Loading...
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * Xen PCI Frontend
4 *
5 * Author: Ryan Wilson <hap9@epoch.ncsc.mil>
6 */
7#include <linux/module.h>
8#include <linux/init.h>
9#include <linux/mm.h>
10#include <xen/xenbus.h>
11#include <xen/events.h>
12#include <xen/grant_table.h>
13#include <xen/page.h>
14#include <linux/spinlock.h>
15#include <linux/pci.h>
16#include <linux/msi.h>
17#include <xen/interface/io/pciif.h>
18#include <asm/xen/pci.h>
19#include <linux/interrupt.h>
20#include <linux/atomic.h>
21#include <linux/workqueue.h>
22#include <linux/bitops.h>
23#include <linux/time.h>
24#include <linux/ktime.h>
25#include <xen/platform_pci.h>
26
27#include <asm/xen/swiotlb-xen.h>
28#define INVALID_GRANT_REF (0)
29#define INVALID_EVTCHN (-1)
30
31struct pci_bus_entry {
32 struct list_head list;
33 struct pci_bus *bus;
34};
35
36#define _PDEVB_op_active (0)
37#define PDEVB_op_active (1 << (_PDEVB_op_active))
38
39struct pcifront_device {
40 struct xenbus_device *xdev;
41 struct list_head root_buses;
42
43 int evtchn;
44 int gnt_ref;
45
46 int irq;
47
48 /* Lock this when doing any operations in sh_info */
49 spinlock_t sh_info_lock;
50 struct xen_pci_sharedinfo *sh_info;
51 struct work_struct op_work;
52 unsigned long flags;
53
54};
55
56struct pcifront_sd {
57 struct pci_sysdata sd;
58 struct pcifront_device *pdev;
59};
60
61static inline struct pcifront_device *
62pcifront_get_pdev(struct pcifront_sd *sd)
63{
64 return sd->pdev;
65}
66
67static inline void pcifront_init_sd(struct pcifront_sd *sd,
68 unsigned int domain, unsigned int bus,
69 struct pcifront_device *pdev)
70{
71 /* Because we do not expose that information via XenBus. */
72 sd->sd.node = first_online_node;
73 sd->sd.domain = domain;
74 sd->pdev = pdev;
75}
76
77static DEFINE_SPINLOCK(pcifront_dev_lock);
78static struct pcifront_device *pcifront_dev;
79
80static int errno_to_pcibios_err(int errno)
81{
82 switch (errno) {
83 case XEN_PCI_ERR_success:
84 return PCIBIOS_SUCCESSFUL;
85
86 case XEN_PCI_ERR_dev_not_found:
87 return PCIBIOS_DEVICE_NOT_FOUND;
88
89 case XEN_PCI_ERR_invalid_offset:
90 case XEN_PCI_ERR_op_failed:
91 return PCIBIOS_BAD_REGISTER_NUMBER;
92
93 case XEN_PCI_ERR_not_implemented:
94 return PCIBIOS_FUNC_NOT_SUPPORTED;
95
96 case XEN_PCI_ERR_access_denied:
97 return PCIBIOS_SET_FAILED;
98 }
99 return errno;
100}
101
102static inline void schedule_pcifront_aer_op(struct pcifront_device *pdev)
103{
104 if (test_bit(_XEN_PCIB_active, (unsigned long *)&pdev->sh_info->flags)
105 && !test_and_set_bit(_PDEVB_op_active, &pdev->flags)) {
106 dev_dbg(&pdev->xdev->dev, "schedule aer frontend job\n");
107 schedule_work(&pdev->op_work);
108 }
109}
110
111static int do_pci_op(struct pcifront_device *pdev, struct xen_pci_op *op)
112{
113 int err = 0;
114 struct xen_pci_op *active_op = &pdev->sh_info->op;
115 unsigned long irq_flags;
116 evtchn_port_t port = pdev->evtchn;
117 unsigned irq = pdev->irq;
118 s64 ns, ns_timeout;
119
120 spin_lock_irqsave(&pdev->sh_info_lock, irq_flags);
121
122 memcpy(active_op, op, sizeof(struct xen_pci_op));
123
124 /* Go */
125 wmb();
126 set_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
127 notify_remote_via_evtchn(port);
128
129 /*
130 * We set a poll timeout of 3 seconds but give up on return after
131 * 2 seconds. It is better to time out too late rather than too early
132 * (in the latter case we end up continually re-executing poll() with a
133 * timeout in the past). 1s difference gives plenty of slack for error.
134 */
135 ns_timeout = ktime_get_ns() + 2 * (s64)NSEC_PER_SEC;
136
137 xen_clear_irq_pending(irq);
138
139 while (test_bit(_XEN_PCIF_active,
140 (unsigned long *)&pdev->sh_info->flags)) {
141 xen_poll_irq_timeout(irq, jiffies + 3*HZ);
142 xen_clear_irq_pending(irq);
143 ns = ktime_get_ns();
144 if (ns > ns_timeout) {
145 dev_err(&pdev->xdev->dev,
146 "pciback not responding!!!\n");
147 clear_bit(_XEN_PCIF_active,
148 (unsigned long *)&pdev->sh_info->flags);
149 err = XEN_PCI_ERR_dev_not_found;
150 goto out;
151 }
152 }
153
154 /*
155 * We might lose backend service request since we
156 * reuse same evtchn with pci_conf backend response. So re-schedule
157 * aer pcifront service.
158 */
159 if (test_bit(_XEN_PCIB_active,
160 (unsigned long *)&pdev->sh_info->flags)) {
161 dev_err(&pdev->xdev->dev,
162 "schedule aer pcifront service\n");
163 schedule_pcifront_aer_op(pdev);
164 }
165
166 memcpy(op, active_op, sizeof(struct xen_pci_op));
167
168 err = op->err;
169out:
170 spin_unlock_irqrestore(&pdev->sh_info_lock, irq_flags);
171 return err;
172}
173
174/* Access to this function is spinlocked in drivers/pci/access.c */
175static int pcifront_bus_read(struct pci_bus *bus, unsigned int devfn,
176 int where, int size, u32 *val)
177{
178 int err = 0;
179 struct xen_pci_op op = {
180 .cmd = XEN_PCI_OP_conf_read,
181 .domain = pci_domain_nr(bus),
182 .bus = bus->number,
183 .devfn = devfn,
184 .offset = where,
185 .size = size,
186 };
187 struct pcifront_sd *sd = bus->sysdata;
188 struct pcifront_device *pdev = pcifront_get_pdev(sd);
189
190 dev_dbg(&pdev->xdev->dev,
191 "read dev=%04x:%02x:%02x.%d - offset %x size %d\n",
192 pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
193 PCI_FUNC(devfn), where, size);
194
195 err = do_pci_op(pdev, &op);
196
197 if (likely(!err)) {
198 dev_dbg(&pdev->xdev->dev, "read got back value %x\n",
199 op.value);
200
201 *val = op.value;
202 } else if (err == -ENODEV) {
203 /* No device here, pretend that it just returned 0 */
204 err = 0;
205 *val = 0;
206 }
207
208 return errno_to_pcibios_err(err);
209}
210
211/* Access to this function is spinlocked in drivers/pci/access.c */
212static int pcifront_bus_write(struct pci_bus *bus, unsigned int devfn,
213 int where, int size, u32 val)
214{
215 struct xen_pci_op op = {
216 .cmd = XEN_PCI_OP_conf_write,
217 .domain = pci_domain_nr(bus),
218 .bus = bus->number,
219 .devfn = devfn,
220 .offset = where,
221 .size = size,
222 .value = val,
223 };
224 struct pcifront_sd *sd = bus->sysdata;
225 struct pcifront_device *pdev = pcifront_get_pdev(sd);
226
227 dev_dbg(&pdev->xdev->dev,
228 "write dev=%04x:%02x:%02x.%d - offset %x size %d val %x\n",
229 pci_domain_nr(bus), bus->number,
230 PCI_SLOT(devfn), PCI_FUNC(devfn), where, size, val);
231
232 return errno_to_pcibios_err(do_pci_op(pdev, &op));
233}
234
235static struct pci_ops pcifront_bus_ops = {
236 .read = pcifront_bus_read,
237 .write = pcifront_bus_write,
238};
239
240#ifdef CONFIG_PCI_MSI
241static int pci_frontend_enable_msix(struct pci_dev *dev,
242 int vector[], int nvec)
243{
244 int err;
245 int i;
246 struct xen_pci_op op = {
247 .cmd = XEN_PCI_OP_enable_msix,
248 .domain = pci_domain_nr(dev->bus),
249 .bus = dev->bus->number,
250 .devfn = dev->devfn,
251 .value = nvec,
252 };
253 struct pcifront_sd *sd = dev->bus->sysdata;
254 struct pcifront_device *pdev = pcifront_get_pdev(sd);
255 struct msi_desc *entry;
256
257 if (nvec > SH_INFO_MAX_VEC) {
258 pci_err(dev, "too many vectors (0x%x) for PCI frontend:"
259 " Increase SH_INFO_MAX_VEC\n", nvec);
260 return -EINVAL;
261 }
262
263 i = 0;
264 for_each_pci_msi_entry(entry, dev) {
265 op.msix_entries[i].entry = entry->msi_attrib.entry_nr;
266 /* Vector is useless at this point. */
267 op.msix_entries[i].vector = -1;
268 i++;
269 }
270
271 err = do_pci_op(pdev, &op);
272
273 if (likely(!err)) {
274 if (likely(!op.value)) {
275 /* we get the result */
276 for (i = 0; i < nvec; i++) {
277 if (op.msix_entries[i].vector <= 0) {
278 pci_warn(dev, "MSI-X entry %d is invalid: %d!\n",
279 i, op.msix_entries[i].vector);
280 err = -EINVAL;
281 vector[i] = -1;
282 continue;
283 }
284 vector[i] = op.msix_entries[i].vector;
285 }
286 } else {
287 pr_info("enable msix get value %x\n", op.value);
288 err = op.value;
289 }
290 } else {
291 pci_err(dev, "enable msix get err %x\n", err);
292 }
293 return err;
294}
295
296static void pci_frontend_disable_msix(struct pci_dev *dev)
297{
298 int err;
299 struct xen_pci_op op = {
300 .cmd = XEN_PCI_OP_disable_msix,
301 .domain = pci_domain_nr(dev->bus),
302 .bus = dev->bus->number,
303 .devfn = dev->devfn,
304 };
305 struct pcifront_sd *sd = dev->bus->sysdata;
306 struct pcifront_device *pdev = pcifront_get_pdev(sd);
307
308 err = do_pci_op(pdev, &op);
309
310 /* What should do for error ? */
311 if (err)
312 pci_err(dev, "pci_disable_msix get err %x\n", err);
313}
314
315static int pci_frontend_enable_msi(struct pci_dev *dev, int vector[])
316{
317 int err;
318 struct xen_pci_op op = {
319 .cmd = XEN_PCI_OP_enable_msi,
320 .domain = pci_domain_nr(dev->bus),
321 .bus = dev->bus->number,
322 .devfn = dev->devfn,
323 };
324 struct pcifront_sd *sd = dev->bus->sysdata;
325 struct pcifront_device *pdev = pcifront_get_pdev(sd);
326
327 err = do_pci_op(pdev, &op);
328 if (likely(!err)) {
329 vector[0] = op.value;
330 if (op.value <= 0) {
331 pci_warn(dev, "MSI entry is invalid: %d!\n",
332 op.value);
333 err = -EINVAL;
334 vector[0] = -1;
335 }
336 } else {
337 pci_err(dev, "pci frontend enable msi failed for dev "
338 "%x:%x\n", op.bus, op.devfn);
339 err = -EINVAL;
340 }
341 return err;
342}
343
344static void pci_frontend_disable_msi(struct pci_dev *dev)
345{
346 int err;
347 struct xen_pci_op op = {
348 .cmd = XEN_PCI_OP_disable_msi,
349 .domain = pci_domain_nr(dev->bus),
350 .bus = dev->bus->number,
351 .devfn = dev->devfn,
352 };
353 struct pcifront_sd *sd = dev->bus->sysdata;
354 struct pcifront_device *pdev = pcifront_get_pdev(sd);
355
356 err = do_pci_op(pdev, &op);
357 if (err == XEN_PCI_ERR_dev_not_found) {
358 /* XXX No response from backend, what shall we do? */
359 pr_info("get no response from backend for disable MSI\n");
360 return;
361 }
362 if (err)
363 /* how can pciback notify us fail? */
364 pr_info("get fake response from backend\n");
365}
366
367static struct xen_pci_frontend_ops pci_frontend_ops = {
368 .enable_msi = pci_frontend_enable_msi,
369 .disable_msi = pci_frontend_disable_msi,
370 .enable_msix = pci_frontend_enable_msix,
371 .disable_msix = pci_frontend_disable_msix,
372};
373
374static void pci_frontend_registrar(int enable)
375{
376 if (enable)
377 xen_pci_frontend = &pci_frontend_ops;
378 else
379 xen_pci_frontend = NULL;
380};
381#else
382static inline void pci_frontend_registrar(int enable) { };
383#endif /* CONFIG_PCI_MSI */
384
385/* Claim resources for the PCI frontend as-is, backend won't allow changes */
386static int pcifront_claim_resource(struct pci_dev *dev, void *data)
387{
388 struct pcifront_device *pdev = data;
389 int i;
390 struct resource *r;
391
392 for (i = 0; i < PCI_NUM_RESOURCES; i++) {
393 r = &dev->resource[i];
394
395 if (!r->parent && r->start && r->flags) {
396 dev_info(&pdev->xdev->dev, "claiming resource %s/%d\n",
397 pci_name(dev), i);
398 if (pci_claim_resource(dev, i)) {
399 dev_err(&pdev->xdev->dev, "Could not claim resource %s/%d! "
400 "Device offline. Try using e820_host=1 in the guest config.\n",
401 pci_name(dev), i);
402 }
403 }
404 }
405
406 return 0;
407}
408
409static int pcifront_scan_bus(struct pcifront_device *pdev,
410 unsigned int domain, unsigned int bus,
411 struct pci_bus *b)
412{
413 struct pci_dev *d;
414 unsigned int devfn;
415
416 /* Scan the bus for functions and add.
417 * We omit handling of PCI bridge attachment because pciback prevents
418 * bridges from being exported.
419 */
420 for (devfn = 0; devfn < 0x100; devfn++) {
421 d = pci_get_slot(b, devfn);
422 if (d) {
423 /* Device is already known. */
424 pci_dev_put(d);
425 continue;
426 }
427
428 d = pci_scan_single_device(b, devfn);
429 if (d)
430 dev_info(&pdev->xdev->dev, "New device on "
431 "%04x:%02x:%02x.%d found.\n", domain, bus,
432 PCI_SLOT(devfn), PCI_FUNC(devfn));
433 }
434
435 return 0;
436}
437
438static int pcifront_scan_root(struct pcifront_device *pdev,
439 unsigned int domain, unsigned int bus)
440{
441 struct pci_bus *b;
442 LIST_HEAD(resources);
443 struct pcifront_sd *sd = NULL;
444 struct pci_bus_entry *bus_entry = NULL;
445 int err = 0;
446 static struct resource busn_res = {
447 .start = 0,
448 .end = 255,
449 .flags = IORESOURCE_BUS,
450 };
451
452#ifndef CONFIG_PCI_DOMAINS
453 if (domain != 0) {
454 dev_err(&pdev->xdev->dev,
455 "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
456 dev_err(&pdev->xdev->dev,
457 "Please compile with CONFIG_PCI_DOMAINS\n");
458 err = -EINVAL;
459 goto err_out;
460 }
461#endif
462
463 dev_info(&pdev->xdev->dev, "Creating PCI Frontend Bus %04x:%02x\n",
464 domain, bus);
465
466 bus_entry = kzalloc(sizeof(*bus_entry), GFP_KERNEL);
467 sd = kzalloc(sizeof(*sd), GFP_KERNEL);
468 if (!bus_entry || !sd) {
469 err = -ENOMEM;
470 goto err_out;
471 }
472 pci_add_resource(&resources, &ioport_resource);
473 pci_add_resource(&resources, &iomem_resource);
474 pci_add_resource(&resources, &busn_res);
475 pcifront_init_sd(sd, domain, bus, pdev);
476
477 pci_lock_rescan_remove();
478
479 b = pci_scan_root_bus(&pdev->xdev->dev, bus,
480 &pcifront_bus_ops, sd, &resources);
481 if (!b) {
482 dev_err(&pdev->xdev->dev,
483 "Error creating PCI Frontend Bus!\n");
484 err = -ENOMEM;
485 pci_unlock_rescan_remove();
486 pci_free_resource_list(&resources);
487 goto err_out;
488 }
489
490 bus_entry->bus = b;
491
492 list_add(&bus_entry->list, &pdev->root_buses);
493
494 /* pci_scan_root_bus skips devices which do not have a
495 * devfn==0. The pcifront_scan_bus enumerates all devfn. */
496 err = pcifront_scan_bus(pdev, domain, bus, b);
497
498 /* Claim resources before going "live" with our devices */
499 pci_walk_bus(b, pcifront_claim_resource, pdev);
500
501 /* Create SysFS and notify udev of the devices. Aka: "going live" */
502 pci_bus_add_devices(b);
503
504 pci_unlock_rescan_remove();
505 return err;
506
507err_out:
508 kfree(bus_entry);
509 kfree(sd);
510
511 return err;
512}
513
514static int pcifront_rescan_root(struct pcifront_device *pdev,
515 unsigned int domain, unsigned int bus)
516{
517 int err;
518 struct pci_bus *b;
519
520#ifndef CONFIG_PCI_DOMAINS
521 if (domain != 0) {
522 dev_err(&pdev->xdev->dev,
523 "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
524 dev_err(&pdev->xdev->dev,
525 "Please compile with CONFIG_PCI_DOMAINS\n");
526 return -EINVAL;
527 }
528#endif
529
530 dev_info(&pdev->xdev->dev, "Rescanning PCI Frontend Bus %04x:%02x\n",
531 domain, bus);
532
533 b = pci_find_bus(domain, bus);
534 if (!b)
535 /* If the bus is unknown, create it. */
536 return pcifront_scan_root(pdev, domain, bus);
537
538 err = pcifront_scan_bus(pdev, domain, bus, b);
539
540 /* Claim resources before going "live" with our devices */
541 pci_walk_bus(b, pcifront_claim_resource, pdev);
542
543 /* Create SysFS and notify udev of the devices. Aka: "going live" */
544 pci_bus_add_devices(b);
545
546 return err;
547}
548
549static void free_root_bus_devs(struct pci_bus *bus)
550{
551 struct pci_dev *dev;
552
553 while (!list_empty(&bus->devices)) {
554 dev = container_of(bus->devices.next, struct pci_dev,
555 bus_list);
556 pci_dbg(dev, "removing device\n");
557 pci_stop_and_remove_bus_device(dev);
558 }
559}
560
561static void pcifront_free_roots(struct pcifront_device *pdev)
562{
563 struct pci_bus_entry *bus_entry, *t;
564
565 dev_dbg(&pdev->xdev->dev, "cleaning up root buses\n");
566
567 pci_lock_rescan_remove();
568 list_for_each_entry_safe(bus_entry, t, &pdev->root_buses, list) {
569 list_del(&bus_entry->list);
570
571 free_root_bus_devs(bus_entry->bus);
572
573 kfree(bus_entry->bus->sysdata);
574
575 device_unregister(bus_entry->bus->bridge);
576 pci_remove_bus(bus_entry->bus);
577
578 kfree(bus_entry);
579 }
580 pci_unlock_rescan_remove();
581}
582
583static pci_ers_result_t pcifront_common_process(int cmd,
584 struct pcifront_device *pdev,
585 pci_channel_state_t state)
586{
587 pci_ers_result_t result;
588 struct pci_driver *pdrv;
589 int bus = pdev->sh_info->aer_op.bus;
590 int devfn = pdev->sh_info->aer_op.devfn;
591 int domain = pdev->sh_info->aer_op.domain;
592 struct pci_dev *pcidev;
593 int flag = 0;
594
595 dev_dbg(&pdev->xdev->dev,
596 "pcifront AER process: cmd %x (bus:%x, devfn%x)",
597 cmd, bus, devfn);
598 result = PCI_ERS_RESULT_NONE;
599
600 pcidev = pci_get_domain_bus_and_slot(domain, bus, devfn);
601 if (!pcidev || !pcidev->driver) {
602 dev_err(&pdev->xdev->dev, "device or AER driver is NULL\n");
603 pci_dev_put(pcidev);
604 return result;
605 }
606 pdrv = pcidev->driver;
607
608 if (pdrv) {
609 if (pdrv->err_handler && pdrv->err_handler->error_detected) {
610 pci_dbg(pcidev, "trying to call AER service\n");
611 if (pcidev) {
612 flag = 1;
613 switch (cmd) {
614 case XEN_PCI_OP_aer_detected:
615 result = pdrv->err_handler->
616 error_detected(pcidev, state);
617 break;
618 case XEN_PCI_OP_aer_mmio:
619 result = pdrv->err_handler->
620 mmio_enabled(pcidev);
621 break;
622 case XEN_PCI_OP_aer_slotreset:
623 result = pdrv->err_handler->
624 slot_reset(pcidev);
625 break;
626 case XEN_PCI_OP_aer_resume:
627 pdrv->err_handler->resume(pcidev);
628 break;
629 default:
630 dev_err(&pdev->xdev->dev,
631 "bad request in aer recovery "
632 "operation!\n");
633
634 }
635 }
636 }
637 }
638 if (!flag)
639 result = PCI_ERS_RESULT_NONE;
640
641 return result;
642}
643
644
645static void pcifront_do_aer(struct work_struct *data)
646{
647 struct pcifront_device *pdev =
648 container_of(data, struct pcifront_device, op_work);
649 int cmd = pdev->sh_info->aer_op.cmd;
650 pci_channel_state_t state =
651 (pci_channel_state_t)pdev->sh_info->aer_op.err;
652
653 /*If a pci_conf op is in progress,
654 we have to wait until it is done before service aer op*/
655 dev_dbg(&pdev->xdev->dev,
656 "pcifront service aer bus %x devfn %x\n",
657 pdev->sh_info->aer_op.bus, pdev->sh_info->aer_op.devfn);
658
659 pdev->sh_info->aer_op.err = pcifront_common_process(cmd, pdev, state);
660
661 /* Post the operation to the guest. */
662 wmb();
663 clear_bit(_XEN_PCIB_active, (unsigned long *)&pdev->sh_info->flags);
664 notify_remote_via_evtchn(pdev->evtchn);
665
666 /*in case of we lost an aer request in four lines time_window*/
667 smp_mb__before_atomic();
668 clear_bit(_PDEVB_op_active, &pdev->flags);
669 smp_mb__after_atomic();
670
671 schedule_pcifront_aer_op(pdev);
672
673}
674
675static irqreturn_t pcifront_handler_aer(int irq, void *dev)
676{
677 struct pcifront_device *pdev = dev;
678 schedule_pcifront_aer_op(pdev);
679 return IRQ_HANDLED;
680}
681static int pcifront_connect_and_init_dma(struct pcifront_device *pdev)
682{
683 int err = 0;
684
685 spin_lock(&pcifront_dev_lock);
686
687 if (!pcifront_dev) {
688 dev_info(&pdev->xdev->dev, "Installing PCI frontend\n");
689 pcifront_dev = pdev;
690 } else
691 err = -EEXIST;
692
693 spin_unlock(&pcifront_dev_lock);
694
695 if (!err && !swiotlb_nr_tbl()) {
696 err = pci_xen_swiotlb_init_late();
697 if (err)
698 dev_err(&pdev->xdev->dev, "Could not setup SWIOTLB!\n");
699 }
700 return err;
701}
702
703static void pcifront_disconnect(struct pcifront_device *pdev)
704{
705 spin_lock(&pcifront_dev_lock);
706
707 if (pdev == pcifront_dev) {
708 dev_info(&pdev->xdev->dev,
709 "Disconnecting PCI Frontend Buses\n");
710 pcifront_dev = NULL;
711 }
712
713 spin_unlock(&pcifront_dev_lock);
714}
715static struct pcifront_device *alloc_pdev(struct xenbus_device *xdev)
716{
717 struct pcifront_device *pdev;
718
719 pdev = kzalloc(sizeof(struct pcifront_device), GFP_KERNEL);
720 if (pdev == NULL)
721 goto out;
722
723 pdev->sh_info =
724 (struct xen_pci_sharedinfo *)__get_free_page(GFP_KERNEL);
725 if (pdev->sh_info == NULL) {
726 kfree(pdev);
727 pdev = NULL;
728 goto out;
729 }
730 pdev->sh_info->flags = 0;
731
732 /*Flag for registering PV AER handler*/
733 set_bit(_XEN_PCIB_AERHANDLER, (void *)&pdev->sh_info->flags);
734
735 dev_set_drvdata(&xdev->dev, pdev);
736 pdev->xdev = xdev;
737
738 INIT_LIST_HEAD(&pdev->root_buses);
739
740 spin_lock_init(&pdev->sh_info_lock);
741
742 pdev->evtchn = INVALID_EVTCHN;
743 pdev->gnt_ref = INVALID_GRANT_REF;
744 pdev->irq = -1;
745
746 INIT_WORK(&pdev->op_work, pcifront_do_aer);
747
748 dev_dbg(&xdev->dev, "Allocated pdev @ 0x%p pdev->sh_info @ 0x%p\n",
749 pdev, pdev->sh_info);
750out:
751 return pdev;
752}
753
754static void free_pdev(struct pcifront_device *pdev)
755{
756 dev_dbg(&pdev->xdev->dev, "freeing pdev @ 0x%p\n", pdev);
757
758 pcifront_free_roots(pdev);
759
760 cancel_work_sync(&pdev->op_work);
761
762 if (pdev->irq >= 0)
763 unbind_from_irqhandler(pdev->irq, pdev);
764
765 if (pdev->evtchn != INVALID_EVTCHN)
766 xenbus_free_evtchn(pdev->xdev, pdev->evtchn);
767
768 if (pdev->gnt_ref != INVALID_GRANT_REF)
769 gnttab_end_foreign_access(pdev->gnt_ref, 0 /* r/w page */,
770 (unsigned long)pdev->sh_info);
771 else
772 free_page((unsigned long)pdev->sh_info);
773
774 dev_set_drvdata(&pdev->xdev->dev, NULL);
775
776 kfree(pdev);
777}
778
779static int pcifront_publish_info(struct pcifront_device *pdev)
780{
781 int err = 0;
782 struct xenbus_transaction trans;
783 grant_ref_t gref;
784
785 err = xenbus_grant_ring(pdev->xdev, pdev->sh_info, 1, &gref);
786 if (err < 0)
787 goto out;
788
789 pdev->gnt_ref = gref;
790
791 err = xenbus_alloc_evtchn(pdev->xdev, &pdev->evtchn);
792 if (err)
793 goto out;
794
795 err = bind_evtchn_to_irqhandler(pdev->evtchn, pcifront_handler_aer,
796 0, "pcifront", pdev);
797
798 if (err < 0)
799 return err;
800
801 pdev->irq = err;
802
803do_publish:
804 err = xenbus_transaction_start(&trans);
805 if (err) {
806 xenbus_dev_fatal(pdev->xdev, err,
807 "Error writing configuration for backend "
808 "(start transaction)");
809 goto out;
810 }
811
812 err = xenbus_printf(trans, pdev->xdev->nodename,
813 "pci-op-ref", "%u", pdev->gnt_ref);
814 if (!err)
815 err = xenbus_printf(trans, pdev->xdev->nodename,
816 "event-channel", "%u", pdev->evtchn);
817 if (!err)
818 err = xenbus_printf(trans, pdev->xdev->nodename,
819 "magic", XEN_PCI_MAGIC);
820
821 if (err) {
822 xenbus_transaction_end(trans, 1);
823 xenbus_dev_fatal(pdev->xdev, err,
824 "Error writing configuration for backend");
825 goto out;
826 } else {
827 err = xenbus_transaction_end(trans, 0);
828 if (err == -EAGAIN)
829 goto do_publish;
830 else if (err) {
831 xenbus_dev_fatal(pdev->xdev, err,
832 "Error completing transaction "
833 "for backend");
834 goto out;
835 }
836 }
837
838 xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
839
840 dev_dbg(&pdev->xdev->dev, "publishing successful!\n");
841
842out:
843 return err;
844}
845
846static int pcifront_try_connect(struct pcifront_device *pdev)
847{
848 int err = -EFAULT;
849 int i, num_roots, len;
850 char str[64];
851 unsigned int domain, bus;
852
853
854 /* Only connect once */
855 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
856 XenbusStateInitialised)
857 goto out;
858
859 err = pcifront_connect_and_init_dma(pdev);
860 if (err && err != -EEXIST) {
861 xenbus_dev_fatal(pdev->xdev, err,
862 "Error setting up PCI Frontend");
863 goto out;
864 }
865
866 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
867 "root_num", "%d", &num_roots);
868 if (err == -ENOENT) {
869 xenbus_dev_error(pdev->xdev, err,
870 "No PCI Roots found, trying 0000:00");
871 err = pcifront_scan_root(pdev, 0, 0);
872 if (err) {
873 xenbus_dev_fatal(pdev->xdev, err,
874 "Error scanning PCI root 0000:00");
875 goto out;
876 }
877 num_roots = 0;
878 } else if (err != 1) {
879 if (err == 0)
880 err = -EINVAL;
881 xenbus_dev_fatal(pdev->xdev, err,
882 "Error reading number of PCI roots");
883 goto out;
884 }
885
886 for (i = 0; i < num_roots; i++) {
887 len = snprintf(str, sizeof(str), "root-%d", i);
888 if (unlikely(len >= (sizeof(str) - 1))) {
889 err = -ENOMEM;
890 goto out;
891 }
892
893 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
894 "%x:%x", &domain, &bus);
895 if (err != 2) {
896 if (err >= 0)
897 err = -EINVAL;
898 xenbus_dev_fatal(pdev->xdev, err,
899 "Error reading PCI root %d", i);
900 goto out;
901 }
902
903 err = pcifront_scan_root(pdev, domain, bus);
904 if (err) {
905 xenbus_dev_fatal(pdev->xdev, err,
906 "Error scanning PCI root %04x:%02x",
907 domain, bus);
908 goto out;
909 }
910 }
911
912 err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
913
914out:
915 return err;
916}
917
918static int pcifront_try_disconnect(struct pcifront_device *pdev)
919{
920 int err = 0;
921 enum xenbus_state prev_state;
922
923
924 prev_state = xenbus_read_driver_state(pdev->xdev->nodename);
925
926 if (prev_state >= XenbusStateClosing)
927 goto out;
928
929 if (prev_state == XenbusStateConnected) {
930 pcifront_free_roots(pdev);
931 pcifront_disconnect(pdev);
932 }
933
934 err = xenbus_switch_state(pdev->xdev, XenbusStateClosed);
935
936out:
937
938 return err;
939}
940
941static int pcifront_attach_devices(struct pcifront_device *pdev)
942{
943 int err = -EFAULT;
944 int i, num_roots, len;
945 unsigned int domain, bus;
946 char str[64];
947
948 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
949 XenbusStateReconfiguring)
950 goto out;
951
952 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
953 "root_num", "%d", &num_roots);
954 if (err == -ENOENT) {
955 xenbus_dev_error(pdev->xdev, err,
956 "No PCI Roots found, trying 0000:00");
957 err = pcifront_rescan_root(pdev, 0, 0);
958 if (err) {
959 xenbus_dev_fatal(pdev->xdev, err,
960 "Error scanning PCI root 0000:00");
961 goto out;
962 }
963 num_roots = 0;
964 } else if (err != 1) {
965 if (err == 0)
966 err = -EINVAL;
967 xenbus_dev_fatal(pdev->xdev, err,
968 "Error reading number of PCI roots");
969 goto out;
970 }
971
972 for (i = 0; i < num_roots; i++) {
973 len = snprintf(str, sizeof(str), "root-%d", i);
974 if (unlikely(len >= (sizeof(str) - 1))) {
975 err = -ENOMEM;
976 goto out;
977 }
978
979 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
980 "%x:%x", &domain, &bus);
981 if (err != 2) {
982 if (err >= 0)
983 err = -EINVAL;
984 xenbus_dev_fatal(pdev->xdev, err,
985 "Error reading PCI root %d", i);
986 goto out;
987 }
988
989 err = pcifront_rescan_root(pdev, domain, bus);
990 if (err) {
991 xenbus_dev_fatal(pdev->xdev, err,
992 "Error scanning PCI root %04x:%02x",
993 domain, bus);
994 goto out;
995 }
996 }
997
998 xenbus_switch_state(pdev->xdev, XenbusStateConnected);
999
1000out:
1001 return err;
1002}
1003
1004static int pcifront_detach_devices(struct pcifront_device *pdev)
1005{
1006 int err = 0;
1007 int i, num_devs;
1008 unsigned int domain, bus, slot, func;
1009 struct pci_dev *pci_dev;
1010 char str[64];
1011
1012 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
1013 XenbusStateConnected)
1014 goto out;
1015
1016 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "num_devs", "%d",
1017 &num_devs);
1018 if (err != 1) {
1019 if (err >= 0)
1020 err = -EINVAL;
1021 xenbus_dev_fatal(pdev->xdev, err,
1022 "Error reading number of PCI devices");
1023 goto out;
1024 }
1025
1026 /* Find devices being detached and remove them. */
1027 for (i = 0; i < num_devs; i++) {
1028 int l, state;
1029 l = snprintf(str, sizeof(str), "state-%d", i);
1030 if (unlikely(l >= (sizeof(str) - 1))) {
1031 err = -ENOMEM;
1032 goto out;
1033 }
1034 state = xenbus_read_unsigned(pdev->xdev->otherend, str,
1035 XenbusStateUnknown);
1036
1037 if (state != XenbusStateClosing)
1038 continue;
1039
1040 /* Remove device. */
1041 l = snprintf(str, sizeof(str), "vdev-%d", i);
1042 if (unlikely(l >= (sizeof(str) - 1))) {
1043 err = -ENOMEM;
1044 goto out;
1045 }
1046 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
1047 "%x:%x:%x.%x", &domain, &bus, &slot, &func);
1048 if (err != 4) {
1049 if (err >= 0)
1050 err = -EINVAL;
1051 xenbus_dev_fatal(pdev->xdev, err,
1052 "Error reading PCI device %d", i);
1053 goto out;
1054 }
1055
1056 pci_dev = pci_get_domain_bus_and_slot(domain, bus,
1057 PCI_DEVFN(slot, func));
1058 if (!pci_dev) {
1059 dev_dbg(&pdev->xdev->dev,
1060 "Cannot get PCI device %04x:%02x:%02x.%d\n",
1061 domain, bus, slot, func);
1062 continue;
1063 }
1064 pci_lock_rescan_remove();
1065 pci_stop_and_remove_bus_device(pci_dev);
1066 pci_dev_put(pci_dev);
1067 pci_unlock_rescan_remove();
1068
1069 dev_dbg(&pdev->xdev->dev,
1070 "PCI device %04x:%02x:%02x.%d removed.\n",
1071 domain, bus, slot, func);
1072 }
1073
1074 err = xenbus_switch_state(pdev->xdev, XenbusStateReconfiguring);
1075
1076out:
1077 return err;
1078}
1079
1080static void __ref pcifront_backend_changed(struct xenbus_device *xdev,
1081 enum xenbus_state be_state)
1082{
1083 struct pcifront_device *pdev = dev_get_drvdata(&xdev->dev);
1084
1085 switch (be_state) {
1086 case XenbusStateUnknown:
1087 case XenbusStateInitialising:
1088 case XenbusStateInitWait:
1089 case XenbusStateInitialised:
1090 break;
1091
1092 case XenbusStateConnected:
1093 pcifront_try_connect(pdev);
1094 break;
1095
1096 case XenbusStateClosed:
1097 if (xdev->state == XenbusStateClosed)
1098 break;
1099 fallthrough; /* Missed the backend's CLOSING state */
1100 case XenbusStateClosing:
1101 dev_warn(&xdev->dev, "backend going away!\n");
1102 pcifront_try_disconnect(pdev);
1103 break;
1104
1105 case XenbusStateReconfiguring:
1106 pcifront_detach_devices(pdev);
1107 break;
1108
1109 case XenbusStateReconfigured:
1110 pcifront_attach_devices(pdev);
1111 break;
1112 }
1113}
1114
1115static int pcifront_xenbus_probe(struct xenbus_device *xdev,
1116 const struct xenbus_device_id *id)
1117{
1118 int err = 0;
1119 struct pcifront_device *pdev = alloc_pdev(xdev);
1120
1121 if (pdev == NULL) {
1122 err = -ENOMEM;
1123 xenbus_dev_fatal(xdev, err,
1124 "Error allocating pcifront_device struct");
1125 goto out;
1126 }
1127
1128 err = pcifront_publish_info(pdev);
1129 if (err)
1130 free_pdev(pdev);
1131
1132out:
1133 return err;
1134}
1135
1136static int pcifront_xenbus_remove(struct xenbus_device *xdev)
1137{
1138 struct pcifront_device *pdev = dev_get_drvdata(&xdev->dev);
1139 if (pdev)
1140 free_pdev(pdev);
1141
1142 return 0;
1143}
1144
1145static const struct xenbus_device_id xenpci_ids[] = {
1146 {"pci"},
1147 {""},
1148};
1149
1150static struct xenbus_driver xenpci_driver = {
1151 .name = "pcifront",
1152 .ids = xenpci_ids,
1153 .probe = pcifront_xenbus_probe,
1154 .remove = pcifront_xenbus_remove,
1155 .otherend_changed = pcifront_backend_changed,
1156};
1157
1158static int __init pcifront_init(void)
1159{
1160 if (!xen_pv_domain() || xen_initial_domain())
1161 return -ENODEV;
1162
1163 if (!xen_has_pv_devices())
1164 return -ENODEV;
1165
1166 pci_frontend_registrar(1 /* enable */);
1167
1168 return xenbus_register_frontend(&xenpci_driver);
1169}
1170
1171static void __exit pcifront_cleanup(void)
1172{
1173 xenbus_unregister_driver(&xenpci_driver);
1174 pci_frontend_registrar(0 /* disable */);
1175}
1176module_init(pcifront_init);
1177module_exit(pcifront_cleanup);
1178
1179MODULE_DESCRIPTION("Xen PCI passthrough frontend.");
1180MODULE_LICENSE("GPL");
1181MODULE_ALIAS("xen:pci");
1/*
2 * Xen PCI Frontend.
3 *
4 * Author: Ryan Wilson <hap9@epoch.ncsc.mil>
5 */
6#include <linux/module.h>
7#include <linux/init.h>
8#include <linux/mm.h>
9#include <xen/xenbus.h>
10#include <xen/events.h>
11#include <xen/grant_table.h>
12#include <xen/page.h>
13#include <linux/spinlock.h>
14#include <linux/pci.h>
15#include <linux/msi.h>
16#include <xen/interface/io/pciif.h>
17#include <asm/xen/pci.h>
18#include <linux/interrupt.h>
19#include <linux/atomic.h>
20#include <linux/workqueue.h>
21#include <linux/bitops.h>
22#include <linux/time.h>
23#include <xen/platform_pci.h>
24
25#include <asm/xen/swiotlb-xen.h>
26#define INVALID_GRANT_REF (0)
27#define INVALID_EVTCHN (-1)
28
29struct pci_bus_entry {
30 struct list_head list;
31 struct pci_bus *bus;
32};
33
34#define _PDEVB_op_active (0)
35#define PDEVB_op_active (1 << (_PDEVB_op_active))
36
37struct pcifront_device {
38 struct xenbus_device *xdev;
39 struct list_head root_buses;
40
41 int evtchn;
42 int gnt_ref;
43
44 int irq;
45
46 /* Lock this when doing any operations in sh_info */
47 spinlock_t sh_info_lock;
48 struct xen_pci_sharedinfo *sh_info;
49 struct work_struct op_work;
50 unsigned long flags;
51
52};
53
54struct pcifront_sd {
55 int domain;
56 struct pcifront_device *pdev;
57};
58
59static inline struct pcifront_device *
60pcifront_get_pdev(struct pcifront_sd *sd)
61{
62 return sd->pdev;
63}
64
65static inline void pcifront_init_sd(struct pcifront_sd *sd,
66 unsigned int domain, unsigned int bus,
67 struct pcifront_device *pdev)
68{
69 sd->domain = domain;
70 sd->pdev = pdev;
71}
72
73static DEFINE_SPINLOCK(pcifront_dev_lock);
74static struct pcifront_device *pcifront_dev;
75
76static int verbose_request;
77module_param(verbose_request, int, 0644);
78
79static int errno_to_pcibios_err(int errno)
80{
81 switch (errno) {
82 case XEN_PCI_ERR_success:
83 return PCIBIOS_SUCCESSFUL;
84
85 case XEN_PCI_ERR_dev_not_found:
86 return PCIBIOS_DEVICE_NOT_FOUND;
87
88 case XEN_PCI_ERR_invalid_offset:
89 case XEN_PCI_ERR_op_failed:
90 return PCIBIOS_BAD_REGISTER_NUMBER;
91
92 case XEN_PCI_ERR_not_implemented:
93 return PCIBIOS_FUNC_NOT_SUPPORTED;
94
95 case XEN_PCI_ERR_access_denied:
96 return PCIBIOS_SET_FAILED;
97 }
98 return errno;
99}
100
101static inline void schedule_pcifront_aer_op(struct pcifront_device *pdev)
102{
103 if (test_bit(_XEN_PCIB_active, (unsigned long *)&pdev->sh_info->flags)
104 && !test_and_set_bit(_PDEVB_op_active, &pdev->flags)) {
105 dev_dbg(&pdev->xdev->dev, "schedule aer frontend job\n");
106 schedule_work(&pdev->op_work);
107 }
108}
109
110static int do_pci_op(struct pcifront_device *pdev, struct xen_pci_op *op)
111{
112 int err = 0;
113 struct xen_pci_op *active_op = &pdev->sh_info->op;
114 unsigned long irq_flags;
115 evtchn_port_t port = pdev->evtchn;
116 unsigned irq = pdev->irq;
117 s64 ns, ns_timeout;
118 struct timeval tv;
119
120 spin_lock_irqsave(&pdev->sh_info_lock, irq_flags);
121
122 memcpy(active_op, op, sizeof(struct xen_pci_op));
123
124 /* Go */
125 wmb();
126 set_bit(_XEN_PCIF_active, (unsigned long *)&pdev->sh_info->flags);
127 notify_remote_via_evtchn(port);
128
129 /*
130 * We set a poll timeout of 3 seconds but give up on return after
131 * 2 seconds. It is better to time out too late rather than too early
132 * (in the latter case we end up continually re-executing poll() with a
133 * timeout in the past). 1s difference gives plenty of slack for error.
134 */
135 do_gettimeofday(&tv);
136 ns_timeout = timeval_to_ns(&tv) + 2 * (s64)NSEC_PER_SEC;
137
138 xen_clear_irq_pending(irq);
139
140 while (test_bit(_XEN_PCIF_active,
141 (unsigned long *)&pdev->sh_info->flags)) {
142 xen_poll_irq_timeout(irq, jiffies + 3*HZ);
143 xen_clear_irq_pending(irq);
144 do_gettimeofday(&tv);
145 ns = timeval_to_ns(&tv);
146 if (ns > ns_timeout) {
147 dev_err(&pdev->xdev->dev,
148 "pciback not responding!!!\n");
149 clear_bit(_XEN_PCIF_active,
150 (unsigned long *)&pdev->sh_info->flags);
151 err = XEN_PCI_ERR_dev_not_found;
152 goto out;
153 }
154 }
155
156 /*
157 * We might lose backend service request since we
158 * reuse same evtchn with pci_conf backend response. So re-schedule
159 * aer pcifront service.
160 */
161 if (test_bit(_XEN_PCIB_active,
162 (unsigned long *)&pdev->sh_info->flags)) {
163 dev_err(&pdev->xdev->dev,
164 "schedule aer pcifront service\n");
165 schedule_pcifront_aer_op(pdev);
166 }
167
168 memcpy(op, active_op, sizeof(struct xen_pci_op));
169
170 err = op->err;
171out:
172 spin_unlock_irqrestore(&pdev->sh_info_lock, irq_flags);
173 return err;
174}
175
176/* Access to this function is spinlocked in drivers/pci/access.c */
177static int pcifront_bus_read(struct pci_bus *bus, unsigned int devfn,
178 int where, int size, u32 *val)
179{
180 int err = 0;
181 struct xen_pci_op op = {
182 .cmd = XEN_PCI_OP_conf_read,
183 .domain = pci_domain_nr(bus),
184 .bus = bus->number,
185 .devfn = devfn,
186 .offset = where,
187 .size = size,
188 };
189 struct pcifront_sd *sd = bus->sysdata;
190 struct pcifront_device *pdev = pcifront_get_pdev(sd);
191
192 if (verbose_request)
193 dev_info(&pdev->xdev->dev,
194 "read dev=%04x:%02x:%02x.%d - offset %x size %d\n",
195 pci_domain_nr(bus), bus->number, PCI_SLOT(devfn),
196 PCI_FUNC(devfn), where, size);
197
198 err = do_pci_op(pdev, &op);
199
200 if (likely(!err)) {
201 if (verbose_request)
202 dev_info(&pdev->xdev->dev, "read got back value %x\n",
203 op.value);
204
205 *val = op.value;
206 } else if (err == -ENODEV) {
207 /* No device here, pretend that it just returned 0 */
208 err = 0;
209 *val = 0;
210 }
211
212 return errno_to_pcibios_err(err);
213}
214
215/* Access to this function is spinlocked in drivers/pci/access.c */
216static int pcifront_bus_write(struct pci_bus *bus, unsigned int devfn,
217 int where, int size, u32 val)
218{
219 struct xen_pci_op op = {
220 .cmd = XEN_PCI_OP_conf_write,
221 .domain = pci_domain_nr(bus),
222 .bus = bus->number,
223 .devfn = devfn,
224 .offset = where,
225 .size = size,
226 .value = val,
227 };
228 struct pcifront_sd *sd = bus->sysdata;
229 struct pcifront_device *pdev = pcifront_get_pdev(sd);
230
231 if (verbose_request)
232 dev_info(&pdev->xdev->dev,
233 "write dev=%04x:%02x:%02x.%d - "
234 "offset %x size %d val %x\n",
235 pci_domain_nr(bus), bus->number,
236 PCI_SLOT(devfn), PCI_FUNC(devfn), where, size, val);
237
238 return errno_to_pcibios_err(do_pci_op(pdev, &op));
239}
240
241static struct pci_ops pcifront_bus_ops = {
242 .read = pcifront_bus_read,
243 .write = pcifront_bus_write,
244};
245
246#ifdef CONFIG_PCI_MSI
247static int pci_frontend_enable_msix(struct pci_dev *dev,
248 int vector[], int nvec)
249{
250 int err;
251 int i;
252 struct xen_pci_op op = {
253 .cmd = XEN_PCI_OP_enable_msix,
254 .domain = pci_domain_nr(dev->bus),
255 .bus = dev->bus->number,
256 .devfn = dev->devfn,
257 .value = nvec,
258 };
259 struct pcifront_sd *sd = dev->bus->sysdata;
260 struct pcifront_device *pdev = pcifront_get_pdev(sd);
261 struct msi_desc *entry;
262
263 if (nvec > SH_INFO_MAX_VEC) {
264 dev_err(&dev->dev, "too much vector for pci frontend: %x."
265 " Increase SH_INFO_MAX_VEC.\n", nvec);
266 return -EINVAL;
267 }
268
269 i = 0;
270 list_for_each_entry(entry, &dev->msi_list, list) {
271 op.msix_entries[i].entry = entry->msi_attrib.entry_nr;
272 /* Vector is useless at this point. */
273 op.msix_entries[i].vector = -1;
274 i++;
275 }
276
277 err = do_pci_op(pdev, &op);
278
279 if (likely(!err)) {
280 if (likely(!op.value)) {
281 /* we get the result */
282 for (i = 0; i < nvec; i++) {
283 if (op.msix_entries[i].vector <= 0) {
284 dev_warn(&dev->dev, "MSI-X entry %d is invalid: %d!\n",
285 i, op.msix_entries[i].vector);
286 err = -EINVAL;
287 vector[i] = -1;
288 continue;
289 }
290 vector[i] = op.msix_entries[i].vector;
291 }
292 } else {
293 printk(KERN_DEBUG "enable msix get value %x\n",
294 op.value);
295 err = op.value;
296 }
297 } else {
298 dev_err(&dev->dev, "enable msix get err %x\n", err);
299 }
300 return err;
301}
302
303static void pci_frontend_disable_msix(struct pci_dev *dev)
304{
305 int err;
306 struct xen_pci_op op = {
307 .cmd = XEN_PCI_OP_disable_msix,
308 .domain = pci_domain_nr(dev->bus),
309 .bus = dev->bus->number,
310 .devfn = dev->devfn,
311 };
312 struct pcifront_sd *sd = dev->bus->sysdata;
313 struct pcifront_device *pdev = pcifront_get_pdev(sd);
314
315 err = do_pci_op(pdev, &op);
316
317 /* What should do for error ? */
318 if (err)
319 dev_err(&dev->dev, "pci_disable_msix get err %x\n", err);
320}
321
322static int pci_frontend_enable_msi(struct pci_dev *dev, int vector[])
323{
324 int err;
325 struct xen_pci_op op = {
326 .cmd = XEN_PCI_OP_enable_msi,
327 .domain = pci_domain_nr(dev->bus),
328 .bus = dev->bus->number,
329 .devfn = dev->devfn,
330 };
331 struct pcifront_sd *sd = dev->bus->sysdata;
332 struct pcifront_device *pdev = pcifront_get_pdev(sd);
333
334 err = do_pci_op(pdev, &op);
335 if (likely(!err)) {
336 vector[0] = op.value;
337 if (op.value <= 0) {
338 dev_warn(&dev->dev, "MSI entry is invalid: %d!\n",
339 op.value);
340 err = -EINVAL;
341 vector[0] = -1;
342 }
343 } else {
344 dev_err(&dev->dev, "pci frontend enable msi failed for dev "
345 "%x:%x\n", op.bus, op.devfn);
346 err = -EINVAL;
347 }
348 return err;
349}
350
351static void pci_frontend_disable_msi(struct pci_dev *dev)
352{
353 int err;
354 struct xen_pci_op op = {
355 .cmd = XEN_PCI_OP_disable_msi,
356 .domain = pci_domain_nr(dev->bus),
357 .bus = dev->bus->number,
358 .devfn = dev->devfn,
359 };
360 struct pcifront_sd *sd = dev->bus->sysdata;
361 struct pcifront_device *pdev = pcifront_get_pdev(sd);
362
363 err = do_pci_op(pdev, &op);
364 if (err == XEN_PCI_ERR_dev_not_found) {
365 /* XXX No response from backend, what shall we do? */
366 printk(KERN_DEBUG "get no response from backend for disable MSI\n");
367 return;
368 }
369 if (err)
370 /* how can pciback notify us fail? */
371 printk(KERN_DEBUG "get fake response frombackend\n");
372}
373
374static struct xen_pci_frontend_ops pci_frontend_ops = {
375 .enable_msi = pci_frontend_enable_msi,
376 .disable_msi = pci_frontend_disable_msi,
377 .enable_msix = pci_frontend_enable_msix,
378 .disable_msix = pci_frontend_disable_msix,
379};
380
381static void pci_frontend_registrar(int enable)
382{
383 if (enable)
384 xen_pci_frontend = &pci_frontend_ops;
385 else
386 xen_pci_frontend = NULL;
387};
388#else
389static inline void pci_frontend_registrar(int enable) { };
390#endif /* CONFIG_PCI_MSI */
391
392/* Claim resources for the PCI frontend as-is, backend won't allow changes */
393static int pcifront_claim_resource(struct pci_dev *dev, void *data)
394{
395 struct pcifront_device *pdev = data;
396 int i;
397 struct resource *r;
398
399 for (i = 0; i < PCI_NUM_RESOURCES; i++) {
400 r = &dev->resource[i];
401
402 if (!r->parent && r->start && r->flags) {
403 dev_info(&pdev->xdev->dev, "claiming resource %s/%d\n",
404 pci_name(dev), i);
405 if (pci_claim_resource(dev, i)) {
406 dev_err(&pdev->xdev->dev, "Could not claim resource %s/%d! "
407 "Device offline. Try using e820_host=1 in the guest config.\n",
408 pci_name(dev), i);
409 }
410 }
411 }
412
413 return 0;
414}
415
416static int pcifront_scan_bus(struct pcifront_device *pdev,
417 unsigned int domain, unsigned int bus,
418 struct pci_bus *b)
419{
420 struct pci_dev *d;
421 unsigned int devfn;
422
423 /* Scan the bus for functions and add.
424 * We omit handling of PCI bridge attachment because pciback prevents
425 * bridges from being exported.
426 */
427 for (devfn = 0; devfn < 0x100; devfn++) {
428 d = pci_get_slot(b, devfn);
429 if (d) {
430 /* Device is already known. */
431 pci_dev_put(d);
432 continue;
433 }
434
435 d = pci_scan_single_device(b, devfn);
436 if (d)
437 dev_info(&pdev->xdev->dev, "New device on "
438 "%04x:%02x:%02x.%d found.\n", domain, bus,
439 PCI_SLOT(devfn), PCI_FUNC(devfn));
440 }
441
442 return 0;
443}
444
445static int pcifront_scan_root(struct pcifront_device *pdev,
446 unsigned int domain, unsigned int bus)
447{
448 struct pci_bus *b;
449 struct pcifront_sd *sd = NULL;
450 struct pci_bus_entry *bus_entry = NULL;
451 int err = 0;
452
453#ifndef CONFIG_PCI_DOMAINS
454 if (domain != 0) {
455 dev_err(&pdev->xdev->dev,
456 "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
457 dev_err(&pdev->xdev->dev,
458 "Please compile with CONFIG_PCI_DOMAINS\n");
459 err = -EINVAL;
460 goto err_out;
461 }
462#endif
463
464 dev_info(&pdev->xdev->dev, "Creating PCI Frontend Bus %04x:%02x\n",
465 domain, bus);
466
467 bus_entry = kmalloc(sizeof(*bus_entry), GFP_KERNEL);
468 sd = kmalloc(sizeof(*sd), GFP_KERNEL);
469 if (!bus_entry || !sd) {
470 err = -ENOMEM;
471 goto err_out;
472 }
473 pcifront_init_sd(sd, domain, bus, pdev);
474
475 pci_lock_rescan_remove();
476
477 b = pci_scan_bus_parented(&pdev->xdev->dev, bus,
478 &pcifront_bus_ops, sd);
479 if (!b) {
480 dev_err(&pdev->xdev->dev,
481 "Error creating PCI Frontend Bus!\n");
482 err = -ENOMEM;
483 pci_unlock_rescan_remove();
484 goto err_out;
485 }
486
487 bus_entry->bus = b;
488
489 list_add(&bus_entry->list, &pdev->root_buses);
490
491 /* pci_scan_bus_parented skips devices which do not have a have
492 * devfn==0. The pcifront_scan_bus enumerates all devfn. */
493 err = pcifront_scan_bus(pdev, domain, bus, b);
494
495 /* Claim resources before going "live" with our devices */
496 pci_walk_bus(b, pcifront_claim_resource, pdev);
497
498 /* Create SysFS and notify udev of the devices. Aka: "going live" */
499 pci_bus_add_devices(b);
500
501 pci_unlock_rescan_remove();
502 return err;
503
504err_out:
505 kfree(bus_entry);
506 kfree(sd);
507
508 return err;
509}
510
511static int pcifront_rescan_root(struct pcifront_device *pdev,
512 unsigned int domain, unsigned int bus)
513{
514 int err;
515 struct pci_bus *b;
516
517#ifndef CONFIG_PCI_DOMAINS
518 if (domain != 0) {
519 dev_err(&pdev->xdev->dev,
520 "PCI Root in non-zero PCI Domain! domain=%d\n", domain);
521 dev_err(&pdev->xdev->dev,
522 "Please compile with CONFIG_PCI_DOMAINS\n");
523 return -EINVAL;
524 }
525#endif
526
527 dev_info(&pdev->xdev->dev, "Rescanning PCI Frontend Bus %04x:%02x\n",
528 domain, bus);
529
530 b = pci_find_bus(domain, bus);
531 if (!b)
532 /* If the bus is unknown, create it. */
533 return pcifront_scan_root(pdev, domain, bus);
534
535 err = pcifront_scan_bus(pdev, domain, bus, b);
536
537 /* Claim resources before going "live" with our devices */
538 pci_walk_bus(b, pcifront_claim_resource, pdev);
539
540 /* Create SysFS and notify udev of the devices. Aka: "going live" */
541 pci_bus_add_devices(b);
542
543 return err;
544}
545
546static void free_root_bus_devs(struct pci_bus *bus)
547{
548 struct pci_dev *dev;
549
550 while (!list_empty(&bus->devices)) {
551 dev = container_of(bus->devices.next, struct pci_dev,
552 bus_list);
553 dev_dbg(&dev->dev, "removing device\n");
554 pci_stop_and_remove_bus_device(dev);
555 }
556}
557
558static void pcifront_free_roots(struct pcifront_device *pdev)
559{
560 struct pci_bus_entry *bus_entry, *t;
561
562 dev_dbg(&pdev->xdev->dev, "cleaning up root buses\n");
563
564 pci_lock_rescan_remove();
565 list_for_each_entry_safe(bus_entry, t, &pdev->root_buses, list) {
566 list_del(&bus_entry->list);
567
568 free_root_bus_devs(bus_entry->bus);
569
570 kfree(bus_entry->bus->sysdata);
571
572 device_unregister(bus_entry->bus->bridge);
573 pci_remove_bus(bus_entry->bus);
574
575 kfree(bus_entry);
576 }
577 pci_unlock_rescan_remove();
578}
579
580static pci_ers_result_t pcifront_common_process(int cmd,
581 struct pcifront_device *pdev,
582 pci_channel_state_t state)
583{
584 pci_ers_result_t result;
585 struct pci_driver *pdrv;
586 int bus = pdev->sh_info->aer_op.bus;
587 int devfn = pdev->sh_info->aer_op.devfn;
588 struct pci_dev *pcidev;
589 int flag = 0;
590
591 dev_dbg(&pdev->xdev->dev,
592 "pcifront AER process: cmd %x (bus:%x, devfn%x)",
593 cmd, bus, devfn);
594 result = PCI_ERS_RESULT_NONE;
595
596 pcidev = pci_get_bus_and_slot(bus, devfn);
597 if (!pcidev || !pcidev->driver) {
598 dev_err(&pdev->xdev->dev, "device or AER driver is NULL\n");
599 if (pcidev)
600 pci_dev_put(pcidev);
601 return result;
602 }
603 pdrv = pcidev->driver;
604
605 if (pdrv) {
606 if (pdrv->err_handler && pdrv->err_handler->error_detected) {
607 dev_dbg(&pcidev->dev,
608 "trying to call AER service\n");
609 if (pcidev) {
610 flag = 1;
611 switch (cmd) {
612 case XEN_PCI_OP_aer_detected:
613 result = pdrv->err_handler->
614 error_detected(pcidev, state);
615 break;
616 case XEN_PCI_OP_aer_mmio:
617 result = pdrv->err_handler->
618 mmio_enabled(pcidev);
619 break;
620 case XEN_PCI_OP_aer_slotreset:
621 result = pdrv->err_handler->
622 slot_reset(pcidev);
623 break;
624 case XEN_PCI_OP_aer_resume:
625 pdrv->err_handler->resume(pcidev);
626 break;
627 default:
628 dev_err(&pdev->xdev->dev,
629 "bad request in aer recovery "
630 "operation!\n");
631
632 }
633 }
634 }
635 }
636 if (!flag)
637 result = PCI_ERS_RESULT_NONE;
638
639 return result;
640}
641
642
643static void pcifront_do_aer(struct work_struct *data)
644{
645 struct pcifront_device *pdev =
646 container_of(data, struct pcifront_device, op_work);
647 int cmd = pdev->sh_info->aer_op.cmd;
648 pci_channel_state_t state =
649 (pci_channel_state_t)pdev->sh_info->aer_op.err;
650
651 /*If a pci_conf op is in progress,
652 we have to wait until it is done before service aer op*/
653 dev_dbg(&pdev->xdev->dev,
654 "pcifront service aer bus %x devfn %x\n",
655 pdev->sh_info->aer_op.bus, pdev->sh_info->aer_op.devfn);
656
657 pdev->sh_info->aer_op.err = pcifront_common_process(cmd, pdev, state);
658
659 /* Post the operation to the guest. */
660 wmb();
661 clear_bit(_XEN_PCIB_active, (unsigned long *)&pdev->sh_info->flags);
662 notify_remote_via_evtchn(pdev->evtchn);
663
664 /*in case of we lost an aer request in four lines time_window*/
665 smp_mb__before_clear_bit();
666 clear_bit(_PDEVB_op_active, &pdev->flags);
667 smp_mb__after_clear_bit();
668
669 schedule_pcifront_aer_op(pdev);
670
671}
672
673static irqreturn_t pcifront_handler_aer(int irq, void *dev)
674{
675 struct pcifront_device *pdev = dev;
676 schedule_pcifront_aer_op(pdev);
677 return IRQ_HANDLED;
678}
679static int pcifront_connect_and_init_dma(struct pcifront_device *pdev)
680{
681 int err = 0;
682
683 spin_lock(&pcifront_dev_lock);
684
685 if (!pcifront_dev) {
686 dev_info(&pdev->xdev->dev, "Installing PCI frontend\n");
687 pcifront_dev = pdev;
688 } else
689 err = -EEXIST;
690
691 spin_unlock(&pcifront_dev_lock);
692
693 if (!err && !swiotlb_nr_tbl()) {
694 err = pci_xen_swiotlb_init_late();
695 if (err)
696 dev_err(&pdev->xdev->dev, "Could not setup SWIOTLB!\n");
697 }
698 return err;
699}
700
701static void pcifront_disconnect(struct pcifront_device *pdev)
702{
703 spin_lock(&pcifront_dev_lock);
704
705 if (pdev == pcifront_dev) {
706 dev_info(&pdev->xdev->dev,
707 "Disconnecting PCI Frontend Buses\n");
708 pcifront_dev = NULL;
709 }
710
711 spin_unlock(&pcifront_dev_lock);
712}
713static struct pcifront_device *alloc_pdev(struct xenbus_device *xdev)
714{
715 struct pcifront_device *pdev;
716
717 pdev = kzalloc(sizeof(struct pcifront_device), GFP_KERNEL);
718 if (pdev == NULL)
719 goto out;
720
721 pdev->sh_info =
722 (struct xen_pci_sharedinfo *)__get_free_page(GFP_KERNEL);
723 if (pdev->sh_info == NULL) {
724 kfree(pdev);
725 pdev = NULL;
726 goto out;
727 }
728 pdev->sh_info->flags = 0;
729
730 /*Flag for registering PV AER handler*/
731 set_bit(_XEN_PCIB_AERHANDLER, (void *)&pdev->sh_info->flags);
732
733 dev_set_drvdata(&xdev->dev, pdev);
734 pdev->xdev = xdev;
735
736 INIT_LIST_HEAD(&pdev->root_buses);
737
738 spin_lock_init(&pdev->sh_info_lock);
739
740 pdev->evtchn = INVALID_EVTCHN;
741 pdev->gnt_ref = INVALID_GRANT_REF;
742 pdev->irq = -1;
743
744 INIT_WORK(&pdev->op_work, pcifront_do_aer);
745
746 dev_dbg(&xdev->dev, "Allocated pdev @ 0x%p pdev->sh_info @ 0x%p\n",
747 pdev, pdev->sh_info);
748out:
749 return pdev;
750}
751
752static void free_pdev(struct pcifront_device *pdev)
753{
754 dev_dbg(&pdev->xdev->dev, "freeing pdev @ 0x%p\n", pdev);
755
756 pcifront_free_roots(pdev);
757
758 cancel_work_sync(&pdev->op_work);
759
760 if (pdev->irq >= 0)
761 unbind_from_irqhandler(pdev->irq, pdev);
762
763 if (pdev->evtchn != INVALID_EVTCHN)
764 xenbus_free_evtchn(pdev->xdev, pdev->evtchn);
765
766 if (pdev->gnt_ref != INVALID_GRANT_REF)
767 gnttab_end_foreign_access(pdev->gnt_ref, 0 /* r/w page */,
768 (unsigned long)pdev->sh_info);
769 else
770 free_page((unsigned long)pdev->sh_info);
771
772 dev_set_drvdata(&pdev->xdev->dev, NULL);
773
774 kfree(pdev);
775}
776
777static int pcifront_publish_info(struct pcifront_device *pdev)
778{
779 int err = 0;
780 struct xenbus_transaction trans;
781
782 err = xenbus_grant_ring(pdev->xdev, virt_to_mfn(pdev->sh_info));
783 if (err < 0)
784 goto out;
785
786 pdev->gnt_ref = err;
787
788 err = xenbus_alloc_evtchn(pdev->xdev, &pdev->evtchn);
789 if (err)
790 goto out;
791
792 err = bind_evtchn_to_irqhandler(pdev->evtchn, pcifront_handler_aer,
793 0, "pcifront", pdev);
794
795 if (err < 0)
796 return err;
797
798 pdev->irq = err;
799
800do_publish:
801 err = xenbus_transaction_start(&trans);
802 if (err) {
803 xenbus_dev_fatal(pdev->xdev, err,
804 "Error writing configuration for backend "
805 "(start transaction)");
806 goto out;
807 }
808
809 err = xenbus_printf(trans, pdev->xdev->nodename,
810 "pci-op-ref", "%u", pdev->gnt_ref);
811 if (!err)
812 err = xenbus_printf(trans, pdev->xdev->nodename,
813 "event-channel", "%u", pdev->evtchn);
814 if (!err)
815 err = xenbus_printf(trans, pdev->xdev->nodename,
816 "magic", XEN_PCI_MAGIC);
817
818 if (err) {
819 xenbus_transaction_end(trans, 1);
820 xenbus_dev_fatal(pdev->xdev, err,
821 "Error writing configuration for backend");
822 goto out;
823 } else {
824 err = xenbus_transaction_end(trans, 0);
825 if (err == -EAGAIN)
826 goto do_publish;
827 else if (err) {
828 xenbus_dev_fatal(pdev->xdev, err,
829 "Error completing transaction "
830 "for backend");
831 goto out;
832 }
833 }
834
835 xenbus_switch_state(pdev->xdev, XenbusStateInitialised);
836
837 dev_dbg(&pdev->xdev->dev, "publishing successful!\n");
838
839out:
840 return err;
841}
842
843static int pcifront_try_connect(struct pcifront_device *pdev)
844{
845 int err = -EFAULT;
846 int i, num_roots, len;
847 char str[64];
848 unsigned int domain, bus;
849
850
851 /* Only connect once */
852 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
853 XenbusStateInitialised)
854 goto out;
855
856 err = pcifront_connect_and_init_dma(pdev);
857 if (err && err != -EEXIST) {
858 xenbus_dev_fatal(pdev->xdev, err,
859 "Error setting up PCI Frontend");
860 goto out;
861 }
862
863 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
864 "root_num", "%d", &num_roots);
865 if (err == -ENOENT) {
866 xenbus_dev_error(pdev->xdev, err,
867 "No PCI Roots found, trying 0000:00");
868 err = pcifront_scan_root(pdev, 0, 0);
869 num_roots = 0;
870 } else if (err != 1) {
871 if (err == 0)
872 err = -EINVAL;
873 xenbus_dev_fatal(pdev->xdev, err,
874 "Error reading number of PCI roots");
875 goto out;
876 }
877
878 for (i = 0; i < num_roots; i++) {
879 len = snprintf(str, sizeof(str), "root-%d", i);
880 if (unlikely(len >= (sizeof(str) - 1))) {
881 err = -ENOMEM;
882 goto out;
883 }
884
885 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
886 "%x:%x", &domain, &bus);
887 if (err != 2) {
888 if (err >= 0)
889 err = -EINVAL;
890 xenbus_dev_fatal(pdev->xdev, err,
891 "Error reading PCI root %d", i);
892 goto out;
893 }
894
895 err = pcifront_scan_root(pdev, domain, bus);
896 if (err) {
897 xenbus_dev_fatal(pdev->xdev, err,
898 "Error scanning PCI root %04x:%02x",
899 domain, bus);
900 goto out;
901 }
902 }
903
904 err = xenbus_switch_state(pdev->xdev, XenbusStateConnected);
905
906out:
907 return err;
908}
909
910static int pcifront_try_disconnect(struct pcifront_device *pdev)
911{
912 int err = 0;
913 enum xenbus_state prev_state;
914
915
916 prev_state = xenbus_read_driver_state(pdev->xdev->nodename);
917
918 if (prev_state >= XenbusStateClosing)
919 goto out;
920
921 if (prev_state == XenbusStateConnected) {
922 pcifront_free_roots(pdev);
923 pcifront_disconnect(pdev);
924 }
925
926 err = xenbus_switch_state(pdev->xdev, XenbusStateClosed);
927
928out:
929
930 return err;
931}
932
933static int pcifront_attach_devices(struct pcifront_device *pdev)
934{
935 int err = -EFAULT;
936 int i, num_roots, len;
937 unsigned int domain, bus;
938 char str[64];
939
940 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
941 XenbusStateReconfiguring)
942 goto out;
943
944 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend,
945 "root_num", "%d", &num_roots);
946 if (err == -ENOENT) {
947 xenbus_dev_error(pdev->xdev, err,
948 "No PCI Roots found, trying 0000:00");
949 err = pcifront_rescan_root(pdev, 0, 0);
950 num_roots = 0;
951 } else if (err != 1) {
952 if (err == 0)
953 err = -EINVAL;
954 xenbus_dev_fatal(pdev->xdev, err,
955 "Error reading number of PCI roots");
956 goto out;
957 }
958
959 for (i = 0; i < num_roots; i++) {
960 len = snprintf(str, sizeof(str), "root-%d", i);
961 if (unlikely(len >= (sizeof(str) - 1))) {
962 err = -ENOMEM;
963 goto out;
964 }
965
966 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
967 "%x:%x", &domain, &bus);
968 if (err != 2) {
969 if (err >= 0)
970 err = -EINVAL;
971 xenbus_dev_fatal(pdev->xdev, err,
972 "Error reading PCI root %d", i);
973 goto out;
974 }
975
976 err = pcifront_rescan_root(pdev, domain, bus);
977 if (err) {
978 xenbus_dev_fatal(pdev->xdev, err,
979 "Error scanning PCI root %04x:%02x",
980 domain, bus);
981 goto out;
982 }
983 }
984
985 xenbus_switch_state(pdev->xdev, XenbusStateConnected);
986
987out:
988 return err;
989}
990
991static int pcifront_detach_devices(struct pcifront_device *pdev)
992{
993 int err = 0;
994 int i, num_devs;
995 unsigned int domain, bus, slot, func;
996 struct pci_dev *pci_dev;
997 char str[64];
998
999 if (xenbus_read_driver_state(pdev->xdev->nodename) !=
1000 XenbusStateConnected)
1001 goto out;
1002
1003 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, "num_devs", "%d",
1004 &num_devs);
1005 if (err != 1) {
1006 if (err >= 0)
1007 err = -EINVAL;
1008 xenbus_dev_fatal(pdev->xdev, err,
1009 "Error reading number of PCI devices");
1010 goto out;
1011 }
1012
1013 /* Find devices being detached and remove them. */
1014 for (i = 0; i < num_devs; i++) {
1015 int l, state;
1016 l = snprintf(str, sizeof(str), "state-%d", i);
1017 if (unlikely(l >= (sizeof(str) - 1))) {
1018 err = -ENOMEM;
1019 goto out;
1020 }
1021 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str, "%d",
1022 &state);
1023 if (err != 1)
1024 state = XenbusStateUnknown;
1025
1026 if (state != XenbusStateClosing)
1027 continue;
1028
1029 /* Remove device. */
1030 l = snprintf(str, sizeof(str), "vdev-%d", i);
1031 if (unlikely(l >= (sizeof(str) - 1))) {
1032 err = -ENOMEM;
1033 goto out;
1034 }
1035 err = xenbus_scanf(XBT_NIL, pdev->xdev->otherend, str,
1036 "%x:%x:%x.%x", &domain, &bus, &slot, &func);
1037 if (err != 4) {
1038 if (err >= 0)
1039 err = -EINVAL;
1040 xenbus_dev_fatal(pdev->xdev, err,
1041 "Error reading PCI device %d", i);
1042 goto out;
1043 }
1044
1045 pci_dev = pci_get_domain_bus_and_slot(domain, bus,
1046 PCI_DEVFN(slot, func));
1047 if (!pci_dev) {
1048 dev_dbg(&pdev->xdev->dev,
1049 "Cannot get PCI device %04x:%02x:%02x.%d\n",
1050 domain, bus, slot, func);
1051 continue;
1052 }
1053 pci_lock_rescan_remove();
1054 pci_stop_and_remove_bus_device(pci_dev);
1055 pci_dev_put(pci_dev);
1056 pci_unlock_rescan_remove();
1057
1058 dev_dbg(&pdev->xdev->dev,
1059 "PCI device %04x:%02x:%02x.%d removed.\n",
1060 domain, bus, slot, func);
1061 }
1062
1063 err = xenbus_switch_state(pdev->xdev, XenbusStateReconfiguring);
1064
1065out:
1066 return err;
1067}
1068
1069static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev,
1070 enum xenbus_state be_state)
1071{
1072 struct pcifront_device *pdev = dev_get_drvdata(&xdev->dev);
1073
1074 switch (be_state) {
1075 case XenbusStateUnknown:
1076 case XenbusStateInitialising:
1077 case XenbusStateInitWait:
1078 case XenbusStateInitialised:
1079 break;
1080
1081 case XenbusStateConnected:
1082 pcifront_try_connect(pdev);
1083 break;
1084
1085 case XenbusStateClosed:
1086 if (xdev->state == XenbusStateClosed)
1087 break;
1088 /* Missed the backend's CLOSING state -- fallthrough */
1089 case XenbusStateClosing:
1090 dev_warn(&xdev->dev, "backend going away!\n");
1091 pcifront_try_disconnect(pdev);
1092 break;
1093
1094 case XenbusStateReconfiguring:
1095 pcifront_detach_devices(pdev);
1096 break;
1097
1098 case XenbusStateReconfigured:
1099 pcifront_attach_devices(pdev);
1100 break;
1101 }
1102}
1103
1104static int pcifront_xenbus_probe(struct xenbus_device *xdev,
1105 const struct xenbus_device_id *id)
1106{
1107 int err = 0;
1108 struct pcifront_device *pdev = alloc_pdev(xdev);
1109
1110 if (pdev == NULL) {
1111 err = -ENOMEM;
1112 xenbus_dev_fatal(xdev, err,
1113 "Error allocating pcifront_device struct");
1114 goto out;
1115 }
1116
1117 err = pcifront_publish_info(pdev);
1118 if (err)
1119 free_pdev(pdev);
1120
1121out:
1122 return err;
1123}
1124
1125static int pcifront_xenbus_remove(struct xenbus_device *xdev)
1126{
1127 struct pcifront_device *pdev = dev_get_drvdata(&xdev->dev);
1128 if (pdev)
1129 free_pdev(pdev);
1130
1131 return 0;
1132}
1133
1134static const struct xenbus_device_id xenpci_ids[] = {
1135 {"pci"},
1136 {""},
1137};
1138
1139static DEFINE_XENBUS_DRIVER(xenpci, "pcifront",
1140 .probe = pcifront_xenbus_probe,
1141 .remove = pcifront_xenbus_remove,
1142 .otherend_changed = pcifront_backend_changed,
1143);
1144
1145static int __init pcifront_init(void)
1146{
1147 if (!xen_pv_domain() || xen_initial_domain())
1148 return -ENODEV;
1149
1150 if (!xen_has_pv_devices())
1151 return -ENODEV;
1152
1153 pci_frontend_registrar(1 /* enable */);
1154
1155 return xenbus_register_frontend(&xenpci_driver);
1156}
1157
1158static void __exit pcifront_cleanup(void)
1159{
1160 xenbus_unregister_driver(&xenpci_driver);
1161 pci_frontend_registrar(0 /* disable */);
1162}
1163module_init(pcifront_init);
1164module_exit(pcifront_cleanup);
1165
1166MODULE_DESCRIPTION("Xen PCI passthrough frontend.");
1167MODULE_LICENSE("GPL");
1168MODULE_ALIAS("xen:pci");