xref: /freebsd/sys/dev/ice/if_ice_iflib.c (revision c66ec88fed842fbaad62c30d510644ceb7bd2d71)
1 /* SPDX-License-Identifier: BSD-3-Clause */
2 /*  Copyright (c) 2020, Intel Corporation
3  *  All rights reserved.
4  *
5  *  Redistribution and use in source and binary forms, with or without
6  *  modification, are permitted provided that the following conditions are met:
7  *
8  *   1. Redistributions of source code must retain the above copyright notice,
9  *      this list of conditions and the following disclaimer.
10  *
11  *   2. Redistributions in binary form must reproduce the above copyright
12  *      notice, this list of conditions and the following disclaimer in the
13  *      documentation and/or other materials provided with the distribution.
14  *
15  *   3. Neither the name of the Intel Corporation nor the names of its
16  *      contributors may be used to endorse or promote products derived from
17  *      this software without specific prior written permission.
18  *
19  *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
20  *  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  *  IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  *  ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
23  *  LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  *  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  *  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26  *  INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27  *  CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28  *  ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29  *  POSSIBILITY OF SUCH DAMAGE.
30  */
31 /*$FreeBSD$*/
32 
33 /**
34  * @file if_ice_iflib.c
35  * @brief iflib driver implementation
36  *
37  * Contains the main entry point for the iflib driver implementation. It
38  * implements the various ifdi driver methods, and sets up the module and
39  * driver values to load an iflib driver.
40  */
41 
42 #include "ice_iflib.h"
43 #include "ice_drv_info.h"
44 #include "ice_switch.h"
45 #include "ice_sched.h"
46 
47 #include <sys/module.h>
48 #include <sys/sockio.h>
49 #include <sys/smp.h>
50 #include <dev/pci/pcivar.h>
51 #include <dev/pci/pcireg.h>
52 
53 /*
54  * Device method prototypes
55  */
56 
57 static void *ice_register(device_t);
58 static int  ice_if_attach_pre(if_ctx_t);
59 static int  ice_attach_pre_recovery_mode(struct ice_softc *sc);
60 static int  ice_if_attach_post(if_ctx_t);
61 static void ice_attach_post_recovery_mode(struct ice_softc *sc);
62 static int  ice_if_detach(if_ctx_t);
63 static int  ice_if_tx_queues_alloc(if_ctx_t ctx, caddr_t *vaddrs, uint64_t *paddrs, int ntxqs, int ntxqsets);
64 static int  ice_if_rx_queues_alloc(if_ctx_t ctx, caddr_t *vaddrs, uint64_t *paddrs, int nqs, int nqsets);
65 static int ice_if_msix_intr_assign(if_ctx_t ctx, int msix);
66 static void ice_if_queues_free(if_ctx_t ctx);
67 static int ice_if_mtu_set(if_ctx_t ctx, uint32_t mtu);
68 static void ice_if_intr_enable(if_ctx_t ctx);
69 static void ice_if_intr_disable(if_ctx_t ctx);
70 static int ice_if_rx_queue_intr_enable(if_ctx_t ctx, uint16_t rxqid);
71 static int ice_if_tx_queue_intr_enable(if_ctx_t ctx, uint16_t txqid);
72 static int ice_if_promisc_set(if_ctx_t ctx, int flags);
73 static void ice_if_media_status(if_ctx_t ctx, struct ifmediareq *ifmr);
74 static int ice_if_media_change(if_ctx_t ctx);
75 static void ice_if_init(if_ctx_t ctx);
76 static void ice_if_timer(if_ctx_t ctx, uint16_t qid);
77 static void ice_if_update_admin_status(if_ctx_t ctx);
78 static void ice_if_multi_set(if_ctx_t ctx);
79 static void ice_if_vlan_register(if_ctx_t ctx, u16 vtag);
80 static void ice_if_vlan_unregister(if_ctx_t ctx, u16 vtag);
81 static void ice_if_stop(if_ctx_t ctx);
82 static uint64_t ice_if_get_counter(if_ctx_t ctx, ift_counter counter);
83 static int ice_if_priv_ioctl(if_ctx_t ctx, u_long command, caddr_t data);
84 static int ice_if_i2c_req(if_ctx_t ctx, struct ifi2creq *req);
85 static int ice_if_suspend(if_ctx_t ctx);
86 static int ice_if_resume(if_ctx_t ctx);
87 
88 static int ice_msix_que(void *arg);
89 static int ice_msix_admin(void *arg);
90 
91 /*
92  * Helper function prototypes
93  */
94 static int ice_pci_mapping(struct ice_softc *sc);
95 static void ice_free_pci_mapping(struct ice_softc *sc);
96 static void ice_update_link_status(struct ice_softc *sc, bool update_media);
97 static void ice_init_device_features(struct ice_softc *sc);
98 static void ice_init_tx_tracking(struct ice_vsi *vsi);
99 static void ice_handle_reset_event(struct ice_softc *sc);
100 static void ice_handle_pf_reset_request(struct ice_softc *sc);
101 static void ice_prepare_for_reset(struct ice_softc *sc);
102 static int ice_rebuild_pf_vsi_qmap(struct ice_softc *sc);
103 static void ice_rebuild(struct ice_softc *sc);
104 static void ice_rebuild_recovery_mode(struct ice_softc *sc);
105 static void ice_free_irqvs(struct ice_softc *sc);
106 static void ice_update_rx_mbuf_sz(struct ice_softc *sc);
107 static void ice_poll_for_media_avail(struct ice_softc *sc);
108 static void ice_setup_scctx(struct ice_softc *sc);
109 static int ice_allocate_msix(struct ice_softc *sc);
110 static void ice_admin_timer(void *arg);
111 static void ice_transition_recovery_mode(struct ice_softc *sc);
112 static void ice_transition_safe_mode(struct ice_softc *sc);
113 
114 /*
115  * Device Interface Declaration
116  */
117 
118 /**
119  * @var ice_methods
120  * @brief ice driver method entry points
121  *
122  * List of device methods implementing the generic device interface used by
123  * the device stack to interact with the ice driver. Since this is an iflib
124  * driver, most of the methods point to the generic iflib implementation.
125  */
126 static device_method_t ice_methods[] = {
127 	/* Device interface */
128 	DEVMETHOD(device_register, ice_register),
129 	DEVMETHOD(device_probe,    iflib_device_probe_vendor),
130 	DEVMETHOD(device_attach,   iflib_device_attach),
131 	DEVMETHOD(device_detach,   iflib_device_detach),
132 	DEVMETHOD(device_shutdown, iflib_device_shutdown),
133 	DEVMETHOD(device_suspend,  iflib_device_suspend),
134 	DEVMETHOD(device_resume,   iflib_device_resume),
135 	DEVMETHOD_END
136 };
137 
138 /**
139  * @var ice_iflib_methods
140  * @brief iflib method entry points
141  *
142  * List of device methods used by the iflib stack to interact with this
143  * driver. These are the real main entry points used to interact with this
144  * driver.
145  */
146 static device_method_t ice_iflib_methods[] = {
147 	DEVMETHOD(ifdi_attach_pre, ice_if_attach_pre),
148 	DEVMETHOD(ifdi_attach_post, ice_if_attach_post),
149 	DEVMETHOD(ifdi_detach, ice_if_detach),
150 	DEVMETHOD(ifdi_tx_queues_alloc, ice_if_tx_queues_alloc),
151 	DEVMETHOD(ifdi_rx_queues_alloc, ice_if_rx_queues_alloc),
152 	DEVMETHOD(ifdi_msix_intr_assign, ice_if_msix_intr_assign),
153 	DEVMETHOD(ifdi_queues_free, ice_if_queues_free),
154 	DEVMETHOD(ifdi_mtu_set, ice_if_mtu_set),
155 	DEVMETHOD(ifdi_intr_enable, ice_if_intr_enable),
156 	DEVMETHOD(ifdi_intr_disable, ice_if_intr_disable),
157 	DEVMETHOD(ifdi_rx_queue_intr_enable, ice_if_rx_queue_intr_enable),
158 	DEVMETHOD(ifdi_tx_queue_intr_enable, ice_if_tx_queue_intr_enable),
159 	DEVMETHOD(ifdi_promisc_set, ice_if_promisc_set),
160 	DEVMETHOD(ifdi_media_status, ice_if_media_status),
161 	DEVMETHOD(ifdi_media_change, ice_if_media_change),
162 	DEVMETHOD(ifdi_init, ice_if_init),
163 	DEVMETHOD(ifdi_stop, ice_if_stop),
164 	DEVMETHOD(ifdi_timer, ice_if_timer),
165 	DEVMETHOD(ifdi_update_admin_status, ice_if_update_admin_status),
166 	DEVMETHOD(ifdi_multi_set, ice_if_multi_set),
167 	DEVMETHOD(ifdi_vlan_register, ice_if_vlan_register),
168 	DEVMETHOD(ifdi_vlan_unregister, ice_if_vlan_unregister),
169 	DEVMETHOD(ifdi_get_counter, ice_if_get_counter),
170 	DEVMETHOD(ifdi_priv_ioctl, ice_if_priv_ioctl),
171 	DEVMETHOD(ifdi_i2c_req, ice_if_i2c_req),
172 	DEVMETHOD(ifdi_suspend, ice_if_suspend),
173 	DEVMETHOD(ifdi_resume, ice_if_resume),
174 	DEVMETHOD_END
175 };
176 
177 /**
178  * @var ice_driver
179  * @brief driver structure for the generic device stack
180  *
181  * driver_t definition used to setup the generic device methods.
182  */
183 static driver_t ice_driver = {
184 	.name = "ice",
185 	.methods = ice_methods,
186 	.size = sizeof(struct ice_softc),
187 };
188 
189 /**
190  * @var ice_iflib_driver
191  * @brief driver structure for the iflib stack
192  *
193  * driver_t definition used to setup the iflib device methods.
194  */
195 static driver_t ice_iflib_driver = {
196 	.name = "ice",
197 	.methods = ice_iflib_methods,
198 	.size = sizeof(struct ice_softc),
199 };
200 
201 extern struct if_txrx ice_txrx;
202 extern struct if_txrx ice_recovery_txrx;
203 
204 /**
205  * @var ice_sctx
206  * @brief ice driver shared context
207  *
208  * Structure defining shared values (context) that is used by all instances of
209  * the device. Primarily used to setup details about how the iflib stack
210  * should treat this driver. Also defines the default, minimum, and maximum
211  * number of descriptors in each ring.
212  */
213 static struct if_shared_ctx ice_sctx = {
214 	.isc_magic = IFLIB_MAGIC,
215 	.isc_q_align = PAGE_SIZE,
216 
217 	.isc_tx_maxsize = ICE_MAX_FRAME_SIZE,
218 	/* We could technically set this as high as ICE_MAX_DMA_SEG_SIZE, but
219 	 * that doesn't make sense since that would be larger than the maximum
220 	 * size of a single packet.
221 	 */
222 	.isc_tx_maxsegsize = ICE_MAX_FRAME_SIZE,
223 
224 	/* XXX: This is only used by iflib to ensure that
225 	 * scctx->isc_tx_tso_size_max + the VLAN header is a valid size.
226 	 */
227 	.isc_tso_maxsize = ICE_TSO_SIZE + sizeof(struct ether_vlan_header),
228 	/* XXX: This is used by iflib to set the number of segments in the TSO
229 	 * DMA tag. However, scctx->isc_tx_tso_segsize_max is used to set the
230 	 * related ifnet parameter.
231 	 */
232 	.isc_tso_maxsegsize = ICE_MAX_DMA_SEG_SIZE,
233 
234 	.isc_rx_maxsize = ICE_MAX_FRAME_SIZE,
235 	.isc_rx_nsegments = ICE_MAX_RX_SEGS,
236 	.isc_rx_maxsegsize = ICE_MAX_FRAME_SIZE,
237 
238 	.isc_nfl = 1,
239 	.isc_ntxqs = 1,
240 	.isc_nrxqs = 1,
241 
242 	.isc_admin_intrcnt = 1,
243 	.isc_vendor_info = ice_vendor_info_array,
244 	.isc_driver_version = __DECONST(char *, ice_driver_version),
245 	.isc_driver = &ice_iflib_driver,
246 
247 	/*
248 	 * IFLIB_NEED_SCRATCH ensures that mbufs have scratch space available
249 	 * for hardware checksum offload
250 	 *
251 	 * IFLIB_TSO_INIT_IP ensures that the TSO packets have zeroed out the
252 	 * IP sum field, required by our hardware to calculate valid TSO
253 	 * checksums.
254 	 *
255 	 * IFLIB_ADMIN_ALWAYS_RUN ensures that the administrative task runs
256 	 * even when the interface is down.
257 	 *
258 	 * IFLIB_SKIP_MSIX allows the driver to handle allocating MSI-X
259 	 * vectors manually instead of relying on iflib code to do this.
260 	 */
261 	.isc_flags = IFLIB_NEED_SCRATCH | IFLIB_TSO_INIT_IP |
262 		IFLIB_ADMIN_ALWAYS_RUN | IFLIB_SKIP_MSIX,
263 
264 	.isc_nrxd_min = {ICE_MIN_DESC_COUNT},
265 	.isc_ntxd_min = {ICE_MIN_DESC_COUNT},
266 	.isc_nrxd_max = {ICE_IFLIB_MAX_DESC_COUNT},
267 	.isc_ntxd_max = {ICE_IFLIB_MAX_DESC_COUNT},
268 	.isc_nrxd_default = {ICE_DEFAULT_DESC_COUNT},
269 	.isc_ntxd_default = {ICE_DEFAULT_DESC_COUNT},
270 };
271 
272 /**
273  * @var ice_devclass
274  * @brief ice driver device class
275  *
276  * device class used to setup the ice driver module kobject class.
277  */
278 devclass_t ice_devclass;
279 DRIVER_MODULE(ice, pci, ice_driver, ice_devclass, ice_module_event_handler, 0);
280 
281 MODULE_VERSION(ice, 1);
282 MODULE_DEPEND(ice, pci, 1, 1, 1);
283 MODULE_DEPEND(ice, ether, 1, 1, 1);
284 MODULE_DEPEND(ice, iflib, 1, 1, 1);
285 
286 IFLIB_PNP_INFO(pci, ice, ice_vendor_info_array);
287 
288 /* Static driver-wide sysctls */
289 #include "ice_iflib_sysctls.h"
290 
291 /**
292  * ice_pci_mapping - Map PCI BAR memory
293  * @sc: device private softc
294  *
295  * Map PCI BAR 0 for device operation.
296  */
297 static int
298 ice_pci_mapping(struct ice_softc *sc)
299 {
300 	int rc;
301 
302 	/* Map BAR0 */
303 	rc = ice_map_bar(sc->dev, &sc->bar0, 0);
304 	if (rc)
305 		return rc;
306 
307 	return 0;
308 }
309 
310 /**
311  * ice_free_pci_mapping - Release PCI BAR memory
312  * @sc: device private softc
313  *
314  * Release PCI BARs which were previously mapped by ice_pci_mapping().
315  */
316 static void
317 ice_free_pci_mapping(struct ice_softc *sc)
318 {
319 	/* Free BAR0 */
320 	ice_free_bar(sc->dev, &sc->bar0);
321 }
322 
323 /*
324  * Device methods
325  */
326 
327 /**
328  * ice_register - register device method callback
329  * @dev: the device being registered
330  *
331  * Returns a pointer to the shared context structure, which is used by iflib.
332  */
333 static void *
334 ice_register(device_t dev __unused)
335 {
336 	return &ice_sctx;
337 } /* ice_register */
338 
339 /**
340  * ice_setup_scctx - Setup the iflib softc context structure
341  * @sc: the device private structure
342  *
343  * Setup the parameters in if_softc_ctx_t structure used by the iflib stack
344  * when loading.
345  */
346 static void
347 ice_setup_scctx(struct ice_softc *sc)
348 {
349 	if_softc_ctx_t scctx = sc->scctx;
350 	struct ice_hw *hw = &sc->hw;
351 	bool safe_mode, recovery_mode;
352 
353 	safe_mode = ice_is_bit_set(sc->feat_en, ICE_FEATURE_SAFE_MODE);
354 	recovery_mode = ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE);
355 
356 	/*
357 	 * If the driver loads in Safe mode or Recovery mode, limit iflib to
358 	 * a single queue pair.
359 	 */
360 	if (safe_mode || recovery_mode) {
361 		scctx->isc_ntxqsets = scctx->isc_nrxqsets = 1;
362 		scctx->isc_ntxqsets_max = 1;
363 		scctx->isc_nrxqsets_max = 1;
364 	} else {
365 		/*
366 		 * iflib initially sets the isc_ntxqsets and isc_nrxqsets to
367 		 * the values of the override sysctls. Cache these initial
368 		 * values so that the driver can be aware of what the iflib
369 		 * sysctl value is when setting up MSI-X vectors.
370 		 */
371 		sc->ifc_sysctl_ntxqs = scctx->isc_ntxqsets;
372 		sc->ifc_sysctl_nrxqs = scctx->isc_nrxqsets;
373 
374 		if (scctx->isc_ntxqsets == 0)
375 			scctx->isc_ntxqsets = hw->func_caps.common_cap.rss_table_size;
376 		if (scctx->isc_nrxqsets == 0)
377 			scctx->isc_nrxqsets = hw->func_caps.common_cap.rss_table_size;
378 
379 		scctx->isc_ntxqsets_max = hw->func_caps.common_cap.num_txq;
380 		scctx->isc_nrxqsets_max = hw->func_caps.common_cap.num_rxq;
381 
382 		/*
383 		 * Sanity check that the iflib sysctl values are within the
384 		 * maximum supported range.
385 		 */
386 		if (sc->ifc_sysctl_ntxqs > scctx->isc_ntxqsets_max)
387 			sc->ifc_sysctl_ntxqs = scctx->isc_ntxqsets_max;
388 		if (sc->ifc_sysctl_nrxqs > scctx->isc_nrxqsets_max)
389 			sc->ifc_sysctl_nrxqs = scctx->isc_nrxqsets_max;
390 	}
391 
392 	scctx->isc_txqsizes[0] = roundup2(scctx->isc_ntxd[0]
393 	    * sizeof(struct ice_tx_desc), DBA_ALIGN);
394 	scctx->isc_rxqsizes[0] = roundup2(scctx->isc_nrxd[0]
395 	    * sizeof(union ice_32b_rx_flex_desc), DBA_ALIGN);
396 
397 	scctx->isc_tx_nsegments = ICE_MAX_TX_SEGS;
398 	scctx->isc_tx_tso_segments_max = ICE_MAX_TSO_SEGS;
399 	scctx->isc_tx_tso_size_max = ICE_TSO_SIZE;
400 	scctx->isc_tx_tso_segsize_max = ICE_MAX_DMA_SEG_SIZE;
401 
402 	scctx->isc_msix_bar = PCIR_BAR(ICE_MSIX_BAR);
403 	scctx->isc_rss_table_size = hw->func_caps.common_cap.rss_table_size;
404 
405 	/*
406 	 * If the driver loads in recovery mode, disable Tx/Rx functionality
407 	 */
408 	if (recovery_mode)
409 		scctx->isc_txrx = &ice_recovery_txrx;
410 	else
411 		scctx->isc_txrx = &ice_txrx;
412 
413 	/*
414 	 * If the driver loads in Safe mode or Recovery mode, disable
415 	 * advanced features including hardware offloads.
416 	 */
417 	if (safe_mode || recovery_mode) {
418 		scctx->isc_capenable = ICE_SAFE_CAPS;
419 		scctx->isc_tx_csum_flags = 0;
420 	} else {
421 		scctx->isc_capenable = ICE_FULL_CAPS;
422 		scctx->isc_tx_csum_flags = ICE_CSUM_OFFLOAD;
423 	}
424 
425 	scctx->isc_capabilities = scctx->isc_capenable;
426 } /* ice_setup_scctx */
427 
428 /**
429  * ice_if_attach_pre - Early device attach logic
430  * @ctx: the iflib context structure
431  *
432  * Called by iflib during the attach process. Earliest main driver entry
433  * point which performs necessary hardware and driver initialization. Called
434  * before the Tx and Rx queues are allocated.
435  */
436 static int
437 ice_if_attach_pre(if_ctx_t ctx)
438 {
439 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
440 	enum ice_fw_modes fw_mode;
441 	enum ice_status status;
442 	if_softc_ctx_t scctx;
443 	struct ice_hw *hw;
444 	device_t dev;
445 	int err;
446 
447 	device_printf(iflib_get_dev(ctx), "Loading the iflib ice driver\n");
448 
449 	sc->ctx = ctx;
450 	sc->media = iflib_get_media(ctx);
451 	sc->sctx = iflib_get_sctx(ctx);
452 	sc->iflib_ctx_lock = iflib_ctx_lock_get(ctx);
453 
454 	dev = sc->dev = iflib_get_dev(ctx);
455 	scctx = sc->scctx = iflib_get_softc_ctx(ctx);
456 
457 	hw = &sc->hw;
458 	hw->back = sc;
459 
460 	snprintf(sc->admin_mtx_name, sizeof(sc->admin_mtx_name),
461 		 "%s:admin", device_get_nameunit(dev));
462 	mtx_init(&sc->admin_mtx, sc->admin_mtx_name, NULL, MTX_DEF);
463 	callout_init_mtx(&sc->admin_timer, &sc->admin_mtx, 0);
464 
465 	ASSERT_CTX_LOCKED(sc);
466 
467 	if (ice_pci_mapping(sc)) {
468 		err = (ENXIO);
469 		goto destroy_admin_timer;
470 	}
471 
472 	/* Save off the PCI information */
473 	ice_save_pci_info(hw, dev);
474 
475 	/* create tunables as early as possible */
476 	ice_add_device_tunables(sc);
477 
478 	/* Setup ControlQ lengths */
479 	ice_set_ctrlq_len(hw);
480 
481 	fw_mode = ice_get_fw_mode(hw);
482 	if (fw_mode == ICE_FW_MODE_REC) {
483 		device_printf(dev, "Firmware recovery mode detected. Limiting functionality. Refer to Intel(R) Ethernet Adapters and Devices User Guide for details on firmware recovery mode.\n");
484 
485 		err = ice_attach_pre_recovery_mode(sc);
486 		if (err)
487 			goto free_pci_mapping;
488 
489 		return (0);
490 	}
491 
492 	/* Initialize the hw data structure */
493 	status = ice_init_hw(hw);
494 	if (status) {
495 		if (status == ICE_ERR_FW_API_VER) {
496 			/* Enter recovery mode, so that the driver remains
497 			 * loaded. This way, if the system administrator
498 			 * cannot update the driver, they may still attempt to
499 			 * downgrade the NVM.
500 			 */
501 			err = ice_attach_pre_recovery_mode(sc);
502 			if (err)
503 				goto free_pci_mapping;
504 
505 			return (0);
506 		} else {
507 			err = EIO;
508 			device_printf(dev, "Unable to initialize hw, err %s aq_err %s\n",
509 				      ice_status_str(status),
510 				      ice_aq_str(hw->adminq.sq_last_status));
511 		}
512 		goto free_pci_mapping;
513 	}
514 
515 	/* Notify firmware of the device driver version */
516 	err = ice_send_version(sc);
517 	if (err)
518 		goto deinit_hw;
519 
520 	ice_load_pkg_file(sc);
521 
522 	err = ice_init_link_events(sc);
523 	if (err) {
524 		device_printf(dev, "ice_init_link_events failed: %s\n",
525 			      ice_err_str(err));
526 		goto deinit_hw;
527 	}
528 
529 	ice_print_nvm_version(sc);
530 
531 	ice_init_device_features(sc);
532 
533 	/* Setup the MAC address */
534 	iflib_set_mac(ctx, hw->port_info->mac.lan_addr);
535 
536 	/* Setup the iflib softc context structure */
537 	ice_setup_scctx(sc);
538 
539 	/* Initialize the Tx queue manager */
540 	err = ice_resmgr_init(&sc->tx_qmgr, hw->func_caps.common_cap.num_txq);
541 	if (err) {
542 		device_printf(dev, "Unable to initialize Tx queue manager: %s\n",
543 			      ice_err_str(err));
544 		goto deinit_hw;
545 	}
546 
547 	/* Initialize the Rx queue manager */
548 	err = ice_resmgr_init(&sc->rx_qmgr, hw->func_caps.common_cap.num_rxq);
549 	if (err) {
550 		device_printf(dev, "Unable to initialize Rx queue manager: %s\n",
551 			      ice_err_str(err));
552 		goto free_tx_qmgr;
553 	}
554 
555 	/* Initialize the interrupt resource manager */
556 	err = ice_alloc_intr_tracking(sc);
557 	if (err)
558 		/* Errors are already printed */
559 		goto free_rx_qmgr;
560 
561 	/* Determine maximum number of VSIs we'll prepare for */
562 	sc->num_available_vsi = min(ICE_MAX_VSI_AVAILABLE,
563 				    hw->func_caps.guar_num_vsi);
564 
565 	if (!sc->num_available_vsi) {
566 		err = EIO;
567 		device_printf(dev, "No VSIs allocated to host\n");
568 		goto free_intr_tracking;
569 	}
570 
571 	/* Allocate storage for the VSI pointers */
572 	sc->all_vsi = (struct ice_vsi **)
573 		malloc(sizeof(struct ice_vsi *) * sc->num_available_vsi,
574 		       M_ICE, M_WAITOK | M_ZERO);
575 	if (!sc->all_vsi) {
576 		err = ENOMEM;
577 		device_printf(dev, "Unable to allocate VSI array\n");
578 		goto free_intr_tracking;
579 	}
580 
581 	/*
582 	 * Prepare the statically allocated primary PF VSI in the softc
583 	 * structure. Other VSIs will be dynamically allocated as needed.
584 	 */
585 	ice_setup_pf_vsi(sc);
586 
587 	err = ice_alloc_vsi_qmap(&sc->pf_vsi, scctx->isc_ntxqsets_max,
588 	    scctx->isc_nrxqsets_max);
589 	if (err) {
590 		device_printf(dev, "Unable to allocate VSI Queue maps\n");
591 		goto free_main_vsi;
592 	}
593 
594 	/* Allocate MSI-X vectors (due to isc_flags IFLIB_SKIP_MSIX) */
595 	err = ice_allocate_msix(sc);
596 	if (err)
597 		goto free_main_vsi;
598 
599 	return 0;
600 
601 free_main_vsi:
602 	/* ice_release_vsi will free the queue maps if they were allocated */
603 	ice_release_vsi(&sc->pf_vsi);
604 	free(sc->all_vsi, M_ICE);
605 	sc->all_vsi = NULL;
606 free_intr_tracking:
607 	ice_free_intr_tracking(sc);
608 free_rx_qmgr:
609 	ice_resmgr_destroy(&sc->rx_qmgr);
610 free_tx_qmgr:
611 	ice_resmgr_destroy(&sc->tx_qmgr);
612 deinit_hw:
613 	ice_deinit_hw(hw);
614 free_pci_mapping:
615 	ice_free_pci_mapping(sc);
616 destroy_admin_timer:
617 	mtx_lock(&sc->admin_mtx);
618 	callout_stop(&sc->admin_timer);
619 	mtx_unlock(&sc->admin_mtx);
620 	mtx_destroy(&sc->admin_mtx);
621 	return err;
622 } /* ice_if_attach_pre */
623 
624 /**
625  * ice_attach_pre_recovery_mode - Limited driver attach_pre for FW recovery
626  * @sc: the device private softc
627  *
628  * Loads the device driver in limited Firmware Recovery mode, intended to
629  * allow users to update the firmware to attempt to recover the device.
630  *
631  * @remark We may enter recovery mode in case either (a) the firmware is
632  * detected to be in an invalid state and must be re-programmed, or (b) the
633  * driver detects that the loaded firmware has a non-compatible API version
634  * that the driver cannot operate with.
635  */
636 static int
637 ice_attach_pre_recovery_mode(struct ice_softc *sc)
638 {
639 	ice_set_state(&sc->state, ICE_STATE_RECOVERY_MODE);
640 
641 	/* Setup the iflib softc context */
642 	ice_setup_scctx(sc);
643 
644 	/* Setup the PF VSI back pointer */
645 	sc->pf_vsi.sc = sc;
646 
647 	/*
648 	 * We still need to allocate MSI-X vectors since we need one vector to
649 	 * run the administrative admin interrupt
650 	 */
651 	return ice_allocate_msix(sc);
652 }
653 
654 /**
655  * ice_update_link_status - notify OS of link state change
656  * @sc: device private softc structure
657  * @update_media: true if we should update media even if link didn't change
658  *
659  * Called to notify iflib core of link status changes. Should be called once
660  * during attach_post, and whenever link status changes during runtime.
661  *
662  * This call only updates the currently supported media types if the link
663  * status changed, or if update_media is set to true.
664  */
665 static void
666 ice_update_link_status(struct ice_softc *sc, bool update_media)
667 {
668 	struct ice_hw *hw = &sc->hw;
669 	enum ice_status status;
670 
671 	/* Never report link up when in recovery mode */
672 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
673 		return;
674 
675 	/* Report link status to iflib only once each time it changes */
676 	if (!ice_testandset_state(&sc->state, ICE_STATE_LINK_STATUS_REPORTED)) {
677 		if (sc->link_up) { /* link is up */
678 			uint64_t baudrate = ice_aq_speed_to_rate(sc->hw.port_info);
679 
680 			ice_set_default_local_lldp_mib(sc);
681 
682 			iflib_link_state_change(sc->ctx, LINK_STATE_UP, baudrate);
683 
684 			ice_link_up_msg(sc);
685 
686 			update_media = true;
687 		} else { /* link is down */
688 			iflib_link_state_change(sc->ctx, LINK_STATE_DOWN, 0);
689 
690 			update_media = true;
691 		}
692 	}
693 
694 	/* Update the supported media types */
695 	if (update_media) {
696 		status = ice_add_media_types(sc, sc->media);
697 		if (status)
698 			device_printf(sc->dev, "Error adding device media types: %s aq_err %s\n",
699 				      ice_status_str(status),
700 				      ice_aq_str(hw->adminq.sq_last_status));
701 	}
702 
703 	/* TODO: notify VFs of link state change */
704 }
705 
706 /**
707  * ice_if_attach_post - Late device attach logic
708  * @ctx: the iflib context structure
709  *
710  * Called by iflib to finish up attaching the device. Performs any attach
711  * logic which must wait until after the Tx and Rx queues have been
712  * allocated.
713  */
714 static int
715 ice_if_attach_post(if_ctx_t ctx)
716 {
717 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
718 	if_t ifp = iflib_get_ifp(ctx);
719 	int err;
720 
721 	ASSERT_CTX_LOCKED(sc);
722 
723 	/* We don't yet support loading if MSI-X is not supported */
724 	if (sc->scctx->isc_intr != IFLIB_INTR_MSIX) {
725 		device_printf(sc->dev, "The ice driver does not support loading without MSI-X\n");
726 		return (ENOTSUP);
727 	}
728 
729 	/* The ifnet structure hasn't yet been initialized when the attach_pre
730 	 * handler is called, so wait until attach_post to setup the
731 	 * isc_max_frame_size.
732 	 */
733 
734 	sc->ifp = ifp;
735 	sc->scctx->isc_max_frame_size = ifp->if_mtu +
736 		ETHER_HDR_LEN + ETHER_CRC_LEN + ETHER_VLAN_ENCAP_LEN;
737 
738 	/*
739 	 * If we are in recovery mode, only perform a limited subset of
740 	 * initialization to support NVM recovery.
741 	 */
742 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE)) {
743 		ice_attach_post_recovery_mode(sc);
744 		return (0);
745 	}
746 
747 	sc->pf_vsi.max_frame_size = sc->scctx->isc_max_frame_size;
748 
749 	err = ice_initialize_vsi(&sc->pf_vsi);
750 	if (err) {
751 		device_printf(sc->dev, "Unable to initialize Main VSI: %s\n",
752 			      ice_err_str(err));
753 		return err;
754 	}
755 
756 	/* Configure the main PF VSI for RSS */
757 	err = ice_config_rss(&sc->pf_vsi);
758 	if (err) {
759 		device_printf(sc->dev,
760 			      "Unable to configure RSS for the main VSI, err %s\n",
761 			      ice_err_str(err));
762 		return err;
763 	}
764 
765 	/* Configure switch to drop transmitted LLDP and PAUSE frames */
766 	err = ice_cfg_pf_ethertype_filters(sc);
767 	if (err)
768 		return err;
769 
770 	ice_get_and_print_bus_info(sc);
771 
772 	ice_set_link_management_mode(sc);
773 
774 	ice_init_saved_phy_cfg(sc);
775 
776 	ice_add_device_sysctls(sc);
777 
778 	/* Get DCBX/LLDP state and start DCBX agent */
779 	ice_init_dcb_setup(sc);
780 
781 	/* Setup link configuration parameters */
782 	ice_init_link_configuration(sc);
783 	ice_update_link_status(sc, true);
784 
785 	/* Configure interrupt causes for the administrative interrupt */
786 	ice_configure_misc_interrupts(sc);
787 
788 	/* Enable ITR 0 right away, so that we can handle admin interrupts */
789 	ice_enable_intr(&sc->hw, sc->irqvs[0].me);
790 
791 	/* Start the admin timer */
792 	mtx_lock(&sc->admin_mtx);
793 	callout_reset(&sc->admin_timer, hz/2, ice_admin_timer, sc);
794 	mtx_unlock(&sc->admin_mtx);
795 
796 	return 0;
797 } /* ice_if_attach_post */
798 
799 /**
800  * ice_attach_post_recovery_mode - Limited driver attach_post for FW recovery
801  * @sc: the device private softc
802  *
803  * Performs minimal work to prepare the driver to recover an NVM in case the
804  * firmware is in recovery mode.
805  */
806 static void
807 ice_attach_post_recovery_mode(struct ice_softc *sc)
808 {
809 	/* Configure interrupt causes for the administrative interrupt */
810 	ice_configure_misc_interrupts(sc);
811 
812 	/* Enable ITR 0 right away, so that we can handle admin interrupts */
813 	ice_enable_intr(&sc->hw, sc->irqvs[0].me);
814 
815 	/* Start the admin timer */
816 	mtx_lock(&sc->admin_mtx);
817 	callout_reset(&sc->admin_timer, hz/2, ice_admin_timer, sc);
818 	mtx_unlock(&sc->admin_mtx);
819 }
820 
821 /**
822  * ice_free_irqvs - Free IRQ vector memory
823  * @sc: the device private softc structure
824  *
825  * Free IRQ vector memory allocated during ice_if_msix_intr_assign.
826  */
827 static void
828 ice_free_irqvs(struct ice_softc *sc)
829 {
830 	struct ice_vsi *vsi = &sc->pf_vsi;
831 	if_ctx_t ctx = sc->ctx;
832 	int i;
833 
834 	/* If the irqvs array is NULL, then there are no vectors to free */
835 	if (sc->irqvs == NULL)
836 		return;
837 
838 	/* Free the IRQ vectors */
839 	for (i = 0; i < sc->num_irq_vectors; i++)
840 		iflib_irq_free(ctx, &sc->irqvs[i].irq);
841 
842 	/* Clear the irqv pointers */
843 	for (i = 0; i < vsi->num_rx_queues; i++)
844 		vsi->rx_queues[i].irqv = NULL;
845 
846 	for (i = 0; i < vsi->num_tx_queues; i++)
847 		vsi->tx_queues[i].irqv = NULL;
848 
849 	/* Release the vector array memory */
850 	free(sc->irqvs, M_ICE);
851 	sc->irqvs = NULL;
852 	sc->num_irq_vectors = 0;
853 }
854 
855 /**
856  * ice_if_detach - Device driver detach logic
857  * @ctx: iflib context structure
858  *
859  * Perform device shutdown logic to detach the device driver.
860  *
861  * Note that there is no guarantee of the ordering of ice_if_queues_free() and
862  * ice_if_detach(). It is possible for the functions to be called in either
863  * order, and they must not assume to have a strict ordering.
864  */
865 static int
866 ice_if_detach(if_ctx_t ctx)
867 {
868 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
869 	struct ice_vsi *vsi = &sc->pf_vsi;
870 	int i;
871 
872 	ASSERT_CTX_LOCKED(sc);
873 
874 	/* Indicate that we're detaching */
875 	ice_set_state(&sc->state, ICE_STATE_DETACHING);
876 
877 	/* Stop the admin timer */
878 	mtx_lock(&sc->admin_mtx);
879 	callout_stop(&sc->admin_timer);
880 	mtx_unlock(&sc->admin_mtx);
881 	mtx_destroy(&sc->admin_mtx);
882 
883 	/* Free allocated media types */
884 	ifmedia_removeall(sc->media);
885 
886 	/* Free the Tx and Rx sysctl contexts, and assign NULL to the node
887 	 * pointers. Note, the calls here and those in ice_if_queues_free()
888 	 * are *BOTH* necessary, as we cannot guarantee which path will be
889 	 * run first
890 	 */
891 	ice_vsi_del_txqs_ctx(vsi);
892 	ice_vsi_del_rxqs_ctx(vsi);
893 
894 	/* Release MSI-X resources */
895 	ice_free_irqvs(sc);
896 
897 	for (i = 0; i < sc->num_available_vsi; i++) {
898 		if (sc->all_vsi[i])
899 			ice_release_vsi(sc->all_vsi[i]);
900 	}
901 
902 	if (sc->all_vsi) {
903 		free(sc->all_vsi, M_ICE);
904 		sc->all_vsi = NULL;
905 	}
906 
907 	/* Release MSI-X memory */
908 	pci_release_msi(sc->dev);
909 
910 	if (sc->msix_table != NULL) {
911 		bus_release_resource(sc->dev, SYS_RES_MEMORY,
912 				     rman_get_rid(sc->msix_table),
913 				     sc->msix_table);
914 		sc->msix_table = NULL;
915 	}
916 
917 	ice_free_intr_tracking(sc);
918 
919 	/* Destroy the queue managers */
920 	ice_resmgr_destroy(&sc->tx_qmgr);
921 	ice_resmgr_destroy(&sc->rx_qmgr);
922 
923 	if (!ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
924 		ice_deinit_hw(&sc->hw);
925 
926 	ice_free_pci_mapping(sc);
927 
928 	return 0;
929 } /* ice_if_detach */
930 
931 /**
932  * ice_if_tx_queues_alloc - Allocate Tx queue memory
933  * @ctx: iflib context structure
934  * @vaddrs: virtual addresses for the queue memory
935  * @paddrs: physical addresses for the queue memory
936  * @ntxqs: the number of Tx queues per set (should always be 1)
937  * @ntxqsets: the number of Tx queue sets to allocate
938  *
939  * Called by iflib to allocate Tx queues for the device. Allocates driver
940  * memory to track each queue, the status arrays used for descriptor
941  * status reporting, and Tx queue sysctls.
942  */
943 static int
944 ice_if_tx_queues_alloc(if_ctx_t ctx, caddr_t *vaddrs, uint64_t *paddrs,
945 		       int __invariant_only ntxqs, int ntxqsets)
946 {
947 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
948 	struct ice_vsi *vsi = &sc->pf_vsi;
949 	struct ice_tx_queue *txq;
950 	int err, i, j;
951 
952 	MPASS(ntxqs == 1);
953 	MPASS(sc->scctx->isc_ntxd[0] <= ICE_MAX_DESC_COUNT);
954 	ASSERT_CTX_LOCKED(sc);
955 
956 	/* Do not bother allocating queues if we're in recovery mode */
957 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
958 		return (0);
959 
960 	/* Allocate queue structure memory */
961 	if (!(vsi->tx_queues =
962 	      (struct ice_tx_queue *) malloc(sizeof(struct ice_tx_queue) * ntxqsets, M_ICE, M_WAITOK | M_ZERO))) {
963 		device_printf(sc->dev, "Unable to allocate Tx queue memory\n");
964 		return (ENOMEM);
965 	}
966 
967 	/* Allocate report status arrays */
968 	for (i = 0, txq = vsi->tx_queues; i < ntxqsets; i++, txq++) {
969 		if (!(txq->tx_rsq =
970 		      (uint16_t *) malloc(sizeof(uint16_t) * sc->scctx->isc_ntxd[0], M_ICE, M_WAITOK))) {
971 			device_printf(sc->dev, "Unable to allocate tx_rsq memory\n");
972 			err = ENOMEM;
973 			goto free_tx_queues;
974 		}
975 		/* Initialize report status array */
976 		for (j = 0; j < sc->scctx->isc_ntxd[0]; j++)
977 			txq->tx_rsq[j] = QIDX_INVALID;
978 	}
979 
980 	/* Assign queues from PF space to the main VSI */
981 	err = ice_resmgr_assign_contiguous(&sc->tx_qmgr, vsi->tx_qmap, ntxqsets);
982 	if (err) {
983 		device_printf(sc->dev, "Unable to assign PF queues: %s\n",
984 			      ice_err_str(err));
985 		goto free_tx_queues;
986 	}
987 	vsi->qmap_type = ICE_RESMGR_ALLOC_CONTIGUOUS;
988 
989 	/* Add Tx queue sysctls context */
990 	ice_vsi_add_txqs_ctx(vsi);
991 
992 	for (i = 0, txq = vsi->tx_queues; i < ntxqsets; i++, txq++) {
993 		txq->me = i;
994 		txq->vsi = vsi;
995 
996 		/* store the queue size for easier access */
997 		txq->desc_count = sc->scctx->isc_ntxd[0];
998 
999 		/* get the virtual and physical address of the hardware queues */
1000 		txq->tail = QTX_COMM_DBELL(vsi->tx_qmap[i]);
1001 		txq->tx_base = (struct ice_tx_desc *)vaddrs[i];
1002 		txq->tx_paddr = paddrs[i];
1003 
1004 		ice_add_txq_sysctls(txq);
1005 	}
1006 
1007 	vsi->num_tx_queues = ntxqsets;
1008 
1009 	return (0);
1010 
1011 free_tx_queues:
1012 	for (i = 0, txq = vsi->tx_queues; i < ntxqsets; i++, txq++) {
1013 		if (txq->tx_rsq != NULL) {
1014 			free(txq->tx_rsq, M_ICE);
1015 			txq->tx_rsq = NULL;
1016 		}
1017 	}
1018 	free(vsi->tx_queues, M_ICE);
1019 	vsi->tx_queues = NULL;
1020 	return err;
1021 }
1022 
1023 /**
1024  * ice_if_rx_queues_alloc - Allocate Rx queue memory
1025  * @ctx: iflib context structure
1026  * @vaddrs: virtual addresses for the queue memory
1027  * @paddrs: physical addresses for the queue memory
1028  * @nrxqs: number of Rx queues per set (should always be 1)
1029  * @nrxqsets: number of Rx queue sets to allocate
1030  *
1031  * Called by iflib to allocate Rx queues for the device. Allocates driver
1032  * memory to track each queue, as well as sets up the Rx queue sysctls.
1033  */
1034 static int
1035 ice_if_rx_queues_alloc(if_ctx_t ctx, caddr_t *vaddrs, uint64_t *paddrs,
1036 		       int __invariant_only nrxqs, int nrxqsets)
1037 {
1038 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1039 	struct ice_vsi *vsi = &sc->pf_vsi;
1040 	struct ice_rx_queue *rxq;
1041 	int err, i;
1042 
1043 	MPASS(nrxqs == 1);
1044 	MPASS(sc->scctx->isc_nrxd[0] <= ICE_MAX_DESC_COUNT);
1045 	ASSERT_CTX_LOCKED(sc);
1046 
1047 	/* Do not bother allocating queues if we're in recovery mode */
1048 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1049 		return (0);
1050 
1051 	/* Allocate queue structure memory */
1052 	if (!(vsi->rx_queues =
1053 	      (struct ice_rx_queue *) malloc(sizeof(struct ice_rx_queue) * nrxqsets, M_ICE, M_WAITOK | M_ZERO))) {
1054 		device_printf(sc->dev, "Unable to allocate Rx queue memory\n");
1055 		return (ENOMEM);
1056 	}
1057 
1058 	/* Assign queues from PF space to the main VSI */
1059 	err = ice_resmgr_assign_contiguous(&sc->rx_qmgr, vsi->rx_qmap, nrxqsets);
1060 	if (err) {
1061 		device_printf(sc->dev, "Unable to assign PF queues: %s\n",
1062 			      ice_err_str(err));
1063 		goto free_rx_queues;
1064 	}
1065 	vsi->qmap_type = ICE_RESMGR_ALLOC_CONTIGUOUS;
1066 
1067 	/* Add Rx queue sysctls context */
1068 	ice_vsi_add_rxqs_ctx(vsi);
1069 
1070 	for (i = 0, rxq = vsi->rx_queues; i < nrxqsets; i++, rxq++) {
1071 		rxq->me = i;
1072 		rxq->vsi = vsi;
1073 
1074 		/* store the queue size for easier access */
1075 		rxq->desc_count = sc->scctx->isc_nrxd[0];
1076 
1077 		/* get the virtual and physical address of the hardware queues */
1078 		rxq->tail = QRX_TAIL(vsi->rx_qmap[i]);
1079 		rxq->rx_base = (union ice_32b_rx_flex_desc *)vaddrs[i];
1080 		rxq->rx_paddr = paddrs[i];
1081 
1082 		ice_add_rxq_sysctls(rxq);
1083 	}
1084 
1085 	vsi->num_rx_queues = nrxqsets;
1086 
1087 	return (0);
1088 
1089 free_rx_queues:
1090 	free(vsi->rx_queues, M_ICE);
1091 	vsi->rx_queues = NULL;
1092 	return err;
1093 }
1094 
1095 /**
1096  * ice_if_queues_free - Free queue memory
1097  * @ctx: the iflib context structure
1098  *
1099  * Free queue memory allocated by ice_if_tx_queues_alloc() and
1100  * ice_if_rx_queues_alloc().
1101  *
1102  * There is no guarantee that ice_if_queues_free() and ice_if_detach() will be
1103  * called in the same order. It's possible for ice_if_queues_free() to be
1104  * called prior to ice_if_detach(), and vice versa.
1105  *
1106  * For this reason, the main VSI is a static member of the ice_softc, which is
1107  * not free'd until after iflib finishes calling both of these functions.
1108  *
1109  * Thus, care must be taken in how we manage the memory being freed by this
1110  * function, and in what tasks it can and must perform.
1111  */
1112 static void
1113 ice_if_queues_free(if_ctx_t ctx)
1114 {
1115 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1116 	struct ice_vsi *vsi = &sc->pf_vsi;
1117 	struct ice_tx_queue *txq;
1118 	int i;
1119 
1120 	/* Free the Tx and Rx sysctl contexts, and assign NULL to the node
1121 	 * pointers. Note, the calls here and those in ice_if_detach()
1122 	 * are *BOTH* necessary, as we cannot guarantee which path will be
1123 	 * run first
1124 	 */
1125 	ice_vsi_del_txqs_ctx(vsi);
1126 	ice_vsi_del_rxqs_ctx(vsi);
1127 
1128 	/* Release MSI-X IRQ vectors, if not yet released in ice_if_detach */
1129 	ice_free_irqvs(sc);
1130 
1131 	if (vsi->tx_queues != NULL) {
1132 		/* free the tx_rsq arrays */
1133 		for (i = 0, txq = vsi->tx_queues; i < vsi->num_tx_queues; i++, txq++) {
1134 			if (txq->tx_rsq != NULL) {
1135 				free(txq->tx_rsq, M_ICE);
1136 				txq->tx_rsq = NULL;
1137 			}
1138 		}
1139 		free(vsi->tx_queues, M_ICE);
1140 		vsi->tx_queues = NULL;
1141 		vsi->num_tx_queues = 0;
1142 	}
1143 	if (vsi->rx_queues != NULL) {
1144 		free(vsi->rx_queues, M_ICE);
1145 		vsi->rx_queues = NULL;
1146 		vsi->num_rx_queues = 0;
1147 	}
1148 }
1149 
1150 /**
1151  * ice_msix_que - Fast interrupt handler for MSI-X receive queues
1152  * @arg: The Rx queue memory
1153  *
1154  * Interrupt filter function for iflib MSI-X interrupts. Called by iflib when
1155  * an MSI-X interrupt for a given queue is triggered. Currently this just asks
1156  * iflib to schedule the main Rx thread.
1157  */
1158 static int
1159 ice_msix_que(void *arg)
1160 {
1161 	struct ice_rx_queue __unused *rxq = (struct ice_rx_queue *)arg;
1162 
1163 	/* TODO: dynamic ITR algorithm?? */
1164 
1165 	return (FILTER_SCHEDULE_THREAD);
1166 }
1167 
1168 /**
1169  * ice_msix_admin - Fast interrupt handler for MSI-X admin interrupt
1170  * @arg: pointer to device softc memory
1171  *
1172  * Called by iflib when an administrative interrupt occurs. Should perform any
1173  * fast logic for handling the interrupt cause, and then indicate whether the
1174  * admin task needs to be queued.
1175  */
1176 static int
1177 ice_msix_admin(void *arg)
1178 {
1179 	struct ice_softc *sc = (struct ice_softc *)arg;
1180 	struct ice_hw *hw = &sc->hw;
1181 	device_t dev = sc->dev;
1182 	u32 oicr;
1183 
1184 	/* There is no safe way to modify the enabled miscellaneous causes of
1185 	 * the OICR vector at runtime, as doing so would be prone to race
1186 	 * conditions. Reading PFINT_OICR will unmask the associated interrupt
1187 	 * causes and allow future interrupts to occur. The admin interrupt
1188 	 * vector will not be re-enabled until after we exit this function,
1189 	 * but any delayed tasks must be resilient against possible "late
1190 	 * arrival" interrupts that occur while we're already handling the
1191 	 * task. This is done by using state bits and serializing these
1192 	 * delayed tasks via the admin status task function.
1193 	 */
1194 	oicr = rd32(hw, PFINT_OICR);
1195 
1196 	/* Processing multiple controlq interrupts on a single vector does not
1197 	 * provide an indication of which controlq triggered the interrupt.
1198 	 * We might try reading the INTEVENT bit of the respective PFINT_*_CTL
1199 	 * registers. However, the INTEVENT bit is not guaranteed to be set as
1200 	 * it gets automatically cleared when the hardware acknowledges the
1201 	 * interrupt.
1202 	 *
1203 	 * This means we don't really have a good indication of whether or
1204 	 * which controlq triggered this interrupt. We'll just notify the
1205 	 * admin task that it should check all the controlqs.
1206 	 */
1207 	ice_set_state(&sc->state, ICE_STATE_CONTROLQ_EVENT_PENDING);
1208 
1209 	if (oicr & PFINT_OICR_VFLR_M) {
1210 		ice_set_state(&sc->state, ICE_STATE_VFLR_PENDING);
1211 	}
1212 
1213 	if (oicr & PFINT_OICR_MAL_DETECT_M) {
1214 		ice_set_state(&sc->state, ICE_STATE_MDD_PENDING);
1215 	}
1216 
1217 	if (oicr & PFINT_OICR_GRST_M) {
1218 		u32 reset;
1219 
1220 		reset = (rd32(hw, GLGEN_RSTAT) & GLGEN_RSTAT_RESET_TYPE_M) >>
1221 			GLGEN_RSTAT_RESET_TYPE_S;
1222 
1223 		if (reset == ICE_RESET_CORER)
1224 			sc->soft_stats.corer_count++;
1225 		else if (reset == ICE_RESET_GLOBR)
1226 			sc->soft_stats.globr_count++;
1227 		else
1228 			sc->soft_stats.empr_count++;
1229 
1230 		/* There are a couple of bits at play for handling resets.
1231 		 * First, the ICE_STATE_RESET_OICR_RECV bit is used to
1232 		 * indicate that the driver has received an OICR with a reset
1233 		 * bit active, indicating that a CORER/GLOBR/EMPR is about to
1234 		 * happen. Second, we set hw->reset_ongoing to indicate that
1235 		 * the hardware is in reset. We will set this back to false as
1236 		 * soon as the driver has determined that the hardware is out
1237 		 * of reset.
1238 		 *
1239 		 * If the driver wishes to trigger a reqest, it can set one of
1240 		 * the ICE_STATE_RESET_*_REQ bits, which will trigger the
1241 		 * correct type of reset.
1242 		 */
1243 		if (!ice_testandset_state(&sc->state, ICE_STATE_RESET_OICR_RECV))
1244 			hw->reset_ongoing = true;
1245 	}
1246 
1247 	if (oicr & PFINT_OICR_ECC_ERR_M) {
1248 		device_printf(dev, "ECC Error detected!\n");
1249 		ice_set_state(&sc->state, ICE_STATE_RESET_PFR_REQ);
1250 	}
1251 
1252 	if (oicr & PFINT_OICR_PE_CRITERR_M) {
1253 		device_printf(dev, "Critical Protocol Engine Error detected!\n");
1254 		ice_set_state(&sc->state, ICE_STATE_RESET_PFR_REQ);
1255 	}
1256 
1257 	if (oicr & PFINT_OICR_PCI_EXCEPTION_M) {
1258 		device_printf(dev, "PCI Exception detected!\n");
1259 		ice_set_state(&sc->state, ICE_STATE_RESET_PFR_REQ);
1260 	}
1261 
1262 	if (oicr & PFINT_OICR_HMC_ERR_M) {
1263 		/* Log the HMC errors, but don't disable the interrupt cause */
1264 		ice_log_hmc_error(hw, dev);
1265 	}
1266 
1267 	return (FILTER_SCHEDULE_THREAD);
1268 }
1269 
1270 /**
1271  * ice_allocate_msix - Allocate MSI-X vectors for the interface
1272  * @sc: the device private softc
1273  *
1274  * Map the MSI-X bar, and then request MSI-X vectors in a two-stage process.
1275  *
1276  * First, determine a suitable total number of vectors based on the number
1277  * of CPUs, RSS buckets, the administrative vector, and other demands such as
1278  * RDMA.
1279  *
1280  * Request the desired amount of vectors, and see how many we obtain. If we
1281  * don't obtain as many as desired, reduce the demands by lowering the number
1282  * of requested queues or reducing the demand from other features such as
1283  * RDMA.
1284  *
1285  * @remark This function is required because the driver sets the
1286  * IFLIB_SKIP_MSIX flag indicating that the driver will manage MSI-X vectors
1287  * manually.
1288  *
1289  * @remark This driver will only use MSI-X vectors. If this is not possible,
1290  * neither MSI or legacy interrupts will be tried.
1291  *
1292  * @post on success this function must set the following scctx parameters:
1293  * isc_vectors, isc_nrxqsets, isc_ntxqsets, and isc_intr.
1294  *
1295  * @returns zero on success or an error code on failure.
1296  */
1297 static int
1298 ice_allocate_msix(struct ice_softc *sc)
1299 {
1300 	bool iflib_override_queue_count = false;
1301 	if_softc_ctx_t scctx = sc->scctx;
1302 	device_t dev = sc->dev;
1303 	cpuset_t cpus;
1304 	int bar, queues, vectors, requested;
1305 	int err = 0;
1306 
1307 	/* Allocate the MSI-X bar */
1308 	bar = scctx->isc_msix_bar;
1309 	sc->msix_table = bus_alloc_resource_any(dev, SYS_RES_MEMORY, &bar, RF_ACTIVE);
1310 	if (!sc->msix_table) {
1311 		device_printf(dev, "Unable to map MSI-X table\n");
1312 		return (ENOMEM);
1313 	}
1314 
1315 	/* Check if the iflib queue count sysctls have been set */
1316 	if (sc->ifc_sysctl_ntxqs || sc->ifc_sysctl_nrxqs)
1317 		iflib_override_queue_count = true;
1318 
1319 	err = bus_get_cpus(dev, INTR_CPUS, sizeof(cpus), &cpus);
1320 	if (err) {
1321 		device_printf(dev, "%s: Unable to fetch the CPU list: %s\n",
1322 			      __func__, ice_err_str(err));
1323 		CPU_COPY(&all_cpus, &cpus);
1324 	}
1325 
1326 	/* Attempt to mimic behavior of iflib_msix_init */
1327 	if (iflib_override_queue_count) {
1328 		/*
1329 		 * If the override sysctls have been set, limit the queues to
1330 		 * the number of logical CPUs.
1331 		 */
1332 		queues = mp_ncpus;
1333 	} else {
1334 		/*
1335 		 * Otherwise, limit the queue count to the CPUs associated
1336 		 * with the NUMA node the device is associated with.
1337 		 */
1338 		queues = CPU_COUNT(&cpus);
1339 	}
1340 
1341 	/* Clamp to the number of RSS buckets */
1342 	queues = imin(queues, rss_getnumbuckets());
1343 
1344 	/*
1345 	 * Clamp the number of queue pairs to the minimum of the requested Tx
1346 	 * and Rx queues.
1347 	 */
1348 	queues = imin(queues, sc->ifc_sysctl_ntxqs ?: scctx->isc_ntxqsets);
1349 	queues = imin(queues, sc->ifc_sysctl_nrxqs ?: scctx->isc_nrxqsets);
1350 
1351 	/*
1352 	 * Determine the number of vectors to request. Note that we also need
1353 	 * to allocate one vector for administrative tasks.
1354 	 */
1355 	requested = queues + 1;
1356 
1357 	vectors = requested;
1358 
1359 	err = pci_alloc_msix(dev, &vectors);
1360 	if (err) {
1361 		device_printf(dev, "Failed to allocate %d MSI-X vectors, err %s\n",
1362 			      vectors, ice_err_str(err));
1363 		goto err_free_msix_table;
1364 	}
1365 
1366 	/* If we don't receive enough vectors, reduce demands */
1367 	if (vectors < requested) {
1368 		int diff = requested - vectors;
1369 
1370 		device_printf(dev, "Requested %d MSI-X vectors, but got only %d\n",
1371 			      requested, vectors);
1372 
1373 		/*
1374 		 * If we still have a difference, we need to reduce the number
1375 		 * of queue pairs.
1376 		 *
1377 		 * However, we still need at least one vector for the admin
1378 		 * interrupt and one queue pair.
1379 		 */
1380 		if (queues <= diff) {
1381 			device_printf(dev, "Unable to allocate sufficient MSI-X vectors\n");
1382 			err = (ERANGE);
1383 			goto err_pci_release_msi;
1384 		}
1385 
1386 		queues -= diff;
1387 	}
1388 
1389 	device_printf(dev, "Using %d Tx and Rx queues\n", queues);
1390 	device_printf(dev, "Using MSI-X interrupts with %d vectors\n",
1391 		      vectors);
1392 
1393 	scctx->isc_vectors = vectors;
1394 	scctx->isc_nrxqsets = queues;
1395 	scctx->isc_ntxqsets = queues;
1396 	scctx->isc_intr = IFLIB_INTR_MSIX;
1397 
1398 	/* Interrupt allocation tracking isn't required in recovery mode,
1399 	 * since neither RDMA nor VFs are enabled.
1400 	 */
1401 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1402 		return (0);
1403 
1404 	/* Keep track of which interrupt indices are being used for what */
1405 	sc->lan_vectors = vectors;
1406 	err = ice_resmgr_assign_contiguous(&sc->imgr, sc->pf_imap, sc->lan_vectors);
1407 	if (err) {
1408 		device_printf(dev, "Unable to assign PF interrupt mapping: %s\n",
1409 			      ice_err_str(err));
1410 		goto err_pci_release_msi;
1411 	}
1412 
1413 	return (0);
1414 
1415 err_pci_release_msi:
1416 	pci_release_msi(dev);
1417 err_free_msix_table:
1418 	if (sc->msix_table != NULL) {
1419 		bus_release_resource(sc->dev, SYS_RES_MEMORY,
1420 				rman_get_rid(sc->msix_table),
1421 				sc->msix_table);
1422 		sc->msix_table = NULL;
1423 	}
1424 
1425 	return (err);
1426 }
1427 
1428 /**
1429  * ice_if_msix_intr_assign - Assign MSI-X interrupt vectors to queues
1430  * @ctx: the iflib context structure
1431  * @msix: the number of vectors we were assigned
1432  *
1433  * Called by iflib to assign MSI-X vectors to queues. Currently requires that
1434  * we get at least the same number of vectors as we have queues, and that we
1435  * always have the same number of Tx and Rx queues.
1436  *
1437  * Tx queues use a softirq instead of using their own hardware interrupt.
1438  */
1439 static int
1440 ice_if_msix_intr_assign(if_ctx_t ctx, int msix)
1441 {
1442 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1443 	struct ice_vsi *vsi = &sc->pf_vsi;
1444 	int err, i, vector;
1445 
1446 	ASSERT_CTX_LOCKED(sc);
1447 
1448 	if (vsi->num_rx_queues != vsi->num_tx_queues) {
1449 		device_printf(sc->dev,
1450 			      "iflib requested %d Tx queues, and %d Rx queues, but the driver isn't able to support a differing number of Tx and Rx queues\n",
1451 			      vsi->num_tx_queues, vsi->num_rx_queues);
1452 		return (EOPNOTSUPP);
1453 	}
1454 
1455 	if (msix < (vsi->num_rx_queues + 1)) {
1456 		device_printf(sc->dev,
1457 			      "Not enough MSI-X vectors to assign one vector to each queue pair\n");
1458 		return (EOPNOTSUPP);
1459 	}
1460 
1461 	/* Save the number of vectors for future use */
1462 	sc->num_irq_vectors = vsi->num_rx_queues + 1;
1463 
1464 	/* Allocate space to store the IRQ vector data */
1465 	if (!(sc->irqvs =
1466 	      (struct ice_irq_vector *) malloc(sizeof(struct ice_irq_vector) * (sc->num_irq_vectors),
1467 					       M_ICE, M_NOWAIT))) {
1468 		device_printf(sc->dev,
1469 			      "Unable to allocate irqv memory\n");
1470 		return (ENOMEM);
1471 	}
1472 
1473 	/* Administrative interrupt events will use vector 0 */
1474 	err = iflib_irq_alloc_generic(ctx, &sc->irqvs[0].irq, 1, IFLIB_INTR_ADMIN,
1475 				      ice_msix_admin, sc, 0, "admin");
1476 	if (err) {
1477 		device_printf(sc->dev,
1478 			      "Failed to register Admin queue handler: %s\n",
1479 			      ice_err_str(err));
1480 		goto free_irqvs;
1481 	}
1482 	sc->irqvs[0].me = 0;
1483 
1484 	/* Do not allocate queue interrupts when in recovery mode */
1485 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1486 		return (0);
1487 
1488 	for (i = 0, vector = 1; i < vsi->num_rx_queues; i++, vector++) {
1489 		struct ice_rx_queue *rxq = &vsi->rx_queues[i];
1490 		struct ice_tx_queue *txq = &vsi->tx_queues[i];
1491 		int rid = vector + 1;
1492 		char irq_name[16];
1493 
1494 		snprintf(irq_name, sizeof(irq_name), "rxq%d", i);
1495 		err = iflib_irq_alloc_generic(ctx, &sc->irqvs[vector].irq, rid,
1496 					      IFLIB_INTR_RXTX, ice_msix_que,
1497 					      rxq, rxq->me, irq_name);
1498 		if (err) {
1499 			device_printf(sc->dev,
1500 				      "Failed to allocate q int %d err: %s\n",
1501 				      i, ice_err_str(err));
1502 			vector--;
1503 			i--;
1504 			goto fail;
1505 		}
1506 		sc->irqvs[vector].me = vector;
1507 		rxq->irqv = &sc->irqvs[vector];
1508 
1509 		bzero(irq_name, sizeof(irq_name));
1510 
1511 		snprintf(irq_name, sizeof(irq_name), "txq%d", i);
1512 		iflib_softirq_alloc_generic(ctx, &sc->irqvs[vector].irq,
1513 					    IFLIB_INTR_TX, txq,
1514 					    txq->me, irq_name);
1515 		txq->irqv = &sc->irqvs[vector];
1516 	}
1517 
1518 	return (0);
1519 fail:
1520 	for (; i >= 0; i--, vector--)
1521 		iflib_irq_free(ctx, &sc->irqvs[vector].irq);
1522 	iflib_irq_free(ctx, &sc->irqvs[0].irq);
1523 free_irqvs:
1524 	free(sc->irqvs, M_ICE);
1525 	sc->irqvs = NULL;
1526 	return err;
1527 }
1528 
1529 /**
1530  * ice_if_mtu_set - Set the device MTU
1531  * @ctx: iflib context structure
1532  * @mtu: the MTU requested
1533  *
1534  * Called by iflib to configure the device's Maximum Transmission Unit (MTU).
1535  *
1536  * @pre assumes the caller holds the iflib CTX lock
1537  */
1538 static int
1539 ice_if_mtu_set(if_ctx_t ctx, uint32_t mtu)
1540 {
1541 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1542 
1543 	ASSERT_CTX_LOCKED(sc);
1544 
1545 	/* Do not support configuration when in recovery mode */
1546 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1547 		return (ENOSYS);
1548 
1549 	if (mtu < ICE_MIN_MTU || mtu > ICE_MAX_MTU)
1550 		return (EINVAL);
1551 
1552 	sc->scctx->isc_max_frame_size = mtu +
1553 		ETHER_HDR_LEN + ETHER_CRC_LEN + ETHER_VLAN_ENCAP_LEN;
1554 
1555 	sc->pf_vsi.max_frame_size = sc->scctx->isc_max_frame_size;
1556 
1557 	return (0);
1558 }
1559 
1560 /**
1561  * ice_if_intr_enable - Enable device interrupts
1562  * @ctx: iflib context structure
1563  *
1564  * Called by iflib to request enabling device interrupts.
1565  */
1566 static void
1567 ice_if_intr_enable(if_ctx_t ctx)
1568 {
1569 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1570 	struct ice_vsi *vsi = &sc->pf_vsi;
1571 	struct ice_hw *hw = &sc->hw;
1572 
1573 	ASSERT_CTX_LOCKED(sc);
1574 
1575 	/* Enable ITR 0 */
1576 	ice_enable_intr(hw, sc->irqvs[0].me);
1577 
1578 	/* Do not enable queue interrupts in recovery mode */
1579 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1580 		return;
1581 
1582 	/* Enable all queue interrupts */
1583 	for (int i = 0; i < vsi->num_rx_queues; i++)
1584 		ice_enable_intr(hw, vsi->rx_queues[i].irqv->me);
1585 }
1586 
1587 /**
1588  * ice_if_intr_disable - Disable device interrupts
1589  * @ctx: iflib context structure
1590  *
1591  * Called by iflib to request disabling device interrupts.
1592  */
1593 static void
1594 ice_if_intr_disable(if_ctx_t ctx)
1595 {
1596 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1597 	struct ice_hw *hw = &sc->hw;
1598 	unsigned int i;
1599 
1600 	ASSERT_CTX_LOCKED(sc);
1601 
1602 	/* IFDI_INTR_DISABLE may be called prior to interrupts actually being
1603 	 * assigned to queues. Instead of assuming that the interrupt
1604 	 * assignment in the rx_queues structure is valid, just disable all
1605 	 * possible interrupts
1606 	 *
1607 	 * Note that we choose not to disable ITR 0 because this handles the
1608 	 * AdminQ interrupts, and we want to keep processing these even when
1609 	 * the interface is offline.
1610 	 */
1611 	for (i = 1; i < hw->func_caps.common_cap.num_msix_vectors; i++)
1612 		ice_disable_intr(hw, i);
1613 }
1614 
1615 /**
1616  * ice_if_rx_queue_intr_enable - Enable a specific Rx queue interrupt
1617  * @ctx: iflib context structure
1618  * @rxqid: the Rx queue to enable
1619  *
1620  * Enable a specific Rx queue interrupt.
1621  *
1622  * This function is not protected by the iflib CTX lock.
1623  */
1624 static int
1625 ice_if_rx_queue_intr_enable(if_ctx_t ctx, uint16_t rxqid)
1626 {
1627 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1628 	struct ice_vsi *vsi = &sc->pf_vsi;
1629 	struct ice_hw *hw = &sc->hw;
1630 
1631 	/* Do not enable queue interrupts in recovery mode */
1632 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1633 		return (ENOSYS);
1634 
1635 	ice_enable_intr(hw, vsi->rx_queues[rxqid].irqv->me);
1636 	return (0);
1637 }
1638 
1639 /**
1640  * ice_if_tx_queue_intr_enable - Enable a specific Tx queue interrupt
1641  * @ctx: iflib context structure
1642  * @txqid: the Tx queue to enable
1643  *
1644  * Enable a specific Tx queue interrupt.
1645  *
1646  * This function is not protected by the iflib CTX lock.
1647  */
1648 static int
1649 ice_if_tx_queue_intr_enable(if_ctx_t ctx, uint16_t txqid)
1650 {
1651 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1652 	struct ice_vsi *vsi = &sc->pf_vsi;
1653 	struct ice_hw *hw = &sc->hw;
1654 
1655 	/* Do not enable queue interrupts in recovery mode */
1656 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1657 		return (ENOSYS);
1658 
1659 	ice_enable_intr(hw, vsi->tx_queues[txqid].irqv->me);
1660 	return (0);
1661 }
1662 
1663 /**
1664  * ice_if_promisc_set - Set device promiscuous mode
1665  * @ctx: iflib context structure
1666  * @flags: promiscuous flags to configure
1667  *
1668  * Called by iflib to configure device promiscuous mode.
1669  *
1670  * @remark Calls to this function will always overwrite the previous setting
1671  */
1672 static int
1673 ice_if_promisc_set(if_ctx_t ctx, int flags)
1674 {
1675 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1676 	struct ice_hw *hw = &sc->hw;
1677 	device_t dev = sc->dev;
1678 	enum ice_status status;
1679 	bool promisc_enable = flags & IFF_PROMISC;
1680 	bool multi_enable = flags & IFF_ALLMULTI;
1681 
1682 	/* Do not support configuration when in recovery mode */
1683 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1684 		return (ENOSYS);
1685 
1686 	if (multi_enable)
1687 		return (EOPNOTSUPP);
1688 
1689 	if (promisc_enable) {
1690 		status = ice_set_vsi_promisc(hw, sc->pf_vsi.idx,
1691 					     ICE_VSI_PROMISC_MASK, 0);
1692 		if (status && status != ICE_ERR_ALREADY_EXISTS) {
1693 			device_printf(dev,
1694 				      "Failed to enable promiscuous mode for PF VSI, err %s aq_err %s\n",
1695 				      ice_status_str(status),
1696 				      ice_aq_str(hw->adminq.sq_last_status));
1697 			return (EIO);
1698 		}
1699 	} else {
1700 		status = ice_clear_vsi_promisc(hw, sc->pf_vsi.idx,
1701 					       ICE_VSI_PROMISC_MASK, 0);
1702 		if (status) {
1703 			device_printf(dev,
1704 				      "Failed to disable promiscuous mode for PF VSI, err %s aq_err %s\n",
1705 				      ice_status_str(status),
1706 				      ice_aq_str(hw->adminq.sq_last_status));
1707 			return (EIO);
1708 		}
1709 	}
1710 
1711 	return (0);
1712 }
1713 
1714 /**
1715  * ice_if_media_change - Change device media
1716  * @ctx: device ctx structure
1717  *
1718  * Called by iflib when a media change is requested. This operation is not
1719  * supported by the hardware, so we just return an error code.
1720  */
1721 static int
1722 ice_if_media_change(if_ctx_t ctx)
1723 {
1724 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1725 
1726 	device_printf(sc->dev, "Media change is not supported.\n");
1727 	return (ENODEV);
1728 }
1729 
1730 /**
1731  * ice_if_media_status - Report current device media
1732  * @ctx: iflib context structure
1733  * @ifmr: ifmedia request structure to update
1734  *
1735  * Updates the provided ifmr with current device media status, including link
1736  * status and media type.
1737  */
1738 static void
1739 ice_if_media_status(if_ctx_t ctx, struct ifmediareq *ifmr)
1740 {
1741 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1742 	struct ice_link_status *li = &sc->hw.port_info->phy.link_info;
1743 
1744 	ifmr->ifm_status = IFM_AVALID;
1745 	ifmr->ifm_active = IFM_ETHER;
1746 
1747 	/* Never report link up or media types when in recovery mode */
1748 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1749 		return;
1750 
1751 	if (!sc->link_up)
1752 		return;
1753 
1754 	ifmr->ifm_status |= IFM_ACTIVE;
1755 	ifmr->ifm_active |= IFM_FDX;
1756 
1757 	if (li->phy_type_low)
1758 		ifmr->ifm_active |= ice_get_phy_type_low(li->phy_type_low);
1759 	else if (li->phy_type_high)
1760 		ifmr->ifm_active |= ice_get_phy_type_high(li->phy_type_high);
1761 	else
1762 		ifmr->ifm_active |= IFM_UNKNOWN;
1763 
1764 	/* Report flow control status as well */
1765 	if (li->an_info & ICE_AQ_LINK_PAUSE_TX)
1766 		ifmr->ifm_active |= IFM_ETH_TXPAUSE;
1767 	if (li->an_info & ICE_AQ_LINK_PAUSE_RX)
1768 		ifmr->ifm_active |= IFM_ETH_RXPAUSE;
1769 }
1770 
1771 /**
1772  * ice_init_tx_tracking - Initialize Tx queue software tracking values
1773  * @vsi: the VSI to initialize
1774  *
1775  * Initialize Tx queue software tracking values, including the Report Status
1776  * queue, and related software tracking values.
1777  */
1778 static void
1779 ice_init_tx_tracking(struct ice_vsi *vsi)
1780 {
1781 	struct ice_tx_queue *txq;
1782 	size_t j;
1783 	int i;
1784 
1785 	for (i = 0, txq = vsi->tx_queues; i < vsi->num_tx_queues; i++, txq++) {
1786 
1787 		txq->tx_rs_cidx = txq->tx_rs_pidx = 0;
1788 
1789 		/* Initialize the last processed descriptor to be the end of
1790 		 * the ring, rather than the start, so that we avoid an
1791 		 * off-by-one error in ice_ift_txd_credits_update for the
1792 		 * first packet.
1793 		 */
1794 		txq->tx_cidx_processed = txq->desc_count - 1;
1795 
1796 		for (j = 0; j < txq->desc_count; j++)
1797 			txq->tx_rsq[j] = QIDX_INVALID;
1798 	}
1799 }
1800 
1801 /**
1802  * ice_update_rx_mbuf_sz - Update the Rx buffer size for all queues
1803  * @sc: the device softc
1804  *
1805  * Called to update the Rx queue mbuf_sz parameter for configuring the receive
1806  * buffer sizes when programming hardware.
1807  */
1808 static void
1809 ice_update_rx_mbuf_sz(struct ice_softc *sc)
1810 {
1811 	uint32_t mbuf_sz = iflib_get_rx_mbuf_sz(sc->ctx);
1812 	struct ice_vsi *vsi = &sc->pf_vsi;
1813 
1814 	MPASS(mbuf_sz <= UINT16_MAX);
1815 	vsi->mbuf_sz = mbuf_sz;
1816 }
1817 
1818 /**
1819  * ice_if_init - Initialize the device
1820  * @ctx: iflib ctx structure
1821  *
1822  * Called by iflib to bring the device up, i.e. ifconfig ice0 up. Initializes
1823  * device filters and prepares the Tx and Rx engines.
1824  *
1825  * @pre assumes the caller holds the iflib CTX lock
1826  */
1827 static void
1828 ice_if_init(if_ctx_t ctx)
1829 {
1830 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1831 	device_t dev = sc->dev;
1832 	int err;
1833 
1834 	ASSERT_CTX_LOCKED(sc);
1835 
1836 	/*
1837 	 * We've seen an issue with 11.3/12.1 where sideband routines are
1838 	 * called after detach is called.  This would call routines after
1839 	 * if_stop, causing issues with the teardown process.  This has
1840 	 * seemingly been fixed in STABLE snapshots, but it seems like a
1841 	 * good idea to have this guard here regardless.
1842 	 */
1843 	if (ice_driver_is_detaching(sc))
1844 		return;
1845 
1846 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1847 		return;
1848 
1849 	if (ice_test_state(&sc->state, ICE_STATE_RESET_FAILED)) {
1850 		device_printf(sc->dev, "request to start interface cannot be completed as the device failed to reset\n");
1851 		return;
1852 	}
1853 
1854 	if (ice_test_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET)) {
1855 		device_printf(sc->dev, "request to start interface while device is prepared for impending reset\n");
1856 		return;
1857 	}
1858 
1859 	ice_update_rx_mbuf_sz(sc);
1860 
1861 	/* Update the MAC address... User might use a LAA */
1862 	err = ice_update_laa_mac(sc);
1863 	if (err) {
1864 		device_printf(dev,
1865 			      "LAA address change failed, err %s\n",
1866 			      ice_err_str(err));
1867 		return;
1868 	}
1869 
1870 	/* Initialize software Tx tracking values */
1871 	ice_init_tx_tracking(&sc->pf_vsi);
1872 
1873 	err = ice_cfg_vsi_for_tx(&sc->pf_vsi);
1874 	if (err) {
1875 		device_printf(dev,
1876 			      "Unable to configure the main VSI for Tx: %s\n",
1877 			      ice_err_str(err));
1878 		return;
1879 	}
1880 
1881 	err = ice_cfg_vsi_for_rx(&sc->pf_vsi);
1882 	if (err) {
1883 		device_printf(dev,
1884 			      "Unable to configure the main VSI for Rx: %s\n",
1885 			      ice_err_str(err));
1886 		goto err_cleanup_tx;
1887 	}
1888 
1889 	err = ice_control_rx_queues(&sc->pf_vsi, true);
1890 	if (err) {
1891 		device_printf(dev,
1892 			      "Unable to enable Rx rings for transmit: %s\n",
1893 			      ice_err_str(err));
1894 		goto err_cleanup_tx;
1895 	}
1896 
1897 	err = ice_cfg_pf_default_mac_filters(sc);
1898 	if (err) {
1899 		device_printf(dev,
1900 			      "Unable to configure default MAC filters: %s\n",
1901 			      ice_err_str(err));
1902 		goto err_stop_rx;
1903 	}
1904 
1905 	/* We use software interrupts for Tx, so we only program the hardware
1906 	 * interrupts for Rx.
1907 	 */
1908 	ice_configure_rxq_interrupts(&sc->pf_vsi);
1909 	ice_configure_rx_itr(&sc->pf_vsi);
1910 
1911 	/* Configure promiscuous mode */
1912 	ice_if_promisc_set(ctx, if_getflags(sc->ifp));
1913 
1914 	ice_set_state(&sc->state, ICE_STATE_DRIVER_INITIALIZED);
1915 	return;
1916 
1917 err_stop_rx:
1918 	ice_control_rx_queues(&sc->pf_vsi, false);
1919 err_cleanup_tx:
1920 	ice_vsi_disable_tx(&sc->pf_vsi);
1921 }
1922 
1923 /**
1924  * ice_poll_for_media_avail - Re-enable link if media is detected
1925  * @sc: device private structure
1926  *
1927  * Intended to be called from the driver's timer function, this function
1928  * sends the Get Link Status AQ command and re-enables HW link if the
1929  * command says that media is available.
1930  *
1931  * If the driver doesn't have the "NO_MEDIA" state set, then this does nothing,
1932  * since media removal events are supposed to be sent to the driver through
1933  * a link status event.
1934  */
1935 static void
1936 ice_poll_for_media_avail(struct ice_softc *sc)
1937 {
1938 	struct ice_hw *hw = &sc->hw;
1939 	struct ice_port_info *pi = hw->port_info;
1940 
1941 	if (ice_test_state(&sc->state, ICE_STATE_NO_MEDIA)) {
1942 		pi->phy.get_link_info = true;
1943 		ice_get_link_status(pi, &sc->link_up);
1944 
1945 		if (pi->phy.link_info.link_info & ICE_AQ_MEDIA_AVAILABLE) {
1946 			enum ice_status status;
1947 
1948 			/* Re-enable link and re-apply user link settings */
1949 			ice_apply_saved_phy_cfg(sc);
1950 
1951 			/* Update the OS about changes in media capability */
1952 			status = ice_add_media_types(sc, sc->media);
1953 			if (status)
1954 				device_printf(sc->dev, "Error adding device media types: %s aq_err %s\n",
1955 					      ice_status_str(status),
1956 					      ice_aq_str(hw->adminq.sq_last_status));
1957 
1958 			ice_clear_state(&sc->state, ICE_STATE_NO_MEDIA);
1959 		}
1960 	}
1961 }
1962 
1963 /**
1964  * ice_if_timer - called by iflib periodically
1965  * @ctx: iflib ctx structure
1966  * @qid: the queue this timer was called for
1967  *
1968  * This callback is triggered by iflib periodically. We use it to update the
1969  * hw statistics.
1970  *
1971  * @remark this function is not protected by the iflib CTX lock.
1972  */
1973 static void
1974 ice_if_timer(if_ctx_t ctx, uint16_t qid)
1975 {
1976 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
1977 	uint64_t prev_link_xoff_rx = sc->stats.cur.link_xoff_rx;
1978 
1979 	if (qid != 0)
1980 		return;
1981 
1982 	/* Do not attempt to update stats when in recovery mode */
1983 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
1984 		return;
1985 
1986 	/* Update device statistics */
1987 	ice_update_pf_stats(sc);
1988 
1989 	/*
1990 	 * For proper watchdog management, the iflib stack needs to know if
1991 	 * we've been paused during the last interval. Check if the
1992 	 * link_xoff_rx stat changed, and set the isc_pause_frames, if so.
1993 	 */
1994 	if (sc->stats.cur.link_xoff_rx != prev_link_xoff_rx)
1995 		sc->scctx->isc_pause_frames = 1;
1996 
1997 	/* Update the primary VSI stats */
1998 	ice_update_vsi_hw_stats(&sc->pf_vsi);
1999 }
2000 
2001 /**
2002  * ice_admin_timer - called periodically to trigger the admin task
2003  * @arg: callout(9) argument pointing to the device private softc structure
2004  *
2005  * Timer function used as part of a callout(9) timer that will periodically
2006  * trigger the admin task, even when the interface is down.
2007  *
2008  * @remark this function is not called by iflib and is not protected by the
2009  * iflib CTX lock.
2010  *
2011  * @remark because this is a callout function, it cannot sleep and should not
2012  * attempt taking the iflib CTX lock.
2013  */
2014 static void
2015 ice_admin_timer(void *arg)
2016 {
2017 	struct ice_softc *sc = (struct ice_softc *)arg;
2018 
2019 	/* Fire off the admin task */
2020 	iflib_admin_intr_deferred(sc->ctx);
2021 
2022 	/* Reschedule the admin timer */
2023 	callout_schedule(&sc->admin_timer, hz/2);
2024 }
2025 
2026 /**
2027  * ice_transition_recovery_mode - Transition to recovery mode
2028  * @sc: the device private softc
2029  *
2030  * Called when the driver detects that the firmware has entered recovery mode
2031  * at run time.
2032  */
2033 static void
2034 ice_transition_recovery_mode(struct ice_softc *sc)
2035 {
2036 	struct ice_vsi *vsi = &sc->pf_vsi;
2037 	int i;
2038 
2039 	device_printf(sc->dev, "Firmware recovery mode detected. Limiting functionality. Refer to Intel(R) Ethernet Adapters and Devices User Guide for details on firmware recovery mode.\n");
2040 
2041 	/* Tell the stack that the link has gone down */
2042 	iflib_link_state_change(sc->ctx, LINK_STATE_DOWN, 0);
2043 
2044 	/* Request that the device be re-initialized */
2045 	ice_request_stack_reinit(sc);
2046 
2047 	ice_clear_bit(ICE_FEATURE_SRIOV, sc->feat_en);
2048 	ice_clear_bit(ICE_FEATURE_SRIOV, sc->feat_cap);
2049 
2050 	ice_vsi_del_txqs_ctx(vsi);
2051 	ice_vsi_del_rxqs_ctx(vsi);
2052 
2053 	for (i = 0; i < sc->num_available_vsi; i++) {
2054 		if (sc->all_vsi[i])
2055 			ice_release_vsi(sc->all_vsi[i]);
2056 	}
2057 	sc->num_available_vsi = 0;
2058 
2059 	if (sc->all_vsi) {
2060 		free(sc->all_vsi, M_ICE);
2061 		sc->all_vsi = NULL;
2062 	}
2063 
2064 	/* Destroy the interrupt manager */
2065 	ice_resmgr_destroy(&sc->imgr);
2066 	/* Destroy the queue managers */
2067 	ice_resmgr_destroy(&sc->tx_qmgr);
2068 	ice_resmgr_destroy(&sc->rx_qmgr);
2069 
2070 	ice_deinit_hw(&sc->hw);
2071 }
2072 
2073 /**
2074  * ice_transition_safe_mode - Transition to safe mode
2075  * @sc: the device private softc
2076  *
2077  * Called when the driver attempts to reload the DDP package during a device
2078  * reset, and the new download fails. If so, we must transition to safe mode
2079  * at run time.
2080  *
2081  * @remark although safe mode normally allocates only a single queue, we can't
2082  * change the number of queues dynamically when using iflib. Due to this, we
2083  * do not attempt to reduce the number of queues.
2084  */
2085 static void
2086 ice_transition_safe_mode(struct ice_softc *sc)
2087 {
2088 	/* Indicate that we are in Safe mode */
2089 	ice_set_bit(ICE_FEATURE_SAFE_MODE, sc->feat_cap);
2090 	ice_set_bit(ICE_FEATURE_SAFE_MODE, sc->feat_en);
2091 
2092 	ice_clear_bit(ICE_FEATURE_SRIOV, sc->feat_en);
2093 	ice_clear_bit(ICE_FEATURE_SRIOV, sc->feat_cap);
2094 
2095 	ice_clear_bit(ICE_FEATURE_RSS, sc->feat_cap);
2096 	ice_clear_bit(ICE_FEATURE_RSS, sc->feat_en);
2097 }
2098 
2099 /**
2100  * ice_if_update_admin_status - update admin status
2101  * @ctx: iflib ctx structure
2102  *
2103  * Called by iflib to update the admin status. For our purposes, this means
2104  * check the adminq, and update the link status. It's ultimately triggered by
2105  * our admin interrupt, or by the ice_if_timer periodically.
2106  *
2107  * @pre assumes the caller holds the iflib CTX lock
2108  */
2109 static void
2110 ice_if_update_admin_status(if_ctx_t ctx)
2111 {
2112 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2113 	enum ice_fw_modes fw_mode;
2114 	bool reschedule = false;
2115 	u16 pending = 0;
2116 
2117 	ASSERT_CTX_LOCKED(sc);
2118 
2119 	/* Check if the firmware entered recovery mode at run time */
2120 	fw_mode = ice_get_fw_mode(&sc->hw);
2121 	if (fw_mode == ICE_FW_MODE_REC) {
2122 		if (!ice_testandset_state(&sc->state, ICE_STATE_RECOVERY_MODE)) {
2123 			/* If we just entered recovery mode, log a warning to
2124 			 * the system administrator and deinit driver state
2125 			 * that is no longer functional.
2126 			 */
2127 			ice_transition_recovery_mode(sc);
2128 		}
2129 	} else if (fw_mode == ICE_FW_MODE_ROLLBACK) {
2130 		if (!ice_testandset_state(&sc->state, ICE_STATE_ROLLBACK_MODE)) {
2131 			/* Rollback mode isn't fatal, but we don't want to
2132 			 * repeatedly post a message about it.
2133 			 */
2134 			ice_print_rollback_msg(&sc->hw);
2135 		}
2136 	}
2137 
2138 	/* Handle global reset events */
2139 	ice_handle_reset_event(sc);
2140 
2141 	/* Handle PF reset requests */
2142 	ice_handle_pf_reset_request(sc);
2143 
2144 	/* Handle MDD events */
2145 	ice_handle_mdd_event(sc);
2146 
2147 	if (ice_test_state(&sc->state, ICE_STATE_RESET_FAILED) ||
2148 	    ice_test_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET) ||
2149 	    ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE)) {
2150 		/*
2151 		 * If we know the control queues are disabled, skip processing
2152 		 * the control queues entirely.
2153 		 */
2154 		;
2155 	} else if (ice_testandclear_state(&sc->state, ICE_STATE_CONTROLQ_EVENT_PENDING)) {
2156 		ice_process_ctrlq(sc, ICE_CTL_Q_ADMIN, &pending);
2157 		if (pending > 0)
2158 			reschedule = true;
2159 
2160 		ice_process_ctrlq(sc, ICE_CTL_Q_MAILBOX, &pending);
2161 		if (pending > 0)
2162 			reschedule = true;
2163 	}
2164 
2165 	/* Poll for link up */
2166 	ice_poll_for_media_avail(sc);
2167 
2168 	/* Check and update link status */
2169 	ice_update_link_status(sc, false);
2170 
2171 	/*
2172 	 * If there are still messages to process, we need to reschedule
2173 	 * ourselves. Otherwise, we can just re-enable the interrupt. We'll be
2174 	 * woken up at the next interrupt or timer event.
2175 	 */
2176 	if (reschedule) {
2177 		ice_set_state(&sc->state, ICE_STATE_CONTROLQ_EVENT_PENDING);
2178 		iflib_admin_intr_deferred(ctx);
2179 	} else {
2180 		ice_enable_intr(&sc->hw, sc->irqvs[0].me);
2181 	}
2182 }
2183 
2184 /**
2185  * ice_prepare_for_reset - Prepare device for an impending reset
2186  * @sc: The device private softc
2187  *
2188  * Prepare the driver for an impending reset, shutting down VSIs, clearing the
2189  * scheduler setup, and shutting down controlqs. Uses the
2190  * ICE_STATE_PREPARED_FOR_RESET to indicate whether we've already prepared the
2191  * driver for reset or not.
2192  */
2193 static void
2194 ice_prepare_for_reset(struct ice_softc *sc)
2195 {
2196 	struct ice_hw *hw = &sc->hw;
2197 
2198 	/* If we're already prepared, there's nothing to do */
2199 	if (ice_testandset_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET))
2200 		return;
2201 
2202 	log(LOG_INFO, "%s: preparing to reset device logic\n", sc->ifp->if_xname);
2203 
2204 	/* In recovery mode, hardware is not initialized */
2205 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
2206 		return;
2207 
2208 	/* Release the main PF VSI queue mappings */
2209 	ice_resmgr_release_map(&sc->tx_qmgr, sc->pf_vsi.tx_qmap,
2210 				    sc->pf_vsi.num_tx_queues);
2211 	ice_resmgr_release_map(&sc->rx_qmgr, sc->pf_vsi.rx_qmap,
2212 				    sc->pf_vsi.num_rx_queues);
2213 
2214 	ice_clear_hw_tbls(hw);
2215 
2216 	if (hw->port_info)
2217 		ice_sched_clear_port(hw->port_info);
2218 
2219 	ice_shutdown_all_ctrlq(hw);
2220 }
2221 
2222 /**
2223  * ice_rebuild_pf_vsi_qmap - Rebuild the main PF VSI queue mapping
2224  * @sc: the device softc pointer
2225  *
2226  * Loops over the Tx and Rx queues for the main PF VSI and reassigns the queue
2227  * mapping after a reset occurred.
2228  */
2229 static int
2230 ice_rebuild_pf_vsi_qmap(struct ice_softc *sc)
2231 {
2232 	struct ice_vsi *vsi = &sc->pf_vsi;
2233 	struct ice_tx_queue *txq;
2234 	struct ice_rx_queue *rxq;
2235 	int err, i;
2236 
2237 	/* Re-assign Tx queues from PF space to the main VSI */
2238 	err = ice_resmgr_assign_contiguous(&sc->tx_qmgr, vsi->tx_qmap,
2239 					    vsi->num_tx_queues);
2240 	if (err) {
2241 		device_printf(sc->dev, "Unable to re-assign PF Tx queues: %s\n",
2242 			      ice_err_str(err));
2243 		return (err);
2244 	}
2245 
2246 	/* Re-assign Rx queues from PF space to this VSI */
2247 	err = ice_resmgr_assign_contiguous(&sc->rx_qmgr, vsi->rx_qmap,
2248 					    vsi->num_rx_queues);
2249 	if (err) {
2250 		device_printf(sc->dev, "Unable to re-assign PF Rx queues: %s\n",
2251 			      ice_err_str(err));
2252 		goto err_release_tx_queues;
2253 	}
2254 
2255 	vsi->qmap_type = ICE_RESMGR_ALLOC_CONTIGUOUS;
2256 
2257 	/* Re-assign Tx queue tail pointers */
2258 	for (i = 0, txq = vsi->tx_queues; i < vsi->num_tx_queues; i++, txq++)
2259 		txq->tail = QTX_COMM_DBELL(vsi->tx_qmap[i]);
2260 
2261 	/* Re-assign Rx queue tail pointers */
2262 	for (i = 0, rxq = vsi->rx_queues; i < vsi->num_rx_queues; i++, rxq++)
2263 		rxq->tail = QRX_TAIL(vsi->rx_qmap[i]);
2264 
2265 	return (0);
2266 
2267 err_release_tx_queues:
2268 	ice_resmgr_release_map(&sc->tx_qmgr, sc->pf_vsi.tx_qmap,
2269 				   sc->pf_vsi.num_tx_queues);
2270 
2271 	return (err);
2272 }
2273 
2274 /* determine if the iflib context is active */
2275 #define CTX_ACTIVE(ctx) ((if_getdrvflags(iflib_get_ifp(ctx)) & IFF_DRV_RUNNING))
2276 
2277 /**
2278  * ice_rebuild_recovery_mode - Rebuild driver state while in recovery mode
2279  * @sc: The device private softc
2280  *
2281  * Handle a driver rebuild while in recovery mode. This will only rebuild the
2282  * limited functionality supported while in recovery mode.
2283  */
2284 static void
2285 ice_rebuild_recovery_mode(struct ice_softc *sc)
2286 {
2287 	device_t dev = sc->dev;
2288 
2289 	/* enable PCIe bus master */
2290 	pci_enable_busmaster(dev);
2291 
2292 	/* Configure interrupt causes for the administrative interrupt */
2293 	ice_configure_misc_interrupts(sc);
2294 
2295 	/* Enable ITR 0 right away, so that we can handle admin interrupts */
2296 	ice_enable_intr(&sc->hw, sc->irqvs[0].me);
2297 
2298 	/* Now that the rebuild is finished, we're no longer prepared to reset */
2299 	ice_clear_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET);
2300 
2301 	log(LOG_INFO, "%s: device rebuild successful\n", sc->ifp->if_xname);
2302 
2303 	/* In order to completely restore device functionality, the iflib core
2304 	 * needs to be reset. We need to request an iflib reset. Additionally,
2305 	 * because the state of IFC_DO_RESET is cached within task_fn_admin in
2306 	 * the iflib core, we also want re-run the admin task so that iflib
2307 	 * resets immediately instead of waiting for the next interrupt.
2308 	 */
2309 	ice_request_stack_reinit(sc);
2310 
2311 	return;
2312 }
2313 
2314 /**
2315  * ice_rebuild - Rebuild driver state post reset
2316  * @sc: The device private softc
2317  *
2318  * Restore driver state after a reset occurred. Restart the controlqs, setup
2319  * the hardware port, and re-enable the VSIs.
2320  */
2321 static void
2322 ice_rebuild(struct ice_softc *sc)
2323 {
2324 	struct ice_hw *hw = &sc->hw;
2325 	device_t dev = sc->dev;
2326 	enum ice_status status;
2327 	int err;
2328 
2329 	sc->rebuild_ticks = ticks;
2330 
2331 	/* If we're rebuilding, then a reset has succeeded. */
2332 	ice_clear_state(&sc->state, ICE_STATE_RESET_FAILED);
2333 
2334 	/*
2335 	 * If the firmware is in recovery mode, only restore the limited
2336 	 * functionality supported by recovery mode.
2337 	 */
2338 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE)) {
2339 		ice_rebuild_recovery_mode(sc);
2340 		return;
2341 	}
2342 
2343 	/* enable PCIe bus master */
2344 	pci_enable_busmaster(dev);
2345 
2346 	status = ice_init_all_ctrlq(hw);
2347 	if (status) {
2348 		device_printf(dev, "failed to re-init controlqs, err %s\n",
2349 			      ice_status_str(status));
2350 		goto err_shutdown_ctrlq;
2351 	}
2352 
2353 	/* Query the allocated resources for Tx scheduler */
2354 	status = ice_sched_query_res_alloc(hw);
2355 	if (status) {
2356 		device_printf(dev,
2357 			      "Failed to query scheduler resources, err %s aq_err %s\n",
2358 			      ice_status_str(status),
2359 			      ice_aq_str(hw->adminq.sq_last_status));
2360 		goto err_shutdown_ctrlq;
2361 	}
2362 
2363 	err = ice_send_version(sc);
2364 	if (err)
2365 		goto err_shutdown_ctrlq;
2366 
2367 	err = ice_init_link_events(sc);
2368 	if (err) {
2369 		device_printf(dev, "ice_init_link_events failed: %s\n",
2370 			      ice_err_str(err));
2371 		goto err_shutdown_ctrlq;
2372 	}
2373 
2374 	status = ice_clear_pf_cfg(hw);
2375 	if (status) {
2376 		device_printf(dev, "failed to clear PF configuration, err %s\n",
2377 			      ice_status_str(status));
2378 		goto err_shutdown_ctrlq;
2379 	}
2380 
2381 	ice_clear_pxe_mode(hw);
2382 
2383 	status = ice_get_caps(hw);
2384 	if (status) {
2385 		device_printf(dev, "failed to get capabilities, err %s\n",
2386 			      ice_status_str(status));
2387 		goto err_shutdown_ctrlq;
2388 	}
2389 
2390 	status = ice_sched_init_port(hw->port_info);
2391 	if (status) {
2392 		device_printf(dev, "failed to initialize port, err %s\n",
2393 			      ice_status_str(status));
2394 		goto err_sched_cleanup;
2395 	}
2396 
2397 	/* If we previously loaded the package, it needs to be reloaded now */
2398 	if (!ice_is_bit_set(sc->feat_en, ICE_FEATURE_SAFE_MODE)) {
2399 		status = ice_init_pkg(hw, hw->pkg_copy, hw->pkg_size);
2400 		if (status) {
2401 			ice_log_pkg_init(sc, &status);
2402 
2403 			ice_transition_safe_mode(sc);
2404 		}
2405 	}
2406 
2407 	ice_reset_pf_stats(sc);
2408 
2409 	err = ice_rebuild_pf_vsi_qmap(sc);
2410 	if (err) {
2411 		device_printf(sc->dev, "Unable to re-assign main VSI queues, err %s\n",
2412 			      ice_err_str(err));
2413 		goto err_sched_cleanup;
2414 	}
2415 	err = ice_initialize_vsi(&sc->pf_vsi);
2416 	if (err) {
2417 		device_printf(sc->dev, "Unable to re-initialize Main VSI, err %s\n",
2418 			      ice_err_str(err));
2419 		goto err_release_queue_allocations;
2420 	}
2421 
2422 	/* Replay all VSI configuration */
2423 	err = ice_replay_all_vsi_cfg(sc);
2424 	if (err)
2425 		goto err_deinit_pf_vsi;
2426 
2427 	/* Reconfigure the main PF VSI for RSS */
2428 	err = ice_config_rss(&sc->pf_vsi);
2429 	if (err) {
2430 		device_printf(sc->dev,
2431 			      "Unable to reconfigure RSS for the main VSI, err %s\n",
2432 			      ice_err_str(err));
2433 		goto err_deinit_pf_vsi;
2434 	}
2435 
2436 	/* Refresh link status */
2437 	ice_clear_state(&sc->state, ICE_STATE_LINK_STATUS_REPORTED);
2438 	sc->hw.port_info->phy.get_link_info = true;
2439 	ice_get_link_status(sc->hw.port_info, &sc->link_up);
2440 	ice_update_link_status(sc, true);
2441 
2442 	/* Configure interrupt causes for the administrative interrupt */
2443 	ice_configure_misc_interrupts(sc);
2444 
2445 	/* Enable ITR 0 right away, so that we can handle admin interrupts */
2446 	ice_enable_intr(&sc->hw, sc->irqvs[0].me);
2447 
2448 	/* Now that the rebuild is finished, we're no longer prepared to reset */
2449 	ice_clear_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET);
2450 
2451 	log(LOG_INFO, "%s: device rebuild successful\n", sc->ifp->if_xname);
2452 
2453 	/* In order to completely restore device functionality, the iflib core
2454 	 * needs to be reset. We need to request an iflib reset. Additionally,
2455 	 * because the state of IFC_DO_RESET is cached within task_fn_admin in
2456 	 * the iflib core, we also want re-run the admin task so that iflib
2457 	 * resets immediately instead of waiting for the next interrupt.
2458 	 */
2459 	ice_request_stack_reinit(sc);
2460 
2461 	return;
2462 
2463 err_deinit_pf_vsi:
2464 	ice_deinit_vsi(&sc->pf_vsi);
2465 err_release_queue_allocations:
2466 	ice_resmgr_release_map(&sc->tx_qmgr, sc->pf_vsi.tx_qmap,
2467 				    sc->pf_vsi.num_tx_queues);
2468 	ice_resmgr_release_map(&sc->rx_qmgr, sc->pf_vsi.rx_qmap,
2469 				    sc->pf_vsi.num_rx_queues);
2470 err_sched_cleanup:
2471 	ice_sched_cleanup_all(hw);
2472 err_shutdown_ctrlq:
2473 	ice_shutdown_all_ctrlq(hw);
2474 	ice_set_state(&sc->state, ICE_STATE_RESET_FAILED);
2475 	device_printf(dev, "Driver rebuild failed, please reload the device driver\n");
2476 }
2477 
2478 /**
2479  * ice_handle_reset_event - Handle reset events triggered by OICR
2480  * @sc: The device private softc
2481  *
2482  * Handle reset events triggered by an OICR notification. This includes CORER,
2483  * GLOBR, and EMPR resets triggered by software on this or any other PF or by
2484  * firmware.
2485  *
2486  * @pre assumes the iflib context lock is held, and will unlock it while
2487  * waiting for the hardware to finish reset.
2488  */
2489 static void
2490 ice_handle_reset_event(struct ice_softc *sc)
2491 {
2492 	struct ice_hw *hw = &sc->hw;
2493 	enum ice_status status;
2494 	device_t dev = sc->dev;
2495 
2496 	/* When a CORER, GLOBR, or EMPR is about to happen, the hardware will
2497 	 * trigger an OICR interrupt. Our OICR handler will determine when
2498 	 * this occurs and set the ICE_STATE_RESET_OICR_RECV bit as
2499 	 * appropriate.
2500 	 */
2501 	if (!ice_testandclear_state(&sc->state, ICE_STATE_RESET_OICR_RECV))
2502 		return;
2503 
2504 	ice_prepare_for_reset(sc);
2505 
2506 	/*
2507 	 * Release the iflib context lock and wait for the device to finish
2508 	 * resetting.
2509 	 */
2510 	IFLIB_CTX_UNLOCK(sc);
2511 	status = ice_check_reset(hw);
2512 	IFLIB_CTX_LOCK(sc);
2513 	if (status) {
2514 		device_printf(dev, "Device never came out of reset, err %s\n",
2515 			      ice_status_str(status));
2516 		ice_set_state(&sc->state, ICE_STATE_RESET_FAILED);
2517 		return;
2518 	}
2519 
2520 	/* We're done with the reset, so we can rebuild driver state */
2521 	sc->hw.reset_ongoing = false;
2522 	ice_rebuild(sc);
2523 
2524 	/* In the unlikely event that a PF reset request occurs at the same
2525 	 * time as a global reset, clear the request now. This avoids
2526 	 * resetting a second time right after we reset due to a global event.
2527 	 */
2528 	if (ice_testandclear_state(&sc->state, ICE_STATE_RESET_PFR_REQ))
2529 		device_printf(dev, "Ignoring PFR request that occurred while a reset was ongoing\n");
2530 }
2531 
2532 /**
2533  * ice_handle_pf_reset_request - Initiate PF reset requested by software
2534  * @sc: The device private softc
2535  *
2536  * Initiate a PF reset requested by software. We handle this in the admin task
2537  * so that only one thread actually handles driver preparation and cleanup,
2538  * rather than having multiple threads possibly attempt to run this code
2539  * simultaneously.
2540  *
2541  * @pre assumes the iflib context lock is held and will unlock it while
2542  * waiting for the PF reset to complete.
2543  */
2544 static void
2545 ice_handle_pf_reset_request(struct ice_softc *sc)
2546 {
2547 	struct ice_hw *hw = &sc->hw;
2548 	enum ice_status status;
2549 
2550 	/* Check for PF reset requests */
2551 	if (!ice_testandclear_state(&sc->state, ICE_STATE_RESET_PFR_REQ))
2552 		return;
2553 
2554 	/* Make sure we're prepared for reset */
2555 	ice_prepare_for_reset(sc);
2556 
2557 	/*
2558 	 * Release the iflib context lock and wait for the device to finish
2559 	 * resetting.
2560 	 */
2561 	IFLIB_CTX_UNLOCK(sc);
2562 	status = ice_reset(hw, ICE_RESET_PFR);
2563 	IFLIB_CTX_LOCK(sc);
2564 	if (status) {
2565 		device_printf(sc->dev, "device PF reset failed, err %s\n",
2566 			      ice_status_str(status));
2567 		ice_set_state(&sc->state, ICE_STATE_RESET_FAILED);
2568 		return;
2569 	}
2570 
2571 	sc->soft_stats.pfr_count++;
2572 	ice_rebuild(sc);
2573 }
2574 
2575 /**
2576  * ice_init_device_features - Init device driver features
2577  * @sc: driver softc structure
2578  *
2579  * @pre assumes that the function capabilities bits have been set up by
2580  * ice_init_hw().
2581  */
2582 static void
2583 ice_init_device_features(struct ice_softc *sc)
2584 {
2585 	/*
2586 	 * A failed pkg file download triggers safe mode, disabling advanced
2587 	 * device feature support
2588 	 */
2589 	if (ice_is_bit_set(sc->feat_en, ICE_FEATURE_SAFE_MODE))
2590 		return;
2591 
2592 	/* Set capabilities that all devices support */
2593 	ice_set_bit(ICE_FEATURE_SRIOV, sc->feat_cap);
2594 	ice_set_bit(ICE_FEATURE_RSS, sc->feat_cap);
2595 	ice_set_bit(ICE_FEATURE_LENIENT_LINK_MODE, sc->feat_cap);
2596 	ice_set_bit(ICE_FEATURE_DEFAULT_OVERRIDE, sc->feat_cap);
2597 
2598 	/* Disable features due to hardware limitations... */
2599 	if (!sc->hw.func_caps.common_cap.rss_table_size)
2600 		ice_clear_bit(ICE_FEATURE_RSS, sc->feat_cap);
2601 
2602 	/* Disable capabilities not supported by the OS */
2603 	ice_disable_unsupported_features(sc->feat_cap);
2604 
2605 	/* RSS is always enabled for iflib */
2606 	if (ice_is_bit_set(sc->feat_cap, ICE_FEATURE_RSS))
2607 		ice_set_bit(ICE_FEATURE_RSS, sc->feat_en);
2608 }
2609 
2610 /**
2611  * ice_if_multi_set - Callback to update Multicast filters in HW
2612  * @ctx: iflib ctx structure
2613  *
2614  * Called by iflib in response to SIOCDELMULTI and SIOCADDMULTI. Must search
2615  * the if_multiaddrs list and determine which filters have been added or
2616  * removed from the list, and update HW programming to reflect the new list.
2617  *
2618  * @pre assumes the caller holds the iflib CTX lock
2619  */
2620 static void
2621 ice_if_multi_set(if_ctx_t ctx)
2622 {
2623 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2624 	int err;
2625 
2626 	ASSERT_CTX_LOCKED(sc);
2627 
2628 	/* Do not handle multicast configuration in recovery mode */
2629 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
2630 		return;
2631 
2632 	err = ice_sync_multicast_filters(sc);
2633 	if (err) {
2634 		device_printf(sc->dev,
2635 			      "Failed to synchronize multicast filter list: %s\n",
2636 			      ice_err_str(err));
2637 		return;
2638 	}
2639 }
2640 
2641 /**
2642  * ice_if_vlan_register - Register a VLAN with the hardware
2643  * @ctx: iflib ctx pointer
2644  * @vtag: VLAN to add
2645  *
2646  * Programs the main PF VSI with a hardware filter for the given VLAN.
2647  *
2648  * @pre assumes the caller holds the iflib CTX lock
2649  */
2650 static void
2651 ice_if_vlan_register(if_ctx_t ctx, u16 vtag)
2652 {
2653 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2654 	enum ice_status status;
2655 
2656 	ASSERT_CTX_LOCKED(sc);
2657 
2658 	/* Do not handle VLAN configuration in recovery mode */
2659 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
2660 		return;
2661 
2662 	status = ice_add_vlan_hw_filter(&sc->pf_vsi, vtag);
2663 	if (status) {
2664 		device_printf(sc->dev,
2665 			      "Failure adding VLAN %d to main VSI, err %s aq_err %s\n",
2666 			      vtag, ice_status_str(status),
2667 			      ice_aq_str(sc->hw.adminq.sq_last_status));
2668 	}
2669 }
2670 
2671 /**
2672  * ice_if_vlan_unregister - Remove a VLAN filter from the hardware
2673  * @ctx: iflib ctx pointer
2674  * @vtag: VLAN to add
2675  *
2676  * Removes the previously programmed VLAN filter from the main PF VSI.
2677  *
2678  * @pre assumes the caller holds the iflib CTX lock
2679  */
2680 static void
2681 ice_if_vlan_unregister(if_ctx_t ctx, u16 vtag)
2682 {
2683 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2684 	enum ice_status status;
2685 
2686 	ASSERT_CTX_LOCKED(sc);
2687 
2688 	/* Do not handle VLAN configuration in recovery mode */
2689 	if (ice_test_state(&sc->state, ICE_STATE_RECOVERY_MODE))
2690 		return;
2691 
2692 	status = ice_remove_vlan_hw_filter(&sc->pf_vsi, vtag);
2693 	if (status) {
2694 		device_printf(sc->dev,
2695 			      "Failure removing VLAN %d from main VSI, err %s aq_err %s\n",
2696 			      vtag, ice_status_str(status),
2697 			      ice_aq_str(sc->hw.adminq.sq_last_status));
2698 	}
2699 }
2700 
2701 /**
2702  * ice_if_stop - Stop the device
2703  * @ctx: iflib context structure
2704  *
2705  * Called by iflib to stop the device and bring it down. (i.e. ifconfig ice0
2706  * down)
2707  *
2708  * @pre assumes the caller holds the iflib CTX lock
2709  */
2710 static void
2711 ice_if_stop(if_ctx_t ctx)
2712 {
2713 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2714 
2715 	ASSERT_CTX_LOCKED(sc);
2716 
2717 	/*
2718 	 * The iflib core may call IFDI_STOP prior to the first call to
2719 	 * IFDI_INIT. This will cause us to attempt to remove MAC filters we
2720 	 * don't have, and disable Tx queues which aren't yet configured.
2721 	 * Although it is likely these extra operations are harmless, they do
2722 	 * cause spurious warning messages to be displayed, which may confuse
2723 	 * users.
2724 	 *
2725 	 * To avoid these messages, we use a state bit indicating if we've
2726 	 * been initialized. It will be set when ice_if_init is called, and
2727 	 * cleared here in ice_if_stop.
2728 	 */
2729 	if (!ice_testandclear_state(&sc->state, ICE_STATE_DRIVER_INITIALIZED))
2730 		return;
2731 
2732 	if (ice_test_state(&sc->state, ICE_STATE_RESET_FAILED)) {
2733 		device_printf(sc->dev, "request to stop interface cannot be completed as the device failed to reset\n");
2734 		return;
2735 	}
2736 
2737 	if (ice_test_state(&sc->state, ICE_STATE_PREPARED_FOR_RESET)) {
2738 		device_printf(sc->dev, "request to stop interface while device is prepared for impending reset\n");
2739 		return;
2740 	}
2741 
2742 	/* Remove the MAC filters, stop Tx, and stop Rx. We don't check the
2743 	 * return of these functions because there's nothing we can really do
2744 	 * if they fail, and the functions already print error messages.
2745 	 * Just try to shut down as much as we can.
2746 	 */
2747 	ice_rm_pf_default_mac_filters(sc);
2748 
2749 	/* Dissociate the Tx and Rx queues from the interrupts */
2750 	ice_flush_txq_interrupts(&sc->pf_vsi);
2751 	ice_flush_rxq_interrupts(&sc->pf_vsi);
2752 
2753 	/* Disable the Tx and Rx queues */
2754 	ice_vsi_disable_tx(&sc->pf_vsi);
2755 	ice_control_rx_queues(&sc->pf_vsi, false);
2756 }
2757 
2758 /**
2759  * ice_if_get_counter - Get current value of an ifnet statistic
2760  * @ctx: iflib context pointer
2761  * @counter: ifnet counter to read
2762  *
2763  * Reads the current value of an ifnet counter for the device.
2764  *
2765  * This function is not protected by the iflib CTX lock.
2766  */
2767 static uint64_t
2768 ice_if_get_counter(if_ctx_t ctx, ift_counter counter)
2769 {
2770 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2771 
2772 	/* Return the counter for the main PF VSI */
2773 	return ice_get_ifnet_counter(&sc->pf_vsi, counter);
2774 }
2775 
2776 /**
2777  * ice_request_stack_reinit - Request that iflib re-initialize
2778  * @sc: the device private softc
2779  *
2780  * Request that the device be brought down and up, to re-initialize. For
2781  * example, this may be called when a device reset occurs, or when Tx and Rx
2782  * queues need to be re-initialized.
2783  *
2784  * This is required because the iflib state is outside the driver, and must be
2785  * re-initialized if we need to resart Tx and Rx queues.
2786  */
2787 void
2788 ice_request_stack_reinit(struct ice_softc *sc)
2789 {
2790 	if (CTX_ACTIVE(sc->ctx)) {
2791 		iflib_request_reset(sc->ctx);
2792 		iflib_admin_intr_deferred(sc->ctx);
2793 	}
2794 }
2795 
2796 /**
2797  * ice_driver_is_detaching - Check if the driver is detaching/unloading
2798  * @sc: device private softc
2799  *
2800  * Returns true if the driver is detaching, false otherwise.
2801  *
2802  * @remark on newer kernels, take advantage of iflib_in_detach in order to
2803  * report detachment correctly as early as possible.
2804  *
2805  * @remark this function is used by various code paths that want to avoid
2806  * running if the driver is about to be removed. This includes sysctls and
2807  * other driver access points. Note that it does not fully resolve
2808  * detach-based race conditions as it is possible for a thread to race with
2809  * iflib_in_detach.
2810  */
2811 bool
2812 ice_driver_is_detaching(struct ice_softc *sc)
2813 {
2814 	return (ice_test_state(&sc->state, ICE_STATE_DETACHING) ||
2815 		iflib_in_detach(sc->ctx));
2816 }
2817 
2818 /**
2819  * ice_if_priv_ioctl - Device private ioctl handler
2820  * @ctx: iflib context pointer
2821  * @command: The ioctl command issued
2822  * @data: ioctl specific data
2823  *
2824  * iflib callback for handling custom driver specific ioctls.
2825  *
2826  * @pre Assumes that the iflib context lock is held.
2827  */
2828 static int
2829 ice_if_priv_ioctl(if_ctx_t ctx, u_long command, caddr_t data)
2830 {
2831 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2832 	struct ifdrv *ifd;
2833 	device_t dev = sc->dev;
2834 
2835 	if (data == NULL)
2836 		return (EINVAL);
2837 
2838 	ASSERT_CTX_LOCKED(sc);
2839 
2840 	/* Make sure the command type is valid */
2841 	switch (command) {
2842 	case SIOCSDRVSPEC:
2843 	case SIOCGDRVSPEC:
2844 		/* Accepted commands */
2845 		break;
2846 	case SIOCGPRIVATE_0:
2847 		/*
2848 		 * Although we do not support this ioctl command, it's
2849 		 * expected that iflib will forward it to the IFDI_PRIV_IOCTL
2850 		 * handler. Do not print a message in this case
2851 		 */
2852 		return (ENOTSUP);
2853 	default:
2854 		/*
2855 		 * If we get a different command for this function, it's
2856 		 * definitely unexpected, so log a message indicating what
2857 		 * command we got for debugging purposes.
2858 		 */
2859 		device_printf(dev, "%s: unexpected ioctl command %08lx\n",
2860 			      __func__, command);
2861 		return (EINVAL);
2862 	}
2863 
2864 	ifd = (struct ifdrv *)data;
2865 
2866 	switch (ifd->ifd_cmd) {
2867 	case ICE_NVM_ACCESS:
2868 		return ice_handle_nvm_access_ioctl(sc, ifd);
2869 	default:
2870 		return EINVAL;
2871 	}
2872 }
2873 
2874 /**
2875  * ice_if_i2c_req - I2C request handler for iflib
2876  * @ctx: iflib context pointer
2877  * @req: The I2C parameters to use
2878  *
2879  * Read from the port's I2C eeprom using the parameters from the ioctl.
2880  *
2881  * @remark The iflib-only part is pretty simple.
2882  */
2883 static int
2884 ice_if_i2c_req(if_ctx_t ctx, struct ifi2creq *req)
2885 {
2886 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2887 
2888 	return ice_handle_i2c_req(sc, req);
2889 }
2890 
2891 /**
2892  * ice_if_suspend - PCI device suspend handler for iflib
2893  * @ctx: iflib context pointer
2894  *
2895  * Deinitializes the driver and clears HW resources in preparation for
2896  * suspend or an FLR.
2897  *
2898  * @returns 0; this return value is ignored
2899  */
2900 static int
2901 ice_if_suspend(if_ctx_t ctx)
2902 {
2903 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2904 
2905 	/* At least a PFR is always going to happen after this;
2906 	 * either via FLR or during the D3->D0 transition.
2907 	 */
2908 	ice_clear_state(&sc->state, ICE_STATE_RESET_PFR_REQ);
2909 
2910 	ice_prepare_for_reset(sc);
2911 
2912 	return (0);
2913 }
2914 
2915 /**
2916  * ice_if_resume - PCI device resume handler for iflib
2917  * @ctx: iflib context pointer
2918  *
2919  * Reinitializes the driver and the HW after PCI resume or after
2920  * an FLR. An init is performed by iflib after this function is finished.
2921  *
2922  * @returns 0; this return value is ignored
2923  */
2924 static int
2925 ice_if_resume(if_ctx_t ctx)
2926 {
2927 	struct ice_softc *sc = (struct ice_softc *)iflib_get_softc(ctx);
2928 
2929 	ice_rebuild(sc);
2930 
2931 	return (0);
2932 }
2933 
2934