B20
#define SK_TXRBS2_LVL 0x0B24
#define SK_TXRBS2_CTLTST 0x0B28
/* Block 23 -- Async. Transmit RAMbuffer 2 */
#define SK_TXRBA2_START 0x0B80
#define SK_TXRBA2_END 0x0B84
#define SK_TXRBA2_WR_PTR 0x0B88
#define SK_TXRBA2_RD_PTR 0x0B8C
#define SK_TXRBA2_PKTCNT 0x0BA0
#define SK_TXRBA2_LVL 0x0BA4
#define SK_TXRBA2_CTLTST 0x0BA8
#define SK_RBCTL_RESET 0x00000001
#define SK_RBCTL_UNRESET 0x00000002
#define SK_RBCTL_OFF 0x00000004
#define SK_RBCTL_ON 0x00000008
#define SK_RBCTL_STORENFWD_OFF 0x00000010
#define SK_RBCTL_STORENFWD_ON 0x00000020
/* Block 24 -- RX MAC FIFO 1 regisrers and LINK_SYNC counter */
#define SK_RXF1_END 0x0C00
#define SK_RXF1_WPTR 0x0C04
#define SK_RXF1_RPTR 0x0C0C
#define SK_RXF1_PKTCNT 0x0C10
#define SK_RXF1_LVL 0x0C14
#define SK_RXF1_MACCTL 0x0C18
#define SK_RXF1_CTL 0x0C1C
#define SK_RXLED1_CNTINIT 0x0C20
#define SK_RXLED1_COUNTER 0x0C24
#define SK_RXLED1_CTL 0x0C28
#define SK_RXLED1_TST 0x0C29
#define SK_LINK_SYNC1_CINIT 0x0C30
#define SK_LINK_SYNC1_COUNTER 0x0C34
#define SK_LINK_SYNC1_CTL 0x0C38
#define SK_LINK_SYNC1_TST 0x0C39
#define SK_LINKLED1_CTL 0x0C3C
#define SK_FIFO_END 0x3F
/* Receive MAC FIFO 1 (Yukon Only) */
#define SK_RXMF1_END 0x0C40
#define SK_RXMF1_THRESHOLD 0x0C44
#define SK_RXMF1_CTRL_TEST 0x0C48
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_RXMF1_FLUSH_MASK 0x0C4C
#define SK_RXMF1_FLUSH_THRESHOLD 0x0C50
#define SK_RXMF1_WRITE_PTR 0x0C60
#define SK_RXMF1_WRITE_LEVEL 0x0C68
#define SK_RXMF1_READ_PTR 0x0C70
#define SK_RXMF1_READ_LEVEL 0x0C78
2006-05-02 02:12:42 +00:00
/* Receive MAC FIFO 1 Control/Test */
#define SK_RFCTL_WR_PTR_TST_ON 0x00004000 /* Write pointer test on*/
#define SK_RFCTL_WR_PTR_TST_OFF 0x00002000 /* Write pointer test off */
#define SK_RFCTL_WR_PTR_STEP 0x00001000 /* Write pointer increment */
#define SK_RFCTL_RD_PTR_TST_ON 0x00000400 /* Read pointer test on */
#define SK_RFCTL_RD_PTR_TST_OFF 0x00000200 /* Read pointer test off */
#define SK_RFCTL_RD_PTR_STEP 0x00000100 /* Read pointer increment */
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_RFCTL_FIFO_FLUSH_OFF 0x00000080 /* RX FIFO Flsuh mode off */
#define SK_RFCTL_FIFO_FLUSH_ON 0x00000040 /* RX FIFO Flush mode on */
#define SK_RFCTL_RX_FIFO_OVER 0x00000020 /* Clear IRQ RX FIFO Overrun */
#define SK_RFCTL_FRAME_RX_DONE 0x00000010 /* Clear IRQ Frame RX Done */
#define SK_RFCTL_OPERATION_ON 0x00000008 /* Operational mode on */
#define SK_RFCTL_OPERATION_OFF 0x00000004 /* Operational mode off */
#define SK_RFCTL_RESET_CLEAR 0x00000002 /* MAC FIFO Reset Clear */
#define SK_RFCTL_RESET_SET 0x00000001 /* MAC FIFO Reset Set */
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_RFCTL_FIFO_THRESHOLD 0x0a /* flush threshold (default) */
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 25 -- RX MAC FIFO 2 regisrers and LINK_SYNC counter */
#define SK_RXF2_END 0x0C80
#define SK_RXF2_WPTR 0x0C84
#define SK_RXF2_RPTR 0x0C8C
#define SK_RXF2_PKTCNT 0x0C90
#define SK_RXF2_LVL 0x0C94
#define SK_RXF2_MACCTL 0x0C98
#define SK_RXF2_CTL 0x0C9C
#define SK_RXLED2_CNTINIT 0x0CA0
#define SK_RXLED2_COUNTER 0x0CA4
#define SK_RXLED2_CTL 0x0CA8
#define SK_RXLED2_TST 0x0CA9
#define SK_LINK_SYNC2_CINIT 0x0CB0
#define SK_LINK_SYNC2_COUNTER 0x0CB4
#define SK_LINK_SYNC2_CTL 0x0CB8
#define SK_LINK_SYNC2_TST 0x0CB9
#define SK_LINKLED2_CTL 0x0CBC
#define SK_RXMACCTL_CLR_IRQ_NOSTS 0x00000001
#define SK_RXMACCTL_CLR_IRQ_NOTSTAMP 0x00000002
#define SK_RXMACCTL_TSTAMP_OFF 0x00000004
#define SK_RXMACCTL_RSTAMP_ON 0x00000008
#define SK_RXMACCTL_FLUSH_OFF 0x00000010
#define SK_RXMACCTL_FLUSH_ON 0x00000020
#define SK_RXMACCTL_PAUSE_OFF 0x00000040
#define SK_RXMACCTL_PAUSE_ON 0x00000080
#define SK_RXMACCTL_AFULL_OFF 0x00000100
#define SK_RXMACCTL_AFULL_ON 0x00000200
#define SK_RXMACCTL_VALIDTIME_PATCH_OFF 0x00000400
#define SK_RXMACCTL_VALIDTIME_PATCH_ON 0x00000800
#define SK_RXMACCTL_RXRDY_PATCH_OFF 0x00001000
#define SK_RXMACCTL_RXRDY_PATCH_ON 0x00002000
#define SK_RXMACCTL_STS_TIMEO 0x00FF0000
#define SK_RXMACCTL_TSTAMP_TIMEO 0xFF000000
#define SK_RXLEDCTL_ENABLE 0x0001
#define SK_RXLEDCTL_COUNTER_STOP 0x0002
#define SK_RXLEDCTL_COUNTER_START 0x0004
#define SK_LINKLED_OFF 0x0001
#define SK_LINKLED_ON 0x0002
#define SK_LINKLED_LINKSYNC_OFF 0x0004
#define SK_LINKLED_LINKSYNC_ON 0x0008
#define SK_LINKLED_BLINK_OFF 0x0010
#define SK_LINKLED_BLINK_ON 0x0020
/* Block 26 -- TX MAC FIFO 1 regisrers */
#define SK_TXF1_END 0x0D00
#define SK_TXF1_WPTR 0x0D04
#define SK_TXF1_RPTR 0x0D0C
#define SK_TXF1_PKTCNT 0x0D10
#define SK_TXF1_LVL 0x0D14
#define SK_TXF1_MACCTL 0x0D18
#define SK_TXF1_CTL 0x0D1C
#define SK_TXLED1_CNTINIT 0x0D20
#define SK_TXLED1_COUNTER 0x0D24
#define SK_TXLED1_CTL 0x0D28
#define SK_TXLED1_TST 0x0D29
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
/* Transmit MAC FIFO 1 (Yukon Only) */
#define SK_TXMF1_END 0x0D40
#define SK_TXMF1_THRESHOLD 0x0D44
#define SK_TXMF1_CTRL_TEST 0x0D48
#define SK_TXMF1_WRITE_PTR 0x0D60
#define SK_TXMF1_WRITE_SHADOW 0x0D64
#define SK_TXMF1_WRITE_LEVEL 0x0D68
#define SK_TXMF1_READ_PTR 0x0D70
#define SK_TXMF1_RESTART_PTR 0x0D74
#define SK_TXMF1_READ_LEVEL 0x0D78
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
/* Transmit MAC FIFO Control/Test */
#define SK_TFCTL_WR_PTR_TST_ON 0x00004000 /* Write pointer test on*/
#define SK_TFCTL_WR_PTR_TST_OFF 0x00002000 /* Write pointer test off */
#define SK_TFCTL_WR_PTR_STEP 0x00001000 /* Write pointer increment */
#define SK_TFCTL_RD_PTR_TST_ON 0x00000400 /* Read pointer test on */
#define SK_TFCTL_RD_PTR_TST_OFF 0x00000200 /* Read pointer test off */
#define SK_TFCTL_RD_PTR_STEP 0x00000100 /* Read pointer increment */
#define SK_TFCTL_TX_FIFO_UNDER 0x00000040 /* Clear IRQ TX FIFO Under */
#define SK_TFCTL_FRAME_TX_DONE 0x00000020 /* Clear IRQ Frame TX Done */
#define SK_TFCTL_IRQ_PARITY_ER 0x00000010 /* Clear IRQ Parity Error */
#define SK_TFCTL_OPERATION_ON 0x00000008 /* Operational mode on */
#define SK_TFCTL_OPERATION_OFF 0x00000004 /* Operational mode off */
#define SK_TFCTL_RESET_CLEAR 0x00000002 /* MAC FIFO Reset Clear */
#define SK_TFCTL_RESET_SET 0x00000001 /* MAC FIFO Reset Set */
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 27 -- TX MAC FIFO 2 regisrers */
#define SK_TXF2_END 0x0D80
#define SK_TXF2_WPTR 0x0D84
#define SK_TXF2_RPTR 0x0D8C
#define SK_TXF2_PKTCNT 0x0D90
#define SK_TXF2_LVL 0x0D94
#define SK_TXF2_MACCTL 0x0D98
#define SK_TXF2_CTL 0x0D9C
#define SK_TXLED2_CNTINIT 0x0DA0
#define SK_TXLED2_COUNTER 0x0DA4
#define SK_TXLED2_CTL 0x0DA8
#define SK_TXLED2_TST 0x0DA9
#define SK_TXMACCTL_XMAC_RESET 0x00000001
#define SK_TXMACCTL_XMAC_UNRESET 0x00000002
#define SK_TXMACCTL_LOOP_OFF 0x00000004
#define SK_TXMACCTL_LOOP_ON 0x00000008
#define SK_TXMACCTL_FLUSH_OFF 0x00000010
#define SK_TXMACCTL_FLUSH_ON 0x00000020
#define SK_TXMACCTL_WAITEMPTY_OFF 0x00000040
#define SK_TXMACCTL_WAITEMPTY_ON 0x00000080
#define SK_TXMACCTL_AFULL_OFF 0x00000100
#define SK_TXMACCTL_AFULL_ON 0x00000200
#define SK_TXMACCTL_TXRDY_PATCH_OFF 0x00000400
#define SK_TXMACCTL_RXRDY_PATCH_ON 0x00000800
#define SK_TXMACCTL_PKT_RECOVERY_OFF 0x00001000
#define SK_TXMACCTL_PKT_RECOVERY_ON 0x00002000
#define SK_TXMACCTL_CLR_IRQ_PERR 0x00008000
#define SK_TXMACCTL_WAITAFTERFLUSH 0x00010000
#define SK_TXLEDCTL_ENABLE 0x0001
#define SK_TXLEDCTL_COUNTER_STOP 0x0002
#define SK_TXLEDCTL_COUNTER_START 0x0004
#define SK_FIFO_RESET 0x00000001
#define SK_FIFO_UNRESET 0x00000002
#define SK_FIFO_OFF 0x00000004
#define SK_FIFO_ON 0x00000008
/* Block 28 -- Descriptor Poll Timer */
#define SK_DPT_INIT 0x0e00 /* Initial value 24 bits */
#define SK_DPT_TIMER 0x0e04 /* Mul of 78.12MHz clk (24b) */
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_DPT_TIMER_MAX 0x00ffffffff /* 214.75ms at 78.125MHz */
#define SK_DPT_TIMER_CTRL 0x0e08 /* Timer Control 16 bits */
#define SK_DPT_TCTL_STOP 0x0001 /* Stop Timer */
#define SK_DPT_TCTL_START 0x0002 /* Start Timer */
#define SK_DPT_TIMER_TEST 0x0e0a /* Timer Test 16 bits */
#define SK_DPT_TTEST_STEP 0x0001 /* Timer Decrement */
#define SK_DPT_TTEST_OFF 0x0002 /* Test Mode Off */
#define SK_DPT_TTEST_ON 0x0004 /* Test Mode On */
/* Block 29 -- reserved */
/* Block 30 -- GMAC/GPHY Control Registers (Yukon Only)*/
#define SK_GMAC_CTRL 0x0f00 /* GMAC Control Register */
#define SK_GPHY_CTRL 0x0f04 /* GPHY Control Register */
#define SK_GMAC_ISR 0x0f08 /* GMAC Interrupt Source Register */
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_GMAC_IMR 0x0f0c /* GMAC Interrupt Mask Register */
#define SK_LINK_CTRL 0x0f10 /* Link Control Register (LCR) */
#define SK_WOL_CTRL 0x0f20 /* Wake on LAN Control Register */
#define SK_MAC_ADDR_LOW 0x0f24 /* Mack Address Registers LOW */
#define SK_MAC_ADDR_HIGH 0x0f28 /* Mack Address Registers HIGH */
#define SK_PAT_READ_PTR 0x0f2c /* Pattern Read Pointer Register */
#define SK_PAT_LEN_REG0 0x0f30 /* Pattern Length Register 0 */
#define SK_PAT_LEN0 0x0f30 /* Pattern Length 0 */
#define SK_PAT_LEN1 0x0f31 /* Pattern Length 1 */
#define SK_PAT_LEN2 0x0f32 /* Pattern Length 2 */
#define SK_PAT_LEN3 0x0f33 /* Pattern Length 3 */
#define SK_PAT_LEN_REG1 0x0f34 /* Pattern Length Register 1 */
#define SK_PAT_LEN4 0x0f34 /* Pattern Length 4 */
#define SK_PAT_LEN5 0x0f35 /* Pattern Length 5 */
#define SK_PAT_LEN6 0x0f36 /* Pattern Length 6 */
#define SK_PAT_LEN7 0x0f37 /* Pattern Length 7 */
#define SK_PAT_CTR_REG0 0x0f38 /* Pattern Counter Register 0 */
#define SK_PAT_CTR0 0x0f38 /* Pattern Counter 0 */
#define SK_PAT_CTR1 0x0f39 /* Pattern Counter 1 */
#define SK_PAT_CTR2 0x0f3a /* Pattern Counter 2 */
#define SK_PAT_CTR3 0x0f3b /* Pattern Counter 3 */
#define SK_PAT_CTR_REG1 0x0f3c /* Pattern Counter Register 1 */
#define SK_PAT_CTR4 0x0f3c /* Pattern Counter 4 */
#define SK_PAT_CTR5 0x0f3d /* Pattern Counter 5 */
#define SK_PAT_CTR6 0x0f3e /* Pattern Counter 6 */
#define SK_PAT_CTR7 0x0f3f /* Pattern Counter 7 */
#define SK_GMAC_LOOP_ON 0x00000020 /* Loopback mode for testing */
#define SK_GMAC_LOOP_OFF 0x00000010 /* purposes */
#define SK_GMAC_PAUSE_ON 0x00000008 /* enable forward of pause */
#define SK_GMAC_PAUSE_OFF 0x00000004 /* signal to GMAC */
#define SK_GMAC_RESET_CLEAR 0x00000002 /* Clear GMAC Reset */
#define SK_GMAC_RESET_SET 0x00000001 /* Set GMAC Reset */
#define SK_GPHY_SEL_BDT 0x10000000 /* Select Bidirectional xfer */
#define SK_GPHY_INT_POL_HI 0x08000000 /* IRQ Polarity Active */
#define SK_GPHY_75_OHM 0x04000000 /* Use 75 Ohm Termination */
#define SK_GPHY_DIS_FC 0x02000000 /* Disable Auto Fiber/Copper */
#define SK_GPHY_DIS_SLEEP 0x01000000 /* Disable Energy Detect */
#define SK_GPHY_HWCFG_M_3 0x00800000 /* HWCFG_MODE[3] */
#define SK_GPHY_HWCFG_M_2 0x00400000 /* HWCFG_MODE[2] */
#define SK_GPHY_HWCFG_M_1 0x00200000 /* HWCFG_MODE[1] */
#define SK_GPHY_HWCFG_M_0 0x00100000 /* HWCFG_MODE[0] */
#define SK_GPHY_ANEG_0 0x00080000 /* ANEG[0] */
#define SK_GPHY_ENA_XC 0x00040000 /* Enable MDI Crossover */
#define SK_GPHY_DIS_125 0x00020000 /* Disable 125MHz Clock */
#define SK_GPHY_ANEG_3 0x00010000 /* ANEG[3] */
#define SK_GPHY_ANEG_2 0x00008000 /* ANEG[2] */
#define SK_GPHY_ANEG_1 0x00004000 /* ANEG[1] */
#define SK_GPHY_ENA_PAUSE 0x00002000 /* Enable Pause */
#define SK_GPHY_PHYADDR_4 0x00001000 /* Bit 4 of Phy Addr */
#define SK_GPHY_PHYADDR_3 0x00000800 /* Bit 3 of Phy Addr */
#define SK_GPHY_PHYADDR_2 0x00000400 /* Bit 2 of Phy Addr */
#define SK_GPHY_PHYADDR_1 0x00000200 /* Bit 1 of Phy Addr */
#define SK_GPHY_PHYADDR_0 0x00000100 /* Bit 0 of Phy Addr */
#define SK_GPHY_RESET_CLEAR 0x00000002 /* Clear GPHY Reset */
#define SK_GPHY_RESET_SET 0x00000001 /* Set GPHY Reset */
#define SK_GPHY_COPPER (SK_GPHY_HWCFG_M_0 | SK_GPHY_HWCFG_M_1 | \
SK_GPHY_HWCFG_M_2 | SK_GPHY_HWCFG_M_3 )
#define SK_GPHY_FIBER (SK_GPHY_HWCFG_M_0 | SK_GPHY_HWCFG_M_1 | \
SK_GPHY_HWCFG_M_2 )
#define SK_GPHY_ANEG_ALL (SK_GPHY_ANEG_0 | SK_GPHY_ANEG_1 | \
SK_GPHY_ANEG_2 | SK_GPHY_ANEG_3 )
#define SK_GMAC_INT_TX_OFLOW 0x20 /* Transmit Counter Overflow */
#define SK_GMAC_INT_RX_OFLOW 0x10 /* Receiver Overflow */
#define SK_GMAC_INT_TX_UNDER 0x08 /* Transmit FIFO Underrun */
#define SK_GMAC_INT_TX_DONE 0x04 /* Transmit Complete */
#define SK_GMAC_INT_RX_OVER 0x02 /* Receive FIFO Overrun */
#define SK_GMAC_INT_RX_DONE 0x01 /* Receive Complete */
#define SK_LINK_RESET_CLEAR 0x0002 /* Link Reset Clear */
#define SK_LINK_RESET_SET 0x0001 /* Link Reset Set */
/* Block 31 -- reserved */
/* Block 32-33 -- Pattern Ram */
#define SK_WOL_PRAM 0x1000
/* Block 0x22 - 0x3f -- reserved */
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 0x40 to 0x4F -- XMAC 1 registers */
#define SK_XMAC1_BASE 0x2000
/* Block 0x50 to 0x5F -- MARV 1 registers */
#define SK_MARV1_BASE 0x2800
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Block 0x60 to 0x6F -- XMAC 2 registers */
#define SK_XMAC2_BASE 0x3000
/* Block 0x70 to 0x7F -- MARV 2 registers */
#define SK_MARV2_BASE 0x3800
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Compute relative offset of an XMAC register in the XMAC window(s). */
#define SK_XMAC_REG(sc, reg) (((reg) * 2) + SK_XMAC1_BASE + \
(((sc)->sk_port) * (SK_XMAC2_BASE - SK_XMAC1_BASE)))
#if 0
#define SK_XM_READ_4(sc, reg) \
((sk_win_read_2(sc->sk_softc, \
SK_XMAC_REG(sc, reg)) & 0xFFFF) | \
((sk_win_read_2(sc->sk_softc, \
SK_XMAC_REG(sc, reg + 2)) & 0xFFFF) << 16))
#define SK_XM_WRITE_4(sc, reg, val) \
sk_win_write_2(sc->sk_softc, SK_XMAC_REG(sc, reg), \
((val) & 0xFFFF)); \
sk_win_write_2(sc->sk_softc, SK_XMAC_REG(sc, reg + 2), \
((val) >> 16) & 0xFFFF)
#else
#define SK_XM_READ_4(sc, reg) \
sk_win_read_4(sc->sk_softc, SK_XMAC_REG(sc, reg))
#define SK_XM_WRITE_4(sc, reg, val) \
sk_win_write_4(sc->sk_softc, SK_XMAC_REG(sc, reg), (val))
#endif
#define SK_XM_READ_2(sc, reg) \
sk_win_read_2(sc->sk_softc, SK_XMAC_REG(sc, reg))
#define SK_XM_WRITE_2(sc, reg, val) \
sk_win_write_2(sc->sk_softc, SK_XMAC_REG(sc, reg), val)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_XM_SETBIT_4(sc, reg, x) \
SK_XM_WRITE_4(sc, reg, (SK_XM_READ_4(sc, reg)) | (x))
#define SK_XM_CLRBIT_4(sc, reg, x) \
SK_XM_WRITE_4(sc, reg, (SK_XM_READ_4(sc, reg)) & ~(x))
#define SK_XM_SETBIT_2(sc, reg, x) \
SK_XM_WRITE_2(sc, reg, (SK_XM_READ_2(sc, reg)) | (x))
#define SK_XM_CLRBIT_2(sc, reg, x) \
SK_XM_WRITE_2(sc, reg, (SK_XM_READ_2(sc, reg)) & ~(x))
/* Compute relative offset of an MARV register in the MARV window(s). */
#define SK_YU_REG(sc, reg) \
((reg) + SK_MARV1_BASE + \
(((sc)->sk_port) * (SK_MARV2_BASE - SK_MARV1_BASE)))
#define SK_YU_READ_4(sc, reg) \
sk_win_read_4((sc)->sk_softc, SK_YU_REG((sc), (reg)))
#define SK_YU_READ_2(sc, reg) \
sk_win_read_2((sc)->sk_softc, SK_YU_REG((sc), (reg)))
#define SK_YU_WRITE_4(sc, reg, val) \
sk_win_write_4((sc)->sk_softc, SK_YU_REG((sc), (reg)), (val))
#define SK_YU_WRITE_2(sc, reg, val) \
sk_win_write_2((sc)->sk_softc, SK_YU_REG((sc), (reg)), (val))
#define SK_YU_SETBIT_4(sc, reg, x) \
SK_YU_WRITE_4(sc, reg, (SK_YU_READ_4(sc, reg)) | (x))
#define SK_YU_CLRBIT_4(sc, reg, x) \
SK_YU_WRITE_4(sc, reg, (SK_YU_READ_4(sc, reg)) & ~(x))
#define SK_YU_SETBIT_2(sc, reg, x) \
SK_YU_WRITE_2(sc, reg, (SK_YU_READ_2(sc, reg)) | (x))
#define SK_YU_CLRBIT_2(sc, reg, x) \
SK_YU_WRITE_2(sc, reg, (SK_YU_READ_2(sc, reg)) & ~(x))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* The default FIFO threshold on the XMAC II is 4 bytes. On
* dual port NICs, this often leads to transmit underruns, so we
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
* bump the threshold a little.
*/
#define SK_XM_TX_FIFOTHRESH 512
#define SK_PCI_VENDOR_ID 0x0000
#define SK_PCI_DEVICE_ID 0x0002
#define SK_PCI_COMMAND 0x0004
#define SK_PCI_STATUS 0x0006
#define SK_PCI_REVID 0x0008
#define SK_PCI_CLASSCODE 0x0009
#define SK_PCI_CACHELEN 0x000C
#define SK_PCI_LATENCY_TIMER 0x000D
#define SK_PCI_HEADER_TYPE 0x000E
#define SK_PCI_LOMEM 0x0010
#define SK_PCI_LOIO 0x0014
#define SK_PCI_SUBVEN_ID 0x002C
#define SK_PCI_SYBSYS_ID 0x002E
#define SK_PCI_BIOSROM 0x0030
#define SK_PCI_INTLINE 0x003C
#define SK_PCI_INTPIN 0x003D
#define SK_PCI_MINGNT 0x003E
#define SK_PCI_MINLAT 0x003F
/* device specific PCI registers */
#define SK_PCI_OURREG1 0x0040
#define SK_PCI_OURREG2 0x0044
#define SK_PCI_CAPID 0x0048 /* 8 bits */
#define SK_PCI_NEXTPTR 0x0049 /* 8 bits */
#define SK_PCI_PWRMGMTCAP 0x004A /* 16 bits */
#define SK_PCI_PWRMGMTCTRL 0x004C /* 16 bits */
#define SK_PCI_PME_EVENT 0x004F
#define SK_PSTATE_MASK 0x0003
#define SK_PSTATE_D0 0x0000
#define SK_PSTATE_D1 0x0001
#define SK_PSTATE_D2 0x0002
#define SK_PSTATE_D3 0x0003
#define SK_PME_EN 0x0010
#define SK_PME_STATUS 0x8000
#define CSR_WRITE_4(sc, reg, val) \
bus_write_4((sc)->sk_res[0], (reg), (val))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define CSR_WRITE_2(sc, reg, val) \
bus_write_2((sc)->sk_res[0], (reg), (val))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define CSR_WRITE_1(sc, reg, val) \
bus_write_1((sc)->sk_res[0], (reg), (val))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define CSR_READ_4(sc, reg) \
bus_read_4((sc)->sk_res[0], (reg))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define CSR_READ_2(sc, reg) \
bus_read_2((sc)->sk_res[0], (reg))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define CSR_READ_1(sc, reg) \
bus_read_1((sc)->sk_res[0], (reg))
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
struct sk_type {
u_int16_t sk_vid;
u_int16_t sk_did;
char *sk_name;
};
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_ADDR_LO(x) ((u_int64_t) (x) & 0xffffffff)
#define SK_ADDR_HI(x) ((u_int64_t) (x) >> 32)
#define SK_RING_ALIGN 64
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* RX queue descriptor data structure */
struct sk_rx_desc {
u_int32_t sk_ctl;
u_int32_t sk_next;
u_int32_t sk_data_lo;
u_int32_t sk_data_hi;
u_int32_t sk_xmac_rxstat;
u_int32_t sk_timestamp;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
u_int32_t sk_csum;
u_int32_t sk_csum_start;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
#define SK_OPCODE_DEFAULT 0x00550000
#define SK_OPCODE_CSUM 0x00560000
#define SK_RXCTL_LEN 0x0000FFFF
#define SK_RXCTL_OPCODE 0x00FF0000
#define SK_RXCTL_TSTAMP_VALID 0x01000000
#define SK_RXCTL_STATUS_VALID 0x02000000
#define SK_RXCTL_DEV0 0x04000000
#define SK_RXCTL_EOF_INTR 0x08000000
#define SK_RXCTL_EOB_INTR 0x10000000
#define SK_RXCTL_LASTFRAG 0x20000000
#define SK_RXCTL_FIRSTFRAG 0x40000000
#define SK_RXCTL_OWN 0x80000000
#define SK_RXSTAT \
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
(SK_RXCTL_EOF_INTR|SK_RXCTL_LASTFRAG|SK_RXCTL_FIRSTFRAG|SK_RXCTL_OWN)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
struct sk_tx_desc {
u_int32_t sk_ctl;
u_int32_t sk_next;
u_int32_t sk_data_lo;
u_int32_t sk_data_hi;
u_int32_t sk_xmac_txstat;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
u_int32_t sk_csum_startval;
u_int32_t sk_csum_start;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
u_int32_t sk_rsvd1;
};
#define SK_TXCTL_LEN 0x0000FFFF
#define SK_TXCTL_OPCODE 0x00FF0000
#define SK_TXCTL_SW 0x01000000
#define SK_TXCTL_NOCRC 0x02000000
#define SK_TXCTL_STORENFWD 0x04000000
#define SK_TXCTL_EOF_INTR 0x08000000
#define SK_TXCTL_EOB_INTR 0x10000000
#define SK_TXCTL_LASTFRAG 0x20000000
#define SK_TXCTL_FIRSTFRAG 0x40000000
#define SK_TXCTL_OWN 0x80000000
#define SK_TXSTAT \
(SK_OPCODE_DEFAULT|SK_TXCTL_EOF_INTR|SK_TXCTL_LASTFRAG|SK_TXCTL_OWN)
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_RXBYTES(x) ((x) & 0x0000FFFF)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_TXBYTES SK_RXBYTES
#define SK_TX_RING_CNT 512
#define SK_RX_RING_CNT 256
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_JUMBO_RX_RING_CNT 256
#define SK_MAXTXSEGS 32
#define SK_MAXRXSEGS 32
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/*
* Jumbo buffer stuff. Note that we must allocate more jumbo
* buffers than there are descriptors in the receive ring. This
* is because we don't know how long it will take for a packet
* to be released after we hand it off to the upper protocol
* layers. To be safe, we allocate 1.5 times the number of
* receive descriptors.
*/
#define SK_JUMBO_FRAMELEN 9018
#define SK_JUMBO_MTU (SK_JUMBO_FRAMELEN-ETHER_HDR_LEN-ETHER_CRC_LEN)
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_MAX_FRAMELEN \
(ETHER_MAX_LEN + ETHER_VLAN_ENCAP_LEN - ETHER_CRC_LEN)
#define SK_MIN_FRAMELEN (ETHER_MIN_LEN - ETHER_CRC_LEN)
#define SK_JSLOTS ((SK_RX_RING_CNT * 3) / 2)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_JRAWLEN (SK_JUMBO_FRAMELEN + ETHER_ALIGN)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_JLEN (SK_JRAWLEN + (sizeof(u_int64_t) - \
(SK_JRAWLEN % sizeof(u_int64_t))))
#define SK_JPAGESZ PAGE_SIZE
#define SK_RESID (SK_JPAGESZ - (SK_JLEN * SK_JSLOTS) % SK_JPAGESZ)
#define SK_JMEM ((SK_JLEN * SK_JSLOTS) + SK_RESID)
struct sk_jpool_entry {
int slot;
SLIST_ENTRY(sk_jpool_entry) jpool_entries;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
struct sk_txdesc {
struct mbuf *tx_m;
bus_dmamap_t tx_dmamap;
STAILQ_ENTRY(sk_txdesc) tx_q;
};
STAILQ_HEAD(sk_txdq, sk_txdesc);
struct sk_rxdesc {
struct mbuf *rx_m;
bus_dmamap_t rx_dmamap;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
struct sk_chain_data {
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
bus_dma_tag_t sk_parent_tag;
bus_dma_tag_t sk_tx_tag;
struct sk_txdesc sk_txdesc[SK_TX_RING_CNT];
struct sk_txdq sk_txfreeq;
struct sk_txdq sk_txbusyq;
bus_dma_tag_t sk_rx_tag;
struct sk_rxdesc sk_rxdesc[SK_RX_RING_CNT];
bus_dma_tag_t sk_tx_ring_tag;
bus_dma_tag_t sk_rx_ring_tag;
bus_dmamap_t sk_tx_ring_map;
bus_dmamap_t sk_rx_ring_map;
bus_dmamap_t sk_rx_sparemap;
bus_dma_tag_t sk_jumbo_rx_tag;
bus_dma_tag_t sk_jumbo_tag;
bus_dmamap_t sk_jumbo_map;
bus_dma_tag_t sk_jumbo_mtag;
caddr_t sk_jslots[SK_JSLOTS];
struct sk_rxdesc sk_jumbo_rxdesc[SK_JUMBO_RX_RING_CNT];
bus_dma_tag_t sk_jumbo_rx_ring_tag;
bus_dmamap_t sk_jumbo_rx_ring_map;
bus_dmamap_t sk_jumbo_rx_sparemap;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
int sk_tx_prod;
int sk_tx_cons;
int sk_tx_cnt;
int sk_rx_cons;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
int sk_jumbo_rx_cons;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
struct sk_ring_data {
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
struct sk_tx_desc *sk_tx_ring;
bus_addr_t sk_tx_ring_paddr;
struct sk_rx_desc *sk_rx_ring;
bus_addr_t sk_rx_ring_paddr;
struct sk_rx_desc *sk_jumbo_rx_ring;
bus_addr_t sk_jumbo_rx_ring_paddr;
void *sk_jumbo_buf;
bus_addr_t sk_jumbo_buf_paddr;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_TX_RING_ADDR(sc, i) \
((sc)->sk_rdata.sk_tx_ring_paddr + sizeof(struct sk_tx_desc) * (i))
#define SK_RX_RING_ADDR(sc, i) \
((sc)->sk_rdata.sk_rx_ring_paddr + sizeof(struct sk_rx_desc) * (i))
#define SK_JUMBO_RX_RING_ADDR(sc, i) \
((sc)->sk_rdata.sk_jumbo_rx_ring_paddr + sizeof(struct sk_rx_desc) * (i))
#define SK_TX_RING_SZ \
(sizeof(struct sk_tx_desc) * SK_TX_RING_CNT)
#define SK_RX_RING_SZ \
(sizeof(struct sk_rx_desc) * SK_RX_RING_CNT)
#define SK_JUMBO_RX_RING_SZ \
(sizeof(struct sk_rx_desc) * SK_JUMBO_RX_RING_CNT)
struct sk_bcom_hack {
int reg;
int val;
};
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_INC(x, y) (x) = (x + 1) % y
/* Forward decl. */
struct sk_if_softc;
/* Softc for the GEnesis controller. */
struct sk_softc {
struct resource *sk_res[2]; /* I/O and IRQ resources */
struct resource_spec *sk_res_spec;
void *sk_intrhand; /* irq handler handle */
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
device_t sk_dev;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
u_int8_t sk_type;
u_int8_t sk_rev;
u_int8_t spare;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
u_int32_t sk_rboff; /* RAMbuffer offset */
u_int32_t sk_ramsize; /* amount of RAM on NIC */
u_int32_t sk_pmd; /* physical media type */
u_int32_t sk_coppertype;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
u_int32_t sk_intrmask;
int sk_int_mod;
int sk_int_ticks;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
int sk_suspended;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
struct sk_if_softc *sk_if[2];
device_t sk_devs[2];
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
struct mtx sk_mii_mtx;
struct mtx sk_mtx;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
Change and clean the mutex lock interface. mtx_enter(lock, type) becomes: mtx_lock(lock) for sleep locks (MTX_DEF-initialized locks) mtx_lock_spin(lock) for spin locks (MTX_SPIN-initialized) similarily, for releasing a lock, we now have: mtx_unlock(lock) for MTX_DEF and mtx_unlock_spin(lock) for MTX_SPIN. We change the caller interface for the two different types of locks because the semantics are entirely different for each case, and this makes it explicitly clear and, at the same time, it rids us of the extra `type' argument. The enter->lock and exit->unlock change has been made with the idea that we're "locking data" and not "entering locked code" in mind. Further, remove all additional "flags" previously passed to the lock acquire/release routines with the exception of two: MTX_QUIET and MTX_NOSWITCH The functionality of these flags is preserved and they can be passed to the lock/unlock routines by calling the corresponding wrappers: mtx_{lock, unlock}_flags(lock, flag(s)) and mtx_{lock, unlock}_spin_flags(lock, flag(s)) for MTX_DEF and MTX_SPIN locks, respectively. Re-inline some lock acq/rel code; in the sleep lock case, we only inline the _obtain_lock()s in order to ensure that the inlined code fits into a cache line. In the spin lock case, we inline recursion and actually only perform a function call if we need to spin. This change has been made with the idea that we generally tend to avoid spin locks and that also the spin locks that we do have and are heavily used (i.e. sched_lock) do recurse, and therefore in an effort to reduce function call overhead for some architectures (such as alpha), we inline recursion for this case. Create a new malloc type for the witness code and retire from using the M_DEV type. The new type is called M_WITNESS and is only declared if WITNESS is enabled. Begin cleaning up some machdep/mutex.h code - specifically updated the "optimized" inlined code in alpha/mutex.h and wrote MTX_LOCK_SPIN and MTX_UNLOCK_SPIN asm macros for the i386/mutex.h as we presently need those. Finally, caught up to the interface changes in all sys code. Contributors: jake, jhb, jasone (in no particular order)
2001-02-09 06:11:45 +00:00
#define SK_LOCK(_sc) mtx_lock(&(_sc)->sk_mtx)
#define SK_UNLOCK(_sc) mtx_unlock(&(_sc)->sk_mtx)
#define SK_LOCK_ASSERT(_sc) mtx_assert(&(_sc)->sk_mtx, MA_OWNED)
#define SK_IF_LOCK(_sc) SK_LOCK((_sc)->sk_softc)
#define SK_IF_UNLOCK(_sc) SK_UNLOCK((_sc)->sk_softc)
#define SK_IF_LOCK_ASSERT(_sc) SK_LOCK_ASSERT((_sc)->sk_softc)
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
#define SK_IF_MII_LOCK(_sc) mtx_lock(&(_sc)->sk_softc->sk_mii_mtx)
#define SK_IF_MII_UNLOCK(_sc) mtx_unlock(&(_sc)->sk_softc->sk_mii_mtx)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
/* Softc for each logical interface */
struct sk_if_softc {
struct ifnet *sk_ifp; /* interface info */
device_t sk_miibus;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
device_t sk_if_dev;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
u_int8_t sk_port; /* port # on controller */
u_int8_t sk_xmac_rev; /* XMAC chip rev (B2 or C1) */
u_int32_t sk_rx_ramstart;
u_int32_t sk_rx_ramend;
u_int32_t sk_tx_ramstart;
u_int32_t sk_tx_ramend;
int sk_phytype;
int sk_phyaddr;
int sk_link;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
struct callout sk_tick_ch;
struct callout sk_watchdog_ch;
int sk_watchdog_timer;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
struct sk_chain_data sk_cdata;
Bring busdmafied sk(4) to all architectures. - MPSAFE. No more recursive lock required. - bus_dma(9) conversion. I think it should work on all architectures. - optimized Rx handler for each normal and jumbo frames. Previously sk(4) used jumbo frame management code to handle normal sized frames. As the handler needs an additional lock to protect jumbo frame management structure from races, it used two lock operations for each received packet. Now sk(4) uses single lock operation for normal frame.(Jumbo frame still needs two lock operations as before.) The hardware supports DMA scatter operations for Rx descriptors such that it's possible to take advantagee of m_cljget(9) for jumbo frames. However, due to a unknown reasons it resulted in poor performance on sparc64. So I dropped m_cljget(9) approach. This should be revisited since it would reduce one lock operation for jumbo frame handling. - Tx TCP/Rx IP checksum offload support. According to the data sheet of SK-NET GENESIS the hardware supports Rx IP/TCP/UDP offload. But I couldn't make it work on my Yukon hardware. So Rx TCP/UDP was disabled at the moment. It seems that newer Yukon chips can support Tx UDP checksum offload too. But I need more documentation first. - Added more wait time in reading VPD data. It seems that ASUS LOM takes a very long time to respond VPD read signal. - Added an additional lock for MII register access callbacks. - Added more strict received packet validation routine. Previously it passed corrupted packets to upper layers under certain conditions. - A new function sk_yukon_tick() to handle auto-negotiation properly. - Interrupt handler now checks shared interrupt source and protects the interrupt handler from NULL pointer dereference which was caused by odd status word value. The status word can returns 0xffffffff if cable is unplugged while Rx/Tx/auto-negotiation is in progress. - suspend/resume support(not tested). - Added Rx/Tx FIFO flush routine for Yukon - Activate Tx descriptor poll timer in order to protect possible loss of SK_TXBMU_TX_START command. Previously the driver continuously issued SK_TXBMU_TX_START when it notices pending Tx descriptors not processed yet in interrupt handler. That approach would add additional PCI write access overhead under high Tx load situations and it might fail if the first SK_TXBMU_TX_START was lost and no interrupt is generated from the first SK_TXBMU_TX_START command. - s/printf/if_printf/, s/printf/device_printf/, Axe sk_unit in softc. - Setting multicast/station address is now safe on strict-alignment architectures. - Fix long standing bug in VLAN header length setup. - Added/corrected register definitions for Yukon. (Register information from Linux skge driver.) - Added Rx status definition for Marvell Yukon/XaQti XMAC. (Rx status register information from Linux skge driver.) - Update if_oerrors if we encounter watchdog error. - callout(9) conversion Special thanks to jkim who let me know RX status differences between Yukon and XaQti XMAC. It seems that there is still occasional watchdog timeout error but I couldn't reproduce it and need more information to analyze it from users. Tested by: bz(amd64), me(i386, sparc64), current ML Frank Behrens frank ! pinky ( sax $ de
2006-04-27 05:59:09 +00:00
struct sk_ring_data sk_rdata;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
struct sk_softc *sk_softc; /* parent controller */
int sk_tx_bmu; /* TX BMU register */
int sk_if_flags;
SLIST_HEAD(__sk_jfreehead, sk_jpool_entry) sk_jfree_listhead;
SLIST_HEAD(__sk_jinusehead, sk_jpool_entry) sk_jinuse_listhead;
struct mtx sk_jlist_mtx;
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
};
#define SK_JLIST_LOCK(_sc) mtx_lock(&(_sc)->sk_jlist_mtx)
#define SK_JLIST_UNLOCK(_sc) mtx_unlock(&(_sc)->sk_jlist_mtx)
This commit adds driver support for the SysKonnect SK-984x series gigabit ethernet adapters. This includes two single port cards (single mode and multimode fiber) and two dual port cards (also single mode and multimode fiber). SysKonnect is currently the only vendor with a dual port gigabit ethernet NIC. The ports on dual port adapters are treated as separate network interfaces. Thus, if you have an SK-9844 dual port SX card, you should have both sk0 and sk1 interfaces attached. Dual port cards are implemented using two XMAC II chips connected to a single SysKonnect GEnesis controller. Hence, dual port cards are really one PCI device, as opposed to two separate PCI devices connected through a PCI to PCI bridge. Note that SysKonnect's drivers use the two ports for failover purposes rather that as two separate interfaces, plus they don't support jumbo frames. This applies to their Linux driver too. :) Support is provided for hardware multicast filtering, BPF and jumbo frames. The SysKonnect cards support TCP checksum offload however this feature is not currently enabled (hopefully it will be once we get checksum offload support). There are still a few things that need to be implemeted, like the ability to communicate with the on-board LM80 voltage/temperature monitor, but I wanted to get the driver under CVS control and into -current so people could bang on it. A big thanks for SysKonnect for making all their programming info for these cards (and for their FDDI and token ring cards) available without NDA (see www.syskonnect.com).
1999-07-09 04:30:09 +00:00
#define SK_TIMEOUT 1000