1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * Copyright 2006 Jake Moilanen <moilanen@austin.ibm.com>, IBM Corp.
4 * Copyright 2006-2007 Michael Ellerman, IBM Corp.
5 */
6
7#include <linux/crash_dump.h>
8#include <linux/device.h>
9#include <linux/irq.h>
10#include <linux/msi.h>
11
12#include <asm/rtas.h>
13#include <asm/hw_irq.h>
14#include <asm/ppc-pci.h>
15#include <asm/machdep.h>
16
17#include "pseries.h"
18
19static int query_token, change_token;
20
21#define RTAS_QUERY_FN		0
22#define RTAS_CHANGE_FN		1
23#define RTAS_RESET_FN		2
24#define RTAS_CHANGE_MSI_FN	3
25#define RTAS_CHANGE_MSIX_FN	4
26#define RTAS_CHANGE_32MSI_FN	5
27
28/* RTAS Helpers */
29
30static int rtas_change_msi(struct pci_dn *pdn, u32 func, u32 num_irqs)
31{
32	u32 addr, seq_num, rtas_ret[3];
33	unsigned long buid;
34	int rc;
35
36	addr = rtas_config_addr(pdn->busno, pdn->devfn, 0);
37	buid = pdn->phb->buid;
38
39	seq_num = 1;
40	do {
41		if (func == RTAS_CHANGE_MSI_FN || func == RTAS_CHANGE_MSIX_FN ||
42		    func == RTAS_CHANGE_32MSI_FN)
43			rc = rtas_call(change_token, 6, 4, rtas_ret, addr,
44					BUID_HI(buid), BUID_LO(buid),
45					func, num_irqs, seq_num);
46		else
47			rc = rtas_call(change_token, 6, 3, rtas_ret, addr,
48					BUID_HI(buid), BUID_LO(buid),
49					func, num_irqs, seq_num);
50
51		seq_num = rtas_ret[1];
52	} while (rtas_busy_delay(rc));
53
54	/*
55	 * If the RTAS call succeeded, return the number of irqs allocated.
56	 * If not, make sure we return a negative error code.
57	 */
58	if (rc == 0)
59		rc = rtas_ret[0];
60	else if (rc > 0)
61		rc = -rc;
62
63	pr_debug("rtas_msi: ibm,change_msi(func=%d,num=%d), got %d rc = %d\n",
64		 func, num_irqs, rtas_ret[0], rc);
65
66	return rc;
67}
68
69static void rtas_disable_msi(struct pci_dev *pdev)
70{
71	struct pci_dn *pdn;
72
73	pdn = pci_get_pdn(pdev);
74	if (!pdn)
75		return;
76
77	/*
78	 * disabling MSI with the explicit interface also disables MSI-X
79	 */
80	if (rtas_change_msi(pdn, RTAS_CHANGE_MSI_FN, 0) != 0) {
81		/*
82		 * may have failed because explicit interface is not
83		 * present
84		 */
85		if (rtas_change_msi(pdn, RTAS_CHANGE_FN, 0) != 0) {
86			pr_debug("rtas_msi: Setting MSIs to 0 failed!\n");
87		}
88	}
89}
90
91static int rtas_query_irq_number(struct pci_dn *pdn, int offset)
92{
93	u32 addr, rtas_ret[2];
94	unsigned long buid;
95	int rc;
96
97	addr = rtas_config_addr(pdn->busno, pdn->devfn, 0);
98	buid = pdn->phb->buid;
99
100	do {
101		rc = rtas_call(query_token, 4, 3, rtas_ret, addr,
102			       BUID_HI(buid), BUID_LO(buid), offset);
103	} while (rtas_busy_delay(rc));
104
105	if (rc) {
106		pr_debug("rtas_msi: error (%d) querying source number\n", rc);
107		return rc;
108	}
109
110	return rtas_ret[0];
111}
112
113static void rtas_teardown_msi_irqs(struct pci_dev *pdev)
114{
115	struct msi_desc *entry;
116
117	for_each_pci_msi_entry(entry, pdev) {
118		if (!entry->irq)
119			continue;
120
121		irq_set_msi_desc(entry->irq, NULL);
122		irq_dispose_mapping(entry->irq);
123	}
124
125	rtas_disable_msi(pdev);
126}
127
128static int check_req(struct pci_dev *pdev, int nvec, char *prop_name)
129{
130	struct device_node *dn;
131	const __be32 *p;
132	u32 req_msi;
133
134	dn = pci_device_to_OF_node(pdev);
135
136	p = of_get_property(dn, prop_name, NULL);
137	if (!p) {
138		pr_debug("rtas_msi: No %s on %pOF\n", prop_name, dn);
139		return -ENOENT;
140	}
141
142	req_msi = be32_to_cpup(p);
143	if (req_msi < nvec) {
144		pr_debug("rtas_msi: %s requests < %d MSIs\n", prop_name, nvec);
145
146		if (req_msi == 0) /* Be paranoid */
147			return -ENOSPC;
148
149		return req_msi;
150	}
151
152	return 0;
153}
154
155static int check_req_msi(struct pci_dev *pdev, int nvec)
156{
157	return check_req(pdev, nvec, "ibm,req#msi");
158}
159
160static int check_req_msix(struct pci_dev *pdev, int nvec)
161{
162	return check_req(pdev, nvec, "ibm,req#msi-x");
163}
164
165/* Quota calculation */
166
167static struct device_node *find_pe_total_msi(struct pci_dev *dev, int *total)
168{
169	struct device_node *dn;
170	const __be32 *p;
171
172	dn = of_node_get(pci_device_to_OF_node(dev));
173	while (dn) {
174		p = of_get_property(dn, "ibm,pe-total-#msi", NULL);
175		if (p) {
176			pr_debug("rtas_msi: found prop on dn %pOF\n",
177				dn);
178			*total = be32_to_cpup(p);
179			return dn;
180		}
181
182		dn = of_get_next_parent(dn);
183	}
184
185	return NULL;
186}
187
188static struct device_node *find_pe_dn(struct pci_dev *dev, int *total)
189{
190	struct device_node *dn;
191	struct eeh_dev *edev;
192
193	/* Found our PE and assume 8 at that point. */
194
195	dn = pci_device_to_OF_node(dev);
196	if (!dn)
197		return NULL;
198
199	/* Get the top level device in the PE */
200	edev = pdn_to_eeh_dev(PCI_DN(dn));
201	if (edev->pe)
202		edev = list_first_entry(&edev->pe->edevs, struct eeh_dev,
203					entry);
204	dn = pci_device_to_OF_node(edev->pdev);
205	if (!dn)
206		return NULL;
207
208	/* We actually want the parent */
209	dn = of_get_parent(dn);
210	if (!dn)
211		return NULL;
212
213	/* Hardcode of 8 for old firmwares */
214	*total = 8;
215	pr_debug("rtas_msi: using PE dn %pOF\n", dn);
216
217	return dn;
218}
219
220struct msi_counts {
221	struct device_node *requestor;
222	int num_devices;
223	int request;
224	int quota;
225	int spare;
226	int over_quota;
227};
228
229static void *count_non_bridge_devices(struct device_node *dn, void *data)
230{
231	struct msi_counts *counts = data;
232	const __be32 *p;
233	u32 class;
234
235	pr_debug("rtas_msi: counting %pOF\n", dn);
236
237	p = of_get_property(dn, "class-code", NULL);
238	class = p ? be32_to_cpup(p) : 0;
239
240	if ((class >> 8) != PCI_CLASS_BRIDGE_PCI)
241		counts->num_devices++;
242
243	return NULL;
244}
245
246static void *count_spare_msis(struct device_node *dn, void *data)
247{
248	struct msi_counts *counts = data;
249	const __be32 *p;
250	int req;
251
252	if (dn == counts->requestor)
253		req = counts->request;
254	else {
255		/* We don't know if a driver will try to use MSI or MSI-X,
256		 * so we just have to punt and use the larger of the two. */
257		req = 0;
258		p = of_get_property(dn, "ibm,req#msi", NULL);
259		if (p)
260			req = be32_to_cpup(p);
261
262		p = of_get_property(dn, "ibm,req#msi-x", NULL);
263		if (p)
264			req = max(req, (int)be32_to_cpup(p));
265	}
266
267	if (req < counts->quota)
268		counts->spare += counts->quota - req;
269	else if (req > counts->quota)
270		counts->over_quota++;
271
272	return NULL;
273}
274
275static int msi_quota_for_device(struct pci_dev *dev, int request)
276{
277	struct device_node *pe_dn;
278	struct msi_counts counts;
279	int total;
280
281	pr_debug("rtas_msi: calc quota for %s, request %d\n", pci_name(dev),
282		  request);
283
284	pe_dn = find_pe_total_msi(dev, &total);
285	if (!pe_dn)
286		pe_dn = find_pe_dn(dev, &total);
287
288	if (!pe_dn) {
289		pr_err("rtas_msi: couldn't find PE for %s\n", pci_name(dev));
290		goto out;
291	}
292
293	pr_debug("rtas_msi: found PE %pOF\n", pe_dn);
294
295	memset(&counts, 0, sizeof(struct msi_counts));
296
297	/* Work out how many devices we have below this PE */
298	pci_traverse_device_nodes(pe_dn, count_non_bridge_devices, &counts);
299
300	if (counts.num_devices == 0) {
301		pr_err("rtas_msi: found 0 devices under PE for %s\n",
302			pci_name(dev));
303		goto out;
304	}
305
306	counts.quota = total / counts.num_devices;
307	if (request <= counts.quota)
308		goto out;
309
310	/* else, we have some more calculating to do */
311	counts.requestor = pci_device_to_OF_node(dev);
312	counts.request = request;
313	pci_traverse_device_nodes(pe_dn, count_spare_msis, &counts);
314
315	/* If the quota isn't an integer multiple of the total, we can
316	 * use the remainder as spare MSIs for anyone that wants them. */
317	counts.spare += total % counts.num_devices;
318
319	/* Divide any spare by the number of over-quota requestors */
320	if (counts.over_quota)
321		counts.quota += counts.spare / counts.over_quota;
322
323	/* And finally clamp the request to the possibly adjusted quota */
324	request = min(counts.quota, request);
325
326	pr_debug("rtas_msi: request clamped to quota %d\n", request);
327out:
328	of_node_put(pe_dn);
329
330	return request;
331}
332
333static int check_msix_entries(struct pci_dev *pdev)
334{
335	struct msi_desc *entry;
336	int expected;
337
338	/* There's no way for us to express to firmware that we want
339	 * a discontiguous, or non-zero based, range of MSI-X entries.
340	 * So we must reject such requests. */
341
342	expected = 0;
343	for_each_pci_msi_entry(entry, pdev) {
344		if (entry->msi_attrib.entry_nr != expected) {
345			pr_debug("rtas_msi: bad MSI-X entries.\n");
346			return -EINVAL;
347		}
348		expected++;
349	}
350
351	return 0;
352}
353
354static void rtas_hack_32bit_msi_gen2(struct pci_dev *pdev)
355{
356	u32 addr_hi, addr_lo;
357
358	/*
359	 * We should only get in here for IODA1 configs. This is based on the
360	 * fact that we using RTAS for MSIs, we don't have the 32 bit MSI RTAS
361	 * support, and we are in a PCIe Gen2 slot.
362	 */
363	dev_info(&pdev->dev,
364		 "rtas_msi: No 32 bit MSI firmware support, forcing 32 bit MSI\n");
365	pci_read_config_dword(pdev, pdev->msi_cap + PCI_MSI_ADDRESS_HI, &addr_hi);
366	addr_lo = 0xffff0000 | ((addr_hi >> (48 - 32)) << 4);
367	pci_write_config_dword(pdev, pdev->msi_cap + PCI_MSI_ADDRESS_LO, addr_lo);
368	pci_write_config_dword(pdev, pdev->msi_cap + PCI_MSI_ADDRESS_HI, 0);
369}
370
371static int rtas_setup_msi_irqs(struct pci_dev *pdev, int nvec_in, int type)
372{
373	struct pci_dn *pdn;
374	int hwirq, virq, i, quota, rc;
375	struct msi_desc *entry;
376	struct msi_msg msg;
377	int nvec = nvec_in;
378	int use_32bit_msi_hack = 0;
379
380	if (type == PCI_CAP_ID_MSIX)
381		rc = check_req_msix(pdev, nvec);
382	else
383		rc = check_req_msi(pdev, nvec);
384
385	if (rc)
386		return rc;
387
388	quota = msi_quota_for_device(pdev, nvec);
389
390	if (quota && quota < nvec)
391		return quota;
392
393	if (type == PCI_CAP_ID_MSIX && check_msix_entries(pdev))
394		return -EINVAL;
395
396	/*
397	 * Firmware currently refuse any non power of two allocation
398	 * so we round up if the quota will allow it.
399	 */
400	if (type == PCI_CAP_ID_MSIX) {
401		int m = roundup_pow_of_two(nvec);
402		quota = msi_quota_for_device(pdev, m);
403
404		if (quota >= m)
405			nvec = m;
406	}
407
408	pdn = pci_get_pdn(pdev);
409
410	/*
411	 * Try the new more explicit firmware interface, if that fails fall
412	 * back to the old interface. The old interface is known to never
413	 * return MSI-Xs.
414	 */
415again:
416	if (type == PCI_CAP_ID_MSI) {
417		if (pdev->no_64bit_msi) {
418			rc = rtas_change_msi(pdn, RTAS_CHANGE_32MSI_FN, nvec);
419			if (rc < 0) {
420				/*
421				 * We only want to run the 32 bit MSI hack below if
422				 * the max bus speed is Gen2 speed
423				 */
424				if (pdev->bus->max_bus_speed != PCIE_SPEED_5_0GT)
425					return rc;
426
427				use_32bit_msi_hack = 1;
428			}
429		} else
430			rc = -1;
431
432		if (rc < 0)
433			rc = rtas_change_msi(pdn, RTAS_CHANGE_MSI_FN, nvec);
434
435		if (rc < 0) {
436			pr_debug("rtas_msi: trying the old firmware call.\n");
437			rc = rtas_change_msi(pdn, RTAS_CHANGE_FN, nvec);
438		}
439
440		if (use_32bit_msi_hack && rc > 0)
441			rtas_hack_32bit_msi_gen2(pdev);
442	} else
443		rc = rtas_change_msi(pdn, RTAS_CHANGE_MSIX_FN, nvec);
444
445	if (rc != nvec) {
446		if (nvec != nvec_in) {
447			nvec = nvec_in;
448			goto again;
449		}
450		pr_debug("rtas_msi: rtas_change_msi() failed\n");
451		return rc;
452	}
453
454	i = 0;
455	for_each_pci_msi_entry(entry, pdev) {
456		hwirq = rtas_query_irq_number(pdn, i++);
457		if (hwirq < 0) {
458			pr_debug("rtas_msi: error (%d) getting hwirq\n", rc);
459			return hwirq;
460		}
461
462		/*
463		 * Depending on the number of online CPUs in the original
464		 * kernel, it is likely for CPU #0 to be offline in a kdump
465		 * kernel. The associated IRQs in the affinity mappings
466		 * provided by irq_create_affinity_masks() are thus not
467		 * started by irq_startup(), as per-design for managed IRQs.
468		 * This can be a problem with multi-queue block devices driven
469		 * by blk-mq : such a non-started IRQ is very likely paired
470		 * with the single queue enforced by blk-mq during kdump (see
471		 * blk_mq_alloc_tag_set()). This causes the device to remain
472		 * silent and likely hangs the guest at some point.
473		 *
474		 * We don't really care for fine-grained affinity when doing
475		 * kdump actually : simply ignore the pre-computed affinity
476		 * masks in this case and let the default mask with all CPUs
477		 * be used when creating the IRQ mappings.
478		 */
479		if (is_kdump_kernel())
480			virq = irq_create_mapping(NULL, hwirq);
481		else
482			virq = irq_create_mapping_affinity(NULL, hwirq,
483							   entry->affinity);
484
485		if (!virq) {
486			pr_debug("rtas_msi: Failed mapping hwirq %d\n", hwirq);
487			return -ENOSPC;
488		}
489
490		dev_dbg(&pdev->dev, "rtas_msi: allocated virq %d\n", virq);
491		irq_set_msi_desc(virq, entry);
492
493		/* Read config space back so we can restore after reset */
494		__pci_read_msi_msg(entry, &msg);
495		entry->msg = msg;
496	}
497
498	return 0;
499}
500
501static void rtas_msi_pci_irq_fixup(struct pci_dev *pdev)
502{
503	/* No LSI -> leave MSIs (if any) configured */
504	if (!pdev->irq) {
505		dev_dbg(&pdev->dev, "rtas_msi: no LSI, nothing to do.\n");
506		return;
507	}
508
509	/* No MSI -> MSIs can't have been assigned by fw, leave LSI */
510	if (check_req_msi(pdev, 1) && check_req_msix(pdev, 1)) {
511		dev_dbg(&pdev->dev, "rtas_msi: no req#msi/x, nothing to do.\n");
512		return;
513	}
514
515	dev_dbg(&pdev->dev, "rtas_msi: disabling existing MSI.\n");
516	rtas_disable_msi(pdev);
517}
518
519static int rtas_msi_init(void)
520{
521	struct pci_controller *phb;
522
523	query_token  = rtas_token("ibm,query-interrupt-source-number");
524	change_token = rtas_token("ibm,change-msi");
525
526	if ((query_token == RTAS_UNKNOWN_SERVICE) ||
527			(change_token == RTAS_UNKNOWN_SERVICE)) {
528		pr_debug("rtas_msi: no RTAS tokens, no MSI support.\n");
529		return -1;
530	}
531
532	pr_debug("rtas_msi: Registering RTAS MSI callbacks.\n");
533
534	WARN_ON(pseries_pci_controller_ops.setup_msi_irqs);
535	pseries_pci_controller_ops.setup_msi_irqs = rtas_setup_msi_irqs;
536	pseries_pci_controller_ops.teardown_msi_irqs = rtas_teardown_msi_irqs;
537
538	list_for_each_entry(phb, &hose_list, list_node) {
539		WARN_ON(phb->controller_ops.setup_msi_irqs);
540		phb->controller_ops.setup_msi_irqs = rtas_setup_msi_irqs;
541		phb->controller_ops.teardown_msi_irqs = rtas_teardown_msi_irqs;
542	}
543
544	WARN_ON(ppc_md.pci_irq_fixup);
545	ppc_md.pci_irq_fixup = rtas_msi_pci_irq_fixup;
546
547	return 0;
548}
549machine_arch_initcall(pseries, rtas_msi_init);
550