xgbe-drv.c 75.0 KB
Newer Older
1 2 3 4 5 6 7 8
/*
 * AMD 10Gb Ethernet driver
 *
 * This file is available to you under your choice of the following two
 * licenses:
 *
 * License 1: GPLv2
 *
9
 * Copyright (c) 2014-2016 Advanced Micro Devices, Inc.
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
 *
 * This file is free software; you may copy, redistribute and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 2 of the License, or (at
 * your option) any later version.
 *
 * This file is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 *
 * This file incorporates work covered by the following copyright and
 * permission notice:
 *     The Synopsys DWC ETHER XGMAC Software Driver and documentation
 *     (hereinafter "Software") is an unsupported proprietary work of Synopsys,
 *     Inc. unless otherwise expressly agreed to in writing between Synopsys
 *     and you.
 *
 *     The Software IS NOT an item of Licensed Software or Licensed Product
 *     under any End User Software License Agreement or Agreement for Licensed
 *     Product with Synopsys or any supplement thereto.  Permission is hereby
 *     granted, free of charge, to any person obtaining a copy of this software
 *     annotated with this license and the Software, to deal in the Software
 *     without restriction, including without limitation the rights to use,
 *     copy, modify, merge, publish, distribute, sublicense, and/or sell copies
 *     of the Software, and to permit persons to whom the Software is furnished
 *     to do so, subject to the following conditions:
 *
 *     The above copyright notice and this permission notice shall be included
 *     in all copies or substantial portions of the Software.
 *
 *     THIS SOFTWARE IS BEING DISTRIBUTED BY SYNOPSYS SOLELY ON AN "AS IS"
 *     BASIS AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
 *     TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
 *     PARTICULAR PURPOSE ARE HEREBY DISCLAIMED. IN NO EVENT SHALL SYNOPSYS
 *     BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 *     CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 *     SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 *     INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 *     CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 *     ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
 *     THE POSSIBILITY OF SUCH DAMAGE.
 *
 *
 * License 2: Modified BSD
 *
59
 * Copyright (c) 2014-2016 Advanced Micro Devices, Inc.
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *     * Redistributions of source code must retain the above copyright
 *       notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright
 *       notice, this list of conditions and the following disclaimer in the
 *       documentation and/or other materials provided with the distribution.
 *     * Neither the name of Advanced Micro Devices, Inc. nor the
 *       names of its contributors may be used to endorse or promote products
 *       derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL <COPYRIGHT HOLDER> BE LIABLE FOR ANY
 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * This file incorporates work covered by the following copyright and
 * permission notice:
 *     The Synopsys DWC ETHER XGMAC Software Driver and documentation
 *     (hereinafter "Software") is an unsupported proprietary work of Synopsys,
 *     Inc. unless otherwise expressly agreed to in writing between Synopsys
 *     and you.
 *
 *     The Software IS NOT an item of Licensed Software or Licensed Product
 *     under any End User Software License Agreement or Agreement for Licensed
 *     Product with Synopsys or any supplement thereto.  Permission is hereby
 *     granted, free of charge, to any person obtaining a copy of this software
 *     annotated with this license and the Software, to deal in the Software
 *     without restriction, including without limitation the rights to use,
 *     copy, modify, merge, publish, distribute, sublicense, and/or sell copies
 *     of the Software, and to permit persons to whom the Software is furnished
 *     to do so, subject to the following conditions:
 *
 *     The above copyright notice and this permission notice shall be included
 *     in all copies or substantial portions of the Software.
 *
 *     THIS SOFTWARE IS BEING DISTRIBUTED BY SYNOPSYS SOLELY ON AN "AS IS"
 *     BASIS AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
 *     TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
 *     PARTICULAR PURPOSE ARE HEREBY DISCLAIMED. IN NO EVENT SHALL SYNOPSYS
 *     BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 *     CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 *     SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 *     INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 *     CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 *     ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
 *     THE POSSIBILITY OF SUCH DAMAGE.
 */

117
#include <linux/module.h>
118 119 120
#include <linux/spinlock.h>
#include <linux/tcp.h>
#include <linux/if_vlan.h>
121
#include <linux/interrupt.h>
122 123
#include <linux/clk.h>
#include <linux/if_ether.h>
124
#include <linux/net_tstamp.h>
125
#include <linux/phy.h>
126
#include <net/vxlan.h>
127 128 129 130

#include "xgbe.h"
#include "xgbe-common.h"

131 132 133 134 135 136 137 138
static unsigned int ecc_sec_info_threshold = 10;
static unsigned int ecc_sec_warn_threshold = 10000;
static unsigned int ecc_sec_period = 600;
static unsigned int ecc_ded_threshold = 2;
static unsigned int ecc_ded_period = 600;

#ifdef CONFIG_AMD_XGBE_HAVE_ECC
/* Only expose the ECC parameters if supported */
139
module_param(ecc_sec_info_threshold, uint, 0644);
140 141 142
MODULE_PARM_DESC(ecc_sec_info_threshold,
		 " ECC corrected error informational threshold setting");

143
module_param(ecc_sec_warn_threshold, uint, 0644);
144 145 146
MODULE_PARM_DESC(ecc_sec_warn_threshold,
		 " ECC corrected error warning threshold setting");

147
module_param(ecc_sec_period, uint, 0644);
148 149
MODULE_PARM_DESC(ecc_sec_period, " ECC corrected error period (in seconds)");

150
module_param(ecc_ded_threshold, uint, 0644);
151 152
MODULE_PARM_DESC(ecc_ded_threshold, " ECC detected error threshold setting");

153
module_param(ecc_ded_period, uint, 0644);
154 155 156
MODULE_PARM_DESC(ecc_ded_period, " ECC detected error period (in seconds)");
#endif

157 158
static int xgbe_one_poll(struct napi_struct *, int);
static int xgbe_all_poll(struct napi_struct *, int);
159
static void xgbe_stop(struct xgbe_prv_data *);
160

161
static void *xgbe_alloc_node(size_t size, int node)
162
{
163
	void *mem;
164

165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185
	mem = kzalloc_node(size, GFP_KERNEL, node);
	if (!mem)
		mem = kzalloc(size, GFP_KERNEL);

	return mem;
}

static void xgbe_free_channels(struct xgbe_prv_data *pdata)
{
	unsigned int i;

	for (i = 0; i < ARRAY_SIZE(pdata->channel); i++) {
		if (!pdata->channel[i])
			continue;

		kfree(pdata->channel[i]->rx_ring);
		kfree(pdata->channel[i]->tx_ring);
		kfree(pdata->channel[i]);

		pdata->channel[i] = NULL;
	}
186

187 188 189 190 191 192 193 194
	pdata->channel_count = 0;
}

static int xgbe_alloc_channels(struct xgbe_prv_data *pdata)
{
	struct xgbe_channel *channel;
	struct xgbe_ring *ring;
	unsigned int count, i;
195
	unsigned int cpu;
196
	int node;
197

198 199
	count = max_t(unsigned int, pdata->tx_ring_count, pdata->rx_ring_count);
	for (i = 0; i < count; i++) {
200 201 202 203 204 205
		/* Attempt to use a CPU on the node the device is on */
		cpu = cpumask_local_spread(i, dev_to_node(pdata->dev));

		/* Set the allocation node based on the returned CPU */
		node = cpu_to_node(cpu);

206 207 208 209
		channel = xgbe_alloc_node(sizeof(*channel), node);
		if (!channel)
			goto err_mem;
		pdata->channel[i] = channel;
210

211
		snprintf(channel->name, sizeof(channel->name), "channel-%u", i);
212 213 214 215
		channel->pdata = pdata;
		channel->queue_index = i;
		channel->dma_regs = pdata->xgmac_regs + DMA_CH_BASE +
				    (DMA_CH_INC * i);
216
		channel->node = node;
217
		cpumask_set_cpu(cpu, &channel->affinity_mask);
218

219 220
		if (pdata->per_channel_irq)
			channel->dma_irq = pdata->channel_irq[i];
221

222
		if (i < pdata->tx_ring_count) {
223 224 225 226 227 228 229 230
			ring = xgbe_alloc_node(sizeof(*ring), node);
			if (!ring)
				goto err_mem;

			spin_lock_init(&ring->lock);
			ring->node = node;

			channel->tx_ring = ring;
231 232 233
		}

		if (i < pdata->rx_ring_count) {
234 235 236 237 238 239 240 241
			ring = xgbe_alloc_node(sizeof(*ring), node);
			if (!ring)
				goto err_mem;

			spin_lock_init(&ring->lock);
			ring->node = node;

			channel->rx_ring = ring;
242 243
		}

244
		netif_dbg(pdata, drv, pdata->netdev,
245
			  "%s: cpu=%u, node=%d\n", channel->name, cpu, node);
246

247 248 249 250
		netif_dbg(pdata, drv, pdata->netdev,
			  "%s: dma_regs=%p, dma_irq=%d, tx=%p, rx=%p\n",
			  channel->name, channel->dma_regs, channel->dma_irq,
			  channel->tx_ring, channel->rx_ring);
251 252 253 254 255 256
	}

	pdata->channel_count = count;

	return 0;

257 258
err_mem:
	xgbe_free_channels(pdata);
259

260
	return -ENOMEM;
261 262
}

263 264 265 266 267
static inline unsigned int xgbe_tx_avail_desc(struct xgbe_ring *ring)
{
	return (ring->rdesc_count - (ring->cur - ring->dirty));
}

268 269 270 271 272
static inline unsigned int xgbe_rx_dirty_desc(struct xgbe_ring *ring)
{
	return (ring->cur - ring->dirty);
}

273 274 275 276 277 278
static int xgbe_maybe_stop_tx_queue(struct xgbe_channel *channel,
				    struct xgbe_ring *ring, unsigned int count)
{
	struct xgbe_prv_data *pdata = channel->pdata;

	if (count > xgbe_tx_avail_desc(ring)) {
279 280
		netif_info(pdata, drv, pdata->netdev,
			   "Tx queue stopped, not enough descriptors available\n");
281 282 283 284 285 286 287 288 289 290 291 292 293 294 295
		netif_stop_subqueue(pdata->netdev, channel->queue_index);
		ring->tx.queue_stopped = 1;

		/* If we haven't notified the hardware because of xmit_more
		 * support, tell it now
		 */
		if (ring->tx.xmit_more)
			pdata->hw_if.tx_start_xmit(channel, ring);

		return NETDEV_TX_BUSY;
	}

	return 0;
}

296 297 298 299 300
static int xgbe_calc_rx_buf_size(struct net_device *netdev, unsigned int mtu)
{
	unsigned int rx_buf_size;

	rx_buf_size = mtu + ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN;
301 302
	rx_buf_size = clamp_val(rx_buf_size, XGBE_RX_MIN_BUF_SIZE, PAGE_SIZE);

303 304
	rx_buf_size = (rx_buf_size + XGBE_RX_BUF_ALIGN - 1) &
		      ~(XGBE_RX_BUF_ALIGN - 1);
305 306 307 308

	return rx_buf_size;
}

309 310
static void xgbe_enable_rx_tx_int(struct xgbe_prv_data *pdata,
				  struct xgbe_channel *channel)
311 312
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
313
	enum xgbe_int int_id;
314 315 316 317 318 319 320 321 322 323 324 325 326 327 328

	if (channel->tx_ring && channel->rx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_TI_RI;
	else if (channel->tx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_TI;
	else if (channel->rx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_RI;
	else
		return;

	hw_if->enable_int(channel, int_id);
}

static void xgbe_enable_rx_tx_ints(struct xgbe_prv_data *pdata)
{
329 330
	unsigned int i;

331 332
	for (i = 0; i < pdata->channel_count; i++)
		xgbe_enable_rx_tx_int(pdata, pdata->channel[i]);
333
}
334

335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350
static void xgbe_disable_rx_tx_int(struct xgbe_prv_data *pdata,
				   struct xgbe_channel *channel)
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	enum xgbe_int int_id;

	if (channel->tx_ring && channel->rx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_TI_RI;
	else if (channel->tx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_TI;
	else if (channel->rx_ring)
		int_id = XGMAC_INT_DMA_CH_SR_RI;
	else
		return;

	hw_if->disable_int(channel, int_id);
351 352 353 354 355 356
}

static void xgbe_disable_rx_tx_ints(struct xgbe_prv_data *pdata)
{
	unsigned int i;

357 358
	for (i = 0; i < pdata->channel_count; i++)
		xgbe_disable_rx_tx_int(pdata, pdata->channel[i]);
359 360
}

361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405
static bool xgbe_ecc_sec(struct xgbe_prv_data *pdata, unsigned long *period,
			 unsigned int *count, const char *area)
{
	if (time_before(jiffies, *period)) {
		(*count)++;
	} else {
		*period = jiffies + (ecc_sec_period * HZ);
		*count = 1;
	}

	if (*count > ecc_sec_info_threshold)
		dev_warn_once(pdata->dev,
			      "%s ECC corrected errors exceed informational threshold\n",
			      area);

	if (*count > ecc_sec_warn_threshold) {
		dev_warn_once(pdata->dev,
			      "%s ECC corrected errors exceed warning threshold\n",
			      area);
		return true;
	}

	return false;
}

static bool xgbe_ecc_ded(struct xgbe_prv_data *pdata, unsigned long *period,
			 unsigned int *count, const char *area)
{
	if (time_before(jiffies, *period)) {
		(*count)++;
	} else {
		*period = jiffies + (ecc_ded_period * HZ);
		*count = 1;
	}

	if (*count > ecc_ded_threshold) {
		netdev_alert(pdata->netdev,
			     "%s ECC detected errors exceed threshold\n",
			     area);
		return true;
	}

	return false;
}

406
static void xgbe_ecc_isr_task(struct tasklet_struct *t)
407
{
408
	struct xgbe_prv_data *pdata = from_tasklet(pdata, t, tasklet_ecc);
409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458
	unsigned int ecc_isr;
	bool stop = false;

	/* Mask status with only the interrupts we care about */
	ecc_isr = XP_IOREAD(pdata, XP_ECC_ISR);
	ecc_isr &= XP_IOREAD(pdata, XP_ECC_IER);
	netif_dbg(pdata, intr, pdata->netdev, "ECC_ISR=%#010x\n", ecc_isr);

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, TX_DED)) {
		stop |= xgbe_ecc_ded(pdata, &pdata->tx_ded_period,
				     &pdata->tx_ded_count, "TX fifo");
	}

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, RX_DED)) {
		stop |= xgbe_ecc_ded(pdata, &pdata->rx_ded_period,
				     &pdata->rx_ded_count, "RX fifo");
	}

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, DESC_DED)) {
		stop |= xgbe_ecc_ded(pdata, &pdata->desc_ded_period,
				     &pdata->desc_ded_count,
				     "descriptor cache");
	}

	if (stop) {
		pdata->hw_if.disable_ecc_ded(pdata);
		schedule_work(&pdata->stopdev_work);
		goto out;
	}

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, TX_SEC)) {
		if (xgbe_ecc_sec(pdata, &pdata->tx_sec_period,
				 &pdata->tx_sec_count, "TX fifo"))
			pdata->hw_if.disable_ecc_sec(pdata, XGBE_ECC_SEC_TX);
	}

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, RX_SEC))
		if (xgbe_ecc_sec(pdata, &pdata->rx_sec_period,
				 &pdata->rx_sec_count, "RX fifo"))
			pdata->hw_if.disable_ecc_sec(pdata, XGBE_ECC_SEC_RX);

	if (XP_GET_BITS(ecc_isr, XP_ECC_ISR, DESC_SEC))
		if (xgbe_ecc_sec(pdata, &pdata->desc_sec_period,
				 &pdata->desc_sec_count, "descriptor cache"))
			pdata->hw_if.disable_ecc_sec(pdata, XGBE_ECC_SEC_DESC);

out:
	/* Clear all ECC interrupts */
	XP_IOWRITE(pdata, XP_ECC_ISR, ecc_isr);

459 460 461
	/* Reissue interrupt if status is not clear */
	if (pdata->vdata->irq_reissue_support)
		XP_IOWRITE(pdata, XP_INT_REISSUE_EN, 1 << 1);
462 463
}

464
static irqreturn_t xgbe_ecc_isr(int irq, void *data)
465 466
{
	struct xgbe_prv_data *pdata = data;
467 468 469 470

	if (pdata->isr_as_tasklet)
		tasklet_schedule(&pdata->tasklet_ecc);
	else
471
		xgbe_ecc_isr_task(&pdata->tasklet_ecc);
472 473 474 475

	return IRQ_HANDLED;
}

476
static void xgbe_isr_task(struct tasklet_struct *t)
477
{
478
	struct xgbe_prv_data *pdata = from_tasklet(pdata, t, tasklet_dev);
479 480 481
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	struct xgbe_channel *channel;
	unsigned int dma_isr, dma_ch_isr;
482
	unsigned int mac_isr, mac_tssr, mac_mdioisr;
483 484 485 486 487 488 489 490 491 492
	unsigned int i;

	/* The DMA interrupt status register also reports MAC and MTL
	 * interrupts. So for polling mode, we just need to check for
	 * this register to be non-zero
	 */
	dma_isr = XGMAC_IOREAD(pdata, DMA_ISR);
	if (!dma_isr)
		goto isr_done;

493
	netif_dbg(pdata, intr, pdata->netdev, "DMA_ISR=%#010x\n", dma_isr);
494 495 496 497 498

	for (i = 0; i < pdata->channel_count; i++) {
		if (!(dma_isr & (1 << i)))
			continue;

499
		channel = pdata->channel[i];
500 501

		dma_ch_isr = XGMAC_DMA_IOREAD(channel, DMA_CH_SR);
502 503
		netif_dbg(pdata, intr, pdata->netdev, "DMA_CH%u_ISR=%#010x\n",
			  i, dma_ch_isr);
504

505 506 507
		/* The TI or RI interrupt bits may still be set even if using
		 * per channel DMA interrupts. Check to be sure those are not
		 * enabled before using the private data napi structure.
508
		 */
509 510 511
		if (!pdata->per_channel_irq &&
		    (XGMAC_GET_BITS(dma_ch_isr, DMA_CH_SR, TI) ||
		     XGMAC_GET_BITS(dma_ch_isr, DMA_CH_SR, RI))) {
512 513 514 515 516
			if (napi_schedule_prep(&pdata->napi)) {
				/* Disable Tx and Rx interrupts */
				xgbe_disable_rx_tx_ints(pdata);

				/* Turn on polling */
517
				__napi_schedule(&pdata->napi);
518
			}
519 520 521 522 523 524 525
		} else {
			/* Don't clear Rx/Tx status if doing per channel DMA
			 * interrupts, these will be cleared by the ISR for
			 * per channel DMA interrupts.
			 */
			XGMAC_SET_BITS(dma_ch_isr, DMA_CH_SR, TI, 0);
			XGMAC_SET_BITS(dma_ch_isr, DMA_CH_SR, RI, 0);
526 527
		}

528 529 530
		if (XGMAC_GET_BITS(dma_ch_isr, DMA_CH_SR, RBU))
			pdata->ext_stats.rx_buffer_unavailable++;

531 532
		/* Restart the device on a Fatal Bus Error */
		if (XGMAC_GET_BITS(dma_ch_isr, DMA_CH_SR, FBE))
533
			schedule_work(&pdata->restart_work);
534

535
		/* Clear interrupt signals */
536 537 538 539 540 541
		XGMAC_DMA_IOWRITE(channel, DMA_CH_SR, dma_ch_isr);
	}

	if (XGMAC_GET_BITS(dma_isr, DMA_ISR, MACIS)) {
		mac_isr = XGMAC_IOREAD(pdata, MAC_ISR);

542 543 544
		netif_dbg(pdata, intr, pdata->netdev, "MAC_ISR=%#010x\n",
			  mac_isr);

545 546 547 548 549
		if (XGMAC_GET_BITS(mac_isr, MAC_ISR, MMCTXIS))
			hw_if->tx_mmc_int(pdata);

		if (XGMAC_GET_BITS(mac_isr, MAC_ISR, MMCRXIS))
			hw_if->rx_mmc_int(pdata);
550 551 552 553

		if (XGMAC_GET_BITS(mac_isr, MAC_ISR, TSIS)) {
			mac_tssr = XGMAC_IOREAD(pdata, MAC_TSSR);

554 555 556
			netif_dbg(pdata, intr, pdata->netdev,
				  "MAC_TSSR=%#010x\n", mac_tssr);

557 558 559 560
			if (XGMAC_GET_BITS(mac_tssr, MAC_TSSR, TXTSC)) {
				/* Read Tx Timestamp to clear interrupt */
				pdata->tx_tstamp =
					hw_if->get_tx_tstamp(pdata);
561 562
				queue_work(pdata->dev_workqueue,
					   &pdata->tx_tstamp_work);
563 564
			}
		}
565 566 567 568 569 570 571 572 573 574 575

		if (XGMAC_GET_BITS(mac_isr, MAC_ISR, SMI)) {
			mac_mdioisr = XGMAC_IOREAD(pdata, MAC_MDIOISR);

			netif_dbg(pdata, intr, pdata->netdev,
				  "MAC_MDIOISR=%#010x\n", mac_mdioisr);

			if (XGMAC_GET_BITS(mac_mdioisr, MAC_MDIOISR,
					   SNGLCOMPINT))
				complete(&pdata->mdio_complete);
		}
576 577
	}

578
isr_done:
579 580
	/* If there is not a separate AN irq, handle it here */
	if (pdata->dev_irq == pdata->an_irq)
581
		pdata->phy_if.an_isr(pdata);
582

583 584
	/* If there is not a separate ECC irq, handle it here */
	if (pdata->vdata->ecc_support && (pdata->dev_irq == pdata->ecc_irq))
585
		xgbe_ecc_isr_task(&pdata->tasklet_ecc);
586

587 588
	/* If there is not a separate I2C irq, handle it here */
	if (pdata->vdata->i2c_support && (pdata->dev_irq == pdata->i2c_irq))
589 590 591 592 593 594 595 596
		pdata->i2c_if.i2c_isr(pdata);

	/* Reissue interrupt if status is not clear */
	if (pdata->vdata->irq_reissue_support) {
		unsigned int reissue_mask;

		reissue_mask = 1 << 0;
		if (!pdata->per_channel_irq)
597
			reissue_mask |= 0xffff << 4;
598 599 600 601 602 603 604 605 606 607 608 609

		XP_IOWRITE(pdata, XP_INT_REISSUE_EN, reissue_mask);
	}
}

static irqreturn_t xgbe_isr(int irq, void *data)
{
	struct xgbe_prv_data *pdata = data;

	if (pdata->isr_as_tasklet)
		tasklet_schedule(&pdata->tasklet_dev);
	else
610
		xgbe_isr_task(&pdata->tasklet_dev);
611

612 613 614
	return IRQ_HANDLED;
}

615 616 617
static irqreturn_t xgbe_dma_isr(int irq, void *data)
{
	struct xgbe_channel *channel = data;
618 619
	struct xgbe_prv_data *pdata = channel->pdata;
	unsigned int dma_status;
620 621 622 623 624 625

	/* Per channel DMA interrupts are enabled, so we use the per
	 * channel napi structure and not the private data napi structure
	 */
	if (napi_schedule_prep(&channel->napi)) {
		/* Disable Tx and Rx interrupts */
626 627 628 629
		if (pdata->channel_irq_mode)
			xgbe_disable_rx_tx_int(pdata, channel);
		else
			disable_irq_nosync(channel->dma_irq);
630 631

		/* Turn on polling */
632
		__napi_schedule_irqoff(&channel->napi);
633 634
	}

635 636 637 638 639 640
	/* Clear Tx/Rx signals */
	dma_status = 0;
	XGMAC_SET_BITS(dma_status, DMA_CH_SR, TI, 1);
	XGMAC_SET_BITS(dma_status, DMA_CH_SR, RI, 1);
	XGMAC_DMA_IOWRITE(channel, DMA_CH_SR, dma_status);

641 642 643
	return IRQ_HANDLED;
}

644
static void xgbe_tx_timer(struct timer_list *t)
645
{
646
	struct xgbe_channel *channel = from_timer(channel, t, tx_timer);
647
	struct xgbe_prv_data *pdata = channel->pdata;
648
	struct napi_struct *napi;
649 650 651

	DBGPR("-->xgbe_tx_timer\n");

652 653 654
	napi = (pdata->per_channel_irq) ? &channel->napi : &pdata->napi;

	if (napi_schedule_prep(napi)) {
655
		/* Disable Tx and Rx interrupts */
656
		if (pdata->per_channel_irq)
657 658 659 660
			if (pdata->channel_irq_mode)
				xgbe_disable_rx_tx_int(pdata, channel);
			else
				disable_irq_nosync(channel->dma_irq);
661 662
		else
			xgbe_disable_rx_tx_ints(pdata);
663 664

		/* Turn on polling */
665
		__napi_schedule(napi);
666 667 668 669 670 671 672
	}

	channel->tx_timer_active = 0;

	DBGPR("<--xgbe_tx_timer\n");
}

673 674 675 676 677 678 679 680 681
static void xgbe_service(struct work_struct *work)
{
	struct xgbe_prv_data *pdata = container_of(work,
						   struct xgbe_prv_data,
						   service_work);

	pdata->phy_if.phy_status(pdata);
}

682
static void xgbe_service_timer(struct timer_list *t)
683
{
684
	struct xgbe_prv_data *pdata = from_timer(pdata, t, service_timer);
685

686
	queue_work(pdata->dev_workqueue, &pdata->service_work);
687 688 689 690 691

	mod_timer(&pdata->service_timer, jiffies + HZ);
}

static void xgbe_init_timers(struct xgbe_prv_data *pdata)
692 693 694 695
{
	struct xgbe_channel *channel;
	unsigned int i;

696
	timer_setup(&pdata->service_timer, xgbe_service_timer, 0);
697

698 699
	for (i = 0; i < pdata->channel_count; i++) {
		channel = pdata->channel[i];
700 701 702
		if (!channel->tx_ring)
			break;

703
		timer_setup(&channel->tx_timer, xgbe_tx_timer, 0);
704
	}
705
}
706

707 708 709
static void xgbe_start_timers(struct xgbe_prv_data *pdata)
{
	mod_timer(&pdata->service_timer, jiffies + HZ);
710 711
}

712
static void xgbe_stop_timers(struct xgbe_prv_data *pdata)
713 714 715 716
{
	struct xgbe_channel *channel;
	unsigned int i;

717
	del_timer_sync(&pdata->service_timer);
718

719 720
	for (i = 0; i < pdata->channel_count; i++) {
		channel = pdata->channel[i];
721 722 723
		if (!channel->tx_ring)
			break;

724
		/* Deactivate the Tx timer */
725
		del_timer_sync(&channel->tx_timer);
726
		channel->tx_timer_active = 0;
727 728 729 730 731 732 733 734 735 736 737 738 739 740
	}
}

void xgbe_get_all_hw_features(struct xgbe_prv_data *pdata)
{
	unsigned int mac_hfr0, mac_hfr1, mac_hfr2;
	struct xgbe_hw_features *hw_feat = &pdata->hw_feat;

	mac_hfr0 = XGMAC_IOREAD(pdata, MAC_HWF0R);
	mac_hfr1 = XGMAC_IOREAD(pdata, MAC_HWF1R);
	mac_hfr2 = XGMAC_IOREAD(pdata, MAC_HWF2R);

	memset(hw_feat, 0, sizeof(*hw_feat));

741 742
	hw_feat->version = XGMAC_IOREAD(pdata, MAC_VR);

743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758
	/* Hardware feature register 0 */
	hw_feat->gmii        = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, GMIISEL);
	hw_feat->vlhash      = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, VLHASH);
	hw_feat->sma         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, SMASEL);
	hw_feat->rwk         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, RWKSEL);
	hw_feat->mgk         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, MGKSEL);
	hw_feat->mmc         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, MMCSEL);
	hw_feat->aoe         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, ARPOFFSEL);
	hw_feat->ts          = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, TSSEL);
	hw_feat->eee         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, EEESEL);
	hw_feat->tx_coe      = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, TXCOESEL);
	hw_feat->rx_coe      = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, RXCOESEL);
	hw_feat->addn_mac    = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R,
					      ADDMACADRSEL);
	hw_feat->ts_src      = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, TSSTSSEL);
	hw_feat->sa_vlan_ins = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, SAVLANINS);
759
	hw_feat->vxn         = XGMAC_GET_BITS(mac_hfr0, MAC_HWF0R, VXN);
760 761 762 763 764 765

	/* Hardware feature register 1 */
	hw_feat->rx_fifo_size  = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R,
						RXFIFOSIZE);
	hw_feat->tx_fifo_size  = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R,
						TXFIFOSIZE);
766
	hw_feat->adv_ts_hi     = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, ADVTHWORD);
767
	hw_feat->dma_width     = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, ADDR64);
768 769 770 771
	hw_feat->dcb           = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, DCBEN);
	hw_feat->sph           = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, SPHEN);
	hw_feat->tso           = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, TSOEN);
	hw_feat->dma_debug     = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, DBGMEMA);
772
	hw_feat->rss           = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, RSSEN);
773
	hw_feat->tc_cnt	       = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R, NUMTC);
774 775 776 777 778 779 780 781 782 783 784 785 786
	hw_feat->hash_table_size = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R,
						  HASHTBLSZ);
	hw_feat->l3l4_filter_num = XGMAC_GET_BITS(mac_hfr1, MAC_HWF1R,
						  L3L4FNUM);

	/* Hardware feature register 2 */
	hw_feat->rx_q_cnt     = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, RXQCNT);
	hw_feat->tx_q_cnt     = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, TXQCNT);
	hw_feat->rx_ch_cnt    = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, RXCHCNT);
	hw_feat->tx_ch_cnt    = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, TXCHCNT);
	hw_feat->pps_out_num  = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, PPSOUTNUM);
	hw_feat->aux_snap_num = XGMAC_GET_BITS(mac_hfr2, MAC_HWF2R, AUXSNAPNUM);

787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
	/* Translate the Hash Table size into actual number */
	switch (hw_feat->hash_table_size) {
	case 0:
		break;
	case 1:
		hw_feat->hash_table_size = 64;
		break;
	case 2:
		hw_feat->hash_table_size = 128;
		break;
	case 3:
		hw_feat->hash_table_size = 256;
		break;
	}

802 803 804 805 806 807 808 809 810 811 812 813 814 815 816
	/* Translate the address width setting into actual number */
	switch (hw_feat->dma_width) {
	case 0:
		hw_feat->dma_width = 32;
		break;
	case 1:
		hw_feat->dma_width = 40;
		break;
	case 2:
		hw_feat->dma_width = 48;
		break;
	default:
		hw_feat->dma_width = 32;
	}

817
	/* The Queue, Channel and TC counts are zero based so increment them
818 819 820 821 822 823
	 * to get the actual number
	 */
	hw_feat->rx_q_cnt++;
	hw_feat->tx_q_cnt++;
	hw_feat->rx_ch_cnt++;
	hw_feat->tx_ch_cnt++;
824
	hw_feat->tc_cnt++;
825

826 827 828 829
	/* Translate the fifo sizes into actual numbers */
	hw_feat->rx_fifo_size = 1 << (hw_feat->rx_fifo_size + 7);
	hw_feat->tx_fifo_size = 1 << (hw_feat->tx_fifo_size + 7);

830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863
	if (netif_msg_probe(pdata)) {
		dev_dbg(pdata->dev, "Hardware features:\n");

		/* Hardware feature register 0 */
		dev_dbg(pdata->dev, "  1GbE support              : %s\n",
			hw_feat->gmii ? "yes" : "no");
		dev_dbg(pdata->dev, "  VLAN hash filter          : %s\n",
			hw_feat->vlhash ? "yes" : "no");
		dev_dbg(pdata->dev, "  MDIO interface            : %s\n",
			hw_feat->sma ? "yes" : "no");
		dev_dbg(pdata->dev, "  Wake-up packet support    : %s\n",
			hw_feat->rwk ? "yes" : "no");
		dev_dbg(pdata->dev, "  Magic packet support      : %s\n",
			hw_feat->mgk ? "yes" : "no");
		dev_dbg(pdata->dev, "  Management counters       : %s\n",
			hw_feat->mmc ? "yes" : "no");
		dev_dbg(pdata->dev, "  ARP offload               : %s\n",
			hw_feat->aoe ? "yes" : "no");
		dev_dbg(pdata->dev, "  IEEE 1588-2008 Timestamp  : %s\n",
			hw_feat->ts ? "yes" : "no");
		dev_dbg(pdata->dev, "  Energy Efficient Ethernet : %s\n",
			hw_feat->eee ? "yes" : "no");
		dev_dbg(pdata->dev, "  TX checksum offload       : %s\n",
			hw_feat->tx_coe ? "yes" : "no");
		dev_dbg(pdata->dev, "  RX checksum offload       : %s\n",
			hw_feat->rx_coe ? "yes" : "no");
		dev_dbg(pdata->dev, "  Additional MAC addresses  : %u\n",
			hw_feat->addn_mac);
		dev_dbg(pdata->dev, "  Timestamp source          : %s\n",
			(hw_feat->ts_src == 1) ? "internal" :
			(hw_feat->ts_src == 2) ? "external" :
			(hw_feat->ts_src == 3) ? "internal/external" : "n/a");
		dev_dbg(pdata->dev, "  SA/VLAN insertion         : %s\n",
			hw_feat->sa_vlan_ins ? "yes" : "no");
864 865
		dev_dbg(pdata->dev, "  VXLAN/NVGRE support       : %s\n",
			hw_feat->vxn ? "yes" : "no");
866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906

		/* Hardware feature register 1 */
		dev_dbg(pdata->dev, "  RX fifo size              : %u\n",
			hw_feat->rx_fifo_size);
		dev_dbg(pdata->dev, "  TX fifo size              : %u\n",
			hw_feat->tx_fifo_size);
		dev_dbg(pdata->dev, "  IEEE 1588 high word       : %s\n",
			hw_feat->adv_ts_hi ? "yes" : "no");
		dev_dbg(pdata->dev, "  DMA width                 : %u\n",
			hw_feat->dma_width);
		dev_dbg(pdata->dev, "  Data Center Bridging      : %s\n",
			hw_feat->dcb ? "yes" : "no");
		dev_dbg(pdata->dev, "  Split header              : %s\n",
			hw_feat->sph ? "yes" : "no");
		dev_dbg(pdata->dev, "  TCP Segmentation Offload  : %s\n",
			hw_feat->tso ? "yes" : "no");
		dev_dbg(pdata->dev, "  Debug memory interface    : %s\n",
			hw_feat->dma_debug ? "yes" : "no");
		dev_dbg(pdata->dev, "  Receive Side Scaling      : %s\n",
			hw_feat->rss ? "yes" : "no");
		dev_dbg(pdata->dev, "  Traffic Class count       : %u\n",
			hw_feat->tc_cnt);
		dev_dbg(pdata->dev, "  Hash table size           : %u\n",
			hw_feat->hash_table_size);
		dev_dbg(pdata->dev, "  L3/L4 Filters             : %u\n",
			hw_feat->l3l4_filter_num);

		/* Hardware feature register 2 */
		dev_dbg(pdata->dev, "  RX queue count            : %u\n",
			hw_feat->rx_q_cnt);
		dev_dbg(pdata->dev, "  TX queue count            : %u\n",
			hw_feat->tx_q_cnt);
		dev_dbg(pdata->dev, "  RX DMA channel count      : %u\n",
			hw_feat->rx_ch_cnt);
		dev_dbg(pdata->dev, "  TX DMA channel count      : %u\n",
			hw_feat->rx_ch_cnt);
		dev_dbg(pdata->dev, "  PPS outputs               : %u\n",
			hw_feat->pps_out_num);
		dev_dbg(pdata->dev, "  Auxiliary snapshot inputs : %u\n",
			hw_feat->aux_snap_num);
	}
907 908
}

909 910
static int xgbe_vxlan_set_port(struct net_device *netdev, unsigned int table,
			       unsigned int entry, struct udp_tunnel_info *ti)
911
{
912
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
913

914 915
	pdata->vxlan_port = be16_to_cpu(ti->port);
	pdata->hw_if.enable_vxlan(pdata);
916

917
	return 0;
918 919
}

920 921
static int xgbe_vxlan_unset_port(struct net_device *netdev, unsigned int table,
				 unsigned int entry, struct udp_tunnel_info *ti)
922
{
923
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
924 925 926 927

	pdata->hw_if.disable_vxlan(pdata);
	pdata->vxlan_port = 0;

928
	return 0;
929 930
}

931 932 933 934 935 936 937 938
static const struct udp_tunnel_nic_info xgbe_udp_tunnels = {
	.set_port	= xgbe_vxlan_set_port,
	.unset_port	= xgbe_vxlan_unset_port,
	.flags		= UDP_TUNNEL_NIC_INFO_OPEN_ONLY,
	.tables		= {
		{ .n_entries = 1, .tunnel_types = UDP_TUNNEL_TYPE_VXLAN, },
	},
};
939

940
const struct udp_tunnel_nic_info *xgbe_get_udp_tunnel_info(void)
941
{
942
	return &xgbe_udp_tunnels;
943 944
}

945 946
static void xgbe_napi_enable(struct xgbe_prv_data *pdata, unsigned int add)
{
947 948 949 950
	struct xgbe_channel *channel;
	unsigned int i;

	if (pdata->per_channel_irq) {
951 952
		for (i = 0; i < pdata->channel_count; i++) {
			channel = pdata->channel[i];
953 954 955 956 957 958 959 960 961 962 963 964 965
			if (add)
				netif_napi_add(pdata->netdev, &channel->napi,
					       xgbe_one_poll, NAPI_POLL_WEIGHT);

			napi_enable(&channel->napi);
		}
	} else {
		if (add)
			netif_napi_add(pdata->netdev, &pdata->napi,
				       xgbe_all_poll, NAPI_POLL_WEIGHT);

		napi_enable(&pdata->napi);
	}
966 967
}

968
static void xgbe_napi_disable(struct xgbe_prv_data *pdata, unsigned int del)
969
{
970 971 972 973
	struct xgbe_channel *channel;
	unsigned int i;

	if (pdata->per_channel_irq) {
974 975
		for (i = 0; i < pdata->channel_count; i++) {
			channel = pdata->channel[i];
976
			napi_disable(&channel->napi);
977

978 979 980 981 982 983 984 985 986
			if (del)
				netif_napi_del(&channel->napi);
		}
	} else {
		napi_disable(&pdata->napi);

		if (del)
			netif_napi_del(&pdata->napi);
	}
987 988
}

989 990 991 992 993 994 995
static int xgbe_request_irqs(struct xgbe_prv_data *pdata)
{
	struct xgbe_channel *channel;
	struct net_device *netdev = pdata->netdev;
	unsigned int i;
	int ret;

996 997
	tasklet_setup(&pdata->tasklet_dev, xgbe_isr_task);
	tasklet_setup(&pdata->tasklet_ecc, xgbe_ecc_isr_task);
998

999
	ret = devm_request_irq(pdata->dev, pdata->dev_irq, xgbe_isr, 0,
1000
			       netdev_name(netdev), pdata);
1001 1002 1003 1004 1005 1006
	if (ret) {
		netdev_alert(netdev, "error requesting irq %d\n",
			     pdata->dev_irq);
		return ret;
	}

1007 1008 1009 1010 1011 1012 1013 1014 1015 1016
	if (pdata->vdata->ecc_support && (pdata->dev_irq != pdata->ecc_irq)) {
		ret = devm_request_irq(pdata->dev, pdata->ecc_irq, xgbe_ecc_isr,
				       0, pdata->ecc_name, pdata);
		if (ret) {
			netdev_alert(netdev, "error requesting ecc irq %d\n",
				     pdata->ecc_irq);
			goto err_dev_irq;
		}
	}

1017 1018 1019
	if (!pdata->per_channel_irq)
		return 0;

1020 1021
	for (i = 0; i < pdata->channel_count; i++) {
		channel = pdata->channel[i];
1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
		snprintf(channel->dma_irq_name,
			 sizeof(channel->dma_irq_name) - 1,
			 "%s-TxRx-%u", netdev_name(netdev),
			 channel->queue_index);

		ret = devm_request_irq(pdata->dev, channel->dma_irq,
				       xgbe_dma_isr, 0,
				       channel->dma_irq_name, channel);
		if (ret) {
			netdev_alert(netdev, "error requesting irq %d\n",
				     channel->dma_irq);
1033
			goto err_dma_irq;
1034
		}
1035 1036 1037

		irq_set_affinity_hint(channel->dma_irq,
				      &channel->affinity_mask);
1038 1039 1040 1041
	}

	return 0;

1042
err_dma_irq:
1043
	/* Using an unsigned int, 'i' will go to UINT_MAX and exit */
1044 1045 1046
	for (i--; i < pdata->channel_count; i--) {
		channel = pdata->channel[i];

1047
		irq_set_affinity_hint(channel->dma_irq, NULL);
1048
		devm_free_irq(pdata->dev, channel->dma_irq, channel);
1049
	}
1050

1051 1052 1053 1054
	if (pdata->vdata->ecc_support && (pdata->dev_irq != pdata->ecc_irq))
		devm_free_irq(pdata->dev, pdata->ecc_irq, pdata);

err_dev_irq:
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066
	devm_free_irq(pdata->dev, pdata->dev_irq, pdata);

	return ret;
}

static void xgbe_free_irqs(struct xgbe_prv_data *pdata)
{
	struct xgbe_channel *channel;
	unsigned int i;

	devm_free_irq(pdata->dev, pdata->dev_irq, pdata);

1067 1068 1069
	if (pdata->vdata->ecc_support && (pdata->dev_irq != pdata->ecc_irq))
		devm_free_irq(pdata->dev, pdata->ecc_irq, pdata);

1070 1071 1072
	if (!pdata->per_channel_irq)
		return;

1073 1074
	for (i = 0; i < pdata->channel_count; i++) {
		channel = pdata->channel[i];
1075 1076

		irq_set_affinity_hint(channel->dma_irq, NULL);
1077
		devm_free_irq(pdata->dev, channel->dma_irq, channel);
1078
	}
1079 1080
}

1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
void xgbe_init_tx_coalesce(struct xgbe_prv_data *pdata)
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;

	DBGPR("-->xgbe_init_tx_coalesce\n");

	pdata->tx_usecs = XGMAC_INIT_DMA_TX_USECS;
	pdata->tx_frames = XGMAC_INIT_DMA_TX_FRAMES;

	hw_if->config_tx_coalesce(pdata);

	DBGPR("<--xgbe_init_tx_coalesce\n");
}

void xgbe_init_rx_coalesce(struct xgbe_prv_data *pdata)
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;

	DBGPR("-->xgbe_init_rx_coalesce\n");

	pdata->rx_riwt = hw_if->usec_to_riwt(pdata, XGMAC_INIT_DMA_RX_USECS);
1102
	pdata->rx_usecs = XGMAC_INIT_DMA_RX_USECS;
1103 1104 1105 1106 1107 1108 1109
	pdata->rx_frames = XGMAC_INIT_DMA_RX_FRAMES;

	hw_if->config_rx_coalesce(pdata);

	DBGPR("<--xgbe_init_rx_coalesce\n");
}

1110
static void xgbe_free_tx_data(struct xgbe_prv_data *pdata)
1111 1112 1113 1114 1115 1116
{
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct xgbe_ring *ring;
	struct xgbe_ring_data *rdata;
	unsigned int i, j;

1117
	DBGPR("-->xgbe_free_tx_data\n");
1118

1119 1120
	for (i = 0; i < pdata->channel_count; i++) {
		ring = pdata->channel[i]->tx_ring;
1121 1122 1123 1124
		if (!ring)
			break;

		for (j = 0; j < ring->rdesc_count; j++) {
1125
			rdata = XGBE_GET_DESC_DATA(ring, j);
1126
			desc_if->unmap_rdata(pdata, rdata);
1127 1128 1129
		}
	}

1130
	DBGPR("<--xgbe_free_tx_data\n");
1131 1132
}

1133
static void xgbe_free_rx_data(struct xgbe_prv_data *pdata)
1134 1135 1136 1137 1138 1139
{
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct xgbe_ring *ring;
	struct xgbe_ring_data *rdata;
	unsigned int i, j;

1140
	DBGPR("-->xgbe_free_rx_data\n");
1141

1142 1143
	for (i = 0; i < pdata->channel_count; i++) {
		ring = pdata->channel[i]->rx_ring;
1144 1145 1146 1147
		if (!ring)
			break;

		for (j = 0; j < ring->rdesc_count; j++) {
1148
			rdata = XGBE_GET_DESC_DATA(ring, j);
1149
			desc_if->unmap_rdata(pdata, rdata);
1150 1151 1152
		}
	}

1153
	DBGPR("<--xgbe_free_rx_data\n");
1154 1155
}

1156
static int xgbe_phy_reset(struct xgbe_prv_data *pdata)
1157 1158 1159 1160
{
	pdata->phy_link = -1;
	pdata->phy_speed = SPEED_UNKNOWN;

1161
	return pdata->phy_if.phy_reset(pdata);
1162 1163
}

1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
int xgbe_powerdown(struct net_device *netdev, unsigned int caller)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	unsigned long flags;

	DBGPR("-->xgbe_powerdown\n");

	if (!netif_running(netdev) ||
	    (caller == XGMAC_IOCTL_CONTEXT && pdata->power_down)) {
		netdev_alert(netdev, "Device is already powered down\n");
		DBGPR("<--xgbe_powerdown\n");
		return -EINVAL;
	}

	spin_lock_irqsave(&pdata->lock, flags);

	if (caller == XGMAC_DRIVER_CONTEXT)
		netif_device_detach(netdev);

	netif_tx_stop_all_queues(netdev);

1186 1187 1188
	xgbe_stop_timers(pdata);
	flush_workqueue(pdata->dev_workqueue);

1189 1190 1191
	hw_if->powerdown_tx(pdata);
	hw_if->powerdown_rx(pdata);

1192 1193
	xgbe_napi_disable(pdata, 0);

1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221
	pdata->power_down = 1;

	spin_unlock_irqrestore(&pdata->lock, flags);

	DBGPR("<--xgbe_powerdown\n");

	return 0;
}

int xgbe_powerup(struct net_device *netdev, unsigned int caller)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	unsigned long flags;

	DBGPR("-->xgbe_powerup\n");

	if (!netif_running(netdev) ||
	    (caller == XGMAC_IOCTL_CONTEXT && !pdata->power_down)) {
		netdev_alert(netdev, "Device is already powered up\n");
		DBGPR("<--xgbe_powerup\n");
		return -EINVAL;
	}

	spin_lock_irqsave(&pdata->lock, flags);

	pdata->power_down = 0;

1222 1223
	xgbe_napi_enable(pdata, 0);

1224 1225 1226 1227 1228 1229 1230 1231
	hw_if->powerup_tx(pdata);
	hw_if->powerup_rx(pdata);

	if (caller == XGMAC_DRIVER_CONTEXT)
		netif_device_attach(netdev);

	netif_tx_start_all_queues(netdev);

1232 1233
	xgbe_start_timers(pdata);

1234 1235 1236 1237 1238 1239 1240
	spin_unlock_irqrestore(&pdata->lock, flags);

	DBGPR("<--xgbe_powerup\n");

	return 0;
}

1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257
static void xgbe_free_memory(struct xgbe_prv_data *pdata)
{
	struct xgbe_desc_if *desc_if = &pdata->desc_if;

	/* Free the ring descriptors and buffers */
	desc_if->free_ring_resources(pdata);

	/* Free the channel and ring structures */
	xgbe_free_channels(pdata);
}

static int xgbe_alloc_memory(struct xgbe_prv_data *pdata)
{
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct net_device *netdev = pdata->netdev;
	int ret;

1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268
	if (pdata->new_tx_ring_count) {
		pdata->tx_ring_count = pdata->new_tx_ring_count;
		pdata->tx_q_count = pdata->tx_ring_count;
		pdata->new_tx_ring_count = 0;
	}

	if (pdata->new_rx_ring_count) {
		pdata->rx_ring_count = pdata->new_rx_ring_count;
		pdata->new_rx_ring_count = 0;
	}

1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
	/* Calculate the Rx buffer size before allocating rings */
	pdata->rx_buf_size = xgbe_calc_rx_buf_size(netdev, netdev->mtu);

	/* Allocate the channel and ring structures */
	ret = xgbe_alloc_channels(pdata);
	if (ret)
		return ret;

	/* Allocate the ring descriptors and buffers */
	ret = desc_if->alloc_ring_resources(pdata);
	if (ret)
		goto err_channels;

	/* Initialize the service and Tx timers */
	xgbe_init_timers(pdata);

	return 0;

err_channels:
	xgbe_free_memory(pdata);

	return ret;
}

1293 1294 1295
static int xgbe_start(struct xgbe_prv_data *pdata)
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
1296
	struct xgbe_phy_if *phy_if = &pdata->phy_if;
1297
	struct net_device *netdev = pdata->netdev;
1298
	unsigned int i;
1299
	int ret;
1300

1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317
	/* Set the number of queues */
	ret = netif_set_real_num_tx_queues(netdev, pdata->tx_ring_count);
	if (ret) {
		netdev_err(netdev, "error setting real tx queue count\n");
		return ret;
	}

	ret = netif_set_real_num_rx_queues(netdev, pdata->rx_ring_count);
	if (ret) {
		netdev_err(netdev, "error setting real rx queue count\n");
		return ret;
	}

	/* Set RSS lookup table data for programming */
	for (i = 0; i < XGBE_RSS_MAX_TABLE_SIZE; i++)
		XGMAC_SET_BITS(pdata->rss_table[i], MAC_RSSDR, DMCH,
			       i % pdata->rx_ring_count);
1318

1319 1320 1321
	ret = hw_if->init(pdata);
	if (ret)
		return ret;
1322

1323 1324 1325 1326 1327 1328
	xgbe_napi_enable(pdata, 1);

	ret = xgbe_request_irqs(pdata);
	if (ret)
		goto err_napi;

1329 1330 1331 1332
	ret = phy_if->phy_start(pdata);
	if (ret)
		goto err_irqs;

1333 1334 1335
	hw_if->enable_tx(pdata);
	hw_if->enable_rx(pdata);

1336
	udp_tunnel_nic_reset_ntf(netdev);
1337

1338 1339
	netif_tx_start_all_queues(netdev);

1340
	xgbe_start_timers(pdata);
1341
	queue_work(pdata->dev_workqueue, &pdata->service_work);
1342

1343 1344
	clear_bit(XGBE_STOPPED, &pdata->dev_state);

1345
	return 0;
1346

1347 1348 1349
err_irqs:
	xgbe_free_irqs(pdata);

1350 1351 1352 1353 1354 1355
err_napi:
	xgbe_napi_disable(pdata, 1);

	hw_if->exit(pdata);

	return ret;
1356 1357 1358 1359 1360
}

static void xgbe_stop(struct xgbe_prv_data *pdata)
{
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
1361
	struct xgbe_phy_if *phy_if = &pdata->phy_if;
L
Lendacky, Thomas 已提交
1362
	struct xgbe_channel *channel;
1363
	struct net_device *netdev = pdata->netdev;
L
Lendacky, Thomas 已提交
1364 1365
	struct netdev_queue *txq;
	unsigned int i;
1366 1367 1368

	DBGPR("-->xgbe_stop\n");

1369 1370 1371
	if (test_bit(XGBE_STOPPED, &pdata->dev_state))
		return;

1372
	netif_tx_stop_all_queues(netdev);
1373
	netif_carrier_off(pdata->netdev);
1374

1375 1376
	xgbe_stop_timers(pdata);
	flush_workqueue(pdata->dev_workqueue);
1377

1378
	xgbe_vxlan_unset_port(netdev, 0, 0, NULL);
1379

1380 1381 1382
	hw_if->disable_tx(pdata);
	hw_if->disable_rx(pdata);

1383 1384
	phy_if->phy_stop(pdata);

1385 1386 1387 1388 1389 1390
	xgbe_free_irqs(pdata);

	xgbe_napi_disable(pdata, 1);

	hw_if->exit(pdata);

1391 1392
	for (i = 0; i < pdata->channel_count; i++) {
		channel = pdata->channel[i];
L
Lendacky, Thomas 已提交
1393 1394 1395 1396 1397 1398 1399
		if (!channel->tx_ring)
			continue;

		txq = netdev_get_tx_queue(netdev, channel->queue_index);
		netdev_tx_reset_queue(txq);
	}

1400 1401
	set_bit(XGBE_STOPPED, &pdata->dev_state);

1402 1403 1404
	DBGPR("<--xgbe_stop\n");
}

1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422
static void xgbe_stopdev(struct work_struct *work)
{
	struct xgbe_prv_data *pdata = container_of(work,
						   struct xgbe_prv_data,
						   stopdev_work);

	rtnl_lock();

	xgbe_stop(pdata);

	xgbe_free_tx_data(pdata);
	xgbe_free_rx_data(pdata);

	rtnl_unlock();

	netdev_alert(pdata->netdev, "device stopped\n");
}

1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436
void xgbe_full_restart_dev(struct xgbe_prv_data *pdata)
{
	/* If not running, "restart" will happen on open */
	if (!netif_running(pdata->netdev))
		return;

	xgbe_stop(pdata);

	xgbe_free_memory(pdata);
	xgbe_alloc_memory(pdata);

	xgbe_start(pdata);
}

1437
void xgbe_restart_dev(struct xgbe_prv_data *pdata)
1438 1439 1440 1441 1442 1443 1444
{
	/* If not running, "restart" will happen on open */
	if (!netif_running(pdata->netdev))
		return;

	xgbe_stop(pdata);

1445 1446
	xgbe_free_tx_data(pdata);
	xgbe_free_rx_data(pdata);
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458

	xgbe_start(pdata);
}

static void xgbe_restart(struct work_struct *work)
{
	struct xgbe_prv_data *pdata = container_of(work,
						   struct xgbe_prv_data,
						   restart_work);

	rtnl_lock();

1459
	xgbe_restart_dev(pdata);
1460 1461 1462 1463

	rtnl_unlock();
}

1464 1465 1466 1467 1468 1469 1470 1471 1472
static void xgbe_tx_tstamp(struct work_struct *work)
{
	struct xgbe_prv_data *pdata = container_of(work,
						   struct xgbe_prv_data,
						   tx_tstamp_work);
	struct skb_shared_hwtstamps hwtstamps;
	u64 nsec;
	unsigned long flags;

1473 1474 1475 1476
	spin_lock_irqsave(&pdata->tstamp_lock, flags);
	if (!pdata->tx_tstamp_skb)
		goto unlock;

1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488
	if (pdata->tx_tstamp) {
		nsec = timecounter_cyc2time(&pdata->tstamp_tc,
					    pdata->tx_tstamp);

		memset(&hwtstamps, 0, sizeof(hwtstamps));
		hwtstamps.hwtstamp = ns_to_ktime(nsec);
		skb_tstamp_tx(pdata->tx_tstamp_skb, &hwtstamps);
	}

	dev_kfree_skb_any(pdata->tx_tstamp_skb);

	pdata->tx_tstamp_skb = NULL;
1489 1490

unlock:
1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530
	spin_unlock_irqrestore(&pdata->tstamp_lock, flags);
}

static int xgbe_get_hwtstamp_settings(struct xgbe_prv_data *pdata,
				      struct ifreq *ifreq)
{
	if (copy_to_user(ifreq->ifr_data, &pdata->tstamp_config,
			 sizeof(pdata->tstamp_config)))
		return -EFAULT;

	return 0;
}

static int xgbe_set_hwtstamp_settings(struct xgbe_prv_data *pdata,
				      struct ifreq *ifreq)
{
	struct hwtstamp_config config;
	unsigned int mac_tscr;

	if (copy_from_user(&config, ifreq->ifr_data, sizeof(config)))
		return -EFAULT;

	mac_tscr = 0;

	switch (config.tx_type) {
	case HWTSTAMP_TX_OFF:
		break;

	case HWTSTAMP_TX_ON:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	default:
		return -ERANGE;
	}

	switch (config.rx_filter) {
	case HWTSTAMP_FILTER_NONE:
		break;

1531
	case HWTSTAMP_FILTER_NTP_ALL:
1532 1533 1534 1535 1536 1537 1538 1539
	case HWTSTAMP_FILTER_ALL:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENALL, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2, UDP, any kind of event packet */
	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
1540
		fallthrough;	/* to PTP v1, UDP, any kind of event packet */
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, SNAPTYPSEL, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2, UDP, Sync packet */
	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
1551
		fallthrough;	/* to PTP v1, UDP, Sync packet */
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561
	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2, UDP, Delay_req packet */
	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
1562
		fallthrough;	/* to PTP v1, UDP, Delay_req packet */
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653
	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSMSTRENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* 802.AS1, Ethernet, any kind of event packet */
	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, AV8021ASMEN, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, SNAPTYPSEL, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* 802.AS1, Ethernet, Sync packet */
	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, AV8021ASMEN, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* 802.AS1, Ethernet, Delay_req packet */
	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, AV8021ASMEN, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSMSTRENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2/802.AS1, any layer, any kind of event packet */
	case HWTSTAMP_FILTER_PTP_V2_EVENT:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, SNAPTYPSEL, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2/802.AS1, any layer, Sync packet */
	case HWTSTAMP_FILTER_PTP_V2_SYNC:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	/* PTP v2/802.AS1, any layer, Delay_req packet */
	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSVER2ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV4ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSIPV6ENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSMSTRENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSEVNTENA, 1);
		XGMAC_SET_BITS(mac_tscr, MAC_TSCR, TSENA, 1);
		break;

	default:
		return -ERANGE;
	}

	pdata->hw_if.config_tstamp(pdata, mac_tscr);

	memcpy(&pdata->tstamp_config, &config, sizeof(config));

	return 0;
}

static void xgbe_prep_tx_tstamp(struct xgbe_prv_data *pdata,
				struct sk_buff *skb,
				struct xgbe_packet_data *packet)
{
	unsigned long flags;

	if (XGMAC_GET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES, PTP)) {
		spin_lock_irqsave(&pdata->tstamp_lock, flags);
		if (pdata->tx_tstamp_skb) {
			/* Another timestamp in progress, ignore this one */
			XGMAC_SET_BITS(packet->attributes,
				       TX_PACKET_ATTRIBUTES, PTP, 0);
		} else {
			pdata->tx_tstamp_skb = skb_get(skb);
			skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
		}
		spin_unlock_irqrestore(&pdata->tstamp_lock, flags);
	}

1654
	skb_tx_timestamp(skb);
1655 1656
}

1657 1658
static void xgbe_prep_vlan(struct sk_buff *skb, struct xgbe_packet_data *packet)
{
1659 1660
	if (skb_vlan_tag_present(skb))
		packet->vlan_ctag = skb_vlan_tag_get(skb);
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674
}

static int xgbe_prep_tso(struct sk_buff *skb, struct xgbe_packet_data *packet)
{
	int ret;

	if (!XGMAC_GET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			    TSO_ENABLE))
		return 0;

	ret = skb_cow_head(skb, 0);
	if (ret)
		return ret;

1675 1676 1677 1678 1679 1680 1681 1682 1683
	if (XGMAC_GET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES, VXLAN)) {
		packet->header_len = skb_inner_transport_offset(skb) +
				     inner_tcp_hdrlen(skb);
		packet->tcp_header_len = inner_tcp_hdrlen(skb);
	} else {
		packet->header_len = skb_transport_offset(skb) +
				     tcp_hdrlen(skb);
		packet->tcp_header_len = tcp_hdrlen(skb);
	}
1684 1685
	packet->tcp_payload_len = skb->len - packet->header_len;
	packet->mss = skb_shinfo(skb)->gso_size;
1686

1687 1688 1689 1690 1691
	DBGPR("  packet->header_len=%u\n", packet->header_len);
	DBGPR("  packet->tcp_header_len=%u, packet->tcp_payload_len=%u\n",
	      packet->tcp_header_len, packet->tcp_payload_len);
	DBGPR("  packet->mss=%u\n", packet->mss);

L
Lendacky, Thomas 已提交
1692 1693 1694 1695 1696 1697
	/* Update the number of packets that will ultimately be transmitted
	 * along with the extra bytes for each extra packet
	 */
	packet->tx_packets = skb_shinfo(skb)->gso_segs;
	packet->tx_bytes += (packet->tx_packets - 1) * packet->header_len;

1698 1699 1700
	return 0;
}

1701
static bool xgbe_is_vxlan(struct sk_buff *skb)
1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
{
	if (!skb->encapsulation)
		return false;

	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return false;

	switch (skb->protocol) {
	case htons(ETH_P_IP):
		if (ip_hdr(skb)->protocol != IPPROTO_UDP)
			return false;
		break;

	case htons(ETH_P_IPV6):
		if (ipv6_hdr(skb)->nexthdr != IPPROTO_UDP)
			return false;
		break;

	default:
		return false;
	}

1724 1725 1726 1727 1728
	if (skb->inner_protocol_type != ENCAP_TYPE_ETHER ||
	    skb->inner_protocol != htons(ETH_P_TEB) ||
	    (skb_inner_mac_header(skb) - skb_transport_header(skb) !=
	     sizeof(struct udphdr) + sizeof(struct vxlanhdr)))
		return false;
1729

1730
	return true;
1731 1732
}

1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745
static int xgbe_is_tso(struct sk_buff *skb)
{
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return 0;

	if (!skb_is_gso(skb))
		return 0;

	DBGPR("  TSO packet to be processed\n");

	return 1;
}

1746 1747
static void xgbe_packet_info(struct xgbe_prv_data *pdata,
			     struct xgbe_ring *ring, struct sk_buff *skb,
1748 1749
			     struct xgbe_packet_data *packet)
{
1750
	skb_frag_t *frag;
1751 1752 1753 1754
	unsigned int context_desc;
	unsigned int len;
	unsigned int i;

1755 1756
	packet->skb = skb;

1757 1758 1759
	context_desc = 0;
	packet->rdesc_count = 0;

L
Lendacky, Thomas 已提交
1760 1761 1762
	packet->tx_packets = 1;
	packet->tx_bytes = skb->len;

1763
	if (xgbe_is_tso(skb)) {
L
Lendacky, Thomas 已提交
1764
		/* TSO requires an extra descriptor if mss is different */
1765 1766 1767 1768 1769
		if (skb_shinfo(skb)->gso_size != ring->tx.cur_mss) {
			context_desc = 1;
			packet->rdesc_count++;
		}

L
Lendacky, Thomas 已提交
1770
		/* TSO requires an extra descriptor for TSO header */
1771 1772 1773 1774 1775 1776 1777 1778 1779 1780
		packet->rdesc_count++;

		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       TSO_ENABLE, 1);
		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       CSUM_ENABLE, 1);
	} else if (skb->ip_summed == CHECKSUM_PARTIAL)
		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       CSUM_ENABLE, 1);

1781
	if (xgbe_is_vxlan(skb))
1782 1783 1784
		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       VXLAN, 1);

1785
	if (skb_vlan_tag_present(skb)) {
1786
		/* VLAN requires an extra descriptor if tag is different */
1787
		if (skb_vlan_tag_get(skb) != ring->tx.cur_vlan_ctag)
1788 1789 1790 1791 1792 1793 1794 1795 1796 1797
			/* We can share with the TSO context descriptor */
			if (!context_desc) {
				context_desc = 1;
				packet->rdesc_count++;
			}

		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       VLAN_CTAG, 1);
	}

1798 1799 1800 1801 1802
	if ((skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) &&
	    (pdata->tstamp_config.tx_type == HWTSTAMP_TX_ON))
		XGMAC_SET_BITS(packet->attributes, TX_PACKET_ATTRIBUTES,
			       PTP, 1);

1803 1804
	for (len = skb_headlen(skb); len;) {
		packet->rdesc_count++;
1805
		len -= min_t(unsigned int, len, XGBE_TX_MAX_BUF_SIZE);
1806 1807 1808 1809 1810 1811
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
		frag = &skb_shinfo(skb)->frags[i];
		for (len = skb_frag_size(frag); len; ) {
			packet->rdesc_count++;
1812
			len -= min_t(unsigned int, len, XGBE_TX_MAX_BUF_SIZE);
1813 1814 1815 1816 1817 1818 1819 1820 1821
		}
	}
}

static int xgbe_open(struct net_device *netdev)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	int ret;

1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847
	/* Create the various names based on netdev name */
	snprintf(pdata->an_name, sizeof(pdata->an_name) - 1, "%s-pcs",
		 netdev_name(netdev));

	snprintf(pdata->ecc_name, sizeof(pdata->ecc_name) - 1, "%s-ecc",
		 netdev_name(netdev));

	snprintf(pdata->i2c_name, sizeof(pdata->i2c_name) - 1, "%s-i2c",
		 netdev_name(netdev));

	/* Create workqueues */
	pdata->dev_workqueue =
		create_singlethread_workqueue(netdev_name(netdev));
	if (!pdata->dev_workqueue) {
		netdev_err(netdev, "device workqueue creation failed\n");
		return -ENOMEM;
	}

	pdata->an_workqueue =
		create_singlethread_workqueue(pdata->an_name);
	if (!pdata->an_workqueue) {
		netdev_err(netdev, "phy workqueue creation failed\n");
		ret = -ENOMEM;
		goto err_dev_wq;
	}

1848 1849
	/* Reset the phy settings */
	ret = xgbe_phy_reset(pdata);
1850
	if (ret)
1851
		goto err_an_wq;
1852

1853 1854
	/* Enable the clocks */
	ret = clk_prepare_enable(pdata->sysclk);
1855
	if (ret) {
1856
		netdev_alert(netdev, "dma clk_prepare_enable failed\n");
1857
		goto err_an_wq;
1858 1859
	}

1860 1861 1862 1863 1864 1865
	ret = clk_prepare_enable(pdata->ptpclk);
	if (ret) {
		netdev_alert(netdev, "ptp clk_prepare_enable failed\n");
		goto err_sysclk;
	}

1866
	INIT_WORK(&pdata->service_work, xgbe_service);
1867
	INIT_WORK(&pdata->restart_work, xgbe_restart);
1868
	INIT_WORK(&pdata->stopdev_work, xgbe_stopdev);
1869
	INIT_WORK(&pdata->tx_tstamp_work, xgbe_tx_tstamp);
1870 1871 1872 1873

	ret = xgbe_alloc_memory(pdata);
	if (ret)
		goto err_ptpclk;
1874 1875 1876

	ret = xgbe_start(pdata);
	if (ret)
1877
		goto err_mem;
1878

1879 1880
	clear_bit(XGBE_DOWN, &pdata->dev_state);

1881 1882
	return 0;

1883 1884
err_mem:
	xgbe_free_memory(pdata);
1885

1886 1887 1888 1889 1890
err_ptpclk:
	clk_disable_unprepare(pdata->ptpclk);

err_sysclk:
	clk_disable_unprepare(pdata->sysclk);
1891

1892 1893 1894 1895 1896 1897
err_an_wq:
	destroy_workqueue(pdata->an_workqueue);

err_dev_wq:
	destroy_workqueue(pdata->dev_workqueue);

1898 1899 1900 1901 1902 1903 1904 1905 1906 1907
	return ret;
}

static int xgbe_close(struct net_device *netdev)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);

	/* Stop the device */
	xgbe_stop(pdata);

1908
	xgbe_free_memory(pdata);
1909

1910 1911 1912
	/* Disable the clocks */
	clk_disable_unprepare(pdata->ptpclk);
	clk_disable_unprepare(pdata->sysclk);
1913

1914 1915 1916 1917
	destroy_workqueue(pdata->an_workqueue);

	destroy_workqueue(pdata->dev_workqueue);

1918
	set_bit(XGBE_DOWN, &pdata->dev_state);
1919

1920 1921 1922
	return 0;
}

1923
static netdev_tx_t xgbe_xmit(struct sk_buff *skb, struct net_device *netdev)
1924 1925 1926 1927 1928 1929 1930
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct xgbe_channel *channel;
	struct xgbe_ring *ring;
	struct xgbe_packet_data *packet;
L
Lendacky, Thomas 已提交
1931
	struct netdev_queue *txq;
1932
	netdev_tx_t ret;
1933 1934 1935

	DBGPR("-->xgbe_xmit: skb->len = %d\n", skb->len);

1936
	channel = pdata->channel[skb->queue_mapping];
L
Lendacky, Thomas 已提交
1937
	txq = netdev_get_tx_queue(netdev, channel->queue_index);
1938 1939 1940 1941 1942 1943
	ring = channel->tx_ring;
	packet = &ring->packet_data;

	ret = NETDEV_TX_OK;

	if (skb->len == 0) {
1944 1945
		netif_err(pdata, tx_err, netdev,
			  "empty skb received from stack\n");
1946 1947 1948 1949 1950 1951
		dev_kfree_skb_any(skb);
		goto tx_netdev_return;
	}

	/* Calculate preliminary packet info */
	memset(packet, 0, sizeof(*packet));
1952
	xgbe_packet_info(pdata, ring, skb, packet);
1953 1954

	/* Check that there are enough descriptors available */
1955 1956
	ret = xgbe_maybe_stop_tx_queue(channel, ring, packet->rdesc_count);
	if (ret)
1957 1958 1959 1960
		goto tx_netdev_return;

	ret = xgbe_prep_tso(skb, packet);
	if (ret) {
1961 1962
		netif_err(pdata, tx_err, netdev,
			  "error processing TSO packet\n");
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972
		dev_kfree_skb_any(skb);
		goto tx_netdev_return;
	}
	xgbe_prep_vlan(skb, packet);

	if (!desc_if->map_tx_skb(channel, skb)) {
		dev_kfree_skb_any(skb);
		goto tx_netdev_return;
	}

1973 1974
	xgbe_prep_tx_tstamp(pdata, skb, packet);

L
Lendacky, Thomas 已提交
1975 1976 1977
	/* Report on the actual number of bytes (to be) sent */
	netdev_tx_sent_queue(txq, packet->tx_bytes);

1978
	/* Configure required descriptor fields for transmission */
1979
	hw_if->dev_xmit(channel);
1980

1981 1982
	if (netif_msg_pktdata(pdata))
		xgbe_print_pkt(netdev, skb, true);
1983

1984 1985 1986 1987 1988
	/* Stop the queue in advance if there may not be enough descriptors */
	xgbe_maybe_stop_tx_queue(channel, ring, XGBE_TX_MAX_DESCS);

	ret = NETDEV_TX_OK;

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
tx_netdev_return:
	return ret;
}

static void xgbe_set_rx_mode(struct net_device *netdev)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;

	DBGPR("-->xgbe_set_rx_mode\n");

2000
	hw_if->config_rx_mode(pdata);
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

	DBGPR("<--xgbe_set_rx_mode\n");
}

static int xgbe_set_mac_address(struct net_device *netdev, void *addr)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	struct sockaddr *saddr = addr;

	DBGPR("-->xgbe_set_mac_address\n");

	if (!is_valid_ether_addr(saddr->sa_data))
		return -EADDRNOTAVAIL;

2016
	eth_hw_addr_set(netdev, saddr->sa_data);
2017 2018 2019 2020 2021 2022 2023 2024

	hw_if->set_mac_address(pdata, netdev->dev_addr);

	DBGPR("<--xgbe_set_mac_address\n");

	return 0;
}

2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045
static int xgbe_ioctl(struct net_device *netdev, struct ifreq *ifreq, int cmd)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	int ret;

	switch (cmd) {
	case SIOCGHWTSTAMP:
		ret = xgbe_get_hwtstamp_settings(pdata, ifreq);
		break;

	case SIOCSHWTSTAMP:
		ret = xgbe_set_hwtstamp_settings(pdata, ifreq);
		break;

	default:
		ret = -EOPNOTSUPP;
	}

	return ret;
}

2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059
static int xgbe_change_mtu(struct net_device *netdev, int mtu)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	int ret;

	DBGPR("-->xgbe_change_mtu\n");

	ret = xgbe_calc_rx_buf_size(netdev, mtu);
	if (ret < 0)
		return ret;

	pdata->rx_buf_size = ret;
	netdev->mtu = mtu;

2060
	xgbe_restart_dev(pdata);
2061 2062 2063 2064 2065 2066

	DBGPR("<--xgbe_change_mtu\n");

	return 0;
}

2067
static void xgbe_tx_timeout(struct net_device *netdev, unsigned int txqueue)
2068 2069 2070 2071
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);

	netdev_warn(netdev, "tx timeout, device restarting\n");
2072
	schedule_work(&pdata->restart_work);
2073 2074
}

2075 2076
static void xgbe_get_stats64(struct net_device *netdev,
			     struct rtnl_link_stats64 *s)
2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_mmc_stats *pstats = &pdata->mmc_stats;

	DBGPR("-->%s\n", __func__);

	pdata->hw_if.read_mmc_stats(pdata);

	s->rx_packets = pstats->rxframecount_gb;
	s->rx_bytes = pstats->rxoctetcount_gb;
	s->rx_errors = pstats->rxframecount_gb -
		       pstats->rxbroadcastframes_g -
		       pstats->rxmulticastframes_g -
		       pstats->rxunicastframes_g;
	s->multicast = pstats->rxmulticastframes_g;
	s->rx_length_errors = pstats->rxlengtherror;
	s->rx_crc_errors = pstats->rxcrcerror;
	s->rx_fifo_errors = pstats->rxfifooverflow;

	s->tx_packets = pstats->txframecount_gb;
	s->tx_bytes = pstats->txoctetcount_gb;
	s->tx_errors = pstats->txframecount_gb - pstats->txframecount_g;
	s->tx_dropped = netdev->stats.tx_dropped;

	DBGPR("<--%s\n", __func__);
}

2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135
static int xgbe_vlan_rx_add_vid(struct net_device *netdev, __be16 proto,
				u16 vid)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;

	DBGPR("-->%s\n", __func__);

	set_bit(vid, pdata->active_vlans);
	hw_if->update_vlan_hash_table(pdata);

	DBGPR("<--%s\n", __func__);

	return 0;
}

static int xgbe_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto,
				 u16 vid)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;

	DBGPR("-->%s\n", __func__);

	clear_bit(vid, pdata->active_vlans);
	hw_if->update_vlan_hash_table(pdata);

	DBGPR("<--%s\n", __func__);

	return 0;
}

2136 2137 2138 2139
#ifdef CONFIG_NET_POLL_CONTROLLER
static void xgbe_poll_controller(struct net_device *netdev)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
2140 2141
	struct xgbe_channel *channel;
	unsigned int i;
2142 2143 2144

	DBGPR("-->xgbe_poll_controller\n");

2145
	if (pdata->per_channel_irq) {
2146 2147
		for (i = 0; i < pdata->channel_count; i++) {
			channel = pdata->channel[i];
2148
			xgbe_dma_isr(channel->dma_irq, channel);
2149
		}
2150 2151 2152 2153 2154
	} else {
		disable_irq(pdata->dev_irq);
		xgbe_isr(pdata->dev_irq, pdata);
		enable_irq(pdata->dev_irq);
	}
2155 2156 2157 2158 2159

	DBGPR("<--xgbe_poll_controller\n");
}
#endif /* End CONFIG_NET_POLL_CONTROLLER */

2160
static int xgbe_setup_tc(struct net_device *netdev, enum tc_setup_type type,
2161
			 void *type_data)
2162 2163
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
2164
	struct tc_mqprio_qopt *mqprio = type_data;
2165
	u8 tc;
2166

2167
	if (type != TC_SETUP_QDISC_MQPRIO)
2168
		return -EOPNOTSUPP;
2169

2170 2171
	mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS;
	tc = mqprio->num_tc;
2172

2173
	if (tc > pdata->hw_feat.tc_cnt)
2174 2175
		return -EINVAL;

2176 2177
	pdata->num_tcs = tc;
	pdata->hw_if.config_tc(pdata);
2178 2179 2180 2181

	return 0;
}

2182 2183 2184 2185
static netdev_features_t xgbe_fix_features(struct net_device *netdev,
					   netdev_features_t features)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
2186
	netdev_features_t vxlan_base;
2187 2188 2189 2190

	vxlan_base = NETIF_F_GSO_UDP_TUNNEL | NETIF_F_RX_UDP_TUNNEL_PORT;

	if (!pdata->hw_feat.vxn)
2191
		return features;
2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224

	/* VXLAN CSUM requires VXLAN base */
	if ((features & NETIF_F_GSO_UDP_TUNNEL_CSUM) &&
	    !(features & NETIF_F_GSO_UDP_TUNNEL)) {
		netdev_notice(netdev,
			      "forcing tx udp tunnel support\n");
		features |= NETIF_F_GSO_UDP_TUNNEL;
	}

	/* Can't do one without doing the other */
	if ((features & vxlan_base) != vxlan_base) {
		netdev_notice(netdev,
			      "forcing both tx and rx udp tunnel support\n");
		features |= vxlan_base;
	}

	if (features & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)) {
		if (!(features & NETIF_F_GSO_UDP_TUNNEL_CSUM)) {
			netdev_notice(netdev,
				      "forcing tx udp tunnel checksumming on\n");
			features |= NETIF_F_GSO_UDP_TUNNEL_CSUM;
		}
	} else {
		if (features & NETIF_F_GSO_UDP_TUNNEL_CSUM) {
			netdev_notice(netdev,
				      "forcing tx udp tunnel checksumming off\n");
			features &= ~NETIF_F_GSO_UDP_TUNNEL_CSUM;
		}
	}

	return features;
}

2225 2226 2227 2228 2229
static int xgbe_set_features(struct net_device *netdev,
			     netdev_features_t features)
{
	struct xgbe_prv_data *pdata = netdev_priv(netdev);
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
2230 2231
	netdev_features_t rxhash, rxcsum, rxvlan, rxvlan_filter;
	int ret = 0;
2232

2233
	rxhash = pdata->netdev_features & NETIF_F_RXHASH;
2234 2235 2236
	rxcsum = pdata->netdev_features & NETIF_F_RXCSUM;
	rxvlan = pdata->netdev_features & NETIF_F_HW_VLAN_CTAG_RX;
	rxvlan_filter = pdata->netdev_features & NETIF_F_HW_VLAN_CTAG_FILTER;
2237

2238 2239 2240 2241 2242 2243 2244
	if ((features & NETIF_F_RXHASH) && !rxhash)
		ret = hw_if->enable_rss(pdata);
	else if (!(features & NETIF_F_RXHASH) && rxhash)
		ret = hw_if->disable_rss(pdata);
	if (ret)
		return ret;

2245
	if ((features & NETIF_F_RXCSUM) && !rxcsum)
2246
		hw_if->enable_rx_csum(pdata);
2247
	else if (!(features & NETIF_F_RXCSUM) && rxcsum)
2248 2249
		hw_if->disable_rx_csum(pdata);

2250
	if ((features & NETIF_F_HW_VLAN_CTAG_RX) && !rxvlan)
2251
		hw_if->enable_rx_vlan_stripping(pdata);
2252
	else if (!(features & NETIF_F_HW_VLAN_CTAG_RX) && rxvlan)
2253
		hw_if->disable_rx_vlan_stripping(pdata);
2254 2255 2256 2257 2258

	if ((features & NETIF_F_HW_VLAN_CTAG_FILTER) && !rxvlan_filter)
		hw_if->enable_rx_vlan_filtering(pdata);
	else if (!(features & NETIF_F_HW_VLAN_CTAG_FILTER) && rxvlan_filter)
		hw_if->disable_rx_vlan_filtering(pdata);
2259 2260 2261 2262 2263 2264 2265 2266

	pdata->netdev_features = features;

	DBGPR("<--xgbe_set_features\n");

	return 0;
}

2267 2268 2269 2270 2271 2272 2273 2274 2275 2276
static netdev_features_t xgbe_features_check(struct sk_buff *skb,
					     struct net_device *netdev,
					     netdev_features_t features)
{
	features = vlan_features_check(skb, features);
	features = vxlan_features_check(skb, features);

	return features;
}

2277 2278 2279 2280 2281 2282 2283
static const struct net_device_ops xgbe_netdev_ops = {
	.ndo_open		= xgbe_open,
	.ndo_stop		= xgbe_close,
	.ndo_start_xmit		= xgbe_xmit,
	.ndo_set_rx_mode	= xgbe_set_rx_mode,
	.ndo_set_mac_address	= xgbe_set_mac_address,
	.ndo_validate_addr	= eth_validate_addr,
2284
	.ndo_eth_ioctl		= xgbe_ioctl,
2285
	.ndo_change_mtu		= xgbe_change_mtu,
2286
	.ndo_tx_timeout		= xgbe_tx_timeout,
2287
	.ndo_get_stats64	= xgbe_get_stats64,
2288 2289
	.ndo_vlan_rx_add_vid	= xgbe_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= xgbe_vlan_rx_kill_vid,
2290 2291 2292
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= xgbe_poll_controller,
#endif
2293
	.ndo_setup_tc		= xgbe_setup_tc,
2294
	.ndo_fix_features	= xgbe_fix_features,
2295
	.ndo_set_features	= xgbe_set_features,
2296
	.ndo_features_check	= xgbe_features_check,
2297 2298
};

2299
const struct net_device_ops *xgbe_get_netdev_ops(void)
2300
{
2301
	return &xgbe_netdev_ops;
2302 2303
}

2304 2305 2306
static void xgbe_rx_refresh(struct xgbe_channel *channel)
{
	struct xgbe_prv_data *pdata = channel->pdata;
2307
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
2308 2309 2310 2311
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct xgbe_ring *ring = channel->rx_ring;
	struct xgbe_ring_data *rdata;

2312 2313 2314 2315 2316 2317 2318 2319 2320
	while (ring->dirty != ring->cur) {
		rdata = XGBE_GET_DESC_DATA(ring, ring->dirty);

		/* Reset rdata values */
		desc_if->unmap_rdata(pdata, rdata);

		if (desc_if->map_rx_buffer(pdata, ring, rdata))
			break;

2321
		hw_if->rx_desc_reset(pdata, rdata, ring->dirty);
2322 2323 2324

		ring->dirty++;
	}
2325

2326 2327 2328
	/* Make sure everything is written before the register write */
	wmb();

2329 2330
	/* Update the Rx Tail Pointer Register with address of
	 * the last cleaned entry */
2331
	rdata = XGBE_GET_DESC_DATA(ring, ring->dirty - 1);
2332 2333 2334 2335
	XGMAC_DMA_IOWRITE(channel, DMA_CH_RDTR_LO,
			  lower_32_bits(rdata->rdesc_dma));
}

2336 2337
static struct sk_buff *xgbe_create_skb(struct xgbe_prv_data *pdata,
				       struct napi_struct *napi,
2338
				       struct xgbe_ring_data *rdata,
2339
				       unsigned int len)
2340 2341 2342 2343
{
	struct sk_buff *skb;
	u8 *packet;

2344
	skb = napi_alloc_skb(napi, rdata->rx.hdr.dma_len);
2345 2346 2347
	if (!skb)
		return NULL;

2348
	/* Pull in the header buffer which may contain just the header
2349 2350
	 * or the header plus data
	 */
2351 2352 2353
	dma_sync_single_range_for_cpu(pdata->dev, rdata->rx.hdr.dma_base,
				      rdata->rx.hdr.dma_off,
				      rdata->rx.hdr.dma_len, DMA_FROM_DEVICE);
2354

2355 2356
	packet = page_address(rdata->rx.hdr.pa.pages) +
		 rdata->rx.hdr.pa.pages_offset;
2357 2358
	skb_copy_to_linear_data(skb, packet, len);
	skb_put(skb, len);
2359 2360 2361 2362

	return skb;
}

2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399
static unsigned int xgbe_rx_buf1_len(struct xgbe_ring_data *rdata,
				     struct xgbe_packet_data *packet)
{
	/* Always zero if not the first descriptor */
	if (!XGMAC_GET_BITS(packet->attributes, RX_PACKET_ATTRIBUTES, FIRST))
		return 0;

	/* First descriptor with split header, return header length */
	if (rdata->rx.hdr_len)
		return rdata->rx.hdr_len;

	/* First descriptor but not the last descriptor and no split header,
	 * so the full buffer was used
	 */
	if (!XGMAC_GET_BITS(packet->attributes, RX_PACKET_ATTRIBUTES, LAST))
		return rdata->rx.hdr.dma_len;

	/* First descriptor and last descriptor and no split header, so
	 * calculate how much of the buffer was used
	 */
	return min_t(unsigned int, rdata->rx.hdr.dma_len, rdata->rx.len);
}

static unsigned int xgbe_rx_buf2_len(struct xgbe_ring_data *rdata,
				     struct xgbe_packet_data *packet,
				     unsigned int len)
{
	/* Always the full buffer if not the last descriptor */
	if (!XGMAC_GET_BITS(packet->attributes, RX_PACKET_ATTRIBUTES, LAST))
		return rdata->rx.buf.dma_len;

	/* Last descriptor so calculate how much of the buffer was used
	 * for the last bit of data
	 */
	return rdata->rx.len - len;
}

2400 2401 2402 2403 2404 2405 2406 2407 2408
static int xgbe_tx_poll(struct xgbe_channel *channel)
{
	struct xgbe_prv_data *pdata = channel->pdata;
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	struct xgbe_desc_if *desc_if = &pdata->desc_if;
	struct xgbe_ring *ring = channel->tx_ring;
	struct xgbe_ring_data *rdata;
	struct xgbe_ring_desc *rdesc;
	struct net_device *netdev = pdata->netdev;
L
Lendacky, Thomas 已提交
2409
	struct netdev_queue *txq;
2410
	int processed = 0;
L
Lendacky, Thomas 已提交
2411
	unsigned int tx_packets = 0, tx_bytes = 0;
2412
	unsigned int cur;
2413 2414 2415 2416 2417 2418 2419

	DBGPR("-->xgbe_tx_poll\n");

	/* Nothing to do if there isn't a Tx ring for this channel */
	if (!ring)
		return 0;

2420
	cur = ring->cur;
2421 2422 2423 2424

	/* Be sure we get ring->cur before accessing descriptor data */
	smp_rmb();

L
Lendacky, Thomas 已提交
2425 2426
	txq = netdev_get_tx_queue(netdev, channel->queue_index);

2427
	while ((processed < XGBE_TX_DESC_MAX_PROC) &&
2428
	       (ring->dirty != cur)) {
2429
		rdata = XGBE_GET_DESC_DATA(ring, ring->dirty);
2430 2431 2432 2433 2434
		rdesc = rdata->rdesc;

		if (!hw_if->tx_complete(rdesc))
			break;

2435 2436
		/* Make sure descriptor fields are read after reading the OWN
		 * bit */
2437
		dma_rmb();
2438

2439 2440
		if (netif_msg_tx_done(pdata))
			xgbe_dump_tx_desc(pdata, ring, ring->dirty, 1, 0);
2441

L
Lendacky, Thomas 已提交
2442 2443 2444 2445 2446
		if (hw_if->is_last_desc(rdesc)) {
			tx_packets += rdata->tx.packets;
			tx_bytes += rdata->tx.bytes;
		}

2447
		/* Free the SKB and reset the descriptor for re-use */
2448
		desc_if->unmap_rdata(pdata, rdata);
2449 2450 2451 2452 2453 2454
		hw_if->tx_desc_reset(rdata);

		processed++;
		ring->dirty++;
	}

L
Lendacky, Thomas 已提交
2455
	if (!processed)
2456
		return 0;
L
Lendacky, Thomas 已提交
2457 2458 2459

	netdev_tx_completed_queue(txq, tx_packets, tx_bytes);

2460
	if ((ring->tx.queue_stopped == 1) &&
2461
	    (xgbe_tx_avail_desc(ring) > XGBE_TX_DESC_MIN_FREE)) {
2462
		ring->tx.queue_stopped = 0;
L
Lendacky, Thomas 已提交
2463
		netif_tx_wake_queue(txq);
2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478
	}

	DBGPR("<--xgbe_tx_poll: processed=%d\n", processed);

	return processed;
}

static int xgbe_rx_poll(struct xgbe_channel *channel, int budget)
{
	struct xgbe_prv_data *pdata = channel->pdata;
	struct xgbe_hw_if *hw_if = &pdata->hw_if;
	struct xgbe_ring *ring = channel->rx_ring;
	struct xgbe_ring_data *rdata;
	struct xgbe_packet_data *packet;
	struct net_device *netdev = pdata->netdev;
2479
	struct napi_struct *napi;
2480
	struct sk_buff *skb;
2481
	struct skb_shared_hwtstamps *hwtstamps;
2482 2483
	unsigned int last, error, context_next, context;
	unsigned int len, buf1_len, buf2_len, max_len;
2484 2485
	unsigned int received = 0;
	int packet_count = 0;
2486 2487 2488 2489 2490 2491 2492

	DBGPR("-->xgbe_rx_poll: budget=%d\n", budget);

	/* Nothing to do if there isn't a Rx ring for this channel */
	if (!ring)
		return 0;

2493
	last = 0;
2494 2495
	context_next = 0;

2496 2497
	napi = (pdata->per_channel_irq) ? &channel->napi : &pdata->napi;

2498
	rdata = XGBE_GET_DESC_DATA(ring, ring->cur);
2499
	packet = &ring->packet_data;
2500
	while (packet_count < budget) {
2501 2502
		DBGPR("  cur = %d\n", ring->cur);

2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513
		/* First time in loop see if we need to restore state */
		if (!received && rdata->state_saved) {
			skb = rdata->state.skb;
			error = rdata->state.error;
			len = rdata->state.len;
		} else {
			memset(packet, 0, sizeof(*packet));
			skb = NULL;
			error = 0;
			len = 0;
		}
2514 2515

read_again:
2516 2517
		rdata = XGBE_GET_DESC_DATA(ring, ring->cur);

2518
		if (xgbe_rx_dirty_desc(ring) > (XGBE_RX_DESC_CNT >> 3))
2519 2520
			xgbe_rx_refresh(channel);

2521 2522 2523 2524 2525 2526
		if (hw_if->dev_read(channel))
			break;

		received++;
		ring->cur++;

2527 2528
		last = XGMAC_GET_BITS(packet->attributes, RX_PACKET_ATTRIBUTES,
				      LAST);
2529 2530 2531 2532 2533 2534
		context_next = XGMAC_GET_BITS(packet->attributes,
					      RX_PACKET_ATTRIBUTES,
					      CONTEXT_NEXT);
		context = XGMAC_GET_BITS(packet->attributes,
					 RX_PACKET_ATTRIBUTES,
					 CONTEXT);
2535 2536

		/* Earlier error, just drain the remaining data */
2537
		if ((!last || context_next) && error)
2538 2539 2540 2541
			goto read_again;

		if (error || packet->errors) {
			if (packet->errors)
2542 2543
				netif_err(pdata, rx_err, netdev,
					  "error in received packet\n");
2544
			dev_kfree_skb(skb);
2545
			goto next_packet;
2546 2547
		}

2548
		if (!context) {
2549 2550 2551 2552 2553
			/* Get the data length in the descriptor buffers */
			buf1_len = xgbe_rx_buf1_len(rdata, packet);
			len += buf1_len;
			buf2_len = xgbe_rx_buf2_len(rdata, packet, len);
			len += buf2_len;
2554

2555 2556 2557 2558 2559 2560 2561 2562
			if (buf2_len > rdata->rx.buf.dma_len) {
				/* Hardware inconsistency within the descriptors
				 * that has resulted in a length underflow.
				 */
				error = 1;
				goto skip_data;
			}

2563
			if (!skb) {
2564
				skb = xgbe_create_skb(pdata, napi, rdata,
2565 2566
						      buf1_len);
				if (!skb) {
2567
					error = 1;
2568 2569 2570 2571 2572
					goto skip_data;
				}
			}

			if (buf2_len) {
2573 2574 2575
				dma_sync_single_range_for_cpu(pdata->dev,
							rdata->rx.buf.dma_base,
							rdata->rx.buf.dma_off,
2576
							rdata->rx.buf.dma_len,
2577 2578 2579
							DMA_FROM_DEVICE);

				skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
2580 2581
						rdata->rx.buf.pa.pages,
						rdata->rx.buf.pa.pages_offset,
2582
						buf2_len,
2583
						rdata->rx.buf.dma_len);
2584
				rdata->rx.buf.pa.pages = NULL;
2585
			}
2586 2587
		}

2588 2589
skip_data:
		if (!last || context_next)
2590 2591
			goto read_again;

2592 2593
		if (!skb || error) {
			dev_kfree_skb(skb);
2594
			goto next_packet;
2595
		}
2596

2597 2598 2599 2600 2601 2602 2603
		/* Be sure we don't exceed the configured MTU */
		max_len = netdev->mtu + ETH_HLEN;
		if (!(netdev->features & NETIF_F_HW_VLAN_CTAG_RX) &&
		    (skb->protocol == htons(ETH_P_8021Q)))
			max_len += VLAN_HLEN;

		if (skb->len > max_len) {
2604 2605
			netif_err(pdata, rx_err, netdev,
				  "packet length exceeds configured MTU\n");
2606
			dev_kfree_skb(skb);
2607
			goto next_packet;
2608 2609
		}

2610 2611
		if (netif_msg_pktdata(pdata))
			xgbe_print_pkt(netdev, skb, false);
2612 2613 2614 2615 2616 2617

		skb_checksum_none_assert(skb);
		if (XGMAC_GET_BITS(packet->attributes,
				   RX_PACKET_ATTRIBUTES, CSUM_DONE))
			skb->ip_summed = CHECKSUM_UNNECESSARY;

2618 2619 2620 2621 2622 2623 2624 2625 2626
		if (XGMAC_GET_BITS(packet->attributes,
				   RX_PACKET_ATTRIBUTES, TNP)) {
			skb->encapsulation = 1;

			if (XGMAC_GET_BITS(packet->attributes,
					   RX_PACKET_ATTRIBUTES, TNPCSUM_DONE))
				skb->csum_level = 1;
		}

2627 2628 2629 2630 2631
		if (XGMAC_GET_BITS(packet->attributes,
				   RX_PACKET_ATTRIBUTES, VLAN_CTAG))
			__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
					       packet->vlan_ctag);

2632 2633 2634 2635 2636 2637 2638 2639 2640 2641
		if (XGMAC_GET_BITS(packet->attributes,
				   RX_PACKET_ATTRIBUTES, RX_TSTAMP)) {
			u64 nsec;

			nsec = timecounter_cyc2time(&pdata->tstamp_tc,
						    packet->rx_tstamp);
			hwtstamps = skb_hwtstamps(skb);
			hwtstamps->hwtstamp = ns_to_ktime(nsec);
		}

2642 2643 2644 2645 2646
		if (XGMAC_GET_BITS(packet->attributes,
				   RX_PACKET_ATTRIBUTES, RSS_HASH))
			skb_set_hash(skb, packet->rss_hash,
				     packet->rss_hash_type);

2647 2648 2649 2650
		skb->dev = netdev;
		skb->protocol = eth_type_trans(skb, netdev);
		skb_record_rx_queue(skb, channel->queue_index);

2651
		napi_gro_receive(napi, skb);
2652 2653 2654

next_packet:
		packet_count++;
2655 2656
	}

2657
	/* Check if we need to save state before leaving */
2658
	if (received && (!last || context_next)) {
2659 2660 2661 2662 2663 2664 2665
		rdata = XGBE_GET_DESC_DATA(ring, ring->cur);
		rdata->state_saved = 1;
		rdata->state.skb = skb;
		rdata->state.len = len;
		rdata->state.error = error;
	}

2666
	DBGPR("<--xgbe_rx_poll: packet_count = %d\n", packet_count);
2667

2668
	return packet_count;
2669 2670
}

2671 2672 2673 2674
static int xgbe_one_poll(struct napi_struct *napi, int budget)
{
	struct xgbe_channel *channel = container_of(napi, struct xgbe_channel,
						    napi);
2675
	struct xgbe_prv_data *pdata = channel->pdata;
2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686
	int processed = 0;

	DBGPR("-->xgbe_one_poll: budget=%d\n", budget);

	/* Cleanup Tx ring first */
	xgbe_tx_poll(channel);

	/* Process Rx ring next */
	processed = xgbe_rx_poll(channel, budget);

	/* If we processed everything, we are done */
2687
	if ((processed < budget) && napi_complete_done(napi, processed)) {
2688
		/* Enable Tx and Rx interrupts */
2689 2690 2691 2692
		if (pdata->channel_irq_mode)
			xgbe_enable_rx_tx_int(pdata, channel);
		else
			enable_irq(channel->dma_irq);
2693 2694 2695 2696 2697 2698 2699 2700
	}

	DBGPR("<--xgbe_one_poll: received = %d\n", processed);

	return processed;
}

static int xgbe_all_poll(struct napi_struct *napi, int budget)
2701 2702 2703 2704
{
	struct xgbe_prv_data *pdata = container_of(napi, struct xgbe_prv_data,
						   napi);
	struct xgbe_channel *channel;
2705 2706
	int ring_budget;
	int processed, last_processed;
2707 2708
	unsigned int i;

2709
	DBGPR("-->xgbe_all_poll: budget=%d\n", budget);
2710 2711

	processed = 0;
2712 2713 2714 2715
	ring_budget = budget / pdata->rx_ring_count;
	do {
		last_processed = processed;

2716 2717 2718
		for (i = 0; i < pdata->channel_count; i++) {
			channel = pdata->channel[i];

2719 2720 2721 2722 2723 2724 2725 2726 2727
			/* Cleanup Tx ring first */
			xgbe_tx_poll(channel);

			/* Process Rx ring next */
			if (ring_budget > (budget - processed))
				ring_budget = budget - processed;
			processed += xgbe_rx_poll(channel, ring_budget);
		}
	} while ((processed < budget) && (processed != last_processed));
2728 2729

	/* If we processed everything, we are done */
2730
	if ((processed < budget) && napi_complete_done(napi, processed)) {
2731 2732 2733 2734
		/* Enable Tx and Rx interrupts */
		xgbe_enable_rx_tx_ints(pdata);
	}

2735
	DBGPR("<--xgbe_all_poll: received = %d\n", processed);
2736 2737 2738 2739

	return processed;
}

2740 2741
void xgbe_dump_tx_desc(struct xgbe_prv_data *pdata, struct xgbe_ring *ring,
		       unsigned int idx, unsigned int count, unsigned int flag)
2742 2743 2744 2745 2746
{
	struct xgbe_ring_data *rdata;
	struct xgbe_ring_desc *rdesc;

	while (count--) {
2747
		rdata = XGBE_GET_DESC_DATA(ring, idx);
2748
		rdesc = rdata->rdesc;
2749 2750 2751 2752 2753 2754 2755
		netdev_dbg(pdata->netdev,
			   "TX_NORMAL_DESC[%d %s] = %08x:%08x:%08x:%08x\n", idx,
			   (flag == 1) ? "QUEUED FOR TX" : "TX BY DEVICE",
			   le32_to_cpu(rdesc->desc0),
			   le32_to_cpu(rdesc->desc1),
			   le32_to_cpu(rdesc->desc2),
			   le32_to_cpu(rdesc->desc3));
2756 2757 2758 2759
		idx++;
	}
}

2760
void xgbe_dump_rx_desc(struct xgbe_prv_data *pdata, struct xgbe_ring *ring,
2761 2762
		       unsigned int idx)
{
2763 2764 2765 2766 2767 2768 2769 2770 2771
	struct xgbe_ring_data *rdata;
	struct xgbe_ring_desc *rdesc;

	rdata = XGBE_GET_DESC_DATA(ring, idx);
	rdesc = rdata->rdesc;
	netdev_dbg(pdata->netdev,
		   "RX_NORMAL_DESC[%d RX BY DEVICE] = %08x:%08x:%08x:%08x\n",
		   idx, le32_to_cpu(rdesc->desc0), le32_to_cpu(rdesc->desc1),
		   le32_to_cpu(rdesc->desc2), le32_to_cpu(rdesc->desc3));
2772 2773 2774 2775 2776 2777
}

void xgbe_print_pkt(struct net_device *netdev, struct sk_buff *skb, bool tx_rx)
{
	struct ethhdr *eth = (struct ethhdr *)skb->data;
	unsigned char buffer[128];
2778
	unsigned int i;
2779

2780
	netdev_dbg(netdev, "\n************** SKB dump ****************\n");
2781

2782 2783
	netdev_dbg(netdev, "%s packet of %d bytes\n",
		   (tx_rx ? "TX" : "RX"), skb->len);
2784

2785 2786
	netdev_dbg(netdev, "Dst MAC addr: %pM\n", eth->h_dest);
	netdev_dbg(netdev, "Src MAC addr: %pM\n", eth->h_source);
2787
	netdev_dbg(netdev, "Protocol: %#06x\n", ntohs(eth->h_proto));
2788

2789 2790 2791 2792 2793 2794
	for (i = 0; i < skb->len; i += 32) {
		unsigned int len = min(skb->len - i, 32U);

		hex_dump_to_buffer(&skb->data[i], len, 32, 1,
				   buffer, sizeof(buffer), false);
		netdev_dbg(netdev, "  %#06x: %s\n", i, buffer);
2795 2796
	}

2797
	netdev_dbg(netdev, "\n************** SKB dump ****************\n");
2798
}