xref: /netbsd-src/sys/dev/pci/if_dge.c (revision 404fbe5fb94ca1e054339640cabb2801ce52dd30)
1 /*	$NetBSD: if_dge.c,v 1.22 2008/11/07 00:20:07 dyoung Exp $ */
2 
3 /*
4  * Copyright (c) 2004, SUNET, Swedish University Computer Network.
5  * All rights reserved.
6  *
7  * Written by Anders Magnusson for SUNET, Swedish University Computer Network.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  *	This product includes software developed for the NetBSD Project by
20  *	SUNET, Swedish University Computer Network.
21  * 4. The name of SUNET may not be used to endorse or promote products
22  *    derived from this software without specific prior written permission.
23  *
24  * THIS SOFTWARE IS PROVIDED BY SUNET ``AS IS'' AND
25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
26  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
27  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL WASABI SYSTEMS, INC
28  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
29  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
30  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
32  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
33  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34  * POSSIBILITY OF SUCH DAMAGE.
35  */
36 
37 /*
38  * Copyright (c) 2001, 2002, 2003 Wasabi Systems, Inc.
39  * All rights reserved.
40  *
41  * Written by Jason R. Thorpe for Wasabi Systems, Inc.
42  *
43  * Redistribution and use in source and binary forms, with or without
44  * modification, are permitted provided that the following conditions
45  * are met:
46  * 1. Redistributions of source code must retain the above copyright
47  *    notice, this list of conditions and the following disclaimer.
48  * 2. Redistributions in binary form must reproduce the above copyright
49  *    notice, this list of conditions and the following disclaimer in the
50  *    documentation and/or other materials provided with the distribution.
51  * 3. All advertising materials mentioning features or use of this software
52  *    must display the following acknowledgement:
53  *	This product includes software developed for the NetBSD Project by
54  *	Wasabi Systems, Inc.
55  * 4. The name of Wasabi Systems, Inc. may not be used to endorse
56  *    or promote products derived from this software without specific prior
57  *    written permission.
58  *
59  * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``AS IS'' AND
60  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
61  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
62  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL WASABI SYSTEMS, INC
63  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
64  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
65  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
66  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
67  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
68  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
69  * POSSIBILITY OF SUCH DAMAGE.
70  */
71 
72 /*
73  * Device driver for the Intel 82597EX Ten Gigabit Ethernet controller.
74  *
75  * TODO (in no specific order):
76  *	HW VLAN support.
77  *	TSE offloading (needs kernel changes...)
78  *	RAIDC (receive interrupt delay adaptation)
79  *	Use memory > 4GB.
80  */
81 
82 #include <sys/cdefs.h>
83 __KERNEL_RCSID(0, "$NetBSD: if_dge.c,v 1.22 2008/11/07 00:20:07 dyoung Exp $");
84 
85 #include "bpfilter.h"
86 #include "rnd.h"
87 
88 #include <sys/param.h>
89 #include <sys/systm.h>
90 #include <sys/callout.h>
91 #include <sys/mbuf.h>
92 #include <sys/malloc.h>
93 #include <sys/kernel.h>
94 #include <sys/socket.h>
95 #include <sys/ioctl.h>
96 #include <sys/errno.h>
97 #include <sys/device.h>
98 #include <sys/queue.h>
99 
100 #include <uvm/uvm_extern.h>		/* for PAGE_SIZE */
101 
102 #if NRND > 0
103 #include <sys/rnd.h>
104 #endif
105 
106 #include <net/if.h>
107 #include <net/if_dl.h>
108 #include <net/if_media.h>
109 #include <net/if_ether.h>
110 
111 #if NBPFILTER > 0
112 #include <net/bpf.h>
113 #endif
114 
115 #include <netinet/in.h>			/* XXX for struct ip */
116 #include <netinet/in_systm.h>		/* XXX for struct ip */
117 #include <netinet/ip.h>			/* XXX for struct ip */
118 #include <netinet/tcp.h>		/* XXX for struct tcphdr */
119 
120 #include <sys/bus.h>
121 #include <sys/intr.h>
122 #include <machine/endian.h>
123 
124 #include <dev/mii/mii.h>
125 #include <dev/mii/miivar.h>
126 #include <dev/mii/mii_bitbang.h>
127 
128 #include <dev/pci/pcireg.h>
129 #include <dev/pci/pcivar.h>
130 #include <dev/pci/pcidevs.h>
131 
132 #include <dev/pci/if_dgereg.h>
133 
134 /*
135  * The receive engine may sometimes become off-by-one when writing back
136  * chained descriptors.	 Avoid this by allocating a large chunk of
137  * memory and use if instead (to avoid chained descriptors).
138  * This only happens with chained descriptors under heavy load.
139  */
140 #define DGE_OFFBYONE_RXBUG
141 
142 #define DGE_EVENT_COUNTERS
143 #define DGE_DEBUG
144 
145 #ifdef DGE_DEBUG
146 #define DGE_DEBUG_LINK		0x01
147 #define DGE_DEBUG_TX		0x02
148 #define DGE_DEBUG_RX		0x04
149 #define DGE_DEBUG_CKSUM		0x08
150 int	dge_debug = 0;
151 
152 #define DPRINTF(x, y)	if (dge_debug & (x)) printf y
153 #else
154 #define DPRINTF(x, y)	/* nothing */
155 #endif /* DGE_DEBUG */
156 
157 /*
158  * Transmit descriptor list size. We allow up to 100 DMA segments per
159  * packet (Intel reports of jumbo frame packets with as
160  * many as 80 DMA segments when using 16k buffers).
161  */
162 #define DGE_NTXSEGS		100
163 #define DGE_IFQUEUELEN		20000
164 #define DGE_TXQUEUELEN		2048
165 #define DGE_TXQUEUELEN_MASK	(DGE_TXQUEUELEN - 1)
166 #define DGE_TXQUEUE_GC		(DGE_TXQUEUELEN / 8)
167 #define DGE_NTXDESC		1024
168 #define DGE_NTXDESC_MASK		(DGE_NTXDESC - 1)
169 #define DGE_NEXTTX(x)		(((x) + 1) & DGE_NTXDESC_MASK)
170 #define DGE_NEXTTXS(x)		(((x) + 1) & DGE_TXQUEUELEN_MASK)
171 
172 /*
173  * Receive descriptor list size.
174  * Packet is of size MCLBYTES, and for jumbo packets buffers may
175  * be chained.	Due to the nature of the card (high-speed), keep this
176  * ring large. With 2k buffers the ring can store 400 jumbo packets,
177  * which at full speed will be received in just under 3ms.
178  */
179 #define DGE_NRXDESC		2048
180 #define DGE_NRXDESC_MASK	(DGE_NRXDESC - 1)
181 #define DGE_NEXTRX(x)		(((x) + 1) & DGE_NRXDESC_MASK)
182 /*
183  * # of descriptors between head and written descriptors.
184  * This is to work-around two erratas.
185  */
186 #define DGE_RXSPACE		10
187 #define DGE_PREVRX(x)		(((x) - DGE_RXSPACE) & DGE_NRXDESC_MASK)
188 /*
189  * Receive descriptor fetch threshholds. These are values recommended
190  * by Intel, do not touch them unless you know what you are doing.
191  */
192 #define RXDCTL_PTHRESH_VAL	128
193 #define RXDCTL_HTHRESH_VAL	16
194 #define RXDCTL_WTHRESH_VAL	16
195 
196 
197 /*
198  * Tweakable parameters; default values.
199  */
200 #define FCRTH	0x30000 /* Send XOFF water mark */
201 #define FCRTL	0x28000 /* Send XON water mark */
202 #define RDTR	0x20	/* Interrupt delay after receive, .8192us units */
203 #define TIDV	0x20	/* Interrupt delay after send, .8192us units */
204 
205 /*
206  * Control structures are DMA'd to the i82597 chip.  We allocate them in
207  * a single clump that maps to a single DMA segment to make serveral things
208  * easier.
209  */
210 struct dge_control_data {
211 	/*
212 	 * The transmit descriptors.
213 	 */
214 	struct dge_tdes wcd_txdescs[DGE_NTXDESC];
215 
216 	/*
217 	 * The receive descriptors.
218 	 */
219 	struct dge_rdes wcd_rxdescs[DGE_NRXDESC];
220 };
221 
222 #define DGE_CDOFF(x)	offsetof(struct dge_control_data, x)
223 #define DGE_CDTXOFF(x)	DGE_CDOFF(wcd_txdescs[(x)])
224 #define DGE_CDRXOFF(x)	DGE_CDOFF(wcd_rxdescs[(x)])
225 
226 /*
227  * The DGE interface have a higher max MTU size than normal jumbo frames.
228  */
229 #define DGE_MAX_MTU	16288	/* Max MTU size for this interface */
230 
231 /*
232  * Software state for transmit jobs.
233  */
234 struct dge_txsoft {
235 	struct mbuf *txs_mbuf;		/* head of our mbuf chain */
236 	bus_dmamap_t txs_dmamap;	/* our DMA map */
237 	int txs_firstdesc;		/* first descriptor in packet */
238 	int txs_lastdesc;		/* last descriptor in packet */
239 	int txs_ndesc;			/* # of descriptors used */
240 };
241 
242 /*
243  * Software state for receive buffers.	Each descriptor gets a
244  * 2k (MCLBYTES) buffer and a DMA map.	For packets which fill
245  * more than one buffer, we chain them together.
246  */
247 struct dge_rxsoft {
248 	struct mbuf *rxs_mbuf;		/* head of our mbuf chain */
249 	bus_dmamap_t rxs_dmamap;	/* our DMA map */
250 };
251 
252 /*
253  * Software state per device.
254  */
255 struct dge_softc {
256 	struct device sc_dev;		/* generic device information */
257 	bus_space_tag_t sc_st;		/* bus space tag */
258 	bus_space_handle_t sc_sh;	/* bus space handle */
259 	bus_dma_tag_t sc_dmat;		/* bus DMA tag */
260 	struct ethercom sc_ethercom;	/* ethernet common data */
261 	void *sc_sdhook;		/* shutdown hook */
262 
263 	int sc_flags;			/* flags; see below */
264 	int sc_bus_speed;		/* PCI/PCIX bus speed */
265 	int sc_pcix_offset;		/* PCIX capability register offset */
266 
267 	pci_chipset_tag_t sc_pc;
268 	pcitag_t sc_pt;
269 	int sc_mmrbc;			/* Max PCIX memory read byte count */
270 
271 	void *sc_ih;			/* interrupt cookie */
272 
273 	struct ifmedia sc_media;
274 
275 	bus_dmamap_t sc_cddmamap;	/* control data DMA map */
276 #define sc_cddma	sc_cddmamap->dm_segs[0].ds_addr
277 
278 	int		sc_align_tweak;
279 
280 	/*
281 	 * Software state for the transmit and receive descriptors.
282 	 */
283 	struct dge_txsoft sc_txsoft[DGE_TXQUEUELEN];
284 	struct dge_rxsoft sc_rxsoft[DGE_NRXDESC];
285 
286 	/*
287 	 * Control data structures.
288 	 */
289 	struct dge_control_data *sc_control_data;
290 #define sc_txdescs	sc_control_data->wcd_txdescs
291 #define sc_rxdescs	sc_control_data->wcd_rxdescs
292 
293 #ifdef DGE_EVENT_COUNTERS
294 	/* Event counters. */
295 	struct evcnt sc_ev_txsstall;	/* Tx stalled due to no txs */
296 	struct evcnt sc_ev_txdstall;	/* Tx stalled due to no txd */
297 	struct evcnt sc_ev_txforceintr; /* Tx interrupts forced */
298 	struct evcnt sc_ev_txdw;	/* Tx descriptor interrupts */
299 	struct evcnt sc_ev_txqe;	/* Tx queue empty interrupts */
300 	struct evcnt sc_ev_rxintr;	/* Rx interrupts */
301 	struct evcnt sc_ev_linkintr;	/* Link interrupts */
302 
303 	struct evcnt sc_ev_rxipsum;	/* IP checksums checked in-bound */
304 	struct evcnt sc_ev_rxtusum;	/* TCP/UDP cksums checked in-bound */
305 	struct evcnt sc_ev_txipsum;	/* IP checksums comp. out-bound */
306 	struct evcnt sc_ev_txtusum;	/* TCP/UDP cksums comp. out-bound */
307 
308 	struct evcnt sc_ev_txctx_init;	/* Tx cksum context cache initialized */
309 	struct evcnt sc_ev_txctx_hit;	/* Tx cksum context cache hit */
310 	struct evcnt sc_ev_txctx_miss;	/* Tx cksum context cache miss */
311 
312 	struct evcnt sc_ev_txseg[DGE_NTXSEGS]; /* Tx packets w/ N segments */
313 	struct evcnt sc_ev_txdrop;	/* Tx packets dropped (too many segs) */
314 #endif /* DGE_EVENT_COUNTERS */
315 
316 	int	sc_txfree;		/* number of free Tx descriptors */
317 	int	sc_txnext;		/* next ready Tx descriptor */
318 
319 	int	sc_txsfree;		/* number of free Tx jobs */
320 	int	sc_txsnext;		/* next free Tx job */
321 	int	sc_txsdirty;		/* dirty Tx jobs */
322 
323 	uint32_t sc_txctx_ipcs;		/* cached Tx IP cksum ctx */
324 	uint32_t sc_txctx_tucs;		/* cached Tx TCP/UDP cksum ctx */
325 
326 	int	sc_rxptr;		/* next ready Rx descriptor/queue ent */
327 	int	sc_rxdiscard;
328 	int	sc_rxlen;
329 	struct mbuf *sc_rxhead;
330 	struct mbuf *sc_rxtail;
331 	struct mbuf **sc_rxtailp;
332 
333 	uint32_t sc_ctrl0;		/* prototype CTRL0 register */
334 	uint32_t sc_icr;		/* prototype interrupt bits */
335 	uint32_t sc_tctl;		/* prototype TCTL register */
336 	uint32_t sc_rctl;		/* prototype RCTL register */
337 
338 	int sc_mchash_type;		/* multicast filter offset */
339 
340 	uint16_t sc_eeprom[EEPROM_SIZE];
341 
342 #if NRND > 0
343 	rndsource_element_t rnd_source; /* random source */
344 #endif
345 #ifdef DGE_OFFBYONE_RXBUG
346 	void *sc_bugbuf;
347 	SLIST_HEAD(, rxbugentry) sc_buglist;
348 	bus_dmamap_t sc_bugmap;
349 	struct rxbugentry *sc_entry;
350 #endif
351 };
352 
353 #define DGE_RXCHAIN_RESET(sc)						\
354 do {									\
355 	(sc)->sc_rxtailp = &(sc)->sc_rxhead;				\
356 	*(sc)->sc_rxtailp = NULL;					\
357 	(sc)->sc_rxlen = 0;						\
358 } while (/*CONSTCOND*/0)
359 
360 #define DGE_RXCHAIN_LINK(sc, m)						\
361 do {									\
362 	*(sc)->sc_rxtailp = (sc)->sc_rxtail = (m);			\
363 	(sc)->sc_rxtailp = &(m)->m_next;				\
364 } while (/*CONSTCOND*/0)
365 
366 /* sc_flags */
367 #define DGE_F_BUS64		0x20	/* bus is 64-bit */
368 #define DGE_F_PCIX		0x40	/* bus is PCI-X */
369 
370 #ifdef DGE_EVENT_COUNTERS
371 #define DGE_EVCNT_INCR(ev)	(ev)->ev_count++
372 #else
373 #define DGE_EVCNT_INCR(ev)	/* nothing */
374 #endif
375 
376 #define CSR_READ(sc, reg)						\
377 	bus_space_read_4((sc)->sc_st, (sc)->sc_sh, (reg))
378 #define CSR_WRITE(sc, reg, val)						\
379 	bus_space_write_4((sc)->sc_st, (sc)->sc_sh, (reg), (val))
380 
381 #define DGE_CDTXADDR(sc, x)	((sc)->sc_cddma + DGE_CDTXOFF((x)))
382 #define DGE_CDRXADDR(sc, x)	((sc)->sc_cddma + DGE_CDRXOFF((x)))
383 
384 #define DGE_CDTXSYNC(sc, x, n, ops)					\
385 do {									\
386 	int __x, __n;							\
387 									\
388 	__x = (x);							\
389 	__n = (n);							\
390 									\
391 	/* If it will wrap around, sync to the end of the ring. */	\
392 	if ((__x + __n) > DGE_NTXDESC) {				\
393 		bus_dmamap_sync((sc)->sc_dmat, (sc)->sc_cddmamap,	\
394 		    DGE_CDTXOFF(__x), sizeof(struct dge_tdes) *		\
395 		    (DGE_NTXDESC - __x), (ops));			\
396 		__n -= (DGE_NTXDESC - __x);				\
397 		__x = 0;						\
398 	}								\
399 									\
400 	/* Now sync whatever is left. */				\
401 	bus_dmamap_sync((sc)->sc_dmat, (sc)->sc_cddmamap,		\
402 	    DGE_CDTXOFF(__x), sizeof(struct dge_tdes) * __n, (ops));	\
403 } while (/*CONSTCOND*/0)
404 
405 #define DGE_CDRXSYNC(sc, x, ops)						\
406 do {									\
407 	bus_dmamap_sync((sc)->sc_dmat, (sc)->sc_cddmamap,		\
408 	   DGE_CDRXOFF((x)), sizeof(struct dge_rdes), (ops));		\
409 } while (/*CONSTCOND*/0)
410 
411 #ifdef DGE_OFFBYONE_RXBUG
412 #define DGE_INIT_RXDESC(sc, x)						\
413 do {									\
414 	struct dge_rxsoft *__rxs = &(sc)->sc_rxsoft[(x)];		\
415 	struct dge_rdes *__rxd = &(sc)->sc_rxdescs[(x)];		\
416 	struct mbuf *__m = __rxs->rxs_mbuf;				\
417 									\
418 	__rxd->dr_baddrl = htole32(sc->sc_bugmap->dm_segs[0].ds_addr +	\
419 	    (mtod((__m), char *) - (char *)sc->sc_bugbuf));		\
420 	__rxd->dr_baddrh = 0;						\
421 	__rxd->dr_len = 0;						\
422 	__rxd->dr_cksum = 0;						\
423 	__rxd->dr_status = 0;						\
424 	__rxd->dr_errors = 0;						\
425 	__rxd->dr_special = 0;						\
426 	DGE_CDRXSYNC((sc), (x), BUS_DMASYNC_PREREAD|BUS_DMASYNC_PREWRITE); \
427 									\
428 	CSR_WRITE((sc), DGE_RDT, (x));					\
429 } while (/*CONSTCOND*/0)
430 #else
431 #define DGE_INIT_RXDESC(sc, x)						\
432 do {									\
433 	struct dge_rxsoft *__rxs = &(sc)->sc_rxsoft[(x)];		\
434 	struct dge_rdes *__rxd = &(sc)->sc_rxdescs[(x)];		\
435 	struct mbuf *__m = __rxs->rxs_mbuf;				\
436 									\
437 	/*								\
438 	 * Note: We scoot the packet forward 2 bytes in the buffer	\
439 	 * so that the payload after the Ethernet header is aligned	\
440 	 * to a 4-byte boundary.					\
441 	 *								\
442 	 * XXX BRAINDAMAGE ALERT!					\
443 	 * The stupid chip uses the same size for every buffer, which	\
444 	 * is set in the Receive Control register.  We are using the 2K \
445 	 * size option, but what we REALLY want is (2K - 2)!  For this	\
446 	 * reason, we can't "scoot" packets longer than the standard	\
447 	 * Ethernet MTU.  On strict-alignment platforms, if the total	\
448 	 * size exceeds (2K - 2) we set align_tweak to 0 and let	\
449 	 * the upper layer copy the headers.				\
450 	 */								\
451 	__m->m_data = __m->m_ext.ext_buf + (sc)->sc_align_tweak;	\
452 									\
453 	__rxd->dr_baddrl =						\
454 	    htole32(__rxs->rxs_dmamap->dm_segs[0].ds_addr +		\
455 		(sc)->sc_align_tweak);					\
456 	__rxd->dr_baddrh = 0;						\
457 	__rxd->dr_len = 0;						\
458 	__rxd->dr_cksum = 0;						\
459 	__rxd->dr_status = 0;						\
460 	__rxd->dr_errors = 0;						\
461 	__rxd->dr_special = 0;						\
462 	DGE_CDRXSYNC((sc), (x), BUS_DMASYNC_PREREAD|BUS_DMASYNC_PREWRITE); \
463 									\
464 	CSR_WRITE((sc), DGE_RDT, (x));					\
465 } while (/*CONSTCOND*/0)
466 #endif
467 
468 #ifdef DGE_OFFBYONE_RXBUG
469 /*
470  * Allocation constants.  Much memory may be used for this.
471  */
472 #ifndef DGE_BUFFER_SIZE
473 #define DGE_BUFFER_SIZE DGE_MAX_MTU
474 #endif
475 #define DGE_NBUFFERS	(4*DGE_NRXDESC)
476 #define DGE_RXMEM	(DGE_NBUFFERS*DGE_BUFFER_SIZE)
477 
478 struct rxbugentry {
479 	SLIST_ENTRY(rxbugentry) rb_entry;
480 	int rb_slot;
481 };
482 
483 static int
484 dge_alloc_rcvmem(struct dge_softc *sc)
485 {
486 	char *ptr, *kva;
487 	bus_dma_segment_t seg;
488 	int i, rseg, state, error;
489 	struct rxbugentry *entry;
490 
491 	state = error = 0;
492 
493 	if (bus_dmamem_alloc(sc->sc_dmat, DGE_RXMEM, PAGE_SIZE, 0,
494 	     &seg, 1, &rseg, BUS_DMA_NOWAIT)) {
495 		aprint_error_dev(&sc->sc_dev, "can't alloc rx buffers\n");
496 		return ENOBUFS;
497 	}
498 
499 	state = 1;
500 	if (bus_dmamem_map(sc->sc_dmat, &seg, rseg, DGE_RXMEM, (void **)&kva,
501 	    BUS_DMA_NOWAIT)) {
502 		aprint_error_dev(&sc->sc_dev, "can't map DMA buffers (%d bytes)\n",
503 		    (int)DGE_RXMEM);
504 		error = ENOBUFS;
505 		goto out;
506 	}
507 
508 	state = 2;
509 	if (bus_dmamap_create(sc->sc_dmat, DGE_RXMEM, 1, DGE_RXMEM, 0,
510 	    BUS_DMA_NOWAIT, &sc->sc_bugmap)) {
511 		aprint_error_dev(&sc->sc_dev, "can't create DMA map\n");
512 		error = ENOBUFS;
513 		goto out;
514 	}
515 
516 	state = 3;
517 	if (bus_dmamap_load(sc->sc_dmat, sc->sc_bugmap,
518 	    kva, DGE_RXMEM, NULL, BUS_DMA_NOWAIT)) {
519 		aprint_error_dev(&sc->sc_dev, "can't load DMA map\n");
520 		error = ENOBUFS;
521 		goto out;
522 	}
523 
524 	state = 4;
525 	sc->sc_bugbuf = (void *)kva;
526 	SLIST_INIT(&sc->sc_buglist);
527 
528 	/*
529 	 * Now divide it up into DGE_BUFFER_SIZE pieces and save the addresses
530 	 * in an array.
531 	 */
532 	ptr = sc->sc_bugbuf;
533 	if ((entry = malloc(sizeof(*entry) * DGE_NBUFFERS,
534 	    M_DEVBUF, M_NOWAIT)) == NULL) {
535 		error = ENOBUFS;
536 		goto out;
537 	}
538 	sc->sc_entry = entry;
539 	for (i = 0; i < DGE_NBUFFERS; i++) {
540 		entry[i].rb_slot = i;
541 		SLIST_INSERT_HEAD(&sc->sc_buglist, &entry[i], rb_entry);
542 	}
543 out:
544 	if (error != 0) {
545 		switch (state) {
546 		case 4:
547 			bus_dmamap_unload(sc->sc_dmat, sc->sc_bugmap);
548 		case 3:
549 			bus_dmamap_destroy(sc->sc_dmat, sc->sc_bugmap);
550 		case 2:
551 			bus_dmamem_unmap(sc->sc_dmat, kva, DGE_RXMEM);
552 		case 1:
553 			bus_dmamem_free(sc->sc_dmat, &seg, rseg);
554 			break;
555 		default:
556 			break;
557 		}
558 	}
559 
560 	return error;
561 }
562 
563 /*
564  * Allocate a jumbo buffer.
565  */
566 static void *
567 dge_getbuf(struct dge_softc *sc)
568 {
569 	struct rxbugentry *entry;
570 
571 	entry = SLIST_FIRST(&sc->sc_buglist);
572 
573 	if (entry == NULL) {
574 		printf("%s: no free RX buffers\n", device_xname(&sc->sc_dev));
575 		return(NULL);
576 	}
577 
578 	SLIST_REMOVE_HEAD(&sc->sc_buglist, rb_entry);
579 	return (char *)sc->sc_bugbuf + entry->rb_slot * DGE_BUFFER_SIZE;
580 }
581 
582 /*
583  * Release a jumbo buffer.
584  */
585 static void
586 dge_freebuf(struct mbuf *m, void *buf, size_t size, void *arg)
587 {
588 	struct rxbugentry *entry;
589 	struct dge_softc *sc;
590 	int i, s;
591 
592 	/* Extract the softc struct pointer. */
593 	sc = (struct dge_softc *)arg;
594 
595 	if (sc == NULL)
596 		panic("dge_freebuf: can't find softc pointer!");
597 
598 	/* calculate the slot this buffer belongs to */
599 
600 	i = ((char *)buf - (char *)sc->sc_bugbuf) / DGE_BUFFER_SIZE;
601 
602 	if ((i < 0) || (i >= DGE_NBUFFERS))
603 		panic("dge_freebuf: asked to free buffer %d!", i);
604 
605 	s = splvm();
606 	entry = sc->sc_entry + i;
607 	SLIST_INSERT_HEAD(&sc->sc_buglist, entry, rb_entry);
608 
609 	if (__predict_true(m != NULL))
610 		pool_cache_put(mb_cache, m);
611 	splx(s);
612 }
613 #endif
614 
615 static void	dge_start(struct ifnet *);
616 static void	dge_watchdog(struct ifnet *);
617 static int	dge_ioctl(struct ifnet *, u_long, void *);
618 static int	dge_init(struct ifnet *);
619 static void	dge_stop(struct ifnet *, int);
620 
621 static void	dge_shutdown(void *);
622 
623 static void	dge_reset(struct dge_softc *);
624 static void	dge_rxdrain(struct dge_softc *);
625 static int	dge_add_rxbuf(struct dge_softc *, int);
626 
627 static void	dge_set_filter(struct dge_softc *);
628 
629 static int	dge_intr(void *);
630 static void	dge_txintr(struct dge_softc *);
631 static void	dge_rxintr(struct dge_softc *);
632 static void	dge_linkintr(struct dge_softc *, uint32_t);
633 
634 static int	dge_match(struct device *, struct cfdata *, void *);
635 static void	dge_attach(struct device *, struct device *, void *);
636 
637 static int	dge_read_eeprom(struct dge_softc *sc);
638 static int	dge_eeprom_clockin(struct dge_softc *sc);
639 static void	dge_eeprom_clockout(struct dge_softc *sc, int bit);
640 static uint16_t	dge_eeprom_word(struct dge_softc *sc, int addr);
641 static int	dge_xgmii_mediachange(struct ifnet *);
642 static void	dge_xgmii_mediastatus(struct ifnet *, struct ifmediareq *);
643 static void	dge_xgmii_reset(struct dge_softc *);
644 static void	dge_xgmii_writereg(struct device *, int, int, int);
645 
646 
647 CFATTACH_DECL(dge, sizeof(struct dge_softc),
648     dge_match, dge_attach, NULL, NULL);
649 
650 #ifdef DGE_EVENT_COUNTERS
651 #if DGE_NTXSEGS > 100
652 #error Update dge_txseg_evcnt_names
653 #endif
654 static char (*dge_txseg_evcnt_names)[DGE_NTXSEGS][8 /* "txseg00" + \0 */];
655 #endif /* DGE_EVENT_COUNTERS */
656 
657 static int
658 dge_match(struct device *parent, struct cfdata *cf, void *aux)
659 {
660 	struct pci_attach_args *pa = aux;
661 
662 	if (PCI_VENDOR(pa->pa_id) == PCI_VENDOR_INTEL &&
663 	    PCI_PRODUCT(pa->pa_id) == PCI_PRODUCT_INTEL_82597EX)
664 		return (1);
665 
666 	return (0);
667 }
668 
669 static void
670 dge_attach(struct device *parent, struct device *self, void *aux)
671 {
672 	struct dge_softc *sc = (void *) self;
673 	struct pci_attach_args *pa = aux;
674 	struct ifnet *ifp = &sc->sc_ethercom.ec_if;
675 	pci_chipset_tag_t pc = pa->pa_pc;
676 	pci_intr_handle_t ih;
677 	const char *intrstr = NULL;
678 	bus_dma_segment_t seg;
679 	int i, rseg, error;
680 	uint8_t enaddr[ETHER_ADDR_LEN];
681 	pcireg_t preg, memtype;
682 	uint32_t reg;
683 
684 	sc->sc_dmat = pa->pa_dmat;
685 	sc->sc_pc = pa->pa_pc;
686 	sc->sc_pt = pa->pa_tag;
687 
688 	preg = PCI_REVISION(pci_conf_read(pc, pa->pa_tag, PCI_CLASS_REG));
689 	aprint_naive(": Ethernet controller\n");
690 	aprint_normal(": Intel i82597EX 10GbE-LR Ethernet, rev. %d\n", preg);
691 
692 	memtype = pci_mapreg_type(pa->pa_pc, pa->pa_tag, DGE_PCI_BAR);
693         if (pci_mapreg_map(pa, DGE_PCI_BAR, memtype, 0,
694             &sc->sc_st, &sc->sc_sh, NULL, NULL)) {
695                 aprint_error_dev(&sc->sc_dev, "unable to map device registers\n");
696                 return;
697         }
698 
699 	/* Enable bus mastering */
700 	preg = pci_conf_read(pc, pa->pa_tag, PCI_COMMAND_STATUS_REG);
701 	preg |= PCI_COMMAND_MASTER_ENABLE;
702 	pci_conf_write(pc, pa->pa_tag, PCI_COMMAND_STATUS_REG, preg);
703 
704 	/*
705 	 * Map and establish our interrupt.
706 	 */
707 	if (pci_intr_map(pa, &ih)) {
708 		aprint_error_dev(&sc->sc_dev, "unable to map interrupt\n");
709 		return;
710 	}
711 	intrstr = pci_intr_string(pc, ih);
712 	sc->sc_ih = pci_intr_establish(pc, ih, IPL_NET, dge_intr, sc);
713 	if (sc->sc_ih == NULL) {
714 		aprint_error_dev(&sc->sc_dev, "unable to establish interrupt");
715 		if (intrstr != NULL)
716 			aprint_normal(" at %s", intrstr);
717 		aprint_normal("\n");
718 		return;
719 	}
720 	aprint_normal_dev(&sc->sc_dev, "interrupting at %s\n", intrstr);
721 
722 	/*
723 	 * Determine a few things about the bus we're connected to.
724 	 */
725 	reg = CSR_READ(sc, DGE_STATUS);
726 	if (reg & STATUS_BUS64)
727 		sc->sc_flags |= DGE_F_BUS64;
728 
729 	sc->sc_flags |= DGE_F_PCIX;
730 	if (pci_get_capability(pa->pa_pc, pa->pa_tag,
731 			       PCI_CAP_PCIX,
732 			       &sc->sc_pcix_offset, NULL) == 0)
733 		aprint_error_dev(&sc->sc_dev, "unable to find PCIX "
734 		    "capability\n");
735 
736 	if (sc->sc_flags & DGE_F_PCIX) {
737 		switch (reg & STATUS_PCIX_MSK) {
738 		case STATUS_PCIX_66:
739 			sc->sc_bus_speed = 66;
740 			break;
741 		case STATUS_PCIX_100:
742 			sc->sc_bus_speed = 100;
743 			break;
744 		case STATUS_PCIX_133:
745 			sc->sc_bus_speed = 133;
746 			break;
747 		default:
748 			aprint_error_dev(&sc->sc_dev,
749 			    "unknown PCIXSPD %d; assuming 66MHz\n",
750 			    reg & STATUS_PCIX_MSK);
751 			sc->sc_bus_speed = 66;
752 		}
753 	} else
754 		sc->sc_bus_speed = (reg & STATUS_BUS64) ? 66 : 33;
755 	aprint_verbose_dev(&sc->sc_dev, "%d-bit %dMHz %s bus\n",
756 	    (sc->sc_flags & DGE_F_BUS64) ? 64 : 32, sc->sc_bus_speed,
757 	    (sc->sc_flags & DGE_F_PCIX) ? "PCIX" : "PCI");
758 
759 	/*
760 	 * Allocate the control data structures, and create and load the
761 	 * DMA map for it.
762 	 */
763 	if ((error = bus_dmamem_alloc(sc->sc_dmat,
764 	    sizeof(struct dge_control_data), PAGE_SIZE, 0, &seg, 1, &rseg,
765 	    0)) != 0) {
766 		aprint_error_dev(&sc->sc_dev,
767 		    "unable to allocate control data, error = %d\n",
768 		    error);
769 		goto fail_0;
770 	}
771 
772 	if ((error = bus_dmamem_map(sc->sc_dmat, &seg, rseg,
773 	    sizeof(struct dge_control_data), (void **)&sc->sc_control_data,
774 	    0)) != 0) {
775 		aprint_error_dev(&sc->sc_dev, "unable to map control data, error = %d\n",
776 		    error);
777 		goto fail_1;
778 	}
779 
780 	if ((error = bus_dmamap_create(sc->sc_dmat,
781 	    sizeof(struct dge_control_data), 1,
782 	    sizeof(struct dge_control_data), 0, 0, &sc->sc_cddmamap)) != 0) {
783 		aprint_error_dev(&sc->sc_dev, "unable to create control data DMA map, "
784 		    "error = %d\n", error);
785 		goto fail_2;
786 	}
787 
788 	if ((error = bus_dmamap_load(sc->sc_dmat, sc->sc_cddmamap,
789 	    sc->sc_control_data, sizeof(struct dge_control_data), NULL,
790 	    0)) != 0) {
791 		aprint_error_dev(&sc->sc_dev,
792 		    "unable to load control data DMA map, error = %d\n",
793 		    error);
794 		goto fail_3;
795 	}
796 
797 #ifdef DGE_OFFBYONE_RXBUG
798 	if (dge_alloc_rcvmem(sc) != 0)
799 		return; /* Already complained */
800 #endif
801 	/*
802 	 * Create the transmit buffer DMA maps.
803 	 */
804 	for (i = 0; i < DGE_TXQUEUELEN; i++) {
805 		if ((error = bus_dmamap_create(sc->sc_dmat, DGE_MAX_MTU,
806 		    DGE_NTXSEGS, MCLBYTES, 0, 0,
807 		    &sc->sc_txsoft[i].txs_dmamap)) != 0) {
808 			aprint_error_dev(&sc->sc_dev, "unable to create Tx DMA map %d, "
809 			    "error = %d\n", i, error);
810 			goto fail_4;
811 		}
812 	}
813 
814 	/*
815 	 * Create the receive buffer DMA maps.
816 	 */
817 	for (i = 0; i < DGE_NRXDESC; i++) {
818 #ifdef DGE_OFFBYONE_RXBUG
819 		if ((error = bus_dmamap_create(sc->sc_dmat, DGE_BUFFER_SIZE, 1,
820 		    DGE_BUFFER_SIZE, 0, 0, &sc->sc_rxsoft[i].rxs_dmamap)) != 0) {
821 #else
822 		if ((error = bus_dmamap_create(sc->sc_dmat, MCLBYTES, 1,
823 		    MCLBYTES, 0, 0, &sc->sc_rxsoft[i].rxs_dmamap)) != 0) {
824 #endif
825 			aprint_error_dev(&sc->sc_dev, "unable to create Rx DMA map %d, "
826 			    "error = %d\n", i, error);
827 			goto fail_5;
828 		}
829 		sc->sc_rxsoft[i].rxs_mbuf = NULL;
830 	}
831 
832 	/*
833 	 * Set bits in ctrl0 register.
834 	 * Should get the software defined pins out of EEPROM?
835 	 */
836 	sc->sc_ctrl0 |= CTRL0_RPE | CTRL0_TPE; /* XON/XOFF */
837 	sc->sc_ctrl0 |= CTRL0_SDP3_DIR | CTRL0_SDP2_DIR | CTRL0_SDP1_DIR |
838 	    CTRL0_SDP0_DIR | CTRL0_SDP3 | CTRL0_SDP2 | CTRL0_SDP0;
839 
840 	/*
841 	 * Reset the chip to a known state.
842 	 */
843 	dge_reset(sc);
844 
845 	/*
846 	 * Reset the PHY.
847 	 */
848 	dge_xgmii_reset(sc);
849 
850 	/*
851 	 * Read in EEPROM data.
852 	 */
853 	if (dge_read_eeprom(sc)) {
854 		aprint_error_dev(&sc->sc_dev, "couldn't read EEPROM\n");
855 		return;
856 	}
857 
858 	/*
859 	 * Get the ethernet address.
860 	 */
861 	enaddr[0] = sc->sc_eeprom[EE_ADDR01] & 0377;
862 	enaddr[1] = sc->sc_eeprom[EE_ADDR01] >> 8;
863 	enaddr[2] = sc->sc_eeprom[EE_ADDR23] & 0377;
864 	enaddr[3] = sc->sc_eeprom[EE_ADDR23] >> 8;
865 	enaddr[4] = sc->sc_eeprom[EE_ADDR45] & 0377;
866 	enaddr[5] = sc->sc_eeprom[EE_ADDR45] >> 8;
867 
868 	aprint_normal_dev(&sc->sc_dev, "Ethernet address %s\n",
869 	    ether_sprintf(enaddr));
870 
871 	/*
872 	 * Setup media stuff.
873 	 */
874         ifmedia_init(&sc->sc_media, IFM_IMASK, dge_xgmii_mediachange,
875             dge_xgmii_mediastatus);
876         ifmedia_add(&sc->sc_media, IFM_ETHER|IFM_10G_LR, 0, NULL);
877         ifmedia_set(&sc->sc_media, IFM_ETHER|IFM_10G_LR);
878 
879 	ifp = &sc->sc_ethercom.ec_if;
880 	strlcpy(ifp->if_xname, device_xname(&sc->sc_dev), IFNAMSIZ);
881 	ifp->if_softc = sc;
882 	ifp->if_flags = IFF_BROADCAST | IFF_SIMPLEX | IFF_MULTICAST;
883 	ifp->if_ioctl = dge_ioctl;
884 	ifp->if_start = dge_start;
885 	ifp->if_watchdog = dge_watchdog;
886 	ifp->if_init = dge_init;
887 	ifp->if_stop = dge_stop;
888 	IFQ_SET_MAXLEN(&ifp->if_snd, max(DGE_IFQUEUELEN, IFQ_MAXLEN));
889 	IFQ_SET_READY(&ifp->if_snd);
890 
891 	sc->sc_ethercom.ec_capabilities |=
892 	    ETHERCAP_JUMBO_MTU | ETHERCAP_VLAN_MTU;
893 
894 	/*
895 	 * We can perform TCPv4 and UDPv4 checkums in-bound.
896 	 */
897 	ifp->if_capabilities |=
898 	    IFCAP_CSUM_IPv4_Tx | IFCAP_CSUM_IPv4_Rx |
899 	    IFCAP_CSUM_TCPv4_Tx | IFCAP_CSUM_TCPv4_Rx |
900 	    IFCAP_CSUM_UDPv4_Tx | IFCAP_CSUM_UDPv4_Rx;
901 
902 	/*
903 	 * Attach the interface.
904 	 */
905 	if_attach(ifp);
906 	ether_ifattach(ifp, enaddr);
907 #if NRND > 0
908 	rnd_attach_source(&sc->rnd_source, device_xname(&sc->sc_dev),
909 	    RND_TYPE_NET, 0);
910 #endif
911 
912 #ifdef DGE_EVENT_COUNTERS
913 	/* Fix segment event naming */
914 	if (dge_txseg_evcnt_names == NULL) {
915 		dge_txseg_evcnt_names =
916 		    malloc(sizeof(*dge_txseg_evcnt_names), M_DEVBUF, M_WAITOK);
917 		for (i = 0; i < DGE_NTXSEGS; i++)
918 			snprintf((*dge_txseg_evcnt_names)[i],
919 			    sizeof((*dge_txseg_evcnt_names)[i]), "txseg%d", i);
920 	}
921 
922 	/* Attach event counters. */
923 	evcnt_attach_dynamic(&sc->sc_ev_txsstall, EVCNT_TYPE_MISC,
924 	    NULL, device_xname(&sc->sc_dev), "txsstall");
925 	evcnt_attach_dynamic(&sc->sc_ev_txdstall, EVCNT_TYPE_MISC,
926 	    NULL, device_xname(&sc->sc_dev), "txdstall");
927 	evcnt_attach_dynamic(&sc->sc_ev_txforceintr, EVCNT_TYPE_MISC,
928 	    NULL, device_xname(&sc->sc_dev), "txforceintr");
929 	evcnt_attach_dynamic(&sc->sc_ev_txdw, EVCNT_TYPE_INTR,
930 	    NULL, device_xname(&sc->sc_dev), "txdw");
931 	evcnt_attach_dynamic(&sc->sc_ev_txqe, EVCNT_TYPE_INTR,
932 	    NULL, device_xname(&sc->sc_dev), "txqe");
933 	evcnt_attach_dynamic(&sc->sc_ev_rxintr, EVCNT_TYPE_INTR,
934 	    NULL, device_xname(&sc->sc_dev), "rxintr");
935 	evcnt_attach_dynamic(&sc->sc_ev_linkintr, EVCNT_TYPE_INTR,
936 	    NULL, device_xname(&sc->sc_dev), "linkintr");
937 
938 	evcnt_attach_dynamic(&sc->sc_ev_rxipsum, EVCNT_TYPE_MISC,
939 	    NULL, device_xname(&sc->sc_dev), "rxipsum");
940 	evcnt_attach_dynamic(&sc->sc_ev_rxtusum, EVCNT_TYPE_MISC,
941 	    NULL, device_xname(&sc->sc_dev), "rxtusum");
942 	evcnt_attach_dynamic(&sc->sc_ev_txipsum, EVCNT_TYPE_MISC,
943 	    NULL, device_xname(&sc->sc_dev), "txipsum");
944 	evcnt_attach_dynamic(&sc->sc_ev_txtusum, EVCNT_TYPE_MISC,
945 	    NULL, device_xname(&sc->sc_dev), "txtusum");
946 
947 	evcnt_attach_dynamic(&sc->sc_ev_txctx_init, EVCNT_TYPE_MISC,
948 	    NULL, device_xname(&sc->sc_dev), "txctx init");
949 	evcnt_attach_dynamic(&sc->sc_ev_txctx_hit, EVCNT_TYPE_MISC,
950 	    NULL, device_xname(&sc->sc_dev), "txctx hit");
951 	evcnt_attach_dynamic(&sc->sc_ev_txctx_miss, EVCNT_TYPE_MISC,
952 	    NULL, device_xname(&sc->sc_dev), "txctx miss");
953 
954 	for (i = 0; i < DGE_NTXSEGS; i++)
955 		evcnt_attach_dynamic(&sc->sc_ev_txseg[i], EVCNT_TYPE_MISC,
956 		    NULL, device_xname(&sc->sc_dev), (*dge_txseg_evcnt_names)[i]);
957 
958 	evcnt_attach_dynamic(&sc->sc_ev_txdrop, EVCNT_TYPE_MISC,
959 	    NULL, device_xname(&sc->sc_dev), "txdrop");
960 
961 #endif /* DGE_EVENT_COUNTERS */
962 
963 	/*
964 	 * Make sure the interface is shutdown during reboot.
965 	 */
966 	sc->sc_sdhook = shutdownhook_establish(dge_shutdown, sc);
967 	if (sc->sc_sdhook == NULL)
968 		aprint_error_dev(&sc->sc_dev, "WARNING: unable to establish shutdown hook\n");
969 	return;
970 
971 	/*
972 	 * Free any resources we've allocated during the failed attach
973 	 * attempt.  Do this in reverse order and fall through.
974 	 */
975  fail_5:
976 	for (i = 0; i < DGE_NRXDESC; i++) {
977 		if (sc->sc_rxsoft[i].rxs_dmamap != NULL)
978 			bus_dmamap_destroy(sc->sc_dmat,
979 			    sc->sc_rxsoft[i].rxs_dmamap);
980 	}
981  fail_4:
982 	for (i = 0; i < DGE_TXQUEUELEN; i++) {
983 		if (sc->sc_txsoft[i].txs_dmamap != NULL)
984 			bus_dmamap_destroy(sc->sc_dmat,
985 			    sc->sc_txsoft[i].txs_dmamap);
986 	}
987 	bus_dmamap_unload(sc->sc_dmat, sc->sc_cddmamap);
988  fail_3:
989 	bus_dmamap_destroy(sc->sc_dmat, sc->sc_cddmamap);
990  fail_2:
991 	bus_dmamem_unmap(sc->sc_dmat, (void *)sc->sc_control_data,
992 	    sizeof(struct dge_control_data));
993  fail_1:
994 	bus_dmamem_free(sc->sc_dmat, &seg, rseg);
995  fail_0:
996 	return;
997 }
998 
999 /*
1000  * dge_shutdown:
1001  *
1002  *	Make sure the interface is stopped at reboot time.
1003  */
1004 static void
1005 dge_shutdown(void *arg)
1006 {
1007 	struct dge_softc *sc = arg;
1008 
1009 	dge_stop(&sc->sc_ethercom.ec_if, 1);
1010 }
1011 
1012 /*
1013  * dge_tx_cksum:
1014  *
1015  *	Set up TCP/IP checksumming parameters for the
1016  *	specified packet.
1017  */
1018 static int
1019 dge_tx_cksum(struct dge_softc *sc, struct dge_txsoft *txs, uint8_t *fieldsp)
1020 {
1021 	struct mbuf *m0 = txs->txs_mbuf;
1022 	struct dge_ctdes *t;
1023 	uint32_t ipcs, tucs;
1024 	struct ether_header *eh;
1025 	int offset, iphl;
1026 	uint8_t fields = 0;
1027 
1028 	/*
1029 	 * XXX It would be nice if the mbuf pkthdr had offset
1030 	 * fields for the protocol headers.
1031 	 */
1032 
1033 	eh = mtod(m0, struct ether_header *);
1034 	switch (htons(eh->ether_type)) {
1035 	case ETHERTYPE_IP:
1036 		offset = ETHER_HDR_LEN;
1037 		break;
1038 
1039 	case ETHERTYPE_VLAN:
1040 		offset = ETHER_HDR_LEN + ETHER_VLAN_ENCAP_LEN;
1041 		break;
1042 
1043 	default:
1044 		/*
1045 		 * Don't support this protocol or encapsulation.
1046 		 */
1047 		*fieldsp = 0;
1048 		return (0);
1049 	}
1050 
1051 	iphl = M_CSUM_DATA_IPv4_IPHL(m0->m_pkthdr.csum_data);
1052 
1053 	/*
1054 	 * NOTE: Even if we're not using the IP or TCP/UDP checksum
1055 	 * offload feature, if we load the context descriptor, we
1056 	 * MUST provide valid values for IPCSS and TUCSS fields.
1057 	 */
1058 
1059 	if (m0->m_pkthdr.csum_flags & M_CSUM_IPv4) {
1060 		DGE_EVCNT_INCR(&sc->sc_ev_txipsum);
1061 		fields |= TDESC_POPTS_IXSM;
1062 		ipcs = DGE_TCPIP_IPCSS(offset) |
1063 		    DGE_TCPIP_IPCSO(offset + offsetof(struct ip, ip_sum)) |
1064 		    DGE_TCPIP_IPCSE(offset + iphl - 1);
1065 	} else if (__predict_true(sc->sc_txctx_ipcs != 0xffffffff)) {
1066 		/* Use the cached value. */
1067 		ipcs = sc->sc_txctx_ipcs;
1068 	} else {
1069 		/* Just initialize it to the likely value anyway. */
1070 		ipcs = DGE_TCPIP_IPCSS(offset) |
1071 		    DGE_TCPIP_IPCSO(offset + offsetof(struct ip, ip_sum)) |
1072 		    DGE_TCPIP_IPCSE(offset + iphl - 1);
1073 	}
1074 	DPRINTF(DGE_DEBUG_CKSUM,
1075 	    ("%s: CKSUM: offset %d ipcs 0x%x\n",
1076 	    device_xname(&sc->sc_dev), offset, ipcs));
1077 
1078 	offset += iphl;
1079 
1080 	if (m0->m_pkthdr.csum_flags & (M_CSUM_TCPv4|M_CSUM_UDPv4)) {
1081 		DGE_EVCNT_INCR(&sc->sc_ev_txtusum);
1082 		fields |= TDESC_POPTS_TXSM;
1083 		tucs = DGE_TCPIP_TUCSS(offset) |
1084 		   DGE_TCPIP_TUCSO(offset + M_CSUM_DATA_IPv4_OFFSET(m0->m_pkthdr.csum_data)) |
1085 		   DGE_TCPIP_TUCSE(0) /* rest of packet */;
1086 	} else if (__predict_true(sc->sc_txctx_tucs != 0xffffffff)) {
1087 		/* Use the cached value. */
1088 		tucs = sc->sc_txctx_tucs;
1089 	} else {
1090 		/* Just initialize it to a valid TCP context. */
1091 		tucs = DGE_TCPIP_TUCSS(offset) |
1092 		    DGE_TCPIP_TUCSO(offset + offsetof(struct tcphdr, th_sum)) |
1093 		    DGE_TCPIP_TUCSE(0) /* rest of packet */;
1094 	}
1095 
1096 	DPRINTF(DGE_DEBUG_CKSUM,
1097 	    ("%s: CKSUM: offset %d tucs 0x%x\n",
1098 	    device_xname(&sc->sc_dev), offset, tucs));
1099 
1100 	if (sc->sc_txctx_ipcs == ipcs &&
1101 	    sc->sc_txctx_tucs == tucs) {
1102 		/* Cached context is fine. */
1103 		DGE_EVCNT_INCR(&sc->sc_ev_txctx_hit);
1104 	} else {
1105 		/* Fill in the context descriptor. */
1106 #ifdef DGE_EVENT_COUNTERS
1107 		if (sc->sc_txctx_ipcs == 0xffffffff &&
1108 		    sc->sc_txctx_tucs == 0xffffffff)
1109 			DGE_EVCNT_INCR(&sc->sc_ev_txctx_init);
1110 		else
1111 			DGE_EVCNT_INCR(&sc->sc_ev_txctx_miss);
1112 #endif
1113 		t = (struct dge_ctdes *)&sc->sc_txdescs[sc->sc_txnext];
1114 		t->dc_tcpip_ipcs = htole32(ipcs);
1115 		t->dc_tcpip_tucs = htole32(tucs);
1116 		t->dc_tcpip_cmdlen = htole32(TDESC_DTYP_CTD);
1117 		t->dc_tcpip_seg = 0;
1118 		DGE_CDTXSYNC(sc, sc->sc_txnext, 1, BUS_DMASYNC_PREWRITE);
1119 
1120 		sc->sc_txctx_ipcs = ipcs;
1121 		sc->sc_txctx_tucs = tucs;
1122 
1123 		sc->sc_txnext = DGE_NEXTTX(sc->sc_txnext);
1124 		txs->txs_ndesc++;
1125 	}
1126 
1127 	*fieldsp = fields;
1128 
1129 	return (0);
1130 }
1131 
1132 /*
1133  * dge_start:		[ifnet interface function]
1134  *
1135  *	Start packet transmission on the interface.
1136  */
1137 static void
1138 dge_start(struct ifnet *ifp)
1139 {
1140 	struct dge_softc *sc = ifp->if_softc;
1141 	struct mbuf *m0;
1142 	struct dge_txsoft *txs;
1143 	bus_dmamap_t dmamap;
1144 	int error, nexttx, lasttx = -1, ofree, seg;
1145 	uint32_t cksumcmd;
1146 	uint8_t cksumfields;
1147 
1148 	if ((ifp->if_flags & (IFF_RUNNING|IFF_OACTIVE)) != IFF_RUNNING)
1149 		return;
1150 
1151 	/*
1152 	 * Remember the previous number of free descriptors.
1153 	 */
1154 	ofree = sc->sc_txfree;
1155 
1156 	/*
1157 	 * Loop through the send queue, setting up transmit descriptors
1158 	 * until we drain the queue, or use up all available transmit
1159 	 * descriptors.
1160 	 */
1161 	for (;;) {
1162 		/* Grab a packet off the queue. */
1163 		IFQ_POLL(&ifp->if_snd, m0);
1164 		if (m0 == NULL)
1165 			break;
1166 
1167 		DPRINTF(DGE_DEBUG_TX,
1168 		    ("%s: TX: have packet to transmit: %p\n",
1169 		    device_xname(&sc->sc_dev), m0));
1170 
1171 		/* Get a work queue entry. */
1172 		if (sc->sc_txsfree < DGE_TXQUEUE_GC) {
1173 			dge_txintr(sc);
1174 			if (sc->sc_txsfree == 0) {
1175 				DPRINTF(DGE_DEBUG_TX,
1176 				    ("%s: TX: no free job descriptors\n",
1177 					device_xname(&sc->sc_dev)));
1178 				DGE_EVCNT_INCR(&sc->sc_ev_txsstall);
1179 				break;
1180 			}
1181 		}
1182 
1183 		txs = &sc->sc_txsoft[sc->sc_txsnext];
1184 		dmamap = txs->txs_dmamap;
1185 
1186 		/*
1187 		 * Load the DMA map.  If this fails, the packet either
1188 		 * didn't fit in the allotted number of segments, or we
1189 		 * were short on resources.  For the too-many-segments
1190 		 * case, we simply report an error and drop the packet,
1191 		 * since we can't sanely copy a jumbo packet to a single
1192 		 * buffer.
1193 		 */
1194 		error = bus_dmamap_load_mbuf(sc->sc_dmat, dmamap, m0,
1195 		    BUS_DMA_WRITE|BUS_DMA_NOWAIT);
1196 		if (error) {
1197 			if (error == EFBIG) {
1198 				DGE_EVCNT_INCR(&sc->sc_ev_txdrop);
1199 				printf("%s: Tx packet consumes too many "
1200 				    "DMA segments, dropping...\n",
1201 				    device_xname(&sc->sc_dev));
1202 				IFQ_DEQUEUE(&ifp->if_snd, m0);
1203 				m_freem(m0);
1204 				continue;
1205 			}
1206 			/*
1207 			 * Short on resources, just stop for now.
1208 			 */
1209 			DPRINTF(DGE_DEBUG_TX,
1210 			    ("%s: TX: dmamap load failed: %d\n",
1211 			    device_xname(&sc->sc_dev), error));
1212 			break;
1213 		}
1214 
1215 		/*
1216 		 * Ensure we have enough descriptors free to describe
1217 		 * the packet.  Note, we always reserve one descriptor
1218 		 * at the end of the ring due to the semantics of the
1219 		 * TDT register, plus one more in the event we need
1220 		 * to re-load checksum offload context.
1221 		 */
1222 		if (dmamap->dm_nsegs > (sc->sc_txfree - 2)) {
1223 			/*
1224 			 * Not enough free descriptors to transmit this
1225 			 * packet.  We haven't committed anything yet,
1226 			 * so just unload the DMA map, put the packet
1227 			 * pack on the queue, and punt.  Notify the upper
1228 			 * layer that there are no more slots left.
1229 			 */
1230 			DPRINTF(DGE_DEBUG_TX,
1231 			    ("%s: TX: need %d descriptors, have %d\n",
1232 			    device_xname(&sc->sc_dev), dmamap->dm_nsegs,
1233 			    sc->sc_txfree - 1));
1234 			ifp->if_flags |= IFF_OACTIVE;
1235 			bus_dmamap_unload(sc->sc_dmat, dmamap);
1236 			DGE_EVCNT_INCR(&sc->sc_ev_txdstall);
1237 			break;
1238 		}
1239 
1240 		IFQ_DEQUEUE(&ifp->if_snd, m0);
1241 
1242 		/*
1243 		 * WE ARE NOW COMMITTED TO TRANSMITTING THE PACKET.
1244 		 */
1245 
1246 		/* Sync the DMA map. */
1247 		bus_dmamap_sync(sc->sc_dmat, dmamap, 0, dmamap->dm_mapsize,
1248 		    BUS_DMASYNC_PREWRITE);
1249 
1250 		DPRINTF(DGE_DEBUG_TX,
1251 		    ("%s: TX: packet has %d DMA segments\n",
1252 		    device_xname(&sc->sc_dev), dmamap->dm_nsegs));
1253 
1254 		DGE_EVCNT_INCR(&sc->sc_ev_txseg[dmamap->dm_nsegs - 1]);
1255 
1256 		/*
1257 		 * Store a pointer to the packet so that we can free it
1258 		 * later.
1259 		 *
1260 		 * Initially, we consider the number of descriptors the
1261 		 * packet uses the number of DMA segments.  This may be
1262 		 * incremented by 1 if we do checksum offload (a descriptor
1263 		 * is used to set the checksum context).
1264 		 */
1265 		txs->txs_mbuf = m0;
1266 		txs->txs_firstdesc = sc->sc_txnext;
1267 		txs->txs_ndesc = dmamap->dm_nsegs;
1268 
1269 		/*
1270 		 * Set up checksum offload parameters for
1271 		 * this packet.
1272 		 */
1273 		if (m0->m_pkthdr.csum_flags &
1274 		    (M_CSUM_IPv4|M_CSUM_TCPv4|M_CSUM_UDPv4)) {
1275 			if (dge_tx_cksum(sc, txs, &cksumfields) != 0) {
1276 				/* Error message already displayed. */
1277 				bus_dmamap_unload(sc->sc_dmat, dmamap);
1278 				continue;
1279 			}
1280 		} else {
1281 			cksumfields = 0;
1282 		}
1283 
1284 		cksumcmd = TDESC_DCMD_IDE | TDESC_DTYP_DATA;
1285 
1286 		/*
1287 		 * Initialize the transmit descriptor.
1288 		 */
1289 		for (nexttx = sc->sc_txnext, seg = 0;
1290 		     seg < dmamap->dm_nsegs;
1291 		     seg++, nexttx = DGE_NEXTTX(nexttx)) {
1292 			/*
1293 			 * Note: we currently only use 32-bit DMA
1294 			 * addresses.
1295 			 */
1296 			sc->sc_txdescs[nexttx].dt_baddrh = 0;
1297 			sc->sc_txdescs[nexttx].dt_baddrl =
1298 			    htole32(dmamap->dm_segs[seg].ds_addr);
1299 			sc->sc_txdescs[nexttx].dt_ctl =
1300 			    htole32(cksumcmd | dmamap->dm_segs[seg].ds_len);
1301 			sc->sc_txdescs[nexttx].dt_status = 0;
1302 			sc->sc_txdescs[nexttx].dt_popts = cksumfields;
1303 			sc->sc_txdescs[nexttx].dt_vlan = 0;
1304 			lasttx = nexttx;
1305 
1306 			DPRINTF(DGE_DEBUG_TX,
1307 			    ("%s: TX: desc %d: low 0x%08lx, len 0x%04lx\n",
1308 			    device_xname(&sc->sc_dev), nexttx,
1309 			    le32toh(dmamap->dm_segs[seg].ds_addr),
1310 			    le32toh(dmamap->dm_segs[seg].ds_len)));
1311 		}
1312 
1313 		KASSERT(lasttx != -1);
1314 
1315 		/*
1316 		 * Set up the command byte on the last descriptor of
1317 		 * the packet.  If we're in the interrupt delay window,
1318 		 * delay the interrupt.
1319 		 */
1320 		sc->sc_txdescs[lasttx].dt_ctl |=
1321 		    htole32(TDESC_DCMD_EOP | TDESC_DCMD_RS);
1322 
1323 		txs->txs_lastdesc = lasttx;
1324 
1325 		DPRINTF(DGE_DEBUG_TX,
1326 		    ("%s: TX: desc %d: cmdlen 0x%08x\n", device_xname(&sc->sc_dev),
1327 		    lasttx, le32toh(sc->sc_txdescs[lasttx].dt_ctl)));
1328 
1329 		/* Sync the descriptors we're using. */
1330 		DGE_CDTXSYNC(sc, sc->sc_txnext, dmamap->dm_nsegs,
1331 		    BUS_DMASYNC_PREREAD|BUS_DMASYNC_PREWRITE);
1332 
1333 		/* Give the packet to the chip. */
1334 		CSR_WRITE(sc, DGE_TDT, nexttx);
1335 
1336 		DPRINTF(DGE_DEBUG_TX,
1337 		    ("%s: TX: TDT -> %d\n", device_xname(&sc->sc_dev), nexttx));
1338 
1339 		DPRINTF(DGE_DEBUG_TX,
1340 		    ("%s: TX: finished transmitting packet, job %d\n",
1341 		    device_xname(&sc->sc_dev), sc->sc_txsnext));
1342 
1343 		/* Advance the tx pointer. */
1344 		sc->sc_txfree -= txs->txs_ndesc;
1345 		sc->sc_txnext = nexttx;
1346 
1347 		sc->sc_txsfree--;
1348 		sc->sc_txsnext = DGE_NEXTTXS(sc->sc_txsnext);
1349 
1350 #if NBPFILTER > 0
1351 		/* Pass the packet to any BPF listeners. */
1352 		if (ifp->if_bpf)
1353 			bpf_mtap(ifp->if_bpf, m0);
1354 #endif /* NBPFILTER > 0 */
1355 	}
1356 
1357 	if (sc->sc_txsfree == 0 || sc->sc_txfree <= 2) {
1358 		/* No more slots; notify upper layer. */
1359 		ifp->if_flags |= IFF_OACTIVE;
1360 	}
1361 
1362 	if (sc->sc_txfree != ofree) {
1363 		/* Set a watchdog timer in case the chip flakes out. */
1364 		ifp->if_timer = 5;
1365 	}
1366 }
1367 
1368 /*
1369  * dge_watchdog:		[ifnet interface function]
1370  *
1371  *	Watchdog timer handler.
1372  */
1373 static void
1374 dge_watchdog(struct ifnet *ifp)
1375 {
1376 	struct dge_softc *sc = ifp->if_softc;
1377 
1378 	/*
1379 	 * Since we're using delayed interrupts, sweep up
1380 	 * before we report an error.
1381 	 */
1382 	dge_txintr(sc);
1383 
1384 	if (sc->sc_txfree != DGE_NTXDESC) {
1385 		printf("%s: device timeout (txfree %d txsfree %d txnext %d)\n",
1386 		    device_xname(&sc->sc_dev), sc->sc_txfree, sc->sc_txsfree,
1387 		    sc->sc_txnext);
1388 		ifp->if_oerrors++;
1389 
1390 		/* Reset the interface. */
1391 		(void) dge_init(ifp);
1392 	}
1393 
1394 	/* Try to get more packets going. */
1395 	dge_start(ifp);
1396 }
1397 
1398 /*
1399  * dge_ioctl:		[ifnet interface function]
1400  *
1401  *	Handle control requests from the operator.
1402  */
1403 static int
1404 dge_ioctl(struct ifnet *ifp, u_long cmd, void *data)
1405 {
1406 	struct dge_softc *sc = ifp->if_softc;
1407 	struct ifreq *ifr = (struct ifreq *) data;
1408 	pcireg_t preg;
1409 	int s, error, mmrbc;
1410 
1411 	s = splnet();
1412 
1413 	switch (cmd) {
1414 	case SIOCSIFMEDIA:
1415 	case SIOCGIFMEDIA:
1416 		error = ifmedia_ioctl(ifp, ifr, &sc->sc_media, cmd);
1417 		break;
1418 
1419 	case SIOCSIFMTU:
1420 		if (ifr->ifr_mtu < ETHERMIN || ifr->ifr_mtu > DGE_MAX_MTU)
1421 			error = EINVAL;
1422 		else if ((error = ifioctl_common(ifp, cmd, data)) != ENETRESET)
1423 			break;
1424 		else if (ifp->if_flags & IFF_UP)
1425 			error = (*ifp->if_init)(ifp);
1426 		else
1427 			error = 0;
1428 		break;
1429 
1430         case SIOCSIFFLAGS:
1431 		if ((error = ifioctl_common(ifp, cmd, data)) != 0)
1432 			break;
1433 		/* extract link flags */
1434 		if ((ifp->if_flags & IFF_LINK0) == 0 &&
1435 		    (ifp->if_flags & IFF_LINK1) == 0)
1436 			mmrbc = PCIX_MMRBC_512;
1437 		else if ((ifp->if_flags & IFF_LINK0) == 0 &&
1438 		    (ifp->if_flags & IFF_LINK1) != 0)
1439 			mmrbc = PCIX_MMRBC_1024;
1440 		else if ((ifp->if_flags & IFF_LINK0) != 0 &&
1441 		    (ifp->if_flags & IFF_LINK1) == 0)
1442 			mmrbc = PCIX_MMRBC_2048;
1443 		else
1444 			mmrbc = PCIX_MMRBC_4096;
1445 		if (mmrbc != sc->sc_mmrbc) {
1446 			preg = pci_conf_read(sc->sc_pc, sc->sc_pt,DGE_PCIX_CMD);
1447 			preg &= ~PCIX_MMRBC_MSK;
1448 			preg |= mmrbc;
1449 			pci_conf_write(sc->sc_pc, sc->sc_pt,DGE_PCIX_CMD, preg);
1450 			sc->sc_mmrbc = mmrbc;
1451 		}
1452                 /* FALLTHROUGH */
1453 	default:
1454 		if ((error = ether_ioctl(ifp, cmd, data)) != ENETRESET)
1455 			break;
1456 
1457 		error = 0;
1458 
1459 		if (cmd == SIOCSIFCAP)
1460 			error = (*ifp->if_init)(ifp);
1461 		else if (cmd != SIOCADDMULTI && cmd != SIOCDELMULTI)
1462 			;
1463 		else if (ifp->if_flags & IFF_RUNNING) {
1464 			/*
1465 			 * Multicast list has changed; set the hardware filter
1466 			 * accordingly.
1467 			 */
1468 			dge_set_filter(sc);
1469 		}
1470 		break;
1471 	}
1472 
1473 	/* Try to get more packets going. */
1474 	dge_start(ifp);
1475 
1476 	splx(s);
1477 	return (error);
1478 }
1479 
1480 /*
1481  * dge_intr:
1482  *
1483  *	Interrupt service routine.
1484  */
1485 static int
1486 dge_intr(void *arg)
1487 {
1488 	struct dge_softc *sc = arg;
1489 	struct ifnet *ifp = &sc->sc_ethercom.ec_if;
1490 	uint32_t icr;
1491 	int wantinit, handled = 0;
1492 
1493 	for (wantinit = 0; wantinit == 0;) {
1494 		icr = CSR_READ(sc, DGE_ICR);
1495 		if ((icr & sc->sc_icr) == 0)
1496 			break;
1497 
1498 #if 0 /*NRND > 0*/
1499 		if (RND_ENABLED(&sc->rnd_source))
1500 			rnd_add_uint32(&sc->rnd_source, icr);
1501 #endif
1502 
1503 		handled = 1;
1504 
1505 #if defined(DGE_DEBUG) || defined(DGE_EVENT_COUNTERS)
1506 		if (icr & (ICR_RXDMT0|ICR_RXT0)) {
1507 			DPRINTF(DGE_DEBUG_RX,
1508 			    ("%s: RX: got Rx intr 0x%08x\n",
1509 			    device_xname(&sc->sc_dev),
1510 			    icr & (ICR_RXDMT0|ICR_RXT0)));
1511 			DGE_EVCNT_INCR(&sc->sc_ev_rxintr);
1512 		}
1513 #endif
1514 		dge_rxintr(sc);
1515 
1516 #if defined(DGE_DEBUG) || defined(DGE_EVENT_COUNTERS)
1517 		if (icr & ICR_TXDW) {
1518 			DPRINTF(DGE_DEBUG_TX,
1519 			    ("%s: TX: got TXDW interrupt\n",
1520 			    device_xname(&sc->sc_dev)));
1521 			DGE_EVCNT_INCR(&sc->sc_ev_txdw);
1522 		}
1523 		if (icr & ICR_TXQE)
1524 			DGE_EVCNT_INCR(&sc->sc_ev_txqe);
1525 #endif
1526 		dge_txintr(sc);
1527 
1528 		if (icr & (ICR_LSC|ICR_RXSEQ)) {
1529 			DGE_EVCNT_INCR(&sc->sc_ev_linkintr);
1530 			dge_linkintr(sc, icr);
1531 		}
1532 
1533 		if (icr & ICR_RXO) {
1534 			printf("%s: Receive overrun\n", device_xname(&sc->sc_dev));
1535 			wantinit = 1;
1536 		}
1537 	}
1538 
1539 	if (handled) {
1540 		if (wantinit)
1541 			dge_init(ifp);
1542 
1543 		/* Try to get more packets going. */
1544 		dge_start(ifp);
1545 	}
1546 
1547 	return (handled);
1548 }
1549 
1550 /*
1551  * dge_txintr:
1552  *
1553  *	Helper; handle transmit interrupts.
1554  */
1555 static void
1556 dge_txintr(struct dge_softc *sc)
1557 {
1558 	struct ifnet *ifp = &sc->sc_ethercom.ec_if;
1559 	struct dge_txsoft *txs;
1560 	uint8_t status;
1561 	int i;
1562 
1563 	ifp->if_flags &= ~IFF_OACTIVE;
1564 
1565 	/*
1566 	 * Go through the Tx list and free mbufs for those
1567 	 * frames which have been transmitted.
1568 	 */
1569 	for (i = sc->sc_txsdirty; sc->sc_txsfree != DGE_TXQUEUELEN;
1570 	     i = DGE_NEXTTXS(i), sc->sc_txsfree++) {
1571 		txs = &sc->sc_txsoft[i];
1572 
1573 		DPRINTF(DGE_DEBUG_TX,
1574 		    ("%s: TX: checking job %d\n", device_xname(&sc->sc_dev), i));
1575 
1576 		DGE_CDTXSYNC(sc, txs->txs_firstdesc, txs->txs_dmamap->dm_nsegs,
1577 		    BUS_DMASYNC_POSTREAD|BUS_DMASYNC_POSTWRITE);
1578 
1579 		status =
1580 		    sc->sc_txdescs[txs->txs_lastdesc].dt_status;
1581 		if ((status & TDESC_STA_DD) == 0) {
1582 			DGE_CDTXSYNC(sc, txs->txs_lastdesc, 1,
1583 			    BUS_DMASYNC_PREREAD);
1584 			break;
1585 		}
1586 
1587 		DPRINTF(DGE_DEBUG_TX,
1588 		    ("%s: TX: job %d done: descs %d..%d\n",
1589 		    device_xname(&sc->sc_dev), i, txs->txs_firstdesc,
1590 		    txs->txs_lastdesc));
1591 
1592 		ifp->if_opackets++;
1593 		sc->sc_txfree += txs->txs_ndesc;
1594 		bus_dmamap_sync(sc->sc_dmat, txs->txs_dmamap,
1595 		    0, txs->txs_dmamap->dm_mapsize, BUS_DMASYNC_POSTWRITE);
1596 		bus_dmamap_unload(sc->sc_dmat, txs->txs_dmamap);
1597 		m_freem(txs->txs_mbuf);
1598 		txs->txs_mbuf = NULL;
1599 	}
1600 
1601 	/* Update the dirty transmit buffer pointer. */
1602 	sc->sc_txsdirty = i;
1603 	DPRINTF(DGE_DEBUG_TX,
1604 	    ("%s: TX: txsdirty -> %d\n", device_xname(&sc->sc_dev), i));
1605 
1606 	/*
1607 	 * If there are no more pending transmissions, cancel the watchdog
1608 	 * timer.
1609 	 */
1610 	if (sc->sc_txsfree == DGE_TXQUEUELEN)
1611 		ifp->if_timer = 0;
1612 }
1613 
1614 /*
1615  * dge_rxintr:
1616  *
1617  *	Helper; handle receive interrupts.
1618  */
1619 static void
1620 dge_rxintr(struct dge_softc *sc)
1621 {
1622 	struct ifnet *ifp = &sc->sc_ethercom.ec_if;
1623 	struct dge_rxsoft *rxs;
1624 	struct mbuf *m;
1625 	int i, len;
1626 	uint8_t status, errors;
1627 
1628 	for (i = sc->sc_rxptr;; i = DGE_NEXTRX(i)) {
1629 		rxs = &sc->sc_rxsoft[i];
1630 
1631 		DPRINTF(DGE_DEBUG_RX,
1632 		    ("%s: RX: checking descriptor %d\n",
1633 		    device_xname(&sc->sc_dev), i));
1634 
1635 		DGE_CDRXSYNC(sc, i, BUS_DMASYNC_POSTREAD|BUS_DMASYNC_POSTWRITE);
1636 
1637 		status = sc->sc_rxdescs[i].dr_status;
1638 		errors = sc->sc_rxdescs[i].dr_errors;
1639 		len = le16toh(sc->sc_rxdescs[i].dr_len);
1640 
1641 		if ((status & RDESC_STS_DD) == 0) {
1642 			/*
1643 			 * We have processed all of the receive descriptors.
1644 			 */
1645 			DGE_CDRXSYNC(sc, i, BUS_DMASYNC_PREREAD);
1646 			break;
1647 		}
1648 
1649 		if (__predict_false(sc->sc_rxdiscard)) {
1650 			DPRINTF(DGE_DEBUG_RX,
1651 			    ("%s: RX: discarding contents of descriptor %d\n",
1652 			    device_xname(&sc->sc_dev), i));
1653 			DGE_INIT_RXDESC(sc, i);
1654 			if (status & RDESC_STS_EOP) {
1655 				/* Reset our state. */
1656 				DPRINTF(DGE_DEBUG_RX,
1657 				    ("%s: RX: resetting rxdiscard -> 0\n",
1658 				    device_xname(&sc->sc_dev)));
1659 				sc->sc_rxdiscard = 0;
1660 			}
1661 			continue;
1662 		}
1663 
1664 		bus_dmamap_sync(sc->sc_dmat, rxs->rxs_dmamap, 0,
1665 		    rxs->rxs_dmamap->dm_mapsize, BUS_DMASYNC_POSTREAD);
1666 
1667 		m = rxs->rxs_mbuf;
1668 
1669 		/*
1670 		 * Add a new receive buffer to the ring.
1671 		 */
1672 		if (dge_add_rxbuf(sc, i) != 0) {
1673 			/*
1674 			 * Failed, throw away what we've done so
1675 			 * far, and discard the rest of the packet.
1676 			 */
1677 			ifp->if_ierrors++;
1678 			bus_dmamap_sync(sc->sc_dmat, rxs->rxs_dmamap, 0,
1679 			    rxs->rxs_dmamap->dm_mapsize, BUS_DMASYNC_PREREAD);
1680 			DGE_INIT_RXDESC(sc, i);
1681 			if ((status & RDESC_STS_EOP) == 0)
1682 				sc->sc_rxdiscard = 1;
1683 			if (sc->sc_rxhead != NULL)
1684 				m_freem(sc->sc_rxhead);
1685 			DGE_RXCHAIN_RESET(sc);
1686 			DPRINTF(DGE_DEBUG_RX,
1687 			    ("%s: RX: Rx buffer allocation failed, "
1688 			    "dropping packet%s\n", device_xname(&sc->sc_dev),
1689 			    sc->sc_rxdiscard ? " (discard)" : ""));
1690 			continue;
1691 		}
1692 		DGE_INIT_RXDESC(sc, DGE_PREVRX(i)); /* Write the descriptor */
1693 
1694 		DGE_RXCHAIN_LINK(sc, m);
1695 
1696 		m->m_len = len;
1697 
1698 		DPRINTF(DGE_DEBUG_RX,
1699 		    ("%s: RX: buffer at %p len %d\n",
1700 		    device_xname(&sc->sc_dev), m->m_data, len));
1701 
1702 		/*
1703 		 * If this is not the end of the packet, keep
1704 		 * looking.
1705 		 */
1706 		if ((status & RDESC_STS_EOP) == 0) {
1707 			sc->sc_rxlen += len;
1708 			DPRINTF(DGE_DEBUG_RX,
1709 			    ("%s: RX: not yet EOP, rxlen -> %d\n",
1710 			    device_xname(&sc->sc_dev), sc->sc_rxlen));
1711 			continue;
1712 		}
1713 
1714 		/*
1715 		 * Okay, we have the entire packet now...
1716 		 */
1717 		*sc->sc_rxtailp = NULL;
1718 		m = sc->sc_rxhead;
1719 		len += sc->sc_rxlen;
1720 
1721 		DGE_RXCHAIN_RESET(sc);
1722 
1723 		DPRINTF(DGE_DEBUG_RX,
1724 		    ("%s: RX: have entire packet, len -> %d\n",
1725 		    device_xname(&sc->sc_dev), len));
1726 
1727 		/*
1728 		 * If an error occurred, update stats and drop the packet.
1729 		 */
1730 		if (errors &
1731 		     (RDESC_ERR_CE|RDESC_ERR_SE|RDESC_ERR_P|RDESC_ERR_RXE)) {
1732 			ifp->if_ierrors++;
1733 			if (errors & RDESC_ERR_SE)
1734 				printf("%s: symbol error\n",
1735 				    device_xname(&sc->sc_dev));
1736 			else if (errors & RDESC_ERR_P)
1737 				printf("%s: parity error\n",
1738 				    device_xname(&sc->sc_dev));
1739 			else if (errors & RDESC_ERR_CE)
1740 				printf("%s: CRC error\n",
1741 				    device_xname(&sc->sc_dev));
1742 			m_freem(m);
1743 			continue;
1744 		}
1745 
1746 		/*
1747 		 * No errors.  Receive the packet.
1748 		 */
1749 		m->m_pkthdr.rcvif = ifp;
1750 		m->m_pkthdr.len = len;
1751 
1752 		/*
1753 		 * Set up checksum info for this packet.
1754 		 */
1755 		if (status & RDESC_STS_IPCS) {
1756 			DGE_EVCNT_INCR(&sc->sc_ev_rxipsum);
1757 			m->m_pkthdr.csum_flags |= M_CSUM_IPv4;
1758 			if (errors & RDESC_ERR_IPE)
1759 				m->m_pkthdr.csum_flags |= M_CSUM_IPv4_BAD;
1760 		}
1761 		if (status & RDESC_STS_TCPCS) {
1762 			/*
1763 			 * Note: we don't know if this was TCP or UDP,
1764 			 * so we just set both bits, and expect the
1765 			 * upper layers to deal.
1766 			 */
1767 			DGE_EVCNT_INCR(&sc->sc_ev_rxtusum);
1768 			m->m_pkthdr.csum_flags |= M_CSUM_TCPv4|M_CSUM_UDPv4;
1769 			if (errors & RDESC_ERR_TCPE)
1770 				m->m_pkthdr.csum_flags |= M_CSUM_TCP_UDP_BAD;
1771 		}
1772 
1773 		ifp->if_ipackets++;
1774 
1775 #if NBPFILTER > 0
1776 		/* Pass this up to any BPF listeners. */
1777 		if (ifp->if_bpf)
1778 			bpf_mtap(ifp->if_bpf, m);
1779 #endif /* NBPFILTER > 0 */
1780 
1781 		/* Pass it on. */
1782 		(*ifp->if_input)(ifp, m);
1783 	}
1784 
1785 	/* Update the receive pointer. */
1786 	sc->sc_rxptr = i;
1787 
1788 	DPRINTF(DGE_DEBUG_RX,
1789 	    ("%s: RX: rxptr -> %d\n", device_xname(&sc->sc_dev), i));
1790 }
1791 
1792 /*
1793  * dge_linkintr:
1794  *
1795  *	Helper; handle link interrupts.
1796  */
1797 static void
1798 dge_linkintr(struct dge_softc *sc, uint32_t icr)
1799 {
1800 	uint32_t status;
1801 
1802 	if (icr & ICR_LSC) {
1803 		status = CSR_READ(sc, DGE_STATUS);
1804 		if (status & STATUS_LINKUP) {
1805 			DPRINTF(DGE_DEBUG_LINK, ("%s: LINK: LSC -> up\n",
1806 			    device_xname(&sc->sc_dev)));
1807 		} else {
1808 			DPRINTF(DGE_DEBUG_LINK, ("%s: LINK: LSC -> down\n",
1809 			    device_xname(&sc->sc_dev)));
1810 		}
1811 	} else if (icr & ICR_RXSEQ) {
1812 		DPRINTF(DGE_DEBUG_LINK,
1813 		    ("%s: LINK: Receive sequence error\n",
1814 		    device_xname(&sc->sc_dev)));
1815 	}
1816 	/* XXX - fix errata */
1817 }
1818 
1819 /*
1820  * dge_reset:
1821  *
1822  *	Reset the i82597 chip.
1823  */
1824 static void
1825 dge_reset(struct dge_softc *sc)
1826 {
1827 	int i;
1828 
1829 	/*
1830 	 * Do a chip reset.
1831 	 */
1832 	CSR_WRITE(sc, DGE_CTRL0, CTRL0_RST | sc->sc_ctrl0);
1833 
1834 	delay(10000);
1835 
1836 	for (i = 0; i < 1000; i++) {
1837 		if ((CSR_READ(sc, DGE_CTRL0) & CTRL0_RST) == 0)
1838 			break;
1839 		delay(20);
1840 	}
1841 
1842 	if (CSR_READ(sc, DGE_CTRL0) & CTRL0_RST)
1843 		printf("%s: WARNING: reset failed to complete\n",
1844 		    device_xname(&sc->sc_dev));
1845         /*
1846          * Reset the EEPROM logic.
1847          * This will cause the chip to reread its default values,
1848 	 * which doesn't happen otherwise (errata).
1849          */
1850         CSR_WRITE(sc, DGE_CTRL1, CTRL1_EE_RST);
1851         delay(10000);
1852 }
1853 
1854 /*
1855  * dge_init:		[ifnet interface function]
1856  *
1857  *	Initialize the interface.  Must be called at splnet().
1858  */
1859 static int
1860 dge_init(struct ifnet *ifp)
1861 {
1862 	struct dge_softc *sc = ifp->if_softc;
1863 	struct dge_rxsoft *rxs;
1864 	int i, error = 0;
1865 	uint32_t reg;
1866 
1867 	/*
1868 	 * *_HDR_ALIGNED_P is constant 1 if __NO_STRICT_ALIGMENT is set.
1869 	 * There is a small but measurable benefit to avoiding the adjusment
1870 	 * of the descriptor so that the headers are aligned, for normal mtu,
1871 	 * on such platforms.  One possibility is that the DMA itself is
1872 	 * slightly more efficient if the front of the entire packet (instead
1873 	 * of the front of the headers) is aligned.
1874 	 *
1875 	 * Note we must always set align_tweak to 0 if we are using
1876 	 * jumbo frames.
1877 	 */
1878 #ifdef __NO_STRICT_ALIGNMENT
1879 	sc->sc_align_tweak = 0;
1880 #else
1881 	if ((ifp->if_mtu + ETHER_HDR_LEN + ETHER_CRC_LEN) > (MCLBYTES - 2))
1882 		sc->sc_align_tweak = 0;
1883 	else
1884 		sc->sc_align_tweak = 2;
1885 #endif /* __NO_STRICT_ALIGNMENT */
1886 
1887 	/* Cancel any pending I/O. */
1888 	dge_stop(ifp, 0);
1889 
1890 	/* Reset the chip to a known state. */
1891 	dge_reset(sc);
1892 
1893 	/* Initialize the transmit descriptor ring. */
1894 	memset(sc->sc_txdescs, 0, sizeof(sc->sc_txdescs));
1895 	DGE_CDTXSYNC(sc, 0, DGE_NTXDESC,
1896 	    BUS_DMASYNC_PREREAD|BUS_DMASYNC_PREWRITE);
1897 	sc->sc_txfree = DGE_NTXDESC;
1898 	sc->sc_txnext = 0;
1899 
1900 	sc->sc_txctx_ipcs = 0xffffffff;
1901 	sc->sc_txctx_tucs = 0xffffffff;
1902 
1903 	CSR_WRITE(sc, DGE_TDBAH, 0);
1904 	CSR_WRITE(sc, DGE_TDBAL, DGE_CDTXADDR(sc, 0));
1905 	CSR_WRITE(sc, DGE_TDLEN, sizeof(sc->sc_txdescs));
1906 	CSR_WRITE(sc, DGE_TDH, 0);
1907 	CSR_WRITE(sc, DGE_TDT, 0);
1908 	CSR_WRITE(sc, DGE_TIDV, TIDV);
1909 
1910 #if 0
1911 	CSR_WRITE(sc, DGE_TXDCTL, TXDCTL_PTHRESH(0) |
1912 	    TXDCTL_HTHRESH(0) | TXDCTL_WTHRESH(0));
1913 #endif
1914 	CSR_WRITE(sc, DGE_RXDCTL,
1915 	    RXDCTL_PTHRESH(RXDCTL_PTHRESH_VAL) |
1916 	    RXDCTL_HTHRESH(RXDCTL_HTHRESH_VAL) |
1917 	    RXDCTL_WTHRESH(RXDCTL_WTHRESH_VAL));
1918 
1919 	/* Initialize the transmit job descriptors. */
1920 	for (i = 0; i < DGE_TXQUEUELEN; i++)
1921 		sc->sc_txsoft[i].txs_mbuf = NULL;
1922 	sc->sc_txsfree = DGE_TXQUEUELEN;
1923 	sc->sc_txsnext = 0;
1924 	sc->sc_txsdirty = 0;
1925 
1926 	/*
1927 	 * Initialize the receive descriptor and receive job
1928 	 * descriptor rings.
1929 	 */
1930 	CSR_WRITE(sc, DGE_RDBAH, 0);
1931 	CSR_WRITE(sc, DGE_RDBAL, DGE_CDRXADDR(sc, 0));
1932 	CSR_WRITE(sc, DGE_RDLEN, sizeof(sc->sc_rxdescs));
1933 	CSR_WRITE(sc, DGE_RDH, DGE_RXSPACE);
1934 	CSR_WRITE(sc, DGE_RDT, 0);
1935 	CSR_WRITE(sc, DGE_RDTR, RDTR | 0x80000000);
1936 	CSR_WRITE(sc, DGE_FCRTL, FCRTL | FCRTL_XONE);
1937 	CSR_WRITE(sc, DGE_FCRTH, FCRTH);
1938 
1939 	for (i = 0; i < DGE_NRXDESC; i++) {
1940 		rxs = &sc->sc_rxsoft[i];
1941 		if (rxs->rxs_mbuf == NULL) {
1942 			if ((error = dge_add_rxbuf(sc, i)) != 0) {
1943 				printf("%s: unable to allocate or map rx "
1944 				    "buffer %d, error = %d\n",
1945 				    device_xname(&sc->sc_dev), i, error);
1946 				/*
1947 				 * XXX Should attempt to run with fewer receive
1948 				 * XXX buffers instead of just failing.
1949 				 */
1950 				dge_rxdrain(sc);
1951 				goto out;
1952 			}
1953 		}
1954 		DGE_INIT_RXDESC(sc, i);
1955 	}
1956 	sc->sc_rxptr = DGE_RXSPACE;
1957 	sc->sc_rxdiscard = 0;
1958 	DGE_RXCHAIN_RESET(sc);
1959 
1960 	if (sc->sc_ethercom.ec_capabilities & ETHERCAP_JUMBO_MTU) {
1961 		sc->sc_ctrl0 |= CTRL0_JFE;
1962 		CSR_WRITE(sc, DGE_MFS, ETHER_MAX_LEN_JUMBO << 16);
1963 	}
1964 
1965 	/* Write the control registers. */
1966 	CSR_WRITE(sc, DGE_CTRL0, sc->sc_ctrl0);
1967 
1968 	/*
1969 	 * Set up checksum offload parameters.
1970 	 */
1971 	reg = CSR_READ(sc, DGE_RXCSUM);
1972 	if (ifp->if_capenable & IFCAP_CSUM_IPv4_Rx)
1973 		reg |= RXCSUM_IPOFL;
1974 	else
1975 		reg &= ~RXCSUM_IPOFL;
1976 	if (ifp->if_capenable & (IFCAP_CSUM_TCPv4_Rx | IFCAP_CSUM_UDPv4_Rx))
1977 		reg |= RXCSUM_IPOFL | RXCSUM_TUOFL;
1978 	else {
1979 		reg &= ~RXCSUM_TUOFL;
1980 		if ((ifp->if_capenable & IFCAP_CSUM_IPv4_Rx) == 0)
1981 			reg &= ~RXCSUM_IPOFL;
1982 	}
1983 	CSR_WRITE(sc, DGE_RXCSUM, reg);
1984 
1985 	/*
1986 	 * Set up the interrupt registers.
1987 	 */
1988 	CSR_WRITE(sc, DGE_IMC, 0xffffffffU);
1989 	sc->sc_icr = ICR_TXDW | ICR_LSC | ICR_RXSEQ | ICR_RXDMT0 |
1990 	    ICR_RXO | ICR_RXT0;
1991 
1992 	CSR_WRITE(sc, DGE_IMS, sc->sc_icr);
1993 
1994 	/*
1995 	 * Set up the transmit control register.
1996 	 */
1997 	sc->sc_tctl = TCTL_TCE|TCTL_TPDE|TCTL_TXEN;
1998 	CSR_WRITE(sc, DGE_TCTL, sc->sc_tctl);
1999 
2000 	/*
2001 	 * Set up the receive control register; we actually program
2002 	 * the register when we set the receive filter.  Use multicast
2003 	 * address offset type 0.
2004 	 */
2005 	sc->sc_mchash_type = 0;
2006 
2007 	sc->sc_rctl = RCTL_RXEN | RCTL_RDMTS_12 | RCTL_RPDA_MC |
2008 	    RCTL_CFF | RCTL_SECRC | RCTL_MO(sc->sc_mchash_type);
2009 
2010 #ifdef DGE_OFFBYONE_RXBUG
2011 	sc->sc_rctl |= RCTL_BSIZE_16k;
2012 #else
2013 	switch(MCLBYTES) {
2014 	case 2048:
2015 		sc->sc_rctl |= RCTL_BSIZE_2k;
2016 		break;
2017 	case 4096:
2018 		sc->sc_rctl |= RCTL_BSIZE_4k;
2019 		break;
2020 	case 8192:
2021 		sc->sc_rctl |= RCTL_BSIZE_8k;
2022 		break;
2023 	case 16384:
2024 		sc->sc_rctl |= RCTL_BSIZE_16k;
2025 		break;
2026 	default:
2027 		panic("dge_init: MCLBYTES %d unsupported", MCLBYTES);
2028 	}
2029 #endif
2030 
2031 	/* Set the receive filter. */
2032 	/* Also sets RCTL */
2033 	dge_set_filter(sc);
2034 
2035 	/* ...all done! */
2036 	ifp->if_flags |= IFF_RUNNING;
2037 	ifp->if_flags &= ~IFF_OACTIVE;
2038 
2039  out:
2040 	if (error)
2041 		printf("%s: interface not running\n", device_xname(&sc->sc_dev));
2042 	return (error);
2043 }
2044 
2045 /*
2046  * dge_rxdrain:
2047  *
2048  *	Drain the receive queue.
2049  */
2050 static void
2051 dge_rxdrain(struct dge_softc *sc)
2052 {
2053 	struct dge_rxsoft *rxs;
2054 	int i;
2055 
2056 	for (i = 0; i < DGE_NRXDESC; i++) {
2057 		rxs = &sc->sc_rxsoft[i];
2058 		if (rxs->rxs_mbuf != NULL) {
2059 			bus_dmamap_unload(sc->sc_dmat, rxs->rxs_dmamap);
2060 			m_freem(rxs->rxs_mbuf);
2061 			rxs->rxs_mbuf = NULL;
2062 		}
2063 	}
2064 }
2065 
2066 /*
2067  * dge_stop:		[ifnet interface function]
2068  *
2069  *	Stop transmission on the interface.
2070  */
2071 static void
2072 dge_stop(struct ifnet *ifp, int disable)
2073 {
2074 	struct dge_softc *sc = ifp->if_softc;
2075 	struct dge_txsoft *txs;
2076 	int i;
2077 
2078 	/* Stop the transmit and receive processes. */
2079 	CSR_WRITE(sc, DGE_TCTL, 0);
2080 	CSR_WRITE(sc, DGE_RCTL, 0);
2081 
2082 	/* Release any queued transmit buffers. */
2083 	for (i = 0; i < DGE_TXQUEUELEN; i++) {
2084 		txs = &sc->sc_txsoft[i];
2085 		if (txs->txs_mbuf != NULL) {
2086 			bus_dmamap_unload(sc->sc_dmat, txs->txs_dmamap);
2087 			m_freem(txs->txs_mbuf);
2088 			txs->txs_mbuf = NULL;
2089 		}
2090 	}
2091 
2092 	/* Mark the interface as down and cancel the watchdog timer. */
2093 	ifp->if_flags &= ~(IFF_RUNNING | IFF_OACTIVE);
2094 	ifp->if_timer = 0;
2095 
2096 	if (disable)
2097 		dge_rxdrain(sc);
2098 }
2099 
2100 /*
2101  * dge_add_rxbuf:
2102  *
2103  *	Add a receive buffer to the indiciated descriptor.
2104  */
2105 static int
2106 dge_add_rxbuf(struct dge_softc *sc, int idx)
2107 {
2108 	struct dge_rxsoft *rxs = &sc->sc_rxsoft[idx];
2109 	struct mbuf *m;
2110 	int error;
2111 #ifdef DGE_OFFBYONE_RXBUG
2112 	void *buf;
2113 #endif
2114 
2115 	MGETHDR(m, M_DONTWAIT, MT_DATA);
2116 	if (m == NULL)
2117 		return (ENOBUFS);
2118 
2119 #ifdef DGE_OFFBYONE_RXBUG
2120 	if ((buf = dge_getbuf(sc)) == NULL)
2121 		return ENOBUFS;
2122 
2123 	m->m_len = m->m_pkthdr.len = DGE_BUFFER_SIZE;
2124 	MEXTADD(m, buf, DGE_BUFFER_SIZE, M_DEVBUF, dge_freebuf, sc);
2125 	m->m_flags |= M_EXT_RW;
2126 
2127 	if (rxs->rxs_mbuf != NULL)
2128 		bus_dmamap_unload(sc->sc_dmat, rxs->rxs_dmamap);
2129 	rxs->rxs_mbuf = m;
2130 
2131 	error = bus_dmamap_load(sc->sc_dmat, rxs->rxs_dmamap, buf,
2132 	    DGE_BUFFER_SIZE, NULL, BUS_DMA_READ|BUS_DMA_NOWAIT);
2133 #else
2134 	MCLGET(m, M_DONTWAIT);
2135 	if ((m->m_flags & M_EXT) == 0) {
2136 		m_freem(m);
2137 		return (ENOBUFS);
2138 	}
2139 
2140 	if (rxs->rxs_mbuf != NULL)
2141 		bus_dmamap_unload(sc->sc_dmat, rxs->rxs_dmamap);
2142 
2143 	rxs->rxs_mbuf = m;
2144 
2145 	m->m_len = m->m_pkthdr.len = m->m_ext.ext_size;
2146 	error = bus_dmamap_load_mbuf(sc->sc_dmat, rxs->rxs_dmamap, m,
2147 	    BUS_DMA_READ|BUS_DMA_NOWAIT);
2148 #endif
2149 	if (error) {
2150 		printf("%s: unable to load rx DMA map %d, error = %d\n",
2151 		    device_xname(&sc->sc_dev), idx, error);
2152 		panic("dge_add_rxbuf");	/* XXX XXX XXX */
2153 	}
2154 	bus_dmamap_sync(sc->sc_dmat, rxs->rxs_dmamap, 0,
2155 	    rxs->rxs_dmamap->dm_mapsize, BUS_DMASYNC_PREREAD);
2156 
2157 	return (0);
2158 }
2159 
2160 /*
2161  * dge_set_ral:
2162  *
2163  *	Set an entry in the receive address list.
2164  */
2165 static void
2166 dge_set_ral(struct dge_softc *sc, const uint8_t *enaddr, int idx)
2167 {
2168 	uint32_t ral_lo, ral_hi;
2169 
2170 	if (enaddr != NULL) {
2171 		ral_lo = enaddr[0] | (enaddr[1] << 8) | (enaddr[2] << 16) |
2172 		    (enaddr[3] << 24);
2173 		ral_hi = enaddr[4] | (enaddr[5] << 8);
2174 		ral_hi |= RAH_AV;
2175 	} else {
2176 		ral_lo = 0;
2177 		ral_hi = 0;
2178 	}
2179 	CSR_WRITE(sc, RA_ADDR(DGE_RAL, idx), ral_lo);
2180 	CSR_WRITE(sc, RA_ADDR(DGE_RAH, idx), ral_hi);
2181 }
2182 
2183 /*
2184  * dge_mchash:
2185  *
2186  *	Compute the hash of the multicast address for the 4096-bit
2187  *	multicast filter.
2188  */
2189 static uint32_t
2190 dge_mchash(struct dge_softc *sc, const uint8_t *enaddr)
2191 {
2192 	static const int lo_shift[4] = { 4, 3, 2, 0 };
2193 	static const int hi_shift[4] = { 4, 5, 6, 8 };
2194 	uint32_t hash;
2195 
2196 	hash = (enaddr[4] >> lo_shift[sc->sc_mchash_type]) |
2197 	    (((uint16_t) enaddr[5]) << hi_shift[sc->sc_mchash_type]);
2198 
2199 	return (hash & 0xfff);
2200 }
2201 
2202 /*
2203  * dge_set_filter:
2204  *
2205  *	Set up the receive filter.
2206  */
2207 static void
2208 dge_set_filter(struct dge_softc *sc)
2209 {
2210 	struct ethercom *ec = &sc->sc_ethercom;
2211 	struct ifnet *ifp = &sc->sc_ethercom.ec_if;
2212 	struct ether_multi *enm;
2213 	struct ether_multistep step;
2214 	uint32_t hash, reg, bit;
2215 	int i;
2216 
2217 	sc->sc_rctl &= ~(RCTL_BAM | RCTL_UPE | RCTL_MPE);
2218 
2219 	if (ifp->if_flags & IFF_BROADCAST)
2220 		sc->sc_rctl |= RCTL_BAM;
2221 	if (ifp->if_flags & IFF_PROMISC) {
2222 		sc->sc_rctl |= RCTL_UPE;
2223 		goto allmulti;
2224 	}
2225 
2226 	/*
2227 	 * Set the station address in the first RAL slot, and
2228 	 * clear the remaining slots.
2229 	 */
2230 	dge_set_ral(sc, CLLADDR(ifp->if_sadl), 0);
2231 	for (i = 1; i < RA_TABSIZE; i++)
2232 		dge_set_ral(sc, NULL, i);
2233 
2234 	/* Clear out the multicast table. */
2235 	for (i = 0; i < MC_TABSIZE; i++)
2236 		CSR_WRITE(sc, DGE_MTA + (i << 2), 0);
2237 
2238 	ETHER_FIRST_MULTI(step, ec, enm);
2239 	while (enm != NULL) {
2240 		if (memcmp(enm->enm_addrlo, enm->enm_addrhi, ETHER_ADDR_LEN)) {
2241 			/*
2242 			 * We must listen to a range of multicast addresses.
2243 			 * For now, just accept all multicasts, rather than
2244 			 * trying to set only those filter bits needed to match
2245 			 * the range.  (At this time, the only use of address
2246 			 * ranges is for IP multicast routing, for which the
2247 			 * range is big enough to require all bits set.)
2248 			 */
2249 			goto allmulti;
2250 		}
2251 
2252 		hash = dge_mchash(sc, enm->enm_addrlo);
2253 
2254 		reg = (hash >> 5) & 0x7f;
2255 		bit = hash & 0x1f;
2256 
2257 		hash = CSR_READ(sc, DGE_MTA + (reg << 2));
2258 		hash |= 1U << bit;
2259 
2260 		CSR_WRITE(sc, DGE_MTA + (reg << 2), hash);
2261 
2262 		ETHER_NEXT_MULTI(step, enm);
2263 	}
2264 
2265 	ifp->if_flags &= ~IFF_ALLMULTI;
2266 	goto setit;
2267 
2268  allmulti:
2269 	ifp->if_flags |= IFF_ALLMULTI;
2270 	sc->sc_rctl |= RCTL_MPE;
2271 
2272  setit:
2273 	CSR_WRITE(sc, DGE_RCTL, sc->sc_rctl);
2274 }
2275 
2276 /*
2277  * Read in the EEPROM info and verify checksum.
2278  */
2279 int
2280 dge_read_eeprom(struct dge_softc *sc)
2281 {
2282 	uint16_t cksum;
2283 	int i;
2284 
2285 	cksum = 0;
2286 	for (i = 0; i < EEPROM_SIZE; i++) {
2287 		sc->sc_eeprom[i] = dge_eeprom_word(sc, i);
2288 		cksum += sc->sc_eeprom[i];
2289 	}
2290 	return cksum != EEPROM_CKSUM;
2291 }
2292 
2293 
2294 /*
2295  * Read a 16-bit word from address addr in the serial EEPROM.
2296  */
2297 uint16_t
2298 dge_eeprom_word(struct dge_softc *sc, int addr)
2299 {
2300 	uint32_t reg;
2301 	uint16_t rval = 0;
2302 	int i;
2303 
2304 	reg = CSR_READ(sc, DGE_EECD) & ~(EECD_SK|EECD_DI|EECD_CS);
2305 
2306 	/* Lower clock pulse (and data in to chip) */
2307 	CSR_WRITE(sc, DGE_EECD, reg);
2308 	/* Select chip */
2309 	CSR_WRITE(sc, DGE_EECD, reg|EECD_CS);
2310 
2311 	/* Send read command */
2312 	dge_eeprom_clockout(sc, 1);
2313 	dge_eeprom_clockout(sc, 1);
2314 	dge_eeprom_clockout(sc, 0);
2315 
2316 	/* Send address */
2317 	for (i = 5; i >= 0; i--)
2318 		dge_eeprom_clockout(sc, (addr >> i) & 1);
2319 
2320 	/* Read data */
2321 	for (i = 0; i < 16; i++) {
2322 		rval <<= 1;
2323 		rval |= dge_eeprom_clockin(sc);
2324 	}
2325 
2326 	/* Deselect chip */
2327 	CSR_WRITE(sc, DGE_EECD, reg);
2328 
2329 	return rval;
2330 }
2331 
2332 /*
2333  * Clock out a single bit to the EEPROM.
2334  */
2335 void
2336 dge_eeprom_clockout(struct dge_softc *sc, int bit)
2337 {
2338 	int reg;
2339 
2340 	reg = CSR_READ(sc, DGE_EECD) & ~(EECD_DI|EECD_SK);
2341 	if (bit)
2342 		reg |= EECD_DI;
2343 
2344 	CSR_WRITE(sc, DGE_EECD, reg);
2345 	delay(2);
2346 	CSR_WRITE(sc, DGE_EECD, reg|EECD_SK);
2347 	delay(2);
2348 	CSR_WRITE(sc, DGE_EECD, reg);
2349 	delay(2);
2350 }
2351 
2352 /*
2353  * Clock in a single bit from EEPROM.
2354  */
2355 int
2356 dge_eeprom_clockin(struct dge_softc *sc)
2357 {
2358 	int reg, rv;
2359 
2360 	reg = CSR_READ(sc, DGE_EECD) & ~(EECD_DI|EECD_DO|EECD_SK);
2361 
2362 	CSR_WRITE(sc, DGE_EECD, reg|EECD_SK); /* Raise clock */
2363 	delay(2);
2364 	rv = (CSR_READ(sc, DGE_EECD) & EECD_DO) != 0; /* Get bit */
2365 	CSR_WRITE(sc, DGE_EECD, reg); /* Lower clock */
2366 	delay(2);
2367 
2368 	return rv;
2369 }
2370 
2371 static void
2372 dge_xgmii_mediastatus(struct ifnet *ifp, struct ifmediareq *ifmr)
2373 {
2374 	struct dge_softc *sc = ifp->if_softc;
2375 
2376 	ifmr->ifm_status = IFM_AVALID;
2377 	ifmr->ifm_active = IFM_ETHER|IFM_10G_LR;
2378 
2379 	if (CSR_READ(sc, DGE_STATUS) & STATUS_LINKUP)
2380 		ifmr->ifm_status |= IFM_ACTIVE;
2381 }
2382 
2383 static inline int
2384 phwait(struct dge_softc *sc, int p, int r, int d, int type)
2385 {
2386         int i, mdic;
2387 
2388         CSR_WRITE(sc, DGE_MDIO,
2389 	    MDIO_PHY(p) | MDIO_REG(r) | MDIO_DEV(d) | type | MDIO_CMD);
2390         for (i = 0; i < 10; i++) {
2391                 delay(10);
2392                 if (((mdic = CSR_READ(sc, DGE_MDIO)) & MDIO_CMD) == 0)
2393                         break;
2394         }
2395         return mdic;
2396 }
2397 
2398 
2399 static void
2400 dge_xgmii_writereg(struct device *self, int phy, int reg, int val)
2401 {
2402 	struct dge_softc *sc = (void *) self;
2403 	int mdic;
2404 
2405 	CSR_WRITE(sc, DGE_MDIRW, val);
2406 	if (((mdic = phwait(sc, phy, reg, 1, MDIO_ADDR)) & MDIO_CMD)) {
2407 		printf("%s: address cycle timeout; phy %d reg %d\n",
2408 		    device_xname(&sc->sc_dev), phy, reg);
2409 		return;
2410 	}
2411 	if (((mdic = phwait(sc, phy, reg, 1, MDIO_WRITE)) & MDIO_CMD)) {
2412 		printf("%s: read cycle timeout; phy %d reg %d\n",
2413 		    device_xname(&sc->sc_dev), phy, reg);
2414 		return;
2415 	}
2416 }
2417 
2418 static void
2419 dge_xgmii_reset(struct dge_softc *sc)
2420 {
2421 	dge_xgmii_writereg((void *)sc, 0, 0, BMCR_RESET);
2422 }
2423 
2424 static int
2425 dge_xgmii_mediachange(struct ifnet *ifp)
2426 {
2427 	return 0;
2428 }
2429