2 * Copyright (c) 1997, 1998-2003
3 * Bill Paul <wpaul@windriver.com>. All rights reserved.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * 3. All advertising materials mentioning features or use of this software
14 * must display the following acknowledgement:
15 * This product includes software developed by Bill Paul.
16 * 4. Neither the name of the author nor the names of any co-contributors
17 * may be used to endorse or promote products derived from this software
18 * without specific prior written permission.
20 * THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED. IN NO EVENT SHALL Bill Paul OR THE VOICES IN HIS HEAD
24 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
25 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
26 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
27 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
29 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
30 * THE POSSIBILITY OF SUCH DAMAGE.
33 #include <sys/cdefs.h>
34 __FBSDID("$FreeBSD$");
37 * RealTek 8139C+/8169/8169S/8110S/8168/8111/8101E PCI NIC driver
39 * Written by Bill Paul <wpaul@windriver.com>
40 * Senior Networking Software Engineer
45 * This driver is designed to support RealTek's next generation of
46 * 10/100 and 10/100/1000 PCI ethernet controllers. There are currently
47 * seven devices in this family: the RTL8139C+, the RTL8169, the RTL8169S,
48 * RTL8110S, the RTL8168, the RTL8111 and the RTL8101E.
50 * The 8139C+ is a 10/100 ethernet chip. It is backwards compatible
51 * with the older 8139 family, however it also supports a special
52 * C+ mode of operation that provides several new performance enhancing
53 * features. These include:
55 * o Descriptor based DMA mechanism. Each descriptor represents
56 * a single packet fragment. Data buffers may be aligned on
61 * o TCP/IP checksum offload for both RX and TX
63 * o High and normal priority transmit DMA rings
65 * o VLAN tag insertion and extraction
67 * o TCP large send (segmentation offload)
69 * Like the 8139, the 8139C+ also has a built-in 10/100 PHY. The C+
70 * programming API is fairly straightforward. The RX filtering, EEPROM
71 * access and PHY access is the same as it is on the older 8139 series
74 * The 8169 is a 64-bit 10/100/1000 gigabit ethernet MAC. It has almost the
75 * same programming API and feature set as the 8139C+ with the following
76 * differences and additions:
82 * o GMII and TBI ports/registers for interfacing with copper
85 * o RX and TX DMA rings can have up to 1024 descriptors
86 * (the 8139C+ allows a maximum of 64)
88 * o Slight differences in register layout from the 8139C+
90 * The TX start and timer interrupt registers are at different locations
91 * on the 8169 than they are on the 8139C+. Also, the status word in the
92 * RX descriptor has a slightly different bit layout. The 8169 does not
93 * have a built-in PHY. Most reference boards use a Marvell 88E1000 'Alaska'
96 * The 8169S/8110S 10/100/1000 devices have built-in copper gigE PHYs
97 * (the 'S' stands for 'single-chip'). These devices have the same
98 * programming API as the older 8169, but also have some vendor-specific
99 * registers for the on-board PHY. The 8110S is a LAN-on-motherboard
100 * part designed to be pin-compatible with the RealTek 8100 10/100 chip.
102 * This driver takes advantage of the RX and TX checksum offload and
103 * VLAN tag insertion/extraction features. It also implements TX
104 * interrupt moderation using the timer interrupt registers, which
105 * significantly reduces TX interrupt load. There is also support
106 * for jumbo frames, however the 8169/8169S/8110S can not transmit
107 * jumbo frames larger than 7440, so the max MTU possible with this
108 * driver is 7422 bytes.
111 #ifdef HAVE_KERNEL_OPTION_HEADERS
112 #include "opt_device_polling.h"
115 #include <sys/param.h>
116 #include <sys/endian.h>
117 #include <sys/systm.h>
118 #include <sys/sockio.h>
119 #include <sys/mbuf.h>
120 #include <sys/malloc.h>
121 #include <sys/module.h>
122 #include <sys/kernel.h>
123 #include <sys/socket.h>
124 #include <sys/lock.h>
125 #include <sys/mutex.h>
126 #include <sys/taskqueue.h>
129 #include <net/if_arp.h>
130 #include <net/ethernet.h>
131 #include <net/if_dl.h>
132 #include <net/if_media.h>
133 #include <net/if_types.h>
134 #include <net/if_vlan_var.h>
138 #include <machine/bus.h>
139 #include <machine/resource.h>
141 #include <sys/rman.h>
143 #include <dev/mii/mii.h>
144 #include <dev/mii/miivar.h>
146 #include <dev/pci/pcireg.h>
147 #include <dev/pci/pcivar.h>
149 #include <pci/if_rlreg.h>
151 MODULE_DEPEND(re, pci, 1, 1, 1);
152 MODULE_DEPEND(re, ether, 1, 1, 1);
153 MODULE_DEPEND(re, miibus, 1, 1, 1);
155 /* "device miibus" required. See GENERIC if you get errors here. */
156 #include "miibus_if.h"
159 static int msi_disable = 0;
160 TUNABLE_INT("hw.re.msi_disable", &msi_disable);
161 static int prefer_iomap = 0;
162 TUNABLE_INT("hw.re.prefer_iomap", &prefer_iomap);
164 #define RE_CSUM_FEATURES (CSUM_IP | CSUM_TCP | CSUM_UDP)
167 * Various supported device vendors/types and their names.
169 static struct rl_type re_devs[] = {
170 { DLINK_VENDORID, DLINK_DEVICEID_528T, 0,
171 "D-Link DGE-528(T) Gigabit Ethernet Adapter" },
172 { RT_VENDORID, RT_DEVICEID_8139, 0,
173 "RealTek 8139C+ 10/100BaseTX" },
174 { RT_VENDORID, RT_DEVICEID_8101E, 0,
175 "RealTek 8101E/8102E/8102EL PCIe 10/100baseTX" },
176 { RT_VENDORID, RT_DEVICEID_8168, 0,
177 "RealTek 8168/8168B/8168C/8168CP/8168D/8111B/8111C/8111CP PCIe "
178 "Gigabit Ethernet" },
179 { RT_VENDORID, RT_DEVICEID_8169, 0,
180 "RealTek 8169/8169S/8169SB(L)/8110S/8110SB(L) Gigabit Ethernet" },
181 { RT_VENDORID, RT_DEVICEID_8169SC, 0,
182 "RealTek 8169SC/8110SC Single-chip Gigabit Ethernet" },
183 { COREGA_VENDORID, COREGA_DEVICEID_CGLAPCIGT, 0,
184 "Corega CG-LAPCIGT (RTL8169S) Gigabit Ethernet" },
185 { LINKSYS_VENDORID, LINKSYS_DEVICEID_EG1032, 0,
186 "Linksys EG1032 (RTL8169S) Gigabit Ethernet" },
187 { USR_VENDORID, USR_DEVICEID_997902, 0,
188 "US Robotics 997902 (RTL8169S) Gigabit Ethernet" }
191 static struct rl_hwrev re_hwrevs[] = {
192 { RL_HWREV_8139, RL_8139, "" },
193 { RL_HWREV_8139A, RL_8139, "A" },
194 { RL_HWREV_8139AG, RL_8139, "A-G" },
195 { RL_HWREV_8139B, RL_8139, "B" },
196 { RL_HWREV_8130, RL_8139, "8130" },
197 { RL_HWREV_8139C, RL_8139, "C" },
198 { RL_HWREV_8139D, RL_8139, "8139D/8100B/8100C" },
199 { RL_HWREV_8139CPLUS, RL_8139CPLUS, "C+"},
200 { RL_HWREV_8168_SPIN1, RL_8169, "8168"},
201 { RL_HWREV_8169, RL_8169, "8169"},
202 { RL_HWREV_8169S, RL_8169, "8169S"},
203 { RL_HWREV_8110S, RL_8169, "8110S"},
204 { RL_HWREV_8169_8110SB, RL_8169, "8169SB/8110SB"},
205 { RL_HWREV_8169_8110SC, RL_8169, "8169SC/8110SC"},
206 { RL_HWREV_8169_8110SBL, RL_8169, "8169SBL/8110SBL"},
207 { RL_HWREV_8169_8110SCE, RL_8169, "8169SC/8110SC"},
208 { RL_HWREV_8100, RL_8139, "8100"},
209 { RL_HWREV_8101, RL_8139, "8101"},
210 { RL_HWREV_8100E, RL_8169, "8100E"},
211 { RL_HWREV_8101E, RL_8169, "8101E"},
212 { RL_HWREV_8102E, RL_8169, "8102E"},
213 { RL_HWREV_8102EL, RL_8169, "8102EL"},
214 { RL_HWREV_8168_SPIN2, RL_8169, "8168"},
215 { RL_HWREV_8168_SPIN3, RL_8169, "8168"},
216 { RL_HWREV_8168C, RL_8169, "8168C/8111C"},
217 { RL_HWREV_8168C_SPIN2, RL_8169, "8168C/8111C"},
218 { RL_HWREV_8168CP, RL_8169, "8168CP/8111CP"},
219 { RL_HWREV_8168D, RL_8169, "8168D"},
223 static int re_probe (device_t);
224 static int re_attach (device_t);
225 static int re_detach (device_t);
227 static int re_encap (struct rl_softc *, struct mbuf **);
229 static void re_dma_map_addr (void *, bus_dma_segment_t *, int, int);
230 static int re_allocmem (device_t, struct rl_softc *);
231 static __inline void re_discard_rxbuf
232 (struct rl_softc *, int);
233 static int re_newbuf (struct rl_softc *, int);
234 static int re_rx_list_init (struct rl_softc *);
235 static int re_tx_list_init (struct rl_softc *);
237 static __inline void re_fixup_rx
240 static int re_rxeof (struct rl_softc *);
241 static void re_txeof (struct rl_softc *);
242 #ifdef DEVICE_POLLING
243 static void re_poll (struct ifnet *, enum poll_cmd, int);
244 static void re_poll_locked (struct ifnet *, enum poll_cmd, int);
246 static int re_intr (void *);
247 static void re_tick (void *);
248 static void re_tx_task (void *, int);
249 static void re_int_task (void *, int);
250 static void re_start (struct ifnet *);
251 static int re_ioctl (struct ifnet *, u_long, caddr_t);
252 static void re_init (void *);
253 static void re_init_locked (struct rl_softc *);
254 static void re_stop (struct rl_softc *);
255 static void re_watchdog (struct rl_softc *);
256 static int re_suspend (device_t);
257 static int re_resume (device_t);
258 static int re_shutdown (device_t);
259 static int re_ifmedia_upd (struct ifnet *);
260 static void re_ifmedia_sts (struct ifnet *, struct ifmediareq *);
262 static void re_eeprom_putbyte (struct rl_softc *, int);
263 static void re_eeprom_getword (struct rl_softc *, int, u_int16_t *);
264 static void re_read_eeprom (struct rl_softc *, caddr_t, int, int);
265 static int re_gmii_readreg (device_t, int, int);
266 static int re_gmii_writereg (device_t, int, int, int);
268 static int re_miibus_readreg (device_t, int, int);
269 static int re_miibus_writereg (device_t, int, int, int);
270 static void re_miibus_statchg (device_t);
272 static void re_set_rxmode (struct rl_softc *);
273 static void re_reset (struct rl_softc *);
274 static void re_setwol (struct rl_softc *);
275 static void re_clrwol (struct rl_softc *);
278 static int re_diag (struct rl_softc *);
281 static device_method_t re_methods[] = {
282 /* Device interface */
283 DEVMETHOD(device_probe, re_probe),
284 DEVMETHOD(device_attach, re_attach),
285 DEVMETHOD(device_detach, re_detach),
286 DEVMETHOD(device_suspend, re_suspend),
287 DEVMETHOD(device_resume, re_resume),
288 DEVMETHOD(device_shutdown, re_shutdown),
291 DEVMETHOD(bus_print_child, bus_generic_print_child),
292 DEVMETHOD(bus_driver_added, bus_generic_driver_added),
295 DEVMETHOD(miibus_readreg, re_miibus_readreg),
296 DEVMETHOD(miibus_writereg, re_miibus_writereg),
297 DEVMETHOD(miibus_statchg, re_miibus_statchg),
302 static driver_t re_driver = {
305 sizeof(struct rl_softc)
308 static devclass_t re_devclass;
310 DRIVER_MODULE(re, pci, re_driver, re_devclass, 0, 0);
311 DRIVER_MODULE(re, cardbus, re_driver, re_devclass, 0, 0);
312 DRIVER_MODULE(miibus, re, miibus_driver, miibus_devclass, 0, 0);
315 CSR_WRITE_1(sc, RL_EECMD, \
316 CSR_READ_1(sc, RL_EECMD) | x)
319 CSR_WRITE_1(sc, RL_EECMD, \
320 CSR_READ_1(sc, RL_EECMD) & ~x)
323 * Send a read command and address to the EEPROM, check for ACK.
326 re_eeprom_putbyte(struct rl_softc *sc, int addr)
330 d = addr | (RL_9346_READ << sc->rl_eewidth);
333 * Feed in each bit and strobe the clock.
336 for (i = 1 << (sc->rl_eewidth + 3); i; i >>= 1) {
338 EE_SET(RL_EE_DATAIN);
340 EE_CLR(RL_EE_DATAIN);
351 * Read a word of data stored in the EEPROM at address 'addr.'
354 re_eeprom_getword(struct rl_softc *sc, int addr, u_int16_t *dest)
360 * Send address of word we want to read.
362 re_eeprom_putbyte(sc, addr);
365 * Start reading bits from EEPROM.
367 for (i = 0x8000; i; i >>= 1) {
370 if (CSR_READ_1(sc, RL_EECMD) & RL_EE_DATAOUT)
380 * Read a sequence of words from the EEPROM.
383 re_read_eeprom(struct rl_softc *sc, caddr_t dest, int off, int cnt)
386 u_int16_t word = 0, *ptr;
388 CSR_SETBIT_1(sc, RL_EECMD, RL_EEMODE_PROGRAM);
392 for (i = 0; i < cnt; i++) {
393 CSR_SETBIT_1(sc, RL_EECMD, RL_EE_SEL);
394 re_eeprom_getword(sc, off + i, &word);
395 CSR_CLRBIT_1(sc, RL_EECMD, RL_EE_SEL);
396 ptr = (u_int16_t *)(dest + (i * 2));
400 CSR_CLRBIT_1(sc, RL_EECMD, RL_EEMODE_PROGRAM);
404 re_gmii_readreg(device_t dev, int phy, int reg)
413 sc = device_get_softc(dev);
415 /* Let the rgephy driver read the GMEDIASTAT register */
417 if (reg == RL_GMEDIASTAT) {
418 rval = CSR_READ_1(sc, RL_GMEDIASTAT);
422 CSR_WRITE_4(sc, RL_PHYAR, reg << 16);
425 for (i = 0; i < RL_PHY_TIMEOUT; i++) {
426 rval = CSR_READ_4(sc, RL_PHYAR);
427 if (rval & RL_PHYAR_BUSY)
432 if (i == RL_PHY_TIMEOUT) {
433 device_printf(sc->rl_dev, "PHY read failed\n");
437 return (rval & RL_PHYAR_PHYDATA);
441 re_gmii_writereg(device_t dev, int phy, int reg, int data)
447 sc = device_get_softc(dev);
449 CSR_WRITE_4(sc, RL_PHYAR, (reg << 16) |
450 (data & RL_PHYAR_PHYDATA) | RL_PHYAR_BUSY);
453 for (i = 0; i < RL_PHY_TIMEOUT; i++) {
454 rval = CSR_READ_4(sc, RL_PHYAR);
455 if (!(rval & RL_PHYAR_BUSY))
460 if (i == RL_PHY_TIMEOUT) {
461 device_printf(sc->rl_dev, "PHY write failed\n");
469 re_miibus_readreg(device_t dev, int phy, int reg)
473 u_int16_t re8139_reg = 0;
475 sc = device_get_softc(dev);
477 if (sc->rl_type == RL_8169) {
478 rval = re_gmii_readreg(dev, phy, reg);
482 /* Pretend the internal PHY is only at address 0 */
488 re8139_reg = RL_BMCR;
491 re8139_reg = RL_BMSR;
494 re8139_reg = RL_ANAR;
497 re8139_reg = RL_ANER;
500 re8139_reg = RL_LPAR;
506 * Allow the rlphy driver to read the media status
507 * register. If we have a link partner which does not
508 * support NWAY, this is the register which will tell
509 * us the results of parallel detection.
512 rval = CSR_READ_1(sc, RL_MEDIASTAT);
515 device_printf(sc->rl_dev, "bad phy register\n");
518 rval = CSR_READ_2(sc, re8139_reg);
519 if (sc->rl_type == RL_8139CPLUS && re8139_reg == RL_BMCR) {
520 /* 8139C+ has different bit layout. */
521 rval &= ~(BMCR_LOOP | BMCR_ISO);
527 re_miibus_writereg(device_t dev, int phy, int reg, int data)
530 u_int16_t re8139_reg = 0;
533 sc = device_get_softc(dev);
535 if (sc->rl_type == RL_8169) {
536 rval = re_gmii_writereg(dev, phy, reg, data);
540 /* Pretend the internal PHY is only at address 0 */
546 re8139_reg = RL_BMCR;
547 if (sc->rl_type == RL_8139CPLUS) {
548 /* 8139C+ has different bit layout. */
549 data &= ~(BMCR_LOOP | BMCR_ISO);
553 re8139_reg = RL_BMSR;
556 re8139_reg = RL_ANAR;
559 re8139_reg = RL_ANER;
562 re8139_reg = RL_LPAR;
569 device_printf(sc->rl_dev, "bad phy register\n");
572 CSR_WRITE_2(sc, re8139_reg, data);
577 re_miibus_statchg(device_t dev)
581 struct mii_data *mii;
583 sc = device_get_softc(dev);
584 mii = device_get_softc(sc->rl_miibus);
586 if (mii == NULL || ifp == NULL ||
587 (ifp->if_drv_flags & IFF_DRV_RUNNING) == 0)
590 sc->rl_flags &= ~RL_FLAG_LINK;
591 if ((mii->mii_media_status & (IFM_ACTIVE | IFM_AVALID)) ==
592 (IFM_ACTIVE | IFM_AVALID)) {
593 switch (IFM_SUBTYPE(mii->mii_media_active)) {
596 sc->rl_flags |= RL_FLAG_LINK;
599 if ((sc->rl_flags & RL_FLAG_FASTETHER) != 0)
601 sc->rl_flags |= RL_FLAG_LINK;
608 * RealTek controllers does not provide any interface to
609 * Tx/Rx MACs for resolved speed, duplex and flow-control
615 * Set the RX configuration and 64-bit multicast hash filter.
618 re_set_rxmode(struct rl_softc *sc)
621 struct ifmultiaddr *ifma;
622 uint32_t hashes[2] = { 0, 0 };
629 rxfilt = RL_RXCFG_CONFIG | RL_RXCFG_RX_INDIV | RL_RXCFG_RX_BROAD;
631 if (ifp->if_flags & (IFF_ALLMULTI | IFF_PROMISC)) {
632 if (ifp->if_flags & IFF_PROMISC)
633 rxfilt |= RL_RXCFG_RX_ALLPHYS;
635 * Unlike other hardwares, we have to explicitly set
636 * RL_RXCFG_RX_MULTI to receive multicast frames in
639 rxfilt |= RL_RXCFG_RX_MULTI;
640 hashes[0] = hashes[1] = 0xffffffff;
645 TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
646 if (ifma->ifma_addr->sa_family != AF_LINK)
648 h = ether_crc32_be(LLADDR((struct sockaddr_dl *)
649 ifma->ifma_addr), ETHER_ADDR_LEN) >> 26;
651 hashes[0] |= (1 << h);
653 hashes[1] |= (1 << (h - 32));
657 if (hashes[0] != 0 || hashes[1] != 0) {
659 * For some unfathomable reason, RealTek decided to
660 * reverse the order of the multicast hash registers
661 * in the PCI Express parts. This means we have to
662 * write the hash pattern in reverse order for those
665 if ((sc->rl_flags & RL_FLAG_PCIE) != 0) {
666 h = bswap32(hashes[0]);
667 hashes[0] = bswap32(hashes[1]);
670 rxfilt |= RL_RXCFG_RX_MULTI;
674 CSR_WRITE_4(sc, RL_MAR0, hashes[0]);
675 CSR_WRITE_4(sc, RL_MAR4, hashes[1]);
676 CSR_WRITE_4(sc, RL_RXCFG, rxfilt);
680 re_reset(struct rl_softc *sc)
686 CSR_WRITE_1(sc, RL_COMMAND, RL_CMD_RESET);
688 for (i = 0; i < RL_TIMEOUT; i++) {
690 if (!(CSR_READ_1(sc, RL_COMMAND) & RL_CMD_RESET))
694 device_printf(sc->rl_dev, "reset never completed!\n");
696 if ((sc->rl_flags & RL_FLAG_MACRESET) != 0)
697 CSR_WRITE_1(sc, 0x82, 1);
698 if (sc->rl_hwrev == RL_HWREV_8169S)
699 re_gmii_writereg(sc->rl_dev, 1, 0x0b, 0);
705 * The following routine is designed to test for a defect on some
706 * 32-bit 8169 cards. Some of these NICs have the REQ64# and ACK64#
707 * lines connected to the bus, however for a 32-bit only card, they
708 * should be pulled high. The result of this defect is that the
709 * NIC will not work right if you plug it into a 64-bit slot: DMA
710 * operations will be done with 64-bit transfers, which will fail
711 * because the 64-bit data lines aren't connected.
713 * There's no way to work around this (short of talking a soldering
714 * iron to the board), however we can detect it. The method we use
715 * here is to put the NIC into digital loopback mode, set the receiver
716 * to promiscuous mode, and then try to send a frame. We then compare
717 * the frame data we sent to what was received. If the data matches,
718 * then the NIC is working correctly, otherwise we know the user has
719 * a defective NIC which has been mistakenly plugged into a 64-bit PCI
720 * slot. In the latter case, there's no way the NIC can work correctly,
721 * so we print out a message on the console and abort the device attach.
725 re_diag(struct rl_softc *sc)
727 struct ifnet *ifp = sc->rl_ifp;
729 struct ether_header *eh;
730 struct rl_desc *cur_rx;
733 int total_len, i, error = 0, phyaddr;
734 u_int8_t dst[] = { 0x00, 'h', 'e', 'l', 'l', 'o' };
735 u_int8_t src[] = { 0x00, 'w', 'o', 'r', 'l', 'd' };
737 /* Allocate a single mbuf */
738 MGETHDR(m0, M_DONTWAIT, MT_DATA);
745 * Initialize the NIC in test mode. This sets the chip up
746 * so that it can send and receive frames, but performs the
747 * following special functions:
748 * - Puts receiver in promiscuous mode
749 * - Enables digital loopback mode
750 * - Leaves interrupts turned off
753 ifp->if_flags |= IFF_PROMISC;
756 sc->rl_flags |= RL_FLAG_LINK;
757 if (sc->rl_type == RL_8169)
762 re_miibus_writereg(sc->rl_dev, phyaddr, MII_BMCR, BMCR_RESET);
763 for (i = 0; i < RL_TIMEOUT; i++) {
764 status = re_miibus_readreg(sc->rl_dev, phyaddr, MII_BMCR);
765 if (!(status & BMCR_RESET))
769 re_miibus_writereg(sc->rl_dev, phyaddr, MII_BMCR, BMCR_LOOP);
770 CSR_WRITE_2(sc, RL_ISR, RL_INTRS);
774 /* Put some data in the mbuf */
776 eh = mtod(m0, struct ether_header *);
777 bcopy ((char *)&dst, eh->ether_dhost, ETHER_ADDR_LEN);
778 bcopy ((char *)&src, eh->ether_shost, ETHER_ADDR_LEN);
779 eh->ether_type = htons(ETHERTYPE_IP);
780 m0->m_pkthdr.len = m0->m_len = ETHER_MIN_LEN - ETHER_CRC_LEN;
783 * Queue the packet, start transmission.
784 * Note: IF_HANDOFF() ultimately calls re_start() for us.
787 CSR_WRITE_2(sc, RL_ISR, 0xFFFF);
789 /* XXX: re_diag must not be called when in ALTQ mode */
790 IF_HANDOFF(&ifp->if_snd, m0, ifp);
794 /* Wait for it to propagate through the chip */
797 for (i = 0; i < RL_TIMEOUT; i++) {
798 status = CSR_READ_2(sc, RL_ISR);
799 CSR_WRITE_2(sc, RL_ISR, status);
800 if ((status & (RL_ISR_TIMEOUT_EXPIRED|RL_ISR_RX_OK)) ==
801 (RL_ISR_TIMEOUT_EXPIRED|RL_ISR_RX_OK))
806 if (i == RL_TIMEOUT) {
807 device_printf(sc->rl_dev,
808 "diagnostic failed, failed to receive packet in"
815 * The packet should have been dumped into the first
816 * entry in the RX DMA ring. Grab it from there.
819 bus_dmamap_sync(sc->rl_ldata.rl_rx_list_tag,
820 sc->rl_ldata.rl_rx_list_map,
821 BUS_DMASYNC_POSTREAD);
822 bus_dmamap_sync(sc->rl_ldata.rl_rx_mtag,
823 sc->rl_ldata.rl_rx_desc[0].rx_dmamap,
824 BUS_DMASYNC_POSTREAD);
825 bus_dmamap_unload(sc->rl_ldata.rl_rx_mtag,
826 sc->rl_ldata.rl_rx_desc[0].rx_dmamap);
828 m0 = sc->rl_ldata.rl_rx_desc[0].rx_m;
829 sc->rl_ldata.rl_rx_desc[0].rx_m = NULL;
830 eh = mtod(m0, struct ether_header *);
832 cur_rx = &sc->rl_ldata.rl_rx_list[0];
833 total_len = RL_RXBYTES(cur_rx);
834 rxstat = le32toh(cur_rx->rl_cmdstat);
836 if (total_len != ETHER_MIN_LEN) {
837 device_printf(sc->rl_dev,
838 "diagnostic failed, received short packet\n");
843 /* Test that the received packet data matches what we sent. */
845 if (bcmp((char *)&eh->ether_dhost, (char *)&dst, ETHER_ADDR_LEN) ||
846 bcmp((char *)&eh->ether_shost, (char *)&src, ETHER_ADDR_LEN) ||
847 ntohs(eh->ether_type) != ETHERTYPE_IP) {
848 device_printf(sc->rl_dev, "WARNING, DMA FAILURE!\n");
849 device_printf(sc->rl_dev, "expected TX data: %6D/%6D/0x%x\n",
850 dst, ":", src, ":", ETHERTYPE_IP);
851 device_printf(sc->rl_dev, "received RX data: %6D/%6D/0x%x\n",
852 eh->ether_dhost, ":", eh->ether_shost, ":",
853 ntohs(eh->ether_type));
854 device_printf(sc->rl_dev, "You may have a defective 32-bit "
855 "NIC plugged into a 64-bit PCI slot.\n");
856 device_printf(sc->rl_dev, "Please re-install the NIC in a "
857 "32-bit slot for proper operation.\n");
858 device_printf(sc->rl_dev, "Read the re(4) man page for more "
864 /* Turn interface off, release resources */
867 sc->rl_flags &= ~RL_FLAG_LINK;
868 ifp->if_flags &= ~IFF_PROMISC;
881 * Probe for a RealTek 8139C+/8169/8110 chip. Check the PCI vendor and device
882 * IDs against our list and return a device name if we find a match.
885 re_probe(device_t dev)
888 uint16_t devid, vendor;
889 uint16_t revid, sdevid;
892 vendor = pci_get_vendor(dev);
893 devid = pci_get_device(dev);
894 revid = pci_get_revid(dev);
895 sdevid = pci_get_subdevice(dev);
897 if (vendor == LINKSYS_VENDORID && devid == LINKSYS_DEVICEID_EG1032) {
898 if (sdevid != LINKSYS_SUBDEVICE_EG1032_REV3) {
900 * Only attach to rev. 3 of the Linksys EG1032 adapter.
901 * Rev. 2 is supported by sk(4).
907 if (vendor == RT_VENDORID && devid == RT_DEVICEID_8139) {
909 /* 8139, let rl(4) take care of this device. */
915 for (i = 0; i < sizeof(re_devs) / sizeof(re_devs[0]); i++, t++) {
916 if (vendor == t->rl_vid && devid == t->rl_did) {
917 device_set_desc(dev, t->rl_name);
918 return (BUS_PROBE_DEFAULT);
926 * Map a single buffer address.
930 re_dma_map_addr(void *arg, bus_dma_segment_t *segs, int nseg, int error)
937 KASSERT(nseg == 1, ("too many DMA segments, %d should be 1", nseg));
939 *addr = segs->ds_addr;
943 re_allocmem(device_t dev, struct rl_softc *sc)
945 bus_size_t rx_list_size, tx_list_size;
949 rx_list_size = sc->rl_ldata.rl_rx_desc_cnt * sizeof(struct rl_desc);
950 tx_list_size = sc->rl_ldata.rl_tx_desc_cnt * sizeof(struct rl_desc);
953 * Allocate the parent bus DMA tag appropriate for PCI.
954 * In order to use DAC, RL_CPLUSCMD_PCI_DAC bit of RL_CPLUS_CMD
955 * register should be set. However some RealTek chips are known
956 * to be buggy on DAC handling, therefore disable DAC by limiting
957 * DMA address space to 32bit. PCIe variants of RealTek chips
958 * may not have the limitation but I took safer path.
960 error = bus_dma_tag_create(bus_get_dma_tag(dev), 1, 0,
961 BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL, NULL,
962 BUS_SPACE_MAXSIZE_32BIT, 0, BUS_SPACE_MAXSIZE_32BIT, 0,
963 NULL, NULL, &sc->rl_parent_tag);
965 device_printf(dev, "could not allocate parent DMA tag\n");
970 * Allocate map for TX mbufs.
972 error = bus_dma_tag_create(sc->rl_parent_tag, 1, 0,
973 BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR, NULL,
974 NULL, MCLBYTES * RL_NTXSEGS, RL_NTXSEGS, 4096, 0,
975 NULL, NULL, &sc->rl_ldata.rl_tx_mtag);
977 device_printf(dev, "could not allocate TX DMA tag\n");
982 * Allocate map for RX mbufs.
985 error = bus_dma_tag_create(sc->rl_parent_tag, sizeof(uint64_t), 0,
986 BUS_SPACE_MAXADDR, BUS_SPACE_MAXADDR, NULL, NULL,
987 MCLBYTES, 1, MCLBYTES, 0, NULL, NULL, &sc->rl_ldata.rl_rx_mtag);
989 device_printf(dev, "could not allocate RX DMA tag\n");
994 * Allocate map for TX descriptor list.
996 error = bus_dma_tag_create(sc->rl_parent_tag, RL_RING_ALIGN,
997 0, BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL,
998 NULL, tx_list_size, 1, tx_list_size, 0,
999 NULL, NULL, &sc->rl_ldata.rl_tx_list_tag);
1001 device_printf(dev, "could not allocate TX DMA ring tag\n");
1005 /* Allocate DMA'able memory for the TX ring */
1007 error = bus_dmamem_alloc(sc->rl_ldata.rl_tx_list_tag,
1008 (void **)&sc->rl_ldata.rl_tx_list,
1009 BUS_DMA_WAITOK | BUS_DMA_COHERENT | BUS_DMA_ZERO,
1010 &sc->rl_ldata.rl_tx_list_map);
1012 device_printf(dev, "could not allocate TX DMA ring\n");
1016 /* Load the map for the TX ring. */
1018 sc->rl_ldata.rl_tx_list_addr = 0;
1019 error = bus_dmamap_load(sc->rl_ldata.rl_tx_list_tag,
1020 sc->rl_ldata.rl_tx_list_map, sc->rl_ldata.rl_tx_list,
1021 tx_list_size, re_dma_map_addr,
1022 &sc->rl_ldata.rl_tx_list_addr, BUS_DMA_NOWAIT);
1023 if (error != 0 || sc->rl_ldata.rl_tx_list_addr == 0) {
1024 device_printf(dev, "could not load TX DMA ring\n");
1028 /* Create DMA maps for TX buffers */
1030 for (i = 0; i < sc->rl_ldata.rl_tx_desc_cnt; i++) {
1031 error = bus_dmamap_create(sc->rl_ldata.rl_tx_mtag, 0,
1032 &sc->rl_ldata.rl_tx_desc[i].tx_dmamap);
1034 device_printf(dev, "could not create DMA map for TX\n");
1040 * Allocate map for RX descriptor list.
1042 error = bus_dma_tag_create(sc->rl_parent_tag, RL_RING_ALIGN,
1043 0, BUS_SPACE_MAXADDR_32BIT, BUS_SPACE_MAXADDR, NULL,
1044 NULL, rx_list_size, 1, rx_list_size, 0,
1045 NULL, NULL, &sc->rl_ldata.rl_rx_list_tag);
1047 device_printf(dev, "could not create RX DMA ring tag\n");
1051 /* Allocate DMA'able memory for the RX ring */
1053 error = bus_dmamem_alloc(sc->rl_ldata.rl_rx_list_tag,
1054 (void **)&sc->rl_ldata.rl_rx_list,
1055 BUS_DMA_WAITOK | BUS_DMA_COHERENT | BUS_DMA_ZERO,
1056 &sc->rl_ldata.rl_rx_list_map);
1058 device_printf(dev, "could not allocate RX DMA ring\n");
1062 /* Load the map for the RX ring. */
1064 sc->rl_ldata.rl_rx_list_addr = 0;
1065 error = bus_dmamap_load(sc->rl_ldata.rl_rx_list_tag,
1066 sc->rl_ldata.rl_rx_list_map, sc->rl_ldata.rl_rx_list,
1067 rx_list_size, re_dma_map_addr,
1068 &sc->rl_ldata.rl_rx_list_addr, BUS_DMA_NOWAIT);
1069 if (error != 0 || sc->rl_ldata.rl_rx_list_addr == 0) {
1070 device_printf(dev, "could not load RX DMA ring\n");
1074 /* Create DMA maps for RX buffers */
1076 error = bus_dmamap_create(sc->rl_ldata.rl_rx_mtag, 0,
1077 &sc->rl_ldata.rl_rx_sparemap);
1079 device_printf(dev, "could not create spare DMA map for RX\n");
1082 for (i = 0; i < sc->rl_ldata.rl_rx_desc_cnt; i++) {
1083 error = bus_dmamap_create(sc->rl_ldata.rl_rx_mtag, 0,
1084 &sc->rl_ldata.rl_rx_desc[i].rx_dmamap);
1086 device_printf(dev, "could not create DMA map for RX\n");
1095 * Attach the interface. Allocate softc structures, do ifmedia
1096 * setup and ethernet/BPF attach.
1099 re_attach(device_t dev)
1101 u_char eaddr[ETHER_ADDR_LEN];
1102 u_int16_t as[ETHER_ADDR_LEN / 2];
1103 struct rl_softc *sc;
1105 struct rl_hwrev *hw_rev;
1107 u_int16_t devid, re_did = 0;
1108 int error = 0, rid, i;
1112 sc = device_get_softc(dev);
1115 mtx_init(&sc->rl_mtx, device_get_nameunit(dev), MTX_NETWORK_LOCK,
1117 callout_init_mtx(&sc->rl_stat_callout, &sc->rl_mtx, 0);
1120 * Map control/status registers.
1122 pci_enable_busmaster(dev);
1124 devid = pci_get_device(dev);
1126 * Prefer memory space register mapping over IO space.
1127 * Because RTL8169SC does not seem to work when memory mapping
1128 * is used always activate io mapping.
1130 if (devid == RT_DEVICEID_8169SC)
1132 if (prefer_iomap == 0) {
1133 sc->rl_res_id = PCIR_BAR(1);
1134 sc->rl_res_type = SYS_RES_MEMORY;
1135 /* RTL8168/8101E seems to use different BARs. */
1136 if (devid == RT_DEVICEID_8168 || devid == RT_DEVICEID_8101E)
1137 sc->rl_res_id = PCIR_BAR(2);
1139 sc->rl_res_id = PCIR_BAR(0);
1140 sc->rl_res_type = SYS_RES_IOPORT;
1142 sc->rl_res = bus_alloc_resource_any(dev, sc->rl_res_type,
1143 &sc->rl_res_id, RF_ACTIVE);
1144 if (sc->rl_res == NULL && prefer_iomap == 0) {
1145 sc->rl_res_id = PCIR_BAR(0);
1146 sc->rl_res_type = SYS_RES_IOPORT;
1147 sc->rl_res = bus_alloc_resource_any(dev, sc->rl_res_type,
1148 &sc->rl_res_id, RF_ACTIVE);
1150 if (sc->rl_res == NULL) {
1151 device_printf(dev, "couldn't map ports/memory\n");
1156 sc->rl_btag = rman_get_bustag(sc->rl_res);
1157 sc->rl_bhandle = rman_get_bushandle(sc->rl_res);
1160 if (pci_find_extcap(dev, PCIY_EXPRESS, ®) == 0) {
1161 sc->rl_flags |= RL_FLAG_PCIE;
1162 msic = pci_msi_count(dev);
1164 device_printf(dev, "MSI count : %d\n", msic);
1166 if (msic > 0 && msi_disable == 0) {
1168 if (pci_alloc_msi(dev, &msic) == 0) {
1169 if (msic == RL_MSI_MESSAGES) {
1170 device_printf(dev, "Using %d MSI messages\n",
1172 sc->rl_flags |= RL_FLAG_MSI;
1173 /* Explicitly set MSI enable bit. */
1174 CSR_WRITE_1(sc, RL_EECMD, RL_EE_MODE);
1175 cfg = CSR_READ_1(sc, RL_CFG2);
1177 CSR_WRITE_1(sc, RL_CFG2, cfg);
1178 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
1180 pci_release_msi(dev);
1184 /* Allocate interrupt */
1185 if ((sc->rl_flags & RL_FLAG_MSI) == 0) {
1187 sc->rl_irq[0] = bus_alloc_resource_any(dev, SYS_RES_IRQ, &rid,
1188 RF_SHAREABLE | RF_ACTIVE);
1189 if (sc->rl_irq[0] == NULL) {
1190 device_printf(dev, "couldn't allocate IRQ resources\n");
1195 for (i = 0, rid = 1; i < RL_MSI_MESSAGES; i++, rid++) {
1196 sc->rl_irq[i] = bus_alloc_resource_any(dev,
1197 SYS_RES_IRQ, &rid, RF_ACTIVE);
1198 if (sc->rl_irq[i] == NULL) {
1200 "couldn't llocate IRQ resources for "
1201 "message %d\n", rid);
1208 if ((sc->rl_flags & RL_FLAG_MSI) == 0) {
1209 CSR_WRITE_1(sc, RL_EECMD, RL_EE_MODE);
1210 cfg = CSR_READ_1(sc, RL_CFG2);
1211 if ((cfg & RL_CFG2_MSI) != 0) {
1212 device_printf(dev, "turning off MSI enable bit.\n");
1213 cfg &= ~RL_CFG2_MSI;
1214 CSR_WRITE_1(sc, RL_CFG2, cfg);
1216 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
1219 /* Reset the adapter. */
1225 hwrev = CSR_READ_4(sc, RL_TXCFG);
1226 switch (hwrev & 0x70000000) {
1229 device_printf(dev, "Chip rev. 0x%08x\n", hwrev & 0xfc800000);
1230 hwrev &= (RL_TXCFG_HWREV | 0x80000000);
1233 device_printf(dev, "Chip rev. 0x%08x\n", hwrev & 0x7c800000);
1234 hwrev &= RL_TXCFG_HWREV;
1237 device_printf(dev, "MAC rev. 0x%08x\n", hwrev & 0x00700000);
1238 while (hw_rev->rl_desc != NULL) {
1239 if (hw_rev->rl_rev == hwrev) {
1240 sc->rl_type = hw_rev->rl_type;
1241 sc->rl_hwrev = hw_rev->rl_rev;
1246 if (hw_rev->rl_desc == NULL) {
1247 device_printf(dev, "Unknown H/W revision: 0x%08x\n", hwrev);
1252 switch (hw_rev->rl_rev) {
1253 case RL_HWREV_8139CPLUS:
1254 sc->rl_flags |= RL_FLAG_NOJUMBO | RL_FLAG_FASTETHER;
1256 case RL_HWREV_8100E:
1257 case RL_HWREV_8101E:
1258 sc->rl_flags |= RL_FLAG_NOJUMBO | RL_FLAG_PHYWAKE |
1261 case RL_HWREV_8102E:
1262 case RL_HWREV_8102EL:
1263 sc->rl_flags |= RL_FLAG_NOJUMBO | RL_FLAG_PHYWAKE |
1264 RL_FLAG_PAR | RL_FLAG_DESCV2 | RL_FLAG_MACSTAT |
1265 RL_FLAG_FASTETHER | RL_FLAG_CMDSTOP;
1267 case RL_HWREV_8168_SPIN1:
1268 case RL_HWREV_8168_SPIN2:
1269 sc->rl_flags |= RL_FLAG_WOLRXENB;
1271 case RL_HWREV_8168_SPIN3:
1272 sc->rl_flags |= RL_FLAG_PHYWAKE | RL_FLAG_MACSTAT;
1274 case RL_HWREV_8168C_SPIN2:
1275 sc->rl_flags |= RL_FLAG_MACSLEEP;
1277 case RL_HWREV_8168C:
1278 if ((hwrev & 0x00700000) == 0x00200000)
1279 sc->rl_flags |= RL_FLAG_MACSLEEP;
1281 case RL_HWREV_8168CP:
1282 case RL_HWREV_8168D:
1283 sc->rl_flags |= RL_FLAG_PHYWAKE | RL_FLAG_PAR |
1284 RL_FLAG_DESCV2 | RL_FLAG_MACSTAT | RL_FLAG_CMDSTOP;
1286 * These controllers support jumbo frame but it seems
1287 * that enabling it requires touching additional magic
1288 * registers. Depending on MAC revisions some
1289 * controllers need to disable checksum offload. So
1290 * disable jumbo frame until I have better idea what
1291 * it really requires to make it support.
1292 * RTL8168C/CP : supports up to 6KB jumbo frame.
1293 * RTL8111C/CP : supports up to 9KB jumbo frame.
1295 sc->rl_flags |= RL_FLAG_NOJUMBO;
1297 case RL_HWREV_8169_8110SB:
1298 case RL_HWREV_8169_8110SBL:
1299 case RL_HWREV_8169_8110SC:
1300 case RL_HWREV_8169_8110SCE:
1301 sc->rl_flags |= RL_FLAG_PHYWAKE;
1304 case RL_HWREV_8169S:
1305 case RL_HWREV_8110S:
1306 sc->rl_flags |= RL_FLAG_MACRESET;
1313 CSR_WRITE_1(sc, RL_EECMD, RL_EE_MODE);
1314 cfg = CSR_READ_1(sc, RL_CFG1);
1316 CSR_WRITE_1(sc, RL_CFG1, cfg);
1317 cfg = CSR_READ_1(sc, RL_CFG5);
1318 cfg &= RL_CFG5_PME_STS;
1319 CSR_WRITE_1(sc, RL_CFG5, cfg);
1320 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
1322 if ((sc->rl_flags & RL_FLAG_PAR) != 0) {
1324 * XXX Should have a better way to extract station
1325 * address from EEPROM.
1327 for (i = 0; i < ETHER_ADDR_LEN; i++)
1328 eaddr[i] = CSR_READ_1(sc, RL_IDR0 + i);
1330 sc->rl_eewidth = RL_9356_ADDR_LEN;
1331 re_read_eeprom(sc, (caddr_t)&re_did, 0, 1);
1332 if (re_did != 0x8129)
1333 sc->rl_eewidth = RL_9346_ADDR_LEN;
1336 * Get station address from the EEPROM.
1338 re_read_eeprom(sc, (caddr_t)as, RL_EE_EADDR, 3);
1339 for (i = 0; i < ETHER_ADDR_LEN / 2; i++)
1340 as[i] = le16toh(as[i]);
1341 bcopy(as, eaddr, sizeof(eaddr));
1344 if (sc->rl_type == RL_8169) {
1345 /* Set RX length mask and number of descriptors. */
1346 sc->rl_rxlenmask = RL_RDESC_STAT_GFRAGLEN;
1347 sc->rl_txstart = RL_GTXSTART;
1348 sc->rl_ldata.rl_tx_desc_cnt = RL_8169_TX_DESC_CNT;
1349 sc->rl_ldata.rl_rx_desc_cnt = RL_8169_RX_DESC_CNT;
1351 /* Set RX length mask and number of descriptors. */
1352 sc->rl_rxlenmask = RL_RDESC_STAT_FRAGLEN;
1353 sc->rl_txstart = RL_TXSTART;
1354 sc->rl_ldata.rl_tx_desc_cnt = RL_8139_TX_DESC_CNT;
1355 sc->rl_ldata.rl_rx_desc_cnt = RL_8139_RX_DESC_CNT;
1358 error = re_allocmem(dev, sc);
1362 ifp = sc->rl_ifp = if_alloc(IFT_ETHER);
1364 device_printf(dev, "can not if_alloc()\n");
1369 /* Take controller out of deep sleep mode. */
1370 if ((sc->rl_flags & RL_FLAG_MACSLEEP) != 0) {
1371 if ((CSR_READ_1(sc, RL_MACDBG) & 0x80) == 0x80)
1372 CSR_WRITE_1(sc, RL_GPIO,
1373 CSR_READ_1(sc, RL_GPIO) | 0x01);
1375 CSR_WRITE_1(sc, RL_GPIO,
1376 CSR_READ_1(sc, RL_GPIO) & ~0x01);
1379 /* Take PHY out of power down mode. */
1380 if ((sc->rl_flags & RL_FLAG_PHYWAKE) != 0) {
1381 re_gmii_writereg(dev, 1, 0x1f, 0);
1382 re_gmii_writereg(dev, 1, 0x0e, 0);
1386 if (mii_phy_probe(dev, &sc->rl_miibus,
1387 re_ifmedia_upd, re_ifmedia_sts)) {
1388 device_printf(dev, "MII without any phy!\n");
1394 if_initname(ifp, device_get_name(dev), device_get_unit(dev));
1395 ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
1396 ifp->if_ioctl = re_ioctl;
1397 ifp->if_start = re_start;
1398 ifp->if_hwassist = RE_CSUM_FEATURES;
1399 ifp->if_capabilities = IFCAP_HWCSUM;
1400 ifp->if_capenable = ifp->if_capabilities;
1401 ifp->if_init = re_init;
1402 IFQ_SET_MAXLEN(&ifp->if_snd, RL_IFQ_MAXLEN);
1403 ifp->if_snd.ifq_drv_maxlen = RL_IFQ_MAXLEN;
1404 IFQ_SET_READY(&ifp->if_snd);
1406 TASK_INIT(&sc->rl_txtask, 1, re_tx_task, ifp);
1407 TASK_INIT(&sc->rl_inttask, 0, re_int_task, sc);
1411 * Still have no idea how to make TSO work on 8168C, 8168CP,
1414 if ((sc->rl_flags & RL_FLAG_DESCV2) == 0) {
1415 ifp->if_hwassist |= CSUM_TSO;
1416 ifp->if_capabilities |= IFCAP_TSO4;
1420 * Call MI attach routine.
1422 ether_ifattach(ifp, eaddr);
1424 /* VLAN capability setup */
1425 ifp->if_capabilities |= IFCAP_VLAN_MTU | IFCAP_VLAN_HWTAGGING;
1426 if (ifp->if_capabilities & IFCAP_HWCSUM)
1427 ifp->if_capabilities |= IFCAP_VLAN_HWCSUM;
1428 /* Enable WOL if PM is supported. */
1429 if (pci_find_extcap(sc->rl_dev, PCIY_PMG, ®) == 0)
1430 ifp->if_capabilities |= IFCAP_WOL;
1431 ifp->if_capenable = ifp->if_capabilities;
1433 * Don't enable TSO by default. Under certain
1434 * circumtances the controller generated corrupted
1435 * packets in TSO size.
1437 ifp->if_hwassist &= ~CSUM_TSO;
1438 ifp->if_capenable &= ~IFCAP_TSO4;
1439 #ifdef DEVICE_POLLING
1440 ifp->if_capabilities |= IFCAP_POLLING;
1443 * Tell the upper layer(s) we support long frames.
1444 * Must appear after the call to ether_ifattach() because
1445 * ether_ifattach() sets ifi_hdrlen to the default value.
1447 ifp->if_data.ifi_hdrlen = sizeof(struct ether_vlan_header);
1451 * Perform hardware diagnostic on the original RTL8169.
1452 * Some 32-bit cards were incorrectly wired and would
1453 * malfunction if plugged into a 64-bit slot.
1456 if (hwrev == RL_HWREV_8169) {
1457 error = re_diag(sc);
1460 "attach aborted due to hardware diag failure\n");
1461 ether_ifdetach(ifp);
1467 /* Hook interrupt last to avoid having to lock softc */
1468 if ((sc->rl_flags & RL_FLAG_MSI) == 0)
1469 error = bus_setup_intr(dev, sc->rl_irq[0],
1470 INTR_TYPE_NET | INTR_MPSAFE, re_intr, NULL, sc,
1471 &sc->rl_intrhand[0]);
1473 for (i = 0; i < RL_MSI_MESSAGES; i++) {
1474 error = bus_setup_intr(dev, sc->rl_irq[i],
1475 INTR_TYPE_NET | INTR_MPSAFE, re_intr, NULL, sc,
1476 &sc->rl_intrhand[i]);
1482 device_printf(dev, "couldn't set up irq\n");
1483 ether_ifdetach(ifp);
1495 * Shutdown hardware and free up resources. This can be called any
1496 * time after the mutex has been initialized. It is called in both
1497 * the error case in attach and the normal detach case so it needs
1498 * to be careful about only freeing resources that have actually been
1502 re_detach(device_t dev)
1504 struct rl_softc *sc;
1508 sc = device_get_softc(dev);
1510 KASSERT(mtx_initialized(&sc->rl_mtx), ("re mutex not initialized"));
1512 /* These should only be active if attach succeeded */
1513 if (device_is_attached(dev)) {
1514 #ifdef DEVICE_POLLING
1515 if (ifp->if_capenable & IFCAP_POLLING)
1516 ether_poll_deregister(ifp);
1524 callout_drain(&sc->rl_stat_callout);
1525 taskqueue_drain(taskqueue_fast, &sc->rl_inttask);
1526 taskqueue_drain(taskqueue_fast, &sc->rl_txtask);
1528 * Force off the IFF_UP flag here, in case someone
1529 * still had a BPF descriptor attached to this
1530 * interface. If they do, ether_ifdetach() will cause
1531 * the BPF code to try and clear the promisc mode
1532 * flag, which will bubble down to re_ioctl(),
1533 * which will try to call re_init() again. This will
1534 * turn the NIC back on and restart the MII ticker,
1535 * which will panic the system when the kernel tries
1536 * to invoke the re_tick() function that isn't there
1539 ifp->if_flags &= ~IFF_UP;
1540 ether_ifdetach(ifp);
1543 device_delete_child(dev, sc->rl_miibus);
1544 bus_generic_detach(dev);
1547 * The rest is resource deallocation, so we should already be
1551 for (i = 0; i < RL_MSI_MESSAGES; i++) {
1552 if (sc->rl_intrhand[i] != NULL) {
1553 bus_teardown_intr(dev, sc->rl_irq[i],
1554 sc->rl_intrhand[i]);
1555 sc->rl_intrhand[i] = NULL;
1560 if ((sc->rl_flags & RL_FLAG_MSI) == 0) {
1561 if (sc->rl_irq[0] != NULL) {
1562 bus_release_resource(dev, SYS_RES_IRQ, 0,
1564 sc->rl_irq[0] = NULL;
1567 for (i = 0, rid = 1; i < RL_MSI_MESSAGES; i++, rid++) {
1568 if (sc->rl_irq[i] != NULL) {
1569 bus_release_resource(dev, SYS_RES_IRQ, rid,
1571 sc->rl_irq[i] = NULL;
1574 pci_release_msi(dev);
1577 bus_release_resource(dev, sc->rl_res_type, sc->rl_res_id,
1580 /* Unload and free the RX DMA ring memory and map */
1582 if (sc->rl_ldata.rl_rx_list_tag) {
1583 bus_dmamap_unload(sc->rl_ldata.rl_rx_list_tag,
1584 sc->rl_ldata.rl_rx_list_map);
1585 bus_dmamem_free(sc->rl_ldata.rl_rx_list_tag,
1586 sc->rl_ldata.rl_rx_list,
1587 sc->rl_ldata.rl_rx_list_map);
1588 bus_dma_tag_destroy(sc->rl_ldata.rl_rx_list_tag);
1591 /* Unload and free the TX DMA ring memory and map */
1593 if (sc->rl_ldata.rl_tx_list_tag) {
1594 bus_dmamap_unload(sc->rl_ldata.rl_tx_list_tag,
1595 sc->rl_ldata.rl_tx_list_map);
1596 bus_dmamem_free(sc->rl_ldata.rl_tx_list_tag,
1597 sc->rl_ldata.rl_tx_list,
1598 sc->rl_ldata.rl_tx_list_map);
1599 bus_dma_tag_destroy(sc->rl_ldata.rl_tx_list_tag);
1602 /* Destroy all the RX and TX buffer maps */
1604 if (sc->rl_ldata.rl_tx_mtag) {
1605 for (i = 0; i < sc->rl_ldata.rl_tx_desc_cnt; i++)
1606 bus_dmamap_destroy(sc->rl_ldata.rl_tx_mtag,
1607 sc->rl_ldata.rl_tx_desc[i].tx_dmamap);
1608 bus_dma_tag_destroy(sc->rl_ldata.rl_tx_mtag);
1610 if (sc->rl_ldata.rl_rx_mtag) {
1611 for (i = 0; i < sc->rl_ldata.rl_rx_desc_cnt; i++)
1612 bus_dmamap_destroy(sc->rl_ldata.rl_rx_mtag,
1613 sc->rl_ldata.rl_rx_desc[i].rx_dmamap);
1614 if (sc->rl_ldata.rl_rx_sparemap)
1615 bus_dmamap_destroy(sc->rl_ldata.rl_rx_mtag,
1616 sc->rl_ldata.rl_rx_sparemap);
1617 bus_dma_tag_destroy(sc->rl_ldata.rl_rx_mtag);
1620 /* Unload and free the stats buffer and map */
1622 if (sc->rl_ldata.rl_stag) {
1623 bus_dmamap_unload(sc->rl_ldata.rl_stag,
1624 sc->rl_ldata.rl_rx_list_map);
1625 bus_dmamem_free(sc->rl_ldata.rl_stag,
1626 sc->rl_ldata.rl_stats,
1627 sc->rl_ldata.rl_smap);
1628 bus_dma_tag_destroy(sc->rl_ldata.rl_stag);
1631 if (sc->rl_parent_tag)
1632 bus_dma_tag_destroy(sc->rl_parent_tag);
1634 mtx_destroy(&sc->rl_mtx);
1639 static __inline void
1640 re_discard_rxbuf(struct rl_softc *sc, int idx)
1642 struct rl_desc *desc;
1643 struct rl_rxdesc *rxd;
1646 rxd = &sc->rl_ldata.rl_rx_desc[idx];
1647 desc = &sc->rl_ldata.rl_rx_list[idx];
1648 desc->rl_vlanctl = 0;
1649 cmdstat = rxd->rx_size;
1650 if (idx == sc->rl_ldata.rl_rx_desc_cnt - 1)
1651 cmdstat |= RL_RDESC_CMD_EOR;
1652 desc->rl_cmdstat = htole32(cmdstat | RL_RDESC_CMD_OWN);
1656 re_newbuf(struct rl_softc *sc, int idx)
1659 struct rl_rxdesc *rxd;
1660 bus_dma_segment_t segs[1];
1662 struct rl_desc *desc;
1666 m = m_getcl(M_DONTWAIT, MT_DATA, M_PKTHDR);
1670 m->m_len = m->m_pkthdr.len = MCLBYTES;
1673 * This is part of an evil trick to deal with non-x86 platforms.
1674 * The RealTek chip requires RX buffers to be aligned on 64-bit
1675 * boundaries, but that will hose non-x86 machines. To get around
1676 * this, we leave some empty space at the start of each buffer
1677 * and for non-x86 hosts, we copy the buffer back six bytes
1678 * to achieve word alignment. This is slightly more efficient
1679 * than allocating a new buffer, copying the contents, and
1680 * discarding the old buffer.
1682 m_adj(m, RE_ETHER_ALIGN);
1684 error = bus_dmamap_load_mbuf_sg(sc->rl_ldata.rl_rx_mtag,
1685 sc->rl_ldata.rl_rx_sparemap, m, segs, &nsegs, BUS_DMA_NOWAIT);
1690 KASSERT(nsegs == 1, ("%s: %d segment returned!", __func__, nsegs));
1692 rxd = &sc->rl_ldata.rl_rx_desc[idx];
1693 if (rxd->rx_m != NULL) {
1694 bus_dmamap_sync(sc->rl_ldata.rl_rx_mtag, rxd->rx_dmamap,
1695 BUS_DMASYNC_POSTREAD);
1696 bus_dmamap_unload(sc->rl_ldata.rl_rx_mtag, rxd->rx_dmamap);
1700 map = rxd->rx_dmamap;
1701 rxd->rx_dmamap = sc->rl_ldata.rl_rx_sparemap;
1702 rxd->rx_size = segs[0].ds_len;
1703 sc->rl_ldata.rl_rx_sparemap = map;
1704 bus_dmamap_sync(sc->rl_ldata.rl_rx_mtag, rxd->rx_dmamap,
1705 BUS_DMASYNC_PREREAD);
1707 desc = &sc->rl_ldata.rl_rx_list[idx];
1708 desc->rl_vlanctl = 0;
1709 desc->rl_bufaddr_lo = htole32(RL_ADDR_LO(segs[0].ds_addr));
1710 desc->rl_bufaddr_hi = htole32(RL_ADDR_HI(segs[0].ds_addr));
1711 cmdstat = segs[0].ds_len;
1712 if (idx == sc->rl_ldata.rl_rx_desc_cnt - 1)
1713 cmdstat |= RL_RDESC_CMD_EOR;
1714 desc->rl_cmdstat = htole32(cmdstat | RL_RDESC_CMD_OWN);
1720 static __inline void
1721 re_fixup_rx(struct mbuf *m)
1724 uint16_t *src, *dst;
1726 src = mtod(m, uint16_t *);
1727 dst = src - (RE_ETHER_ALIGN - ETHER_ALIGN) / sizeof *src;
1729 for (i = 0; i < (m->m_len / sizeof(uint16_t) + 1); i++)
1732 m->m_data -= RE_ETHER_ALIGN - ETHER_ALIGN;
1737 re_tx_list_init(struct rl_softc *sc)
1739 struct rl_desc *desc;
1744 bzero(sc->rl_ldata.rl_tx_list,
1745 sc->rl_ldata.rl_tx_desc_cnt * sizeof(struct rl_desc));
1746 for (i = 0; i < sc->rl_ldata.rl_tx_desc_cnt; i++)
1747 sc->rl_ldata.rl_tx_desc[i].tx_m = NULL;
1749 desc = &sc->rl_ldata.rl_tx_list[sc->rl_ldata.rl_tx_desc_cnt - 1];
1750 desc->rl_cmdstat |= htole32(RL_TDESC_CMD_EOR);
1752 bus_dmamap_sync(sc->rl_ldata.rl_tx_list_tag,
1753 sc->rl_ldata.rl_tx_list_map,
1754 BUS_DMASYNC_PREREAD | BUS_DMASYNC_PREWRITE);
1756 sc->rl_ldata.rl_tx_prodidx = 0;
1757 sc->rl_ldata.rl_tx_considx = 0;
1758 sc->rl_ldata.rl_tx_free = sc->rl_ldata.rl_tx_desc_cnt;
1764 re_rx_list_init(struct rl_softc *sc)
1768 bzero(sc->rl_ldata.rl_rx_list,
1769 sc->rl_ldata.rl_rx_desc_cnt * sizeof(struct rl_desc));
1770 for (i = 0; i < sc->rl_ldata.rl_rx_desc_cnt; i++) {
1771 sc->rl_ldata.rl_rx_desc[i].rx_m = NULL;
1772 if ((error = re_newbuf(sc, i)) != 0)
1776 /* Flush the RX descriptors */
1778 bus_dmamap_sync(sc->rl_ldata.rl_rx_list_tag,
1779 sc->rl_ldata.rl_rx_list_map,
1780 BUS_DMASYNC_PREWRITE|BUS_DMASYNC_PREREAD);
1782 sc->rl_ldata.rl_rx_prodidx = 0;
1783 sc->rl_head = sc->rl_tail = NULL;
1789 * RX handler for C+ and 8169. For the gigE chips, we support
1790 * the reception of jumbo frames that have been fragmented
1791 * across multiple 2K mbuf cluster buffers.
1794 re_rxeof(struct rl_softc *sc)
1799 struct rl_desc *cur_rx;
1800 u_int32_t rxstat, rxvlan;
1807 /* Invalidate the descriptor memory */
1809 bus_dmamap_sync(sc->rl_ldata.rl_rx_list_tag,
1810 sc->rl_ldata.rl_rx_list_map,
1811 BUS_DMASYNC_POSTREAD | BUS_DMASYNC_POSTWRITE);
1813 for (i = sc->rl_ldata.rl_rx_prodidx; maxpkt > 0;
1814 i = RL_RX_DESC_NXT(sc, i)) {
1815 cur_rx = &sc->rl_ldata.rl_rx_list[i];
1816 rxstat = le32toh(cur_rx->rl_cmdstat);
1817 if ((rxstat & RL_RDESC_STAT_OWN) != 0)
1819 total_len = rxstat & sc->rl_rxlenmask;
1820 rxvlan = le32toh(cur_rx->rl_vlanctl);
1821 m = sc->rl_ldata.rl_rx_desc[i].rx_m;
1823 if (!(rxstat & RL_RDESC_STAT_EOF)) {
1824 if (re_newbuf(sc, i) != 0) {
1826 * If this is part of a multi-fragment packet,
1827 * discard all the pieces.
1829 if (sc->rl_head != NULL) {
1830 m_freem(sc->rl_head);
1831 sc->rl_head = sc->rl_tail = NULL;
1833 re_discard_rxbuf(sc, i);
1836 m->m_len = RE_RX_DESC_BUFLEN;
1837 if (sc->rl_head == NULL)
1838 sc->rl_head = sc->rl_tail = m;
1840 m->m_flags &= ~M_PKTHDR;
1841 sc->rl_tail->m_next = m;
1848 * NOTE: for the 8139C+, the frame length field
1849 * is always 12 bits in size, but for the gigE chips,
1850 * it is 13 bits (since the max RX frame length is 16K).
1851 * Unfortunately, all 32 bits in the status word
1852 * were already used, so to make room for the extra
1853 * length bit, RealTek took out the 'frame alignment
1854 * error' bit and shifted the other status bits
1855 * over one slot. The OWN, EOR, FS and LS bits are
1856 * still in the same places. We have already extracted
1857 * the frame length and checked the OWN bit, so rather
1858 * than using an alternate bit mapping, we shift the
1859 * status bits one space to the right so we can evaluate
1860 * them using the 8169 status as though it was in the
1861 * same format as that of the 8139C+.
1863 if (sc->rl_type == RL_8169)
1867 * if total_len > 2^13-1, both _RXERRSUM and _GIANT will be
1868 * set, but if CRC is clear, it will still be a valid frame.
1870 if (rxstat & RL_RDESC_STAT_RXERRSUM && !(total_len > 8191 &&
1871 (rxstat & RL_RDESC_STAT_ERRS) == RL_RDESC_STAT_GIANT)) {
1874 * If this is part of a multi-fragment packet,
1875 * discard all the pieces.
1877 if (sc->rl_head != NULL) {
1878 m_freem(sc->rl_head);
1879 sc->rl_head = sc->rl_tail = NULL;
1881 re_discard_rxbuf(sc, i);
1886 * If allocating a replacement mbuf fails,
1887 * reload the current one.
1890 if (re_newbuf(sc, i) != 0) {
1892 if (sc->rl_head != NULL) {
1893 m_freem(sc->rl_head);
1894 sc->rl_head = sc->rl_tail = NULL;
1896 re_discard_rxbuf(sc, i);
1900 if (sc->rl_head != NULL) {
1901 m->m_len = total_len % RE_RX_DESC_BUFLEN;
1903 m->m_len = RE_RX_DESC_BUFLEN;
1905 * Special case: if there's 4 bytes or less
1906 * in this buffer, the mbuf can be discarded:
1907 * the last 4 bytes is the CRC, which we don't
1908 * care about anyway.
1910 if (m->m_len <= ETHER_CRC_LEN) {
1911 sc->rl_tail->m_len -=
1912 (ETHER_CRC_LEN - m->m_len);
1915 m->m_len -= ETHER_CRC_LEN;
1916 m->m_flags &= ~M_PKTHDR;
1917 sc->rl_tail->m_next = m;
1920 sc->rl_head = sc->rl_tail = NULL;
1921 m->m_pkthdr.len = total_len - ETHER_CRC_LEN;
1923 m->m_pkthdr.len = m->m_len =
1924 (total_len - ETHER_CRC_LEN);
1930 m->m_pkthdr.rcvif = ifp;
1932 /* Do RX checksumming if enabled */
1934 if (ifp->if_capenable & IFCAP_RXCSUM) {
1935 if ((sc->rl_flags & RL_FLAG_DESCV2) == 0) {
1936 /* Check IP header checksum */
1937 if (rxstat & RL_RDESC_STAT_PROTOID)
1938 m->m_pkthdr.csum_flags |=
1940 if (!(rxstat & RL_RDESC_STAT_IPSUMBAD))
1941 m->m_pkthdr.csum_flags |=
1944 /* Check TCP/UDP checksum */
1945 if ((RL_TCPPKT(rxstat) &&
1946 !(rxstat & RL_RDESC_STAT_TCPSUMBAD)) ||
1947 (RL_UDPPKT(rxstat) &&
1948 !(rxstat & RL_RDESC_STAT_UDPSUMBAD))) {
1949 m->m_pkthdr.csum_flags |=
1950 CSUM_DATA_VALID|CSUM_PSEUDO_HDR;
1951 m->m_pkthdr.csum_data = 0xffff;
1955 * RTL8168C/RTL816CP/RTL8111C/RTL8111CP
1957 if ((rxstat & RL_RDESC_STAT_PROTOID) &&
1958 (rxvlan & RL_RDESC_IPV4))
1959 m->m_pkthdr.csum_flags |=
1961 if (!(rxstat & RL_RDESC_STAT_IPSUMBAD) &&
1962 (rxvlan & RL_RDESC_IPV4))
1963 m->m_pkthdr.csum_flags |=
1965 if (((rxstat & RL_RDESC_STAT_TCP) &&
1966 !(rxstat & RL_RDESC_STAT_TCPSUMBAD)) ||
1967 ((rxstat & RL_RDESC_STAT_UDP) &&
1968 !(rxstat & RL_RDESC_STAT_UDPSUMBAD))) {
1969 m->m_pkthdr.csum_flags |=
1970 CSUM_DATA_VALID|CSUM_PSEUDO_HDR;
1971 m->m_pkthdr.csum_data = 0xffff;
1976 if (rxvlan & RL_RDESC_VLANCTL_TAG) {
1977 m->m_pkthdr.ether_vtag =
1978 bswap16((rxvlan & RL_RDESC_VLANCTL_DATA));
1979 m->m_flags |= M_VLANTAG;
1982 (*ifp->if_input)(ifp, m);
1986 /* Flush the RX DMA ring */
1988 bus_dmamap_sync(sc->rl_ldata.rl_rx_list_tag,
1989 sc->rl_ldata.rl_rx_list_map,
1990 BUS_DMASYNC_PREWRITE|BUS_DMASYNC_PREREAD);
1992 sc->rl_ldata.rl_rx_prodidx = i;
2001 re_txeof(struct rl_softc *sc)
2004 struct rl_txdesc *txd;
2008 cons = sc->rl_ldata.rl_tx_considx;
2009 if (cons == sc->rl_ldata.rl_tx_prodidx)
2013 /* Invalidate the TX descriptor list */
2014 bus_dmamap_sync(sc->rl_ldata.rl_tx_list_tag,
2015 sc->rl_ldata.rl_tx_list_map,
2016 BUS_DMASYNC_POSTREAD | BUS_DMASYNC_POSTWRITE);
2018 for (; cons != sc->rl_ldata.rl_tx_prodidx;
2019 cons = RL_TX_DESC_NXT(sc, cons)) {
2020 txstat = le32toh(sc->rl_ldata.rl_tx_list[cons].rl_cmdstat);
2021 if (txstat & RL_TDESC_STAT_OWN)
2024 * We only stash mbufs in the last descriptor
2025 * in a fragment chain, which also happens to
2026 * be the only place where the TX status bits
2029 if (txstat & RL_TDESC_CMD_EOF) {
2030 txd = &sc->rl_ldata.rl_tx_desc[cons];
2031 bus_dmamap_sync(sc->rl_ldata.rl_tx_mtag,
2032 txd->tx_dmamap, BUS_DMASYNC_POSTWRITE);
2033 bus_dmamap_unload(sc->rl_ldata.rl_tx_mtag,
2035 KASSERT(txd->tx_m != NULL,
2036 ("%s: freeing NULL mbufs!", __func__));
2039 if (txstat & (RL_TDESC_STAT_EXCESSCOL|
2040 RL_TDESC_STAT_COLCNT))
2041 ifp->if_collisions++;
2042 if (txstat & RL_TDESC_STAT_TXERRSUM)
2047 sc->rl_ldata.rl_tx_free++;
2048 ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
2050 sc->rl_ldata.rl_tx_considx = cons;
2052 /* No changes made to the TX ring, so no flush needed */
2054 if (sc->rl_ldata.rl_tx_free != sc->rl_ldata.rl_tx_desc_cnt) {
2055 #ifdef RE_TX_MODERATION
2057 * If not all descriptors have been reaped yet, reload
2058 * the timer so that we will eventually get another
2059 * interrupt that will cause us to re-enter this routine.
2060 * This is done in case the transmitter has gone idle.
2062 CSR_WRITE_4(sc, RL_TIMERCNT, 1);
2065 sc->rl_watchdog_timer = 0;
2071 struct rl_softc *sc;
2072 struct mii_data *mii;
2078 mii = device_get_softc(sc->rl_miibus);
2080 if ((sc->rl_flags & RL_FLAG_LINK) == 0)
2081 re_miibus_statchg(sc->rl_dev);
2083 * Reclaim transmitted frames here. Technically it is not
2084 * necessary to do here but it ensures periodic reclamation
2085 * regardless of Tx completion interrupt which seems to be
2086 * lost on PCIe based controllers under certain situations.
2090 callout_reset(&sc->rl_stat_callout, hz, re_tick, sc);
2093 #ifdef DEVICE_POLLING
2095 re_poll(struct ifnet *ifp, enum poll_cmd cmd, int count)
2097 struct rl_softc *sc = ifp->if_softc;
2100 if (ifp->if_drv_flags & IFF_DRV_RUNNING)
2101 re_poll_locked(ifp, cmd, count);
2106 re_poll_locked(struct ifnet *ifp, enum poll_cmd cmd, int count)
2108 struct rl_softc *sc = ifp->if_softc;
2112 sc->rxcycles = count;
2116 if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
2117 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_txtask);
2119 if (cmd == POLL_AND_CHECK_STATUS) { /* also check status register */
2122 status = CSR_READ_2(sc, RL_ISR);
2123 if (status == 0xffff)
2126 CSR_WRITE_2(sc, RL_ISR, status);
2127 if ((status & (RL_ISR_TX_OK | RL_ISR_TX_DESC_UNAVAIL)) &&
2128 (sc->rl_flags & RL_FLAG_PCIE))
2129 CSR_WRITE_1(sc, sc->rl_txstart, RL_TXSTART_START);
2132 * XXX check behaviour on receiver stalls.
2135 if (status & RL_ISR_SYSTEM_ERR)
2139 #endif /* DEVICE_POLLING */
2144 struct rl_softc *sc;
2149 status = CSR_READ_2(sc, RL_ISR);
2150 if (status == 0xFFFF || (status & RL_INTRS_CPLUS) == 0)
2151 return (FILTER_STRAY);
2152 CSR_WRITE_2(sc, RL_IMR, 0);
2154 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_inttask);
2156 return (FILTER_HANDLED);
2160 re_int_task(void *arg, int npending)
2162 struct rl_softc *sc;
2172 status = CSR_READ_2(sc, RL_ISR);
2173 CSR_WRITE_2(sc, RL_ISR, status);
2175 if (sc->suspended ||
2176 (ifp->if_drv_flags & IFF_DRV_RUNNING) == 0) {
2181 #ifdef DEVICE_POLLING
2182 if (ifp->if_capenable & IFCAP_POLLING) {
2188 if (status & (RL_ISR_RX_OK|RL_ISR_RX_ERR|RL_ISR_FIFO_OFLOW))
2189 rval = re_rxeof(sc);
2192 * Some chips will ignore a second TX request issued
2193 * while an existing transmission is in progress. If
2194 * the transmitter goes idle but there are still
2195 * packets waiting to be sent, we need to restart the
2196 * channel here to flush them out. This only seems to
2197 * be required with the PCIe devices.
2199 if ((status & (RL_ISR_TX_OK | RL_ISR_TX_DESC_UNAVAIL)) &&
2200 (sc->rl_flags & RL_FLAG_PCIE))
2201 CSR_WRITE_1(sc, sc->rl_txstart, RL_TXSTART_START);
2203 #ifdef RE_TX_MODERATION
2204 RL_ISR_TIMEOUT_EXPIRED|
2208 RL_ISR_TX_ERR|RL_ISR_TX_DESC_UNAVAIL))
2211 if (status & RL_ISR_SYSTEM_ERR)
2214 if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
2215 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_txtask);
2219 if ((CSR_READ_2(sc, RL_ISR) & RL_INTRS_CPLUS) || rval) {
2220 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_inttask);
2224 CSR_WRITE_2(sc, RL_IMR, RL_INTRS_CPLUS);
2228 re_encap(struct rl_softc *sc, struct mbuf **m_head)
2230 struct rl_txdesc *txd, *txd_last;
2231 bus_dma_segment_t segs[RL_NTXSEGS];
2234 struct rl_desc *desc;
2236 int i, error, ei, si;
2238 uint32_t cmdstat, csum_flags, vlanctl;
2241 M_ASSERTPKTHDR((*m_head));
2244 * With some of the RealTek chips, using the checksum offload
2245 * support in conjunction with the autopadding feature results
2246 * in the transmission of corrupt frames. For example, if we
2247 * need to send a really small IP fragment that's less than 60
2248 * bytes in size, and IP header checksumming is enabled, the
2249 * resulting ethernet frame that appears on the wire will
2250 * have garbled payload. To work around this, if TX IP checksum
2251 * offload is enabled, we always manually pad short frames out
2252 * to the minimum ethernet frame size.
2254 if ((sc->rl_flags & RL_FLAG_DESCV2) == 0 &&
2255 (*m_head)->m_pkthdr.len < RL_IP4CSUMTX_PADLEN &&
2256 ((*m_head)->m_pkthdr.csum_flags & CSUM_IP) != 0) {
2257 padlen = RL_MIN_FRAMELEN - (*m_head)->m_pkthdr.len;
2258 if (M_WRITABLE(*m_head) == 0) {
2259 /* Get a writable copy. */
2260 m_new = m_dup(*m_head, M_DONTWAIT);
2262 if (m_new == NULL) {
2268 if ((*m_head)->m_next != NULL ||
2269 M_TRAILINGSPACE(*m_head) < padlen) {
2270 m_new = m_defrag(*m_head, M_DONTWAIT);
2271 if (m_new == NULL) {
2280 * Manually pad short frames, and zero the pad space
2281 * to avoid leaking data.
2283 bzero(mtod(m_new, char *) + m_new->m_pkthdr.len, padlen);
2284 m_new->m_pkthdr.len += padlen;
2285 m_new->m_len = m_new->m_pkthdr.len;
2289 prod = sc->rl_ldata.rl_tx_prodidx;
2290 txd = &sc->rl_ldata.rl_tx_desc[prod];
2291 error = bus_dmamap_load_mbuf_sg(sc->rl_ldata.rl_tx_mtag, txd->tx_dmamap,
2292 *m_head, segs, &nsegs, BUS_DMA_NOWAIT);
2293 if (error == EFBIG) {
2294 m_new = m_collapse(*m_head, M_DONTWAIT, RL_NTXSEGS);
2295 if (m_new == NULL) {
2301 error = bus_dmamap_load_mbuf_sg(sc->rl_ldata.rl_tx_mtag,
2302 txd->tx_dmamap, *m_head, segs, &nsegs, BUS_DMA_NOWAIT);
2308 } else if (error != 0)
2316 /* Check for number of available descriptors. */
2317 if (sc->rl_ldata.rl_tx_free - nsegs <= 1) {
2318 bus_dmamap_unload(sc->rl_ldata.rl_tx_mtag, txd->tx_dmamap);
2322 bus_dmamap_sync(sc->rl_ldata.rl_tx_mtag, txd->tx_dmamap,
2323 BUS_DMASYNC_PREWRITE);
2326 * Set up checksum offload. Note: checksum offload bits must
2327 * appear in all descriptors of a multi-descriptor transmit
2328 * attempt. This is according to testing done with an 8169
2329 * chip. This is a requirement.
2333 if (((*m_head)->m_pkthdr.csum_flags & CSUM_TSO) != 0)
2334 csum_flags = RL_TDESC_CMD_LGSEND |
2335 ((uint32_t)(*m_head)->m_pkthdr.tso_segsz <<
2336 RL_TDESC_CMD_MSSVAL_SHIFT);
2339 * Unconditionally enable IP checksum if TCP or UDP
2340 * checksum is required. Otherwise, TCP/UDP checksum
2341 * does't make effects.
2343 if (((*m_head)->m_pkthdr.csum_flags & RE_CSUM_FEATURES) != 0) {
2344 if ((sc->rl_flags & RL_FLAG_DESCV2) == 0) {
2345 csum_flags |= RL_TDESC_CMD_IPCSUM;
2346 if (((*m_head)->m_pkthdr.csum_flags &
2348 csum_flags |= RL_TDESC_CMD_TCPCSUM;
2349 if (((*m_head)->m_pkthdr.csum_flags &
2351 csum_flags |= RL_TDESC_CMD_UDPCSUM;
2353 vlanctl |= RL_TDESC_CMD_IPCSUMV2;
2354 if (((*m_head)->m_pkthdr.csum_flags &
2356 vlanctl |= RL_TDESC_CMD_TCPCSUMV2;
2357 if (((*m_head)->m_pkthdr.csum_flags &
2359 vlanctl |= RL_TDESC_CMD_UDPCSUMV2;
2365 * Set up hardware VLAN tagging. Note: vlan tag info must
2366 * appear in all descriptors of a multi-descriptor
2367 * transmission attempt.
2369 if ((*m_head)->m_flags & M_VLANTAG)
2370 vlanctl |= bswap16((*m_head)->m_pkthdr.ether_vtag) |
2371 RL_TDESC_VLANCTL_TAG;
2374 for (i = 0; i < nsegs; i++, prod = RL_TX_DESC_NXT(sc, prod)) {
2375 desc = &sc->rl_ldata.rl_tx_list[prod];
2376 desc->rl_vlanctl = htole32(vlanctl);
2377 desc->rl_bufaddr_lo = htole32(RL_ADDR_LO(segs[i].ds_addr));
2378 desc->rl_bufaddr_hi = htole32(RL_ADDR_HI(segs[i].ds_addr));
2379 cmdstat = segs[i].ds_len;
2381 cmdstat |= RL_TDESC_CMD_OWN;
2382 if (prod == sc->rl_ldata.rl_tx_desc_cnt - 1)
2383 cmdstat |= RL_TDESC_CMD_EOR;
2384 desc->rl_cmdstat = htole32(cmdstat | csum_flags);
2385 sc->rl_ldata.rl_tx_free--;
2387 /* Update producer index. */
2388 sc->rl_ldata.rl_tx_prodidx = prod;
2390 /* Set EOF on the last descriptor. */
2391 ei = RL_TX_DESC_PRV(sc, prod);
2392 desc = &sc->rl_ldata.rl_tx_list[ei];
2393 desc->rl_cmdstat |= htole32(RL_TDESC_CMD_EOF);
2395 desc = &sc->rl_ldata.rl_tx_list[si];
2396 /* Set SOF and transfer ownership of packet to the chip. */
2397 desc->rl_cmdstat |= htole32(RL_TDESC_CMD_OWN | RL_TDESC_CMD_SOF);
2400 * Insure that the map for this transmission
2401 * is placed at the array index of the last descriptor
2402 * in this chain. (Swap last and first dmamaps.)
2404 txd_last = &sc->rl_ldata.rl_tx_desc[ei];
2405 map = txd->tx_dmamap;
2406 txd->tx_dmamap = txd_last->tx_dmamap;
2407 txd_last->tx_dmamap = map;
2408 txd_last->tx_m = *m_head;
2414 re_tx_task(void *arg, int npending)
2423 * Main transmit routine for C+ and gigE NICs.
2426 re_start(struct ifnet *ifp)
2428 struct rl_softc *sc;
2429 struct mbuf *m_head;
2436 if ((ifp->if_drv_flags & (IFF_DRV_RUNNING | IFF_DRV_OACTIVE)) !=
2437 IFF_DRV_RUNNING || (sc->rl_flags & RL_FLAG_LINK) == 0) {
2442 for (queued = 0; !IFQ_DRV_IS_EMPTY(&ifp->if_snd) &&
2443 sc->rl_ldata.rl_tx_free > 1;) {
2444 IFQ_DRV_DEQUEUE(&ifp->if_snd, m_head);
2448 if (re_encap(sc, &m_head) != 0) {
2451 IFQ_DRV_PREPEND(&ifp->if_snd, m_head);
2452 ifp->if_drv_flags |= IFF_DRV_OACTIVE;
2457 * If there's a BPF listener, bounce a copy of this frame
2460 ETHER_BPF_MTAP(ifp, m_head);
2466 #ifdef RE_TX_MODERATION
2467 if (sc->rl_ldata.rl_tx_free != sc->rl_ldata.rl_tx_desc_cnt)
2468 CSR_WRITE_4(sc, RL_TIMERCNT, 1);
2474 /* Flush the TX descriptors */
2476 bus_dmamap_sync(sc->rl_ldata.rl_tx_list_tag,
2477 sc->rl_ldata.rl_tx_list_map,
2478 BUS_DMASYNC_PREWRITE|BUS_DMASYNC_PREREAD);
2480 CSR_WRITE_1(sc, sc->rl_txstart, RL_TXSTART_START);
2482 #ifdef RE_TX_MODERATION
2484 * Use the countdown timer for interrupt moderation.
2485 * 'TX done' interrupts are disabled. Instead, we reset the
2486 * countdown timer, which will begin counting until it hits
2487 * the value in the TIMERINT register, and then trigger an
2488 * interrupt. Each time we write to the TIMERCNT register,
2489 * the timer count is reset to 0.
2491 CSR_WRITE_4(sc, RL_TIMERCNT, 1);
2495 * Set a timeout in case the chip goes out to lunch.
2497 sc->rl_watchdog_timer = 5;
2505 struct rl_softc *sc = xsc;
2513 re_init_locked(struct rl_softc *sc)
2515 struct ifnet *ifp = sc->rl_ifp;
2516 struct mii_data *mii;
2520 uint32_t align_dummy;
2521 u_char eaddr[ETHER_ADDR_LEN];
2526 mii = device_get_softc(sc->rl_miibus);
2529 * Cancel pending I/O and free all RX/TX buffers.
2533 /* Put controller into known state. */
2537 * Enable C+ RX and TX mode, as well as VLAN stripping and
2538 * RX checksum offload. We must configure the C+ register
2539 * before all others.
2541 cfg = RL_CPLUSCMD_PCI_MRW;
2542 if ((ifp->if_capenable & IFCAP_RXCSUM) != 0)
2543 cfg |= RL_CPLUSCMD_RXCSUM_ENB;
2544 if ((ifp->if_capenable & IFCAP_VLAN_HWTAGGING) != 0)
2545 cfg |= RL_CPLUSCMD_VLANSTRIP;
2546 if ((sc->rl_flags & RL_FLAG_MACSTAT) != 0) {
2547 cfg |= RL_CPLUSCMD_MACSTAT_DIS;
2551 cfg |= RL_CPLUSCMD_RXENB | RL_CPLUSCMD_TXENB;
2552 CSR_WRITE_2(sc, RL_CPLUS_CMD, cfg);
2553 if (sc->rl_hwrev == RL_HWREV_8169_8110SC ||
2554 sc->rl_hwrev == RL_HWREV_8169_8110SCE) {
2556 if ((CSR_READ_1(sc, RL_CFG2) & RL_CFG2_PCI66MHZ) != 0)
2558 if (sc->rl_hwrev == RL_HWREV_8169_8110SCE)
2560 CSR_WRITE_4(sc, 0x7c, reg);
2561 /* Disable interrupt mitigation. */
2562 CSR_WRITE_2(sc, 0xe2, 0);
2565 * Disable TSO if interface MTU size is greater than MSS
2566 * allowed in controller.
2568 if (ifp->if_mtu > RL_TSO_MTU && (ifp->if_capenable & IFCAP_TSO4) != 0) {
2569 ifp->if_capenable &= ~IFCAP_TSO4;
2570 ifp->if_hwassist &= ~CSUM_TSO;
2574 * Init our MAC address. Even though the chipset
2575 * documentation doesn't mention it, we need to enter "Config
2576 * register write enable" mode to modify the ID registers.
2578 /* Copy MAC address on stack to align. */
2579 bcopy(IF_LLADDR(ifp), eaddr.eaddr, ETHER_ADDR_LEN);
2580 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_WRITECFG);
2581 CSR_WRITE_4(sc, RL_IDR0,
2582 htole32(*(u_int32_t *)(&eaddr.eaddr[0])));
2583 CSR_WRITE_4(sc, RL_IDR4,
2584 htole32(*(u_int32_t *)(&eaddr.eaddr[4])));
2585 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
2588 * For C+ mode, initialize the RX descriptors and mbufs.
2590 re_rx_list_init(sc);
2591 re_tx_list_init(sc);
2594 * Load the addresses of the RX and TX lists into the chip.
2597 CSR_WRITE_4(sc, RL_RXLIST_ADDR_HI,
2598 RL_ADDR_HI(sc->rl_ldata.rl_rx_list_addr));
2599 CSR_WRITE_4(sc, RL_RXLIST_ADDR_LO,
2600 RL_ADDR_LO(sc->rl_ldata.rl_rx_list_addr));
2602 CSR_WRITE_4(sc, RL_TXLIST_ADDR_HI,
2603 RL_ADDR_HI(sc->rl_ldata.rl_tx_list_addr));
2604 CSR_WRITE_4(sc, RL_TXLIST_ADDR_LO,
2605 RL_ADDR_LO(sc->rl_ldata.rl_tx_list_addr));
2608 * Enable transmit and receive.
2610 CSR_WRITE_1(sc, RL_COMMAND, RL_CMD_TX_ENB|RL_CMD_RX_ENB);
2613 * Set the initial TX configuration.
2615 if (sc->rl_testmode) {
2616 if (sc->rl_type == RL_8169)
2617 CSR_WRITE_4(sc, RL_TXCFG,
2618 RL_TXCFG_CONFIG|RL_LOOPTEST_ON);
2620 CSR_WRITE_4(sc, RL_TXCFG,
2621 RL_TXCFG_CONFIG|RL_LOOPTEST_ON_CPLUS);
2623 CSR_WRITE_4(sc, RL_TXCFG, RL_TXCFG_CONFIG);
2625 CSR_WRITE_1(sc, RL_EARLY_TX_THRESH, 16);
2628 * Set the initial RX configuration.
2632 #ifdef DEVICE_POLLING
2634 * Disable interrupts if we are polling.
2636 if (ifp->if_capenable & IFCAP_POLLING)
2637 CSR_WRITE_2(sc, RL_IMR, 0);
2638 else /* otherwise ... */
2642 * Enable interrupts.
2644 if (sc->rl_testmode)
2645 CSR_WRITE_2(sc, RL_IMR, 0);
2647 CSR_WRITE_2(sc, RL_IMR, RL_INTRS_CPLUS);
2648 CSR_WRITE_2(sc, RL_ISR, RL_INTRS_CPLUS);
2650 /* Set initial TX threshold */
2651 sc->rl_txthresh = RL_TX_THRESH_INIT;
2653 /* Start RX/TX process. */
2654 CSR_WRITE_4(sc, RL_MISSEDPKT, 0);
2656 /* Enable receiver and transmitter. */
2657 CSR_WRITE_1(sc, RL_COMMAND, RL_CMD_TX_ENB|RL_CMD_RX_ENB);
2660 #ifdef RE_TX_MODERATION
2662 * Initialize the timer interrupt register so that
2663 * a timer interrupt will be generated once the timer
2664 * reaches a certain number of ticks. The timer is
2665 * reloaded on each transmit. This gives us TX interrupt
2666 * moderation, which dramatically improves TX frame rate.
2668 if (sc->rl_type == RL_8169)
2669 CSR_WRITE_4(sc, RL_TIMERINT_8169, 0x800);
2671 CSR_WRITE_4(sc, RL_TIMERINT, 0x400);
2675 * For 8169 gigE NICs, set the max allowed RX packet
2676 * size so we can receive jumbo frames.
2678 if (sc->rl_type == RL_8169)
2679 CSR_WRITE_2(sc, RL_MAXRXPKTLEN, 16383);
2681 if (sc->rl_testmode)
2686 CSR_WRITE_1(sc, RL_CFG1, CSR_READ_1(sc, RL_CFG1) | RL_CFG1_DRVLOAD);
2688 ifp->if_drv_flags |= IFF_DRV_RUNNING;
2689 ifp->if_drv_flags &= ~IFF_DRV_OACTIVE;
2691 sc->rl_flags &= ~RL_FLAG_LINK;
2692 sc->rl_watchdog_timer = 0;
2693 callout_reset(&sc->rl_stat_callout, hz, re_tick, sc);
2697 * Set media options.
2700 re_ifmedia_upd(struct ifnet *ifp)
2702 struct rl_softc *sc;
2703 struct mii_data *mii;
2707 mii = device_get_softc(sc->rl_miibus);
2709 error = mii_mediachg(mii);
2716 * Report current media status.
2719 re_ifmedia_sts(struct ifnet *ifp, struct ifmediareq *ifmr)
2721 struct rl_softc *sc;
2722 struct mii_data *mii;
2725 mii = device_get_softc(sc->rl_miibus);
2730 ifmr->ifm_active = mii->mii_media_active;
2731 ifmr->ifm_status = mii->mii_media_status;
2735 re_ioctl(struct ifnet *ifp, u_long command, caddr_t data)
2737 struct rl_softc *sc = ifp->if_softc;
2738 struct ifreq *ifr = (struct ifreq *) data;
2739 struct mii_data *mii;
2744 if (ifr->ifr_mtu < ETHERMIN || ifr->ifr_mtu > RL_JUMBO_MTU) {
2748 if ((sc->rl_flags & RL_FLAG_NOJUMBO) != 0 &&
2749 ifr->ifr_mtu > RL_MAX_FRAMELEN) {
2754 if (ifp->if_mtu != ifr->ifr_mtu)
2755 ifp->if_mtu = ifr->ifr_mtu;
2756 if (ifp->if_mtu > RL_TSO_MTU &&
2757 (ifp->if_capenable & IFCAP_TSO4) != 0) {
2758 ifp->if_capenable &= ~IFCAP_TSO4;
2759 ifp->if_hwassist &= ~CSUM_TSO;
2765 if ((ifp->if_flags & IFF_UP) != 0) {
2766 if ((ifp->if_drv_flags & IFF_DRV_RUNNING) != 0) {
2767 if (((ifp->if_flags ^ sc->rl_if_flags)
2768 & (IFF_PROMISC | IFF_ALLMULTI)) != 0)
2773 if ((ifp->if_drv_flags & IFF_DRV_RUNNING) != 0)
2776 sc->rl_if_flags = ifp->if_flags;
2787 mii = device_get_softc(sc->rl_miibus);
2788 error = ifmedia_ioctl(ifp, ifr, &mii->mii_media, command);
2794 mask = ifr->ifr_reqcap ^ ifp->if_capenable;
2796 #ifdef DEVICE_POLLING
2797 if (mask & IFCAP_POLLING) {
2798 if (ifr->ifr_reqcap & IFCAP_POLLING) {
2799 error = ether_poll_register(re_poll, ifp);
2803 /* Disable interrupts */
2804 CSR_WRITE_2(sc, RL_IMR, 0x0000);
2805 ifp->if_capenable |= IFCAP_POLLING;
2808 error = ether_poll_deregister(ifp);
2809 /* Enable interrupts. */
2811 CSR_WRITE_2(sc, RL_IMR, RL_INTRS_CPLUS);
2812 ifp->if_capenable &= ~IFCAP_POLLING;
2816 #endif /* DEVICE_POLLING */
2817 if (mask & IFCAP_HWCSUM) {
2818 ifp->if_capenable ^= IFCAP_HWCSUM;
2819 if (ifp->if_capenable & IFCAP_TXCSUM)
2820 ifp->if_hwassist |= RE_CSUM_FEATURES;
2822 ifp->if_hwassist &= ~RE_CSUM_FEATURES;
2825 if (mask & IFCAP_VLAN_HWTAGGING) {
2826 ifp->if_capenable ^= IFCAP_VLAN_HWTAGGING;
2829 if (mask & IFCAP_TSO4) {
2830 ifp->if_capenable ^= IFCAP_TSO4;
2831 if ((IFCAP_TSO4 & ifp->if_capenable) &&
2832 (IFCAP_TSO4 & ifp->if_capabilities))
2833 ifp->if_hwassist |= CSUM_TSO;
2835 ifp->if_hwassist &= ~CSUM_TSO;
2836 if (ifp->if_mtu > RL_TSO_MTU &&
2837 (ifp->if_capenable & IFCAP_TSO4) != 0) {
2838 ifp->if_capenable &= ~IFCAP_TSO4;
2839 ifp->if_hwassist &= ~CSUM_TSO;
2842 if ((mask & IFCAP_WOL) != 0 &&
2843 (ifp->if_capabilities & IFCAP_WOL) != 0) {
2844 if ((mask & IFCAP_WOL_UCAST) != 0)
2845 ifp->if_capenable ^= IFCAP_WOL_UCAST;
2846 if ((mask & IFCAP_WOL_MCAST) != 0)
2847 ifp->if_capenable ^= IFCAP_WOL_MCAST;
2848 if ((mask & IFCAP_WOL_MAGIC) != 0)
2849 ifp->if_capenable ^= IFCAP_WOL_MAGIC;
2851 if (reinit && ifp->if_drv_flags & IFF_DRV_RUNNING)
2853 VLAN_CAPABILITIES(ifp);
2857 error = ether_ioctl(ifp, command, data);
2865 re_watchdog(struct rl_softc *sc)
2871 if (sc->rl_watchdog_timer == 0 || --sc->rl_watchdog_timer != 0)
2876 if (sc->rl_ldata.rl_tx_free == sc->rl_ldata.rl_tx_desc_cnt) {
2877 if_printf(ifp, "watchdog timeout (missed Tx interrupts) "
2879 if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
2880 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_txtask);
2884 if_printf(ifp, "watchdog timeout\n");
2889 if (!IFQ_DRV_IS_EMPTY(&ifp->if_snd))
2890 taskqueue_enqueue_fast(taskqueue_fast, &sc->rl_txtask);
2894 * Stop the adapter and free any mbufs allocated to the
2898 re_stop(struct rl_softc *sc)
2902 struct rl_txdesc *txd;
2903 struct rl_rxdesc *rxd;
2909 sc->rl_watchdog_timer = 0;
2910 callout_stop(&sc->rl_stat_callout);
2911 ifp->if_drv_flags &= ~(IFF_DRV_RUNNING | IFF_DRV_OACTIVE);
2913 if ((sc->rl_flags & RL_FLAG_CMDSTOP) != 0)
2914 CSR_WRITE_1(sc, RL_COMMAND, RL_CMD_STOPREQ | RL_CMD_TX_ENB |
2917 CSR_WRITE_1(sc, RL_COMMAND, 0x00);
2919 CSR_WRITE_2(sc, RL_IMR, 0x0000);
2920 CSR_WRITE_2(sc, RL_ISR, 0xFFFF);
2922 if (sc->rl_head != NULL) {
2923 m_freem(sc->rl_head);
2924 sc->rl_head = sc->rl_tail = NULL;
2927 /* Free the TX list buffers. */
2929 for (i = 0; i < sc->rl_ldata.rl_tx_desc_cnt; i++) {
2930 txd = &sc->rl_ldata.rl_tx_desc[i];
2931 if (txd->tx_m != NULL) {
2932 bus_dmamap_sync(sc->rl_ldata.rl_tx_mtag,
2933 txd->tx_dmamap, BUS_DMASYNC_POSTWRITE);
2934 bus_dmamap_unload(sc->rl_ldata.rl_tx_mtag,
2941 /* Free the RX list buffers. */
2943 for (i = 0; i < sc->rl_ldata.rl_rx_desc_cnt; i++) {
2944 rxd = &sc->rl_ldata.rl_rx_desc[i];
2945 if (rxd->rx_m != NULL) {
2946 bus_dmamap_sync(sc->rl_ldata.rl_tx_mtag,
2947 rxd->rx_dmamap, BUS_DMASYNC_POSTREAD);
2948 bus_dmamap_unload(sc->rl_ldata.rl_rx_mtag,
2957 * Device suspend routine. Stop the interface and save some PCI
2958 * settings in case the BIOS doesn't restore them properly on
2962 re_suspend(device_t dev)
2964 struct rl_softc *sc;
2966 sc = device_get_softc(dev);
2978 * Device resume routine. Restore some PCI settings in case the BIOS
2979 * doesn't, re-enable busmastering, and restart the interface if
2983 re_resume(device_t dev)
2985 struct rl_softc *sc;
2988 sc = device_get_softc(dev);
2993 /* Take controller out of sleep mode. */
2994 if ((sc->rl_flags & RL_FLAG_MACSLEEP) != 0) {
2995 if ((CSR_READ_1(sc, RL_MACDBG) & 0x80) == 0x80)
2996 CSR_WRITE_1(sc, RL_GPIO,
2997 CSR_READ_1(sc, RL_GPIO) | 0x01);
3000 /* reinitialize interface if necessary */
3001 if (ifp->if_flags & IFF_UP)
3005 * Clear WOL matching such that normal Rx filtering
3006 * wouldn't interfere with WOL patterns.
3016 * Stop all chip I/O so that the kernel's probe routines don't
3017 * get confused by errant DMAs when rebooting.
3020 re_shutdown(device_t dev)
3022 struct rl_softc *sc;
3024 sc = device_get_softc(dev);
3029 * Mark interface as down since otherwise we will panic if
3030 * interrupt comes in later on, which can happen in some
3033 sc->rl_ifp->if_flags &= ~IFF_UP;
3041 re_setwol(struct rl_softc *sc)
3050 if (pci_find_extcap(sc->rl_dev, PCIY_PMG, &pmc) != 0)
3054 /* Put controller into sleep mode. */
3055 if ((sc->rl_flags & RL_FLAG_MACSLEEP) != 0) {
3056 if ((CSR_READ_1(sc, RL_MACDBG) & 0x80) == 0x80)
3057 CSR_WRITE_1(sc, RL_GPIO,
3058 CSR_READ_1(sc, RL_GPIO) & ~0x01);
3060 if ((ifp->if_capenable & IFCAP_WOL) != 0 &&
3061 (sc->rl_flags & RL_FLAG_WOLRXENB) != 0)
3062 CSR_WRITE_1(sc, RL_COMMAND, RL_CMD_RX_ENB);
3063 /* Enable config register write. */
3064 CSR_WRITE_1(sc, RL_EECMD, RL_EE_MODE);
3067 v = CSR_READ_1(sc, RL_CFG1);
3069 if ((ifp->if_capenable & IFCAP_WOL) != 0)
3071 CSR_WRITE_1(sc, RL_CFG1, v);
3073 v = CSR_READ_1(sc, RL_CFG3);
3074 v &= ~(RL_CFG3_WOL_LINK | RL_CFG3_WOL_MAGIC);
3075 if ((ifp->if_capenable & IFCAP_WOL_MAGIC) != 0)
3076 v |= RL_CFG3_WOL_MAGIC;
3077 CSR_WRITE_1(sc, RL_CFG3, v);
3079 /* Config register write done. */
3080 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
3082 v = CSR_READ_1(sc, RL_CFG5);
3083 v &= ~(RL_CFG5_WOL_BCAST | RL_CFG5_WOL_MCAST | RL_CFG5_WOL_UCAST);
3084 v &= ~RL_CFG5_WOL_LANWAKE;
3085 if ((ifp->if_capenable & IFCAP_WOL_UCAST) != 0)
3086 v |= RL_CFG5_WOL_UCAST;
3087 if ((ifp->if_capenable & IFCAP_WOL_MCAST) != 0)
3088 v |= RL_CFG5_WOL_MCAST | RL_CFG5_WOL_BCAST;
3089 if ((ifp->if_capenable & IFCAP_WOL) != 0)
3090 v |= RL_CFG5_WOL_LANWAKE;
3091 CSR_WRITE_1(sc, RL_CFG5, v);
3094 * It seems that hardware resets its link speed to 100Mbps in
3095 * power down mode so switching to 100Mbps in driver is not
3099 /* Request PME if WOL is requested. */
3100 pmstat = pci_read_config(sc->rl_dev, pmc + PCIR_POWER_STATUS, 2);
3101 pmstat &= ~(PCIM_PSTAT_PME | PCIM_PSTAT_PMEENABLE);
3102 if ((ifp->if_capenable & IFCAP_WOL) != 0)
3103 pmstat |= PCIM_PSTAT_PME | PCIM_PSTAT_PMEENABLE;
3104 pci_write_config(sc->rl_dev, pmc + PCIR_POWER_STATUS, pmstat, 2);
3108 re_clrwol(struct rl_softc *sc)
3115 if (pci_find_extcap(sc->rl_dev, PCIY_PMG, &pmc) != 0)
3118 /* Enable config register write. */
3119 CSR_WRITE_1(sc, RL_EECMD, RL_EE_MODE);
3121 v = CSR_READ_1(sc, RL_CFG3);
3122 v &= ~(RL_CFG3_WOL_LINK | RL_CFG3_WOL_MAGIC);
3123 CSR_WRITE_1(sc, RL_CFG3, v);
3125 /* Config register write done. */
3126 CSR_WRITE_1(sc, RL_EECMD, RL_EEMODE_OFF);
3128 v = CSR_READ_1(sc, RL_CFG5);
3129 v &= ~(RL_CFG5_WOL_BCAST | RL_CFG5_WOL_MCAST | RL_CFG5_WOL_UCAST);
3130 v &= ~RL_CFG5_WOL_LANWAKE;
3131 CSR_WRITE_1(sc, RL_CFG5, v);