]>
Commit | Line | Data |
---|---|---|
9f27ee59 JF |
1 | /* |
2 | * blkfront.c | |
3 | * | |
4 | * XenLinux virtual block device driver. | |
5 | * | |
6 | * Copyright (c) 2003-2004, Keir Fraser & Steve Hand | |
7 | * Modifications by Mark A. Williamson are (c) Intel Research Cambridge | |
8 | * Copyright (c) 2004, Christian Limpach | |
9 | * Copyright (c) 2004, Andrew Warfield | |
10 | * Copyright (c) 2005, Christopher Clark | |
11 | * Copyright (c) 2005, XenSource Ltd | |
12 | * | |
13 | * This program is free software; you can redistribute it and/or | |
14 | * modify it under the terms of the GNU General Public License version 2 | |
15 | * as published by the Free Software Foundation; or, when distributed | |
16 | * separately from the Linux kernel or incorporated into other | |
17 | * software packages, subject to the following license: | |
18 | * | |
19 | * Permission is hereby granted, free of charge, to any person obtaining a copy | |
20 | * of this source file (the "Software"), to deal in the Software without | |
21 | * restriction, including without limitation the rights to use, copy, modify, | |
22 | * merge, publish, distribute, sublicense, and/or sell copies of the Software, | |
23 | * and to permit persons to whom the Software is furnished to do so, subject to | |
24 | * the following conditions: | |
25 | * | |
26 | * The above copyright notice and this permission notice shall be included in | |
27 | * all copies or substantial portions of the Software. | |
28 | * | |
29 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR | |
30 | * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, | |
31 | * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE | |
32 | * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER | |
33 | * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING | |
34 | * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS | |
35 | * IN THE SOFTWARE. | |
36 | */ | |
37 | ||
38 | #include <linux/interrupt.h> | |
39 | #include <linux/blkdev.h> | |
907c3eb1 | 40 | #include <linux/blk-mq.h> |
597592d9 | 41 | #include <linux/hdreg.h> |
440a01a7 | 42 | #include <linux/cdrom.h> |
9f27ee59 | 43 | #include <linux/module.h> |
5a0e3ad6 | 44 | #include <linux/slab.h> |
2a48fc0a | 45 | #include <linux/mutex.h> |
9e973e64 | 46 | #include <linux/scatterlist.h> |
34ae2e47 | 47 | #include <linux/bitmap.h> |
155b7edb | 48 | #include <linux/list.h> |
9f27ee59 | 49 | |
1ccbf534 | 50 | #include <xen/xen.h> |
9f27ee59 JF |
51 | #include <xen/xenbus.h> |
52 | #include <xen/grant_table.h> | |
53 | #include <xen/events.h> | |
54 | #include <xen/page.h> | |
c1c5413a | 55 | #include <xen/platform_pci.h> |
9f27ee59 JF |
56 | |
57 | #include <xen/interface/grant_table.h> | |
58 | #include <xen/interface/io/blkif.h> | |
3e334239 | 59 | #include <xen/interface/io/protocols.h> |
9f27ee59 JF |
60 | |
61 | #include <asm/xen/hypervisor.h> | |
62 | ||
6cc56833 JG |
63 | /* |
64 | * The minimal size of segment supported by the block framework is PAGE_SIZE. | |
65 | * When Linux is using a different page size than Xen, it may not be possible | |
66 | * to put all the data in a single segment. | |
67 | * This can happen when the backend doesn't support indirect descriptor and | |
68 | * therefore the maximum amount of data that a request can carry is | |
69 | * BLKIF_MAX_SEGMENTS_PER_REQUEST * XEN_PAGE_SIZE = 44KB | |
70 | * | |
71 | * Note that we only support one extra request. So the Linux page size | |
72 | * should be <= ( 2 * BLKIF_MAX_SEGMENTS_PER_REQUEST * XEN_PAGE_SIZE) = | |
73 | * 88KB. | |
74 | */ | |
75 | #define HAS_EXTRA_REQ (BLKIF_MAX_SEGMENTS_PER_REQUEST < XEN_PFN_PER_PAGE) | |
76 | ||
9f27ee59 JF |
77 | enum blkif_state { |
78 | BLKIF_STATE_DISCONNECTED, | |
79 | BLKIF_STATE_CONNECTED, | |
80 | BLKIF_STATE_SUSPENDED, | |
81 | }; | |
82 | ||
0a8704a5 RPM |
83 | struct grant { |
84 | grant_ref_t gref; | |
a7a6df22 | 85 | struct page *page; |
155b7edb | 86 | struct list_head node; |
0a8704a5 RPM |
87 | }; |
88 | ||
6cc56833 JG |
89 | enum blk_req_status { |
90 | REQ_WAITING, | |
91 | REQ_DONE, | |
92 | REQ_ERROR, | |
93 | REQ_EOPNOTSUPP, | |
94 | }; | |
95 | ||
9f27ee59 JF |
96 | struct blk_shadow { |
97 | struct blkif_request req; | |
a945b980 | 98 | struct request *request; |
402b27f9 RPM |
99 | struct grant **grants_used; |
100 | struct grant **indirect_grants; | |
b7649158 | 101 | struct scatterlist *sg; |
c004a6fe | 102 | unsigned int num_sg; |
6cc56833 JG |
103 | enum blk_req_status status; |
104 | ||
105 | #define NO_ASSOCIATED_ID ~0UL | |
106 | /* | |
107 | * Id of the sibling if we ever need 2 requests when handling a | |
108 | * block I/O request | |
109 | */ | |
110 | unsigned long associated_id; | |
402b27f9 RPM |
111 | }; |
112 | ||
2609587c | 113 | struct blkif_req { |
31c4ccc3 | 114 | blk_status_t error; |
2609587c CH |
115 | }; |
116 | ||
117 | static inline struct blkif_req *blkif_req(struct request *rq) | |
118 | { | |
119 | return blk_mq_rq_to_pdu(rq); | |
120 | } | |
121 | ||
2a48fc0a | 122 | static DEFINE_MUTEX(blkfront_mutex); |
83d5cde4 | 123 | static const struct block_device_operations xlvbd_block_fops; |
9f27ee59 | 124 | |
402b27f9 RPM |
125 | /* |
126 | * Maximum number of segments in indirect requests, the actual value used by | |
127 | * the frontend driver is the minimum of this value and the value provided | |
128 | * by the backend driver. | |
129 | */ | |
130 | ||
131 | static unsigned int xen_blkif_max_segments = 32; | |
14e710fe JB |
132 | module_param_named(max_indirect_segments, xen_blkif_max_segments, uint, |
133 | S_IRUGO); | |
134 | MODULE_PARM_DESC(max_indirect_segments, | |
135 | "Maximum amount of segments in indirect requests (default is 32)"); | |
402b27f9 | 136 | |
28d949bc BL |
137 | static unsigned int xen_blkif_max_queues = 4; |
138 | module_param_named(max_queues, xen_blkif_max_queues, uint, S_IRUGO); | |
139 | MODULE_PARM_DESC(max_queues, "Maximum number of hardware queues/rings used per virtual disk"); | |
140 | ||
86839c56 BL |
141 | /* |
142 | * Maximum order of pages to be used for the shared ring between front and | |
143 | * backend, 4KB page granularity is used. | |
144 | */ | |
145 | static unsigned int xen_blkif_max_ring_order; | |
146 | module_param_named(max_ring_page_order, xen_blkif_max_ring_order, int, S_IRUGO); | |
147 | MODULE_PARM_DESC(max_ring_page_order, "Maximum order of pages to be used for the shared ring"); | |
148 | ||
c004a6fe JG |
149 | #define BLK_RING_SIZE(info) \ |
150 | __CONST_RING_SIZE(blkif, XEN_PAGE_SIZE * (info)->nr_ring_pages) | |
151 | ||
152 | #define BLK_MAX_RING_SIZE \ | |
9cce2914 | 153 | __CONST_RING_SIZE(blkif, XEN_PAGE_SIZE * XENBUS_MAX_RING_GRANTS) |
c004a6fe | 154 | |
86839c56 | 155 | /* |
6f03a7ff KRW |
156 | * ring-ref%u i=(-1UL) would take 11 characters + 'ring-ref' is 8, so 19 |
157 | * characters are enough. Define to 20 to keep consistent with backend. | |
86839c56 BL |
158 | */ |
159 | #define RINGREF_NAME_LEN (20) | |
28d949bc BL |
160 | /* |
161 | * queue-%u would take 7 + 10(UINT_MAX) = 17 characters. | |
162 | */ | |
163 | #define QUEUE_NAME_LEN (17) | |
9f27ee59 | 164 | |
81f35161 BL |
165 | /* |
166 | * Per-ring info. | |
167 | * Every blkfront device can associate with one or more blkfront_ring_info, | |
168 | * depending on how many hardware queues/rings to be used. | |
169 | */ | |
170 | struct blkfront_ring_info { | |
11659569 BL |
171 | /* Lock to protect data in every ring buffer. */ |
172 | spinlock_t ring_lock; | |
81f35161 BL |
173 | struct blkif_front_ring ring; |
174 | unsigned int ring_ref[XENBUS_MAX_RING_GRANTS]; | |
175 | unsigned int evtchn, irq; | |
176 | struct work_struct work; | |
177 | struct gnttab_free_callback callback; | |
178 | struct blk_shadow shadow[BLK_MAX_RING_SIZE]; | |
179 | struct list_head indirect_pages; | |
73716df7 BL |
180 | struct list_head grants; |
181 | unsigned int persistent_gnts_c; | |
81f35161 BL |
182 | unsigned long shadow_free; |
183 | struct blkfront_info *dev_info; | |
184 | }; | |
185 | ||
9f27ee59 JF |
186 | /* |
187 | * We have one of these per vbd, whether ide, scsi or 'other'. They | |
188 | * hang in private_data off the gendisk structure. We may end up | |
189 | * putting all kinds of interesting stuff here :-) | |
190 | */ | |
191 | struct blkfront_info | |
192 | { | |
b70f5fa0 | 193 | struct mutex mutex; |
9f27ee59 | 194 | struct xenbus_device *xbdev; |
9f27ee59 | 195 | struct gendisk *gd; |
172335ad BL |
196 | u16 sector_size; |
197 | unsigned int physical_sector_size; | |
9f27ee59 JF |
198 | int vdevice; |
199 | blkif_vdev_t handle; | |
200 | enum blkif_state connected; | |
3df0e505 | 201 | /* Number of pages per ring buffer. */ |
86839c56 | 202 | unsigned int nr_ring_pages; |
9f27ee59 | 203 | struct request_queue *rq; |
b32728ff JB |
204 | unsigned int feature_flush:1; |
205 | unsigned int feature_fua:1; | |
5ea42986 KRW |
206 | unsigned int feature_discard:1; |
207 | unsigned int feature_secdiscard:1; | |
b32728ff | 208 | unsigned int feature_persistent:1; |
ed30bf31 LD |
209 | unsigned int discard_granularity; |
210 | unsigned int discard_alignment; | |
c004a6fe | 211 | /* Number of 4KB segments handled */ |
402b27f9 | 212 | unsigned int max_indirect_segments; |
1d78d705 | 213 | int is_ready; |
907c3eb1 | 214 | struct blk_mq_tag_set tag_set; |
3df0e505 BL |
215 | struct blkfront_ring_info *rinfo; |
216 | unsigned int nr_rings; | |
7b427a59 BL |
217 | /* Save uncomplete reqs and bios for migration. */ |
218 | struct list_head requests; | |
219 | struct bio_list bio_list; | |
9f27ee59 JF |
220 | }; |
221 | ||
0e345826 JB |
222 | static unsigned int nr_minors; |
223 | static unsigned long *minors; | |
224 | static DEFINE_SPINLOCK(minor_lock); | |
225 | ||
9f27ee59 JF |
226 | #define GRANT_INVALID_REF 0 |
227 | ||
228 | #define PARTS_PER_DISK 16 | |
9246b5f0 | 229 | #define PARTS_PER_EXT_DISK 256 |
9f27ee59 JF |
230 | |
231 | #define BLKIF_MAJOR(dev) ((dev)>>8) | |
232 | #define BLKIF_MINOR(dev) ((dev) & 0xff) | |
233 | ||
9246b5f0 CL |
234 | #define EXT_SHIFT 28 |
235 | #define EXTENDED (1<<EXT_SHIFT) | |
236 | #define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED)) | |
237 | #define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED)) | |
c80a4209 SS |
238 | #define EMULATED_HD_DISK_MINOR_OFFSET (0) |
239 | #define EMULATED_HD_DISK_NAME_OFFSET (EMULATED_HD_DISK_MINOR_OFFSET / 256) | |
196cfe2a SB |
240 | #define EMULATED_SD_DISK_MINOR_OFFSET (0) |
241 | #define EMULATED_SD_DISK_NAME_OFFSET (EMULATED_SD_DISK_MINOR_OFFSET / 256) | |
9f27ee59 | 242 | |
9246b5f0 | 243 | #define DEV_NAME "xvd" /* name in /dev */ |
9f27ee59 | 244 | |
c004a6fe JG |
245 | /* |
246 | * Grants are always the same size as a Xen page (i.e 4KB). | |
247 | * A physical segment is always the same size as a Linux page. | |
248 | * Number of grants per physical segment | |
249 | */ | |
250 | #define GRANTS_PER_PSEG (PAGE_SIZE / XEN_PAGE_SIZE) | |
251 | ||
252 | #define GRANTS_PER_INDIRECT_FRAME \ | |
253 | (XEN_PAGE_SIZE / sizeof(struct blkif_request_segment)) | |
254 | ||
255 | #define PSEGS_PER_INDIRECT_FRAME \ | |
256 | (GRANTS_INDIRECT_FRAME / GRANTS_PSEGS) | |
257 | ||
258 | #define INDIRECT_GREFS(_grants) \ | |
259 | DIV_ROUND_UP(_grants, GRANTS_PER_INDIRECT_FRAME) | |
260 | ||
261 | #define GREFS(_psegs) ((_psegs) * GRANTS_PER_PSEG) | |
402b27f9 | 262 | |
81f35161 | 263 | static int blkfront_setup_indirect(struct blkfront_ring_info *rinfo); |
3df0e505 | 264 | static void blkfront_gather_backend_features(struct blkfront_info *info); |
7b2b3579 | 265 | static int negotiate_mq(struct blkfront_info *info); |
402b27f9 | 266 | |
81f35161 | 267 | static int get_id_from_freelist(struct blkfront_ring_info *rinfo) |
9f27ee59 | 268 | { |
81f35161 BL |
269 | unsigned long free = rinfo->shadow_free; |
270 | ||
271 | BUG_ON(free >= BLK_RING_SIZE(rinfo->dev_info)); | |
272 | rinfo->shadow_free = rinfo->shadow[free].req.u.rw.id; | |
273 | rinfo->shadow[free].req.u.rw.id = 0x0fffffee; /* debug */ | |
9f27ee59 JF |
274 | return free; |
275 | } | |
276 | ||
81f35161 | 277 | static int add_id_to_freelist(struct blkfront_ring_info *rinfo, |
6f03a7ff | 278 | unsigned long id) |
9f27ee59 | 279 | { |
81f35161 | 280 | if (rinfo->shadow[id].req.u.rw.id != id) |
6878c32e | 281 | return -EINVAL; |
81f35161 | 282 | if (rinfo->shadow[id].request == NULL) |
6878c32e | 283 | return -EINVAL; |
81f35161 BL |
284 | rinfo->shadow[id].req.u.rw.id = rinfo->shadow_free; |
285 | rinfo->shadow[id].request = NULL; | |
286 | rinfo->shadow_free = id; | |
6878c32e | 287 | return 0; |
9f27ee59 JF |
288 | } |
289 | ||
81f35161 | 290 | static int fill_grant_buffer(struct blkfront_ring_info *rinfo, int num) |
9c1e050c | 291 | { |
81f35161 | 292 | struct blkfront_info *info = rinfo->dev_info; |
9c1e050c RPM |
293 | struct page *granted_page; |
294 | struct grant *gnt_list_entry, *n; | |
295 | int i = 0; | |
296 | ||
6f03a7ff | 297 | while (i < num) { |
9c1e050c RPM |
298 | gnt_list_entry = kzalloc(sizeof(struct grant), GFP_NOIO); |
299 | if (!gnt_list_entry) | |
300 | goto out_of_memory; | |
301 | ||
bfe11d6d RPM |
302 | if (info->feature_persistent) { |
303 | granted_page = alloc_page(GFP_NOIO); | |
304 | if (!granted_page) { | |
305 | kfree(gnt_list_entry); | |
306 | goto out_of_memory; | |
307 | } | |
a7a6df22 | 308 | gnt_list_entry->page = granted_page; |
9c1e050c RPM |
309 | } |
310 | ||
9c1e050c | 311 | gnt_list_entry->gref = GRANT_INVALID_REF; |
73716df7 | 312 | list_add(&gnt_list_entry->node, &rinfo->grants); |
9c1e050c RPM |
313 | i++; |
314 | } | |
315 | ||
316 | return 0; | |
317 | ||
318 | out_of_memory: | |
319 | list_for_each_entry_safe(gnt_list_entry, n, | |
73716df7 | 320 | &rinfo->grants, node) { |
9c1e050c | 321 | list_del(&gnt_list_entry->node); |
bfe11d6d | 322 | if (info->feature_persistent) |
a7a6df22 | 323 | __free_page(gnt_list_entry->page); |
9c1e050c RPM |
324 | kfree(gnt_list_entry); |
325 | i--; | |
326 | } | |
327 | BUG_ON(i != 0); | |
328 | return -ENOMEM; | |
329 | } | |
330 | ||
73716df7 | 331 | static struct grant *get_free_grant(struct blkfront_ring_info *rinfo) |
9c1e050c RPM |
332 | { |
333 | struct grant *gnt_list_entry; | |
9c1e050c | 334 | |
73716df7 BL |
335 | BUG_ON(list_empty(&rinfo->grants)); |
336 | gnt_list_entry = list_first_entry(&rinfo->grants, struct grant, | |
4f503fbd | 337 | node); |
9c1e050c RPM |
338 | list_del(&gnt_list_entry->node); |
339 | ||
4f503fbd | 340 | if (gnt_list_entry->gref != GRANT_INVALID_REF) |
73716df7 | 341 | rinfo->persistent_gnts_c--; |
4f503fbd JG |
342 | |
343 | return gnt_list_entry; | |
344 | } | |
345 | ||
346 | static inline void grant_foreign_access(const struct grant *gnt_list_entry, | |
347 | const struct blkfront_info *info) | |
348 | { | |
349 | gnttab_page_grant_foreign_access_ref_one(gnt_list_entry->gref, | |
350 | info->xbdev->otherend_id, | |
351 | gnt_list_entry->page, | |
352 | 0); | |
353 | } | |
354 | ||
355 | static struct grant *get_grant(grant_ref_t *gref_head, | |
356 | unsigned long gfn, | |
73716df7 | 357 | struct blkfront_ring_info *rinfo) |
4f503fbd | 358 | { |
73716df7 BL |
359 | struct grant *gnt_list_entry = get_free_grant(rinfo); |
360 | struct blkfront_info *info = rinfo->dev_info; | |
4f503fbd JG |
361 | |
362 | if (gnt_list_entry->gref != GRANT_INVALID_REF) | |
9c1e050c | 363 | return gnt_list_entry; |
4f503fbd JG |
364 | |
365 | /* Assign a gref to this page */ | |
366 | gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head); | |
367 | BUG_ON(gnt_list_entry->gref == -ENOSPC); | |
368 | if (info->feature_persistent) | |
369 | grant_foreign_access(gnt_list_entry, info); | |
370 | else { | |
371 | /* Grant access to the GFN passed by the caller */ | |
372 | gnttab_grant_foreign_access_ref(gnt_list_entry->gref, | |
373 | info->xbdev->otherend_id, | |
374 | gfn, 0); | |
9c1e050c RPM |
375 | } |
376 | ||
4f503fbd JG |
377 | return gnt_list_entry; |
378 | } | |
379 | ||
380 | static struct grant *get_indirect_grant(grant_ref_t *gref_head, | |
73716df7 | 381 | struct blkfront_ring_info *rinfo) |
4f503fbd | 382 | { |
73716df7 BL |
383 | struct grant *gnt_list_entry = get_free_grant(rinfo); |
384 | struct blkfront_info *info = rinfo->dev_info; | |
4f503fbd JG |
385 | |
386 | if (gnt_list_entry->gref != GRANT_INVALID_REF) | |
387 | return gnt_list_entry; | |
388 | ||
9c1e050c RPM |
389 | /* Assign a gref to this page */ |
390 | gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head); | |
391 | BUG_ON(gnt_list_entry->gref == -ENOSPC); | |
bfe11d6d | 392 | if (!info->feature_persistent) { |
4f503fbd JG |
393 | struct page *indirect_page; |
394 | ||
395 | /* Fetch a pre-allocated page to use for indirect grefs */ | |
73716df7 BL |
396 | BUG_ON(list_empty(&rinfo->indirect_pages)); |
397 | indirect_page = list_first_entry(&rinfo->indirect_pages, | |
4f503fbd JG |
398 | struct page, lru); |
399 | list_del(&indirect_page->lru); | |
400 | gnt_list_entry->page = indirect_page; | |
bfe11d6d | 401 | } |
4f503fbd JG |
402 | grant_foreign_access(gnt_list_entry, info); |
403 | ||
9c1e050c RPM |
404 | return gnt_list_entry; |
405 | } | |
406 | ||
6878c32e KRW |
407 | static const char *op_name(int op) |
408 | { | |
409 | static const char *const names[] = { | |
410 | [BLKIF_OP_READ] = "read", | |
411 | [BLKIF_OP_WRITE] = "write", | |
412 | [BLKIF_OP_WRITE_BARRIER] = "barrier", | |
413 | [BLKIF_OP_FLUSH_DISKCACHE] = "flush", | |
414 | [BLKIF_OP_DISCARD] = "discard" }; | |
415 | ||
416 | if (op < 0 || op >= ARRAY_SIZE(names)) | |
417 | return "unknown"; | |
418 | ||
419 | if (!names[op]) | |
420 | return "reserved"; | |
421 | ||
422 | return names[op]; | |
423 | } | |
0e345826 JB |
424 | static int xlbd_reserve_minors(unsigned int minor, unsigned int nr) |
425 | { | |
426 | unsigned int end = minor + nr; | |
427 | int rc; | |
428 | ||
429 | if (end > nr_minors) { | |
430 | unsigned long *bitmap, *old; | |
431 | ||
f094148a | 432 | bitmap = kcalloc(BITS_TO_LONGS(end), sizeof(*bitmap), |
0e345826 JB |
433 | GFP_KERNEL); |
434 | if (bitmap == NULL) | |
435 | return -ENOMEM; | |
436 | ||
437 | spin_lock(&minor_lock); | |
438 | if (end > nr_minors) { | |
439 | old = minors; | |
440 | memcpy(bitmap, minors, | |
441 | BITS_TO_LONGS(nr_minors) * sizeof(*bitmap)); | |
442 | minors = bitmap; | |
443 | nr_minors = BITS_TO_LONGS(end) * BITS_PER_LONG; | |
444 | } else | |
445 | old = bitmap; | |
446 | spin_unlock(&minor_lock); | |
447 | kfree(old); | |
448 | } | |
449 | ||
450 | spin_lock(&minor_lock); | |
451 | if (find_next_bit(minors, end, minor) >= end) { | |
34ae2e47 | 452 | bitmap_set(minors, minor, nr); |
0e345826 JB |
453 | rc = 0; |
454 | } else | |
455 | rc = -EBUSY; | |
456 | spin_unlock(&minor_lock); | |
457 | ||
458 | return rc; | |
459 | } | |
460 | ||
461 | static void xlbd_release_minors(unsigned int minor, unsigned int nr) | |
462 | { | |
463 | unsigned int end = minor + nr; | |
464 | ||
465 | BUG_ON(end > nr_minors); | |
466 | spin_lock(&minor_lock); | |
34ae2e47 | 467 | bitmap_clear(minors, minor, nr); |
0e345826 JB |
468 | spin_unlock(&minor_lock); |
469 | } | |
470 | ||
9f27ee59 JF |
471 | static void blkif_restart_queue_callback(void *arg) |
472 | { | |
81f35161 BL |
473 | struct blkfront_ring_info *rinfo = (struct blkfront_ring_info *)arg; |
474 | schedule_work(&rinfo->work); | |
9f27ee59 JF |
475 | } |
476 | ||
afe42d7d | 477 | static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg) |
597592d9 IC |
478 | { |
479 | /* We don't have real geometry info, but let's at least return | |
480 | values consistent with the size of the device */ | |
481 | sector_t nsect = get_capacity(bd->bd_disk); | |
482 | sector_t cylinders = nsect; | |
483 | ||
484 | hg->heads = 0xff; | |
485 | hg->sectors = 0x3f; | |
486 | sector_div(cylinders, hg->heads * hg->sectors); | |
487 | hg->cylinders = cylinders; | |
488 | if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect) | |
489 | hg->cylinders = 0xffff; | |
490 | return 0; | |
491 | } | |
492 | ||
a63c848b | 493 | static int blkif_ioctl(struct block_device *bdev, fmode_t mode, |
62aa0054 | 494 | unsigned command, unsigned long argument) |
440a01a7 | 495 | { |
a63c848b | 496 | struct blkfront_info *info = bdev->bd_disk->private_data; |
440a01a7 CL |
497 | int i; |
498 | ||
499 | dev_dbg(&info->xbdev->dev, "command: 0x%x, argument: 0x%lx\n", | |
500 | command, (long)argument); | |
501 | ||
502 | switch (command) { | |
503 | case CDROMMULTISESSION: | |
504 | dev_dbg(&info->xbdev->dev, "FIXME: support multisession CDs later\n"); | |
505 | for (i = 0; i < sizeof(struct cdrom_multisession); i++) | |
506 | if (put_user(0, (char __user *)(argument + i))) | |
507 | return -EFAULT; | |
508 | return 0; | |
509 | ||
510 | case CDROM_GET_CAPABILITY: { | |
511 | struct gendisk *gd = info->gd; | |
512 | if (gd->flags & GENHD_FL_CD) | |
513 | return 0; | |
514 | return -EINVAL; | |
515 | } | |
516 | ||
517 | default: | |
518 | /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n", | |
519 | command);*/ | |
520 | return -EINVAL; /* same return as native Linux */ | |
521 | } | |
522 | ||
523 | return 0; | |
524 | } | |
525 | ||
2e073969 JG |
526 | static unsigned long blkif_ring_get_request(struct blkfront_ring_info *rinfo, |
527 | struct request *req, | |
528 | struct blkif_request **ring_req) | |
529 | { | |
530 | unsigned long id; | |
531 | ||
532 | *ring_req = RING_GET_REQUEST(&rinfo->ring, rinfo->ring.req_prod_pvt); | |
533 | rinfo->ring.req_prod_pvt++; | |
534 | ||
535 | id = get_id_from_freelist(rinfo); | |
536 | rinfo->shadow[id].request = req; | |
6cc56833 JG |
537 | rinfo->shadow[id].status = REQ_WAITING; |
538 | rinfo->shadow[id].associated_id = NO_ASSOCIATED_ID; | |
2e073969 JG |
539 | |
540 | (*ring_req)->u.rw.id = id; | |
541 | ||
542 | return id; | |
543 | } | |
544 | ||
81f35161 | 545 | static int blkif_queue_discard_req(struct request *req, struct blkfront_ring_info *rinfo) |
9f27ee59 | 546 | { |
81f35161 | 547 | struct blkfront_info *info = rinfo->dev_info; |
9f27ee59 | 548 | struct blkif_request *ring_req; |
9f27ee59 | 549 | unsigned long id; |
33204663 JG |
550 | |
551 | /* Fill out a communications ring structure. */ | |
2e073969 | 552 | id = blkif_ring_get_request(rinfo, req, &ring_req); |
33204663 JG |
553 | |
554 | ring_req->operation = BLKIF_OP_DISCARD; | |
555 | ring_req->u.discard.nr_sectors = blk_rq_sectors(req); | |
556 | ring_req->u.discard.id = id; | |
557 | ring_req->u.discard.sector_number = (blkif_sector_t)blk_rq_pos(req); | |
288dab8a | 558 | if (req_op(req) == REQ_OP_SECURE_ERASE && info->feature_secdiscard) |
33204663 JG |
559 | ring_req->u.discard.flag = BLKIF_DISCARD_SECURE; |
560 | else | |
561 | ring_req->u.discard.flag = 0; | |
562 | ||
33204663 | 563 | /* Keep a private copy so we can reissue requests when recovering. */ |
81f35161 | 564 | rinfo->shadow[id].req = *ring_req; |
33204663 JG |
565 | |
566 | return 0; | |
567 | } | |
568 | ||
c004a6fe JG |
569 | struct setup_rw_req { |
570 | unsigned int grant_idx; | |
571 | struct blkif_request_segment *segments; | |
81f35161 | 572 | struct blkfront_ring_info *rinfo; |
c004a6fe JG |
573 | struct blkif_request *ring_req; |
574 | grant_ref_t gref_head; | |
575 | unsigned int id; | |
576 | /* Only used when persistent grant is used and it's a read request */ | |
577 | bool need_copy; | |
578 | unsigned int bvec_off; | |
579 | char *bvec_data; | |
6cc56833 JG |
580 | |
581 | bool require_extra_req; | |
582 | struct blkif_request *extra_ring_req; | |
c004a6fe JG |
583 | }; |
584 | ||
585 | static void blkif_setup_rw_req_grant(unsigned long gfn, unsigned int offset, | |
586 | unsigned int len, void *data) | |
587 | { | |
588 | struct setup_rw_req *setup = data; | |
589 | int n, ref; | |
590 | struct grant *gnt_list_entry; | |
9f27ee59 | 591 | unsigned int fsect, lsect; |
c004a6fe JG |
592 | /* Convenient aliases */ |
593 | unsigned int grant_idx = setup->grant_idx; | |
594 | struct blkif_request *ring_req = setup->ring_req; | |
81f35161 | 595 | struct blkfront_ring_info *rinfo = setup->rinfo; |
6cc56833 JG |
596 | /* |
597 | * We always use the shadow of the first request to store the list | |
598 | * of grant associated to the block I/O request. This made the | |
599 | * completion more easy to handle even if the block I/O request is | |
600 | * split. | |
601 | */ | |
81f35161 | 602 | struct blk_shadow *shadow = &rinfo->shadow[setup->id]; |
c004a6fe | 603 | |
6cc56833 JG |
604 | if (unlikely(setup->require_extra_req && |
605 | grant_idx >= BLKIF_MAX_SEGMENTS_PER_REQUEST)) { | |
606 | /* | |
607 | * We are using the second request, setup grant_idx | |
608 | * to be the index of the segment array. | |
609 | */ | |
610 | grant_idx -= BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
611 | ring_req = setup->extra_ring_req; | |
612 | } | |
613 | ||
c004a6fe JG |
614 | if ((ring_req->operation == BLKIF_OP_INDIRECT) && |
615 | (grant_idx % GRANTS_PER_INDIRECT_FRAME == 0)) { | |
616 | if (setup->segments) | |
617 | kunmap_atomic(setup->segments); | |
618 | ||
619 | n = grant_idx / GRANTS_PER_INDIRECT_FRAME; | |
73716df7 | 620 | gnt_list_entry = get_indirect_grant(&setup->gref_head, rinfo); |
c004a6fe JG |
621 | shadow->indirect_grants[n] = gnt_list_entry; |
622 | setup->segments = kmap_atomic(gnt_list_entry->page); | |
623 | ring_req->u.indirect.indirect_grefs[n] = gnt_list_entry->gref; | |
624 | } | |
625 | ||
73716df7 | 626 | gnt_list_entry = get_grant(&setup->gref_head, gfn, rinfo); |
c004a6fe | 627 | ref = gnt_list_entry->gref; |
6cc56833 JG |
628 | /* |
629 | * All the grants are stored in the shadow of the first | |
630 | * request. Therefore we have to use the global index. | |
631 | */ | |
632 | shadow->grants_used[setup->grant_idx] = gnt_list_entry; | |
c004a6fe JG |
633 | |
634 | if (setup->need_copy) { | |
635 | void *shared_data; | |
636 | ||
637 | shared_data = kmap_atomic(gnt_list_entry->page); | |
638 | /* | |
639 | * this does not wipe data stored outside the | |
640 | * range sg->offset..sg->offset+sg->length. | |
641 | * Therefore, blkback *could* see data from | |
642 | * previous requests. This is OK as long as | |
643 | * persistent grants are shared with just one | |
644 | * domain. It may need refactoring if this | |
645 | * changes | |
646 | */ | |
647 | memcpy(shared_data + offset, | |
648 | setup->bvec_data + setup->bvec_off, | |
649 | len); | |
650 | ||
651 | kunmap_atomic(shared_data); | |
652 | setup->bvec_off += len; | |
653 | } | |
654 | ||
655 | fsect = offset >> 9; | |
656 | lsect = fsect + (len >> 9) - 1; | |
657 | if (ring_req->operation != BLKIF_OP_INDIRECT) { | |
658 | ring_req->u.rw.seg[grant_idx] = | |
659 | (struct blkif_request_segment) { | |
660 | .gref = ref, | |
661 | .first_sect = fsect, | |
662 | .last_sect = lsect }; | |
663 | } else { | |
664 | setup->segments[grant_idx % GRANTS_PER_INDIRECT_FRAME] = | |
665 | (struct blkif_request_segment) { | |
666 | .gref = ref, | |
667 | .first_sect = fsect, | |
668 | .last_sect = lsect }; | |
669 | } | |
670 | ||
671 | (setup->grant_idx)++; | |
672 | } | |
673 | ||
6cc56833 JG |
674 | static void blkif_setup_extra_req(struct blkif_request *first, |
675 | struct blkif_request *second) | |
676 | { | |
677 | uint16_t nr_segments = first->u.rw.nr_segments; | |
678 | ||
679 | /* | |
680 | * The second request is only present when the first request uses | |
681 | * all its segments. It's always the continuity of the first one. | |
682 | */ | |
683 | first->u.rw.nr_segments = BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
684 | ||
685 | second->u.rw.nr_segments = nr_segments - BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
686 | second->u.rw.sector_number = first->u.rw.sector_number + | |
687 | (BLKIF_MAX_SEGMENTS_PER_REQUEST * XEN_PAGE_SIZE) / 512; | |
688 | ||
689 | second->u.rw.handle = first->u.rw.handle; | |
690 | second->operation = first->operation; | |
691 | } | |
692 | ||
81f35161 | 693 | static int blkif_queue_rw_req(struct request *req, struct blkfront_ring_info *rinfo) |
9f27ee59 | 694 | { |
81f35161 | 695 | struct blkfront_info *info = rinfo->dev_info; |
6cc56833 JG |
696 | struct blkif_request *ring_req, *extra_ring_req = NULL; |
697 | unsigned long id, extra_id = NO_ASSOCIATED_ID; | |
698 | bool require_extra_req = false; | |
c004a6fe JG |
699 | int i; |
700 | struct setup_rw_req setup = { | |
701 | .grant_idx = 0, | |
702 | .segments = NULL, | |
81f35161 | 703 | .rinfo = rinfo, |
c004a6fe JG |
704 | .need_copy = rq_data_dir(req) && info->feature_persistent, |
705 | }; | |
0a8704a5 RPM |
706 | |
707 | /* | |
708 | * Used to store if we are able to queue the request by just using | |
709 | * existing persistent grants, or if we have to get new grants, | |
710 | * as there are not sufficiently many free. | |
711 | */ | |
bd912ef3 | 712 | bool new_persistent_gnts = false; |
9e973e64 | 713 | struct scatterlist *sg; |
c004a6fe | 714 | int num_sg, max_grefs, num_grant; |
9f27ee59 | 715 | |
c004a6fe | 716 | max_grefs = req->nr_phys_segments * GRANTS_PER_PSEG; |
c47206e2 RPM |
717 | if (max_grefs > BLKIF_MAX_SEGMENTS_PER_REQUEST) |
718 | /* | |
719 | * If we are using indirect segments we need to account | |
720 | * for the indirect grefs used in the request. | |
721 | */ | |
c004a6fe | 722 | max_grefs += INDIRECT_GREFS(max_grefs); |
402b27f9 | 723 | |
bd912ef3 DZ |
724 | /* Check if we have enough persistent grants to allocate a requests */ |
725 | if (rinfo->persistent_gnts_c < max_grefs) { | |
726 | new_persistent_gnts = true; | |
727 | ||
728 | if (gnttab_alloc_grant_references( | |
729 | max_grefs - rinfo->persistent_gnts_c, | |
730 | &setup.gref_head) < 0) { | |
0a8704a5 | 731 | gnttab_request_free_callback( |
81f35161 | 732 | &rinfo->callback, |
0a8704a5 | 733 | blkif_restart_queue_callback, |
81f35161 | 734 | rinfo, |
bd912ef3 | 735 | max_grefs - rinfo->persistent_gnts_c); |
0a8704a5 RPM |
736 | return 1; |
737 | } | |
bd912ef3 | 738 | } |
9f27ee59 JF |
739 | |
740 | /* Fill out a communications ring structure. */ | |
2e073969 | 741 | id = blkif_ring_get_request(rinfo, req, &ring_req); |
9f27ee59 | 742 | |
81f35161 | 743 | num_sg = blk_rq_map_sg(req->q, req, rinfo->shadow[id].sg); |
c004a6fe JG |
744 | num_grant = 0; |
745 | /* Calculate the number of grant used */ | |
81f35161 | 746 | for_each_sg(rinfo->shadow[id].sg, sg, num_sg, i) |
c004a6fe JG |
747 | num_grant += gnttab_count_grant(sg->offset, sg->length); |
748 | ||
6cc56833 JG |
749 | require_extra_req = info->max_indirect_segments == 0 && |
750 | num_grant > BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
751 | BUG_ON(!HAS_EXTRA_REQ && require_extra_req); | |
752 | ||
81f35161 | 753 | rinfo->shadow[id].num_sg = num_sg; |
6cc56833 JG |
754 | if (num_grant > BLKIF_MAX_SEGMENTS_PER_REQUEST && |
755 | likely(!require_extra_req)) { | |
33204663 JG |
756 | /* |
757 | * The indirect operation can only be a BLKIF_OP_READ or | |
758 | * BLKIF_OP_WRITE | |
759 | */ | |
3a5e02ce | 760 | BUG_ON(req_op(req) == REQ_OP_FLUSH || req->cmd_flags & REQ_FUA); |
33204663 JG |
761 | ring_req->operation = BLKIF_OP_INDIRECT; |
762 | ring_req->u.indirect.indirect_op = rq_data_dir(req) ? | |
763 | BLKIF_OP_WRITE : BLKIF_OP_READ; | |
764 | ring_req->u.indirect.sector_number = (blkif_sector_t)blk_rq_pos(req); | |
765 | ring_req->u.indirect.handle = info->handle; | |
c004a6fe | 766 | ring_req->u.indirect.nr_segments = num_grant; |
ed30bf31 | 767 | } else { |
33204663 JG |
768 | ring_req->u.rw.sector_number = (blkif_sector_t)blk_rq_pos(req); |
769 | ring_req->u.rw.handle = info->handle; | |
770 | ring_req->operation = rq_data_dir(req) ? | |
771 | BLKIF_OP_WRITE : BLKIF_OP_READ; | |
3a5e02ce | 772 | if (req_op(req) == REQ_OP_FLUSH || req->cmd_flags & REQ_FUA) { |
402b27f9 | 773 | /* |
33204663 JG |
774 | * Ideally we can do an unordered flush-to-disk. |
775 | * In case the backend onlysupports barriers, use that. | |
776 | * A barrier request a superset of FUA, so we can | |
777 | * implement it the same way. (It's also a FLUSH+FUA, | |
778 | * since it is guaranteed ordered WRT previous writes.) | |
402b27f9 | 779 | */ |
a418090a | 780 | if (info->feature_flush && info->feature_fua) |
33204663 JG |
781 | ring_req->operation = |
782 | BLKIF_OP_WRITE_BARRIER; | |
a418090a | 783 | else if (info->feature_flush) |
33204663 JG |
784 | ring_req->operation = |
785 | BLKIF_OP_FLUSH_DISKCACHE; | |
a418090a | 786 | else |
33204663 | 787 | ring_req->operation = 0; |
402b27f9 | 788 | } |
c004a6fe | 789 | ring_req->u.rw.nr_segments = num_grant; |
6cc56833 JG |
790 | if (unlikely(require_extra_req)) { |
791 | extra_id = blkif_ring_get_request(rinfo, req, | |
792 | &extra_ring_req); | |
793 | /* | |
794 | * Only the first request contains the scatter-gather | |
795 | * list. | |
796 | */ | |
797 | rinfo->shadow[extra_id].num_sg = 0; | |
798 | ||
799 | blkif_setup_extra_req(ring_req, extra_ring_req); | |
800 | ||
801 | /* Link the 2 requests together */ | |
802 | rinfo->shadow[extra_id].associated_id = id; | |
803 | rinfo->shadow[id].associated_id = extra_id; | |
804 | } | |
33204663 | 805 | } |
0a8704a5 | 806 | |
c004a6fe JG |
807 | setup.ring_req = ring_req; |
808 | setup.id = id; | |
6cc56833 JG |
809 | |
810 | setup.require_extra_req = require_extra_req; | |
811 | if (unlikely(require_extra_req)) | |
812 | setup.extra_ring_req = extra_ring_req; | |
813 | ||
81f35161 | 814 | for_each_sg(rinfo->shadow[id].sg, sg, num_sg, i) { |
c004a6fe | 815 | BUG_ON(sg->offset + sg->length > PAGE_SIZE); |
0a8704a5 | 816 | |
c004a6fe JG |
817 | if (setup.need_copy) { |
818 | setup.bvec_off = sg->offset; | |
819 | setup.bvec_data = kmap_atomic(sg_page(sg)); | |
820 | } | |
0a8704a5 | 821 | |
c004a6fe JG |
822 | gnttab_foreach_grant_in_range(sg_page(sg), |
823 | sg->offset, | |
824 | sg->length, | |
825 | blkif_setup_rw_req_grant, | |
826 | &setup); | |
0a8704a5 | 827 | |
c004a6fe JG |
828 | if (setup.need_copy) |
829 | kunmap_atomic(setup.bvec_data); | |
9f27ee59 | 830 | } |
c004a6fe JG |
831 | if (setup.segments) |
832 | kunmap_atomic(setup.segments); | |
9f27ee59 | 833 | |
9f27ee59 | 834 | /* Keep a private copy so we can reissue requests when recovering. */ |
81f35161 | 835 | rinfo->shadow[id].req = *ring_req; |
6cc56833 JG |
836 | if (unlikely(require_extra_req)) |
837 | rinfo->shadow[extra_id].req = *extra_ring_req; | |
9f27ee59 | 838 | |
bd912ef3 | 839 | if (new_persistent_gnts) |
c004a6fe | 840 | gnttab_free_grant_references(setup.gref_head); |
9f27ee59 JF |
841 | |
842 | return 0; | |
843 | } | |
844 | ||
33204663 JG |
845 | /* |
846 | * Generate a Xen blkfront IO request from a blk layer request. Reads | |
847 | * and writes are handled as expected. | |
848 | * | |
849 | * @req: a request struct | |
850 | */ | |
81f35161 | 851 | static int blkif_queue_request(struct request *req, struct blkfront_ring_info *rinfo) |
33204663 | 852 | { |
81f35161 | 853 | if (unlikely(rinfo->dev_info->connected != BLKIF_STATE_CONNECTED)) |
33204663 JG |
854 | return 1; |
855 | ||
c2df40df | 856 | if (unlikely(req_op(req) == REQ_OP_DISCARD || |
288dab8a | 857 | req_op(req) == REQ_OP_SECURE_ERASE)) |
81f35161 | 858 | return blkif_queue_discard_req(req, rinfo); |
33204663 | 859 | else |
81f35161 | 860 | return blkif_queue_rw_req(req, rinfo); |
33204663 | 861 | } |
9f27ee59 | 862 | |
81f35161 | 863 | static inline void flush_requests(struct blkfront_ring_info *rinfo) |
9f27ee59 JF |
864 | { |
865 | int notify; | |
866 | ||
81f35161 | 867 | RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&rinfo->ring, notify); |
9f27ee59 JF |
868 | |
869 | if (notify) | |
81f35161 | 870 | notify_remote_via_irq(rinfo->irq); |
9f27ee59 JF |
871 | } |
872 | ||
ad42d391 VK |
873 | static inline bool blkif_request_flush_invalid(struct request *req, |
874 | struct blkfront_info *info) | |
0f1ca65e | 875 | { |
aebf526b | 876 | return (blk_rq_is_passthrough(req) || |
3a5e02ce | 877 | ((req_op(req) == REQ_OP_FLUSH) && |
a418090a | 878 | !info->feature_flush) || |
ad42d391 | 879 | ((req->cmd_flags & REQ_FUA) && |
a418090a | 880 | !info->feature_fua)); |
0f1ca65e AA |
881 | } |
882 | ||
fc17b653 | 883 | static blk_status_t blkif_queue_rq(struct blk_mq_hw_ctx *hctx, |
6f03a7ff | 884 | const struct blk_mq_queue_data *qd) |
9f27ee59 | 885 | { |
11659569 | 886 | unsigned long flags; |
2a6f71ad BL |
887 | int qid = hctx->queue_num; |
888 | struct blkfront_info *info = hctx->queue->queuedata; | |
889 | struct blkfront_ring_info *rinfo = NULL; | |
9f27ee59 | 890 | |
2a6f71ad BL |
891 | BUG_ON(info->nr_rings <= qid); |
892 | rinfo = &info->rinfo[qid]; | |
907c3eb1 | 893 | blk_mq_start_request(qd->rq); |
11659569 | 894 | spin_lock_irqsave(&rinfo->ring_lock, flags); |
81f35161 | 895 | if (RING_FULL(&rinfo->ring)) |
907c3eb1 | 896 | goto out_busy; |
9f27ee59 | 897 | |
81f35161 | 898 | if (blkif_request_flush_invalid(qd->rq, rinfo->dev_info)) |
907c3eb1 | 899 | goto out_err; |
296b2f6a | 900 | |
81f35161 | 901 | if (blkif_queue_request(qd->rq, rinfo)) |
907c3eb1 | 902 | goto out_busy; |
296b2f6a | 903 | |
81f35161 | 904 | flush_requests(rinfo); |
11659569 | 905 | spin_unlock_irqrestore(&rinfo->ring_lock, flags); |
fc17b653 | 906 | return BLK_STS_OK; |
9f27ee59 | 907 | |
907c3eb1 | 908 | out_err: |
11659569 | 909 | spin_unlock_irqrestore(&rinfo->ring_lock, flags); |
fc17b653 | 910 | return BLK_STS_IOERR; |
9f27ee59 | 911 | |
907c3eb1 | 912 | out_busy: |
907c3eb1 | 913 | blk_mq_stop_hw_queue(hctx); |
4b422cb9 | 914 | spin_unlock_irqrestore(&rinfo->ring_lock, flags); |
fc17b653 | 915 | return BLK_STS_RESOURCE; |
9f27ee59 JF |
916 | } |
917 | ||
2609587c CH |
918 | static void blkif_complete_rq(struct request *rq) |
919 | { | |
920 | blk_mq_end_request(rq, blkif_req(rq)->error); | |
921 | } | |
922 | ||
f363b089 | 923 | static const struct blk_mq_ops blkfront_mq_ops = { |
907c3eb1 | 924 | .queue_rq = blkif_queue_rq, |
2609587c | 925 | .complete = blkif_complete_rq, |
907c3eb1 BL |
926 | }; |
927 | ||
172335ad BL |
928 | static void blkif_set_queue_limits(struct blkfront_info *info) |
929 | { | |
930 | struct request_queue *rq = info->rq; | |
931 | struct gendisk *gd = info->gd; | |
932 | unsigned int segments = info->max_indirect_segments ? : | |
933 | BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
934 | ||
935 | queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq); | |
936 | ||
937 | if (info->feature_discard) { | |
938 | queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, rq); | |
939 | blk_queue_max_discard_sectors(rq, get_capacity(gd)); | |
940 | rq->limits.discard_granularity = info->discard_granularity; | |
941 | rq->limits.discard_alignment = info->discard_alignment; | |
942 | if (info->feature_secdiscard) | |
943 | queue_flag_set_unlocked(QUEUE_FLAG_SECERASE, rq); | |
944 | } | |
945 | ||
946 | /* Hard sector size and max sectors impersonate the equiv. hardware. */ | |
947 | blk_queue_logical_block_size(rq, info->sector_size); | |
948 | blk_queue_physical_block_size(rq, info->physical_sector_size); | |
949 | blk_queue_max_hw_sectors(rq, (segments * XEN_PAGE_SIZE) / 512); | |
950 | ||
951 | /* Each segment in a request is up to an aligned page in size. */ | |
952 | blk_queue_segment_boundary(rq, PAGE_SIZE - 1); | |
953 | blk_queue_max_segment_size(rq, PAGE_SIZE); | |
954 | ||
955 | /* Ensure a merged request will fit in a single I/O ring slot. */ | |
956 | blk_queue_max_segments(rq, segments / GRANTS_PER_PSEG); | |
957 | ||
958 | /* Make sure buffer addresses are sector-aligned. */ | |
959 | blk_queue_dma_alignment(rq, 511); | |
172335ad BL |
960 | } |
961 | ||
402b27f9 | 962 | static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size, |
172335ad | 963 | unsigned int physical_sector_size) |
9f27ee59 | 964 | { |
165125e1 | 965 | struct request_queue *rq; |
ed30bf31 | 966 | struct blkfront_info *info = gd->private_data; |
9f27ee59 | 967 | |
907c3eb1 BL |
968 | memset(&info->tag_set, 0, sizeof(info->tag_set)); |
969 | info->tag_set.ops = &blkfront_mq_ops; | |
28d949bc | 970 | info->tag_set.nr_hw_queues = info->nr_rings; |
6cc56833 JG |
971 | if (HAS_EXTRA_REQ && info->max_indirect_segments == 0) { |
972 | /* | |
973 | * When indirect descriptior is not supported, the I/O request | |
974 | * will be split between multiple request in the ring. | |
975 | * To avoid problems when sending the request, divide by | |
976 | * 2 the depth of the queue. | |
977 | */ | |
978 | info->tag_set.queue_depth = BLK_RING_SIZE(info) / 2; | |
979 | } else | |
980 | info->tag_set.queue_depth = BLK_RING_SIZE(info); | |
907c3eb1 BL |
981 | info->tag_set.numa_node = NUMA_NO_NODE; |
982 | info->tag_set.flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_SG_MERGE; | |
2609587c | 983 | info->tag_set.cmd_size = sizeof(struct blkif_req); |
907c3eb1 BL |
984 | info->tag_set.driver_data = info; |
985 | ||
986 | if (blk_mq_alloc_tag_set(&info->tag_set)) | |
bde21f73 | 987 | return -EINVAL; |
907c3eb1 BL |
988 | rq = blk_mq_init_queue(&info->tag_set); |
989 | if (IS_ERR(rq)) { | |
990 | blk_mq_free_tag_set(&info->tag_set); | |
bde21f73 | 991 | return PTR_ERR(rq); |
907c3eb1 | 992 | } |
9f27ee59 | 993 | |
2a6f71ad | 994 | rq->queuedata = info; |
172335ad BL |
995 | info->rq = gd->queue = rq; |
996 | info->gd = gd; | |
997 | info->sector_size = sector_size; | |
998 | info->physical_sector_size = physical_sector_size; | |
999 | blkif_set_queue_limits(info); | |
9f27ee59 JF |
1000 | |
1001 | return 0; | |
1002 | } | |
1003 | ||
a418090a | 1004 | static const char *flush_info(struct blkfront_info *info) |
fdf9b965 | 1005 | { |
a418090a | 1006 | if (info->feature_flush && info->feature_fua) |
fdf9b965 | 1007 | return "barrier: enabled;"; |
a418090a | 1008 | else if (info->feature_flush) |
fdf9b965 | 1009 | return "flush diskcache: enabled;"; |
a418090a | 1010 | else |
fdf9b965 | 1011 | return "barrier or flush: disabled;"; |
fdf9b965 | 1012 | } |
9f27ee59 | 1013 | |
4913efe4 | 1014 | static void xlvbd_flush(struct blkfront_info *info) |
9f27ee59 | 1015 | { |
a418090a MC |
1016 | blk_queue_write_cache(info->rq, info->feature_flush ? true : false, |
1017 | info->feature_fua ? true : false); | |
fdf9b965 | 1018 | pr_info("blkfront: %s: %s %s %s %s %s\n", |
a418090a | 1019 | info->gd->disk_name, flush_info(info), |
fdf9b965 VK |
1020 | "persistent grants:", info->feature_persistent ? |
1021 | "enabled;" : "disabled;", "indirect descriptors:", | |
1022 | info->max_indirect_segments ? "enabled;" : "disabled;"); | |
9f27ee59 JF |
1023 | } |
1024 | ||
c80a4209 SS |
1025 | static int xen_translate_vdev(int vdevice, int *minor, unsigned int *offset) |
1026 | { | |
1027 | int major; | |
1028 | major = BLKIF_MAJOR(vdevice); | |
1029 | *minor = BLKIF_MINOR(vdevice); | |
1030 | switch (major) { | |
1031 | case XEN_IDE0_MAJOR: | |
1032 | *offset = (*minor / 64) + EMULATED_HD_DISK_NAME_OFFSET; | |
1033 | *minor = ((*minor / 64) * PARTS_PER_DISK) + | |
1034 | EMULATED_HD_DISK_MINOR_OFFSET; | |
1035 | break; | |
1036 | case XEN_IDE1_MAJOR: | |
1037 | *offset = (*minor / 64) + 2 + EMULATED_HD_DISK_NAME_OFFSET; | |
1038 | *minor = (((*minor / 64) + 2) * PARTS_PER_DISK) + | |
1039 | EMULATED_HD_DISK_MINOR_OFFSET; | |
1040 | break; | |
1041 | case XEN_SCSI_DISK0_MAJOR: | |
1042 | *offset = (*minor / PARTS_PER_DISK) + EMULATED_SD_DISK_NAME_OFFSET; | |
1043 | *minor = *minor + EMULATED_SD_DISK_MINOR_OFFSET; | |
1044 | break; | |
1045 | case XEN_SCSI_DISK1_MAJOR: | |
1046 | case XEN_SCSI_DISK2_MAJOR: | |
1047 | case XEN_SCSI_DISK3_MAJOR: | |
1048 | case XEN_SCSI_DISK4_MAJOR: | |
1049 | case XEN_SCSI_DISK5_MAJOR: | |
1050 | case XEN_SCSI_DISK6_MAJOR: | |
1051 | case XEN_SCSI_DISK7_MAJOR: | |
1052 | *offset = (*minor / PARTS_PER_DISK) + | |
1053 | ((major - XEN_SCSI_DISK1_MAJOR + 1) * 16) + | |
1054 | EMULATED_SD_DISK_NAME_OFFSET; | |
1055 | *minor = *minor + | |
1056 | ((major - XEN_SCSI_DISK1_MAJOR + 1) * 16 * PARTS_PER_DISK) + | |
1057 | EMULATED_SD_DISK_MINOR_OFFSET; | |
1058 | break; | |
1059 | case XEN_SCSI_DISK8_MAJOR: | |
1060 | case XEN_SCSI_DISK9_MAJOR: | |
1061 | case XEN_SCSI_DISK10_MAJOR: | |
1062 | case XEN_SCSI_DISK11_MAJOR: | |
1063 | case XEN_SCSI_DISK12_MAJOR: | |
1064 | case XEN_SCSI_DISK13_MAJOR: | |
1065 | case XEN_SCSI_DISK14_MAJOR: | |
1066 | case XEN_SCSI_DISK15_MAJOR: | |
1067 | *offset = (*minor / PARTS_PER_DISK) + | |
1068 | ((major - XEN_SCSI_DISK8_MAJOR + 8) * 16) + | |
1069 | EMULATED_SD_DISK_NAME_OFFSET; | |
1070 | *minor = *minor + | |
1071 | ((major - XEN_SCSI_DISK8_MAJOR + 8) * 16 * PARTS_PER_DISK) + | |
1072 | EMULATED_SD_DISK_MINOR_OFFSET; | |
1073 | break; | |
1074 | case XENVBD_MAJOR: | |
1075 | *offset = *minor / PARTS_PER_DISK; | |
1076 | break; | |
1077 | default: | |
1078 | printk(KERN_WARNING "blkfront: your disk configuration is " | |
1079 | "incorrect, please use an xvd device instead\n"); | |
1080 | return -ENODEV; | |
1081 | } | |
1082 | return 0; | |
1083 | } | |
9f27ee59 | 1084 | |
e77c78c0 JB |
1085 | static char *encode_disk_name(char *ptr, unsigned int n) |
1086 | { | |
1087 | if (n >= 26) | |
1088 | ptr = encode_disk_name(ptr, n / 26 - 1); | |
1089 | *ptr = 'a' + n % 26; | |
1090 | return ptr + 1; | |
1091 | } | |
1092 | ||
9246b5f0 CL |
1093 | static int xlvbd_alloc_gendisk(blkif_sector_t capacity, |
1094 | struct blkfront_info *info, | |
7c4d7d71 SB |
1095 | u16 vdisk_info, u16 sector_size, |
1096 | unsigned int physical_sector_size) | |
9f27ee59 JF |
1097 | { |
1098 | struct gendisk *gd; | |
1099 | int nr_minors = 1; | |
c80a4209 | 1100 | int err; |
9246b5f0 CL |
1101 | unsigned int offset; |
1102 | int minor; | |
1103 | int nr_parts; | |
e77c78c0 | 1104 | char *ptr; |
9f27ee59 JF |
1105 | |
1106 | BUG_ON(info->gd != NULL); | |
1107 | BUG_ON(info->rq != NULL); | |
1108 | ||
9246b5f0 CL |
1109 | if ((info->vdevice>>EXT_SHIFT) > 1) { |
1110 | /* this is above the extended range; something is wrong */ | |
1111 | printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", info->vdevice); | |
1112 | return -ENODEV; | |
1113 | } | |
1114 | ||
1115 | if (!VDEV_IS_EXTENDED(info->vdevice)) { | |
c80a4209 SS |
1116 | err = xen_translate_vdev(info->vdevice, &minor, &offset); |
1117 | if (err) | |
1118 | return err; | |
1119 | nr_parts = PARTS_PER_DISK; | |
9246b5f0 CL |
1120 | } else { |
1121 | minor = BLKIF_MINOR_EXT(info->vdevice); | |
1122 | nr_parts = PARTS_PER_EXT_DISK; | |
c80a4209 | 1123 | offset = minor / nr_parts; |
89153b5c | 1124 | if (xen_hvm_domain() && offset < EMULATED_HD_DISK_NAME_OFFSET + 4) |
c80a4209 SS |
1125 | printk(KERN_WARNING "blkfront: vdevice 0x%x might conflict with " |
1126 | "emulated IDE disks,\n\t choose an xvd device name" | |
1127 | "from xvde on\n", info->vdevice); | |
9246b5f0 | 1128 | } |
e77c78c0 JB |
1129 | if (minor >> MINORBITS) { |
1130 | pr_warn("blkfront: %#x's minor (%#x) out of range; ignoring\n", | |
1131 | info->vdevice, minor); | |
1132 | return -ENODEV; | |
1133 | } | |
9246b5f0 CL |
1134 | |
1135 | if ((minor % nr_parts) == 0) | |
1136 | nr_minors = nr_parts; | |
9f27ee59 | 1137 | |
0e345826 JB |
1138 | err = xlbd_reserve_minors(minor, nr_minors); |
1139 | if (err) | |
1140 | goto out; | |
1141 | err = -ENODEV; | |
1142 | ||
9f27ee59 JF |
1143 | gd = alloc_disk(nr_minors); |
1144 | if (gd == NULL) | |
0e345826 | 1145 | goto release; |
9f27ee59 | 1146 | |
e77c78c0 JB |
1147 | strcpy(gd->disk_name, DEV_NAME); |
1148 | ptr = encode_disk_name(gd->disk_name + sizeof(DEV_NAME) - 1, offset); | |
1149 | BUG_ON(ptr >= gd->disk_name + DISK_NAME_LEN); | |
1150 | if (nr_minors > 1) | |
1151 | *ptr = 0; | |
1152 | else | |
1153 | snprintf(ptr, gd->disk_name + DISK_NAME_LEN - ptr, | |
1154 | "%d", minor & (nr_parts - 1)); | |
9f27ee59 JF |
1155 | |
1156 | gd->major = XENVBD_MAJOR; | |
1157 | gd->first_minor = minor; | |
1158 | gd->fops = &xlvbd_block_fops; | |
1159 | gd->private_data = info; | |
9f27ee59 JF |
1160 | set_capacity(gd, capacity); |
1161 | ||
172335ad | 1162 | if (xlvbd_init_blk_queue(gd, sector_size, physical_sector_size)) { |
9f27ee59 | 1163 | del_gendisk(gd); |
0e345826 | 1164 | goto release; |
9f27ee59 JF |
1165 | } |
1166 | ||
4913efe4 | 1167 | xlvbd_flush(info); |
9f27ee59 JF |
1168 | |
1169 | if (vdisk_info & VDISK_READONLY) | |
1170 | set_disk_ro(gd, 1); | |
1171 | ||
1172 | if (vdisk_info & VDISK_REMOVABLE) | |
1173 | gd->flags |= GENHD_FL_REMOVABLE; | |
1174 | ||
1175 | if (vdisk_info & VDISK_CDROM) | |
1176 | gd->flags |= GENHD_FL_CD; | |
1177 | ||
1178 | return 0; | |
1179 | ||
0e345826 JB |
1180 | release: |
1181 | xlbd_release_minors(minor, nr_minors); | |
9f27ee59 JF |
1182 | out: |
1183 | return err; | |
1184 | } | |
1185 | ||
a66b5aeb DS |
1186 | static void xlvbd_release_gendisk(struct blkfront_info *info) |
1187 | { | |
3df0e505 | 1188 | unsigned int minor, nr_minors, i; |
a66b5aeb DS |
1189 | |
1190 | if (info->rq == NULL) | |
1191 | return; | |
1192 | ||
a66b5aeb | 1193 | /* No more blkif_request(). */ |
907c3eb1 | 1194 | blk_mq_stop_hw_queues(info->rq); |
a66b5aeb | 1195 | |
3df0e505 BL |
1196 | for (i = 0; i < info->nr_rings; i++) { |
1197 | struct blkfront_ring_info *rinfo = &info->rinfo[i]; | |
a66b5aeb | 1198 | |
3df0e505 BL |
1199 | /* No more gnttab callback work. */ |
1200 | gnttab_cancel_free_callback(&rinfo->callback); | |
1201 | ||
1202 | /* Flush gnttab callback work. Must be done with no locks held. */ | |
1203 | flush_work(&rinfo->work); | |
1204 | } | |
a66b5aeb DS |
1205 | |
1206 | del_gendisk(info->gd); | |
1207 | ||
1208 | minor = info->gd->first_minor; | |
1209 | nr_minors = info->gd->minors; | |
1210 | xlbd_release_minors(minor, nr_minors); | |
1211 | ||
1212 | blk_cleanup_queue(info->rq); | |
907c3eb1 | 1213 | blk_mq_free_tag_set(&info->tag_set); |
a66b5aeb DS |
1214 | info->rq = NULL; |
1215 | ||
1216 | put_disk(info->gd); | |
1217 | info->gd = NULL; | |
1218 | } | |
1219 | ||
11659569 BL |
1220 | /* Already hold rinfo->ring_lock. */ |
1221 | static inline void kick_pending_request_queues_locked(struct blkfront_ring_info *rinfo) | |
9f27ee59 | 1222 | { |
81f35161 BL |
1223 | if (!RING_FULL(&rinfo->ring)) |
1224 | blk_mq_start_stopped_hw_queues(rinfo->dev_info->rq, true); | |
9f27ee59 JF |
1225 | } |
1226 | ||
11659569 BL |
1227 | static void kick_pending_request_queues(struct blkfront_ring_info *rinfo) |
1228 | { | |
1229 | unsigned long flags; | |
1230 | ||
1231 | spin_lock_irqsave(&rinfo->ring_lock, flags); | |
1232 | kick_pending_request_queues_locked(rinfo); | |
1233 | spin_unlock_irqrestore(&rinfo->ring_lock, flags); | |
1234 | } | |
1235 | ||
9f27ee59 JF |
1236 | static void blkif_restart_queue(struct work_struct *work) |
1237 | { | |
81f35161 | 1238 | struct blkfront_ring_info *rinfo = container_of(work, struct blkfront_ring_info, work); |
9f27ee59 | 1239 | |
81f35161 BL |
1240 | if (rinfo->dev_info->connected == BLKIF_STATE_CONNECTED) |
1241 | kick_pending_request_queues(rinfo); | |
9f27ee59 JF |
1242 | } |
1243 | ||
3df0e505 | 1244 | static void blkif_free_ring(struct blkfront_ring_info *rinfo) |
9f27ee59 | 1245 | { |
73716df7 | 1246 | struct grant *persistent_gnt, *n; |
3df0e505 | 1247 | struct blkfront_info *info = rinfo->dev_info; |
402b27f9 | 1248 | int i, j, segs; |
0a8704a5 | 1249 | |
bfe11d6d RPM |
1250 | /* |
1251 | * Remove indirect pages, this only happens when using indirect | |
1252 | * descriptors but not persistent grants | |
1253 | */ | |
81f35161 | 1254 | if (!list_empty(&rinfo->indirect_pages)) { |
bfe11d6d RPM |
1255 | struct page *indirect_page, *n; |
1256 | ||
1257 | BUG_ON(info->feature_persistent); | |
81f35161 | 1258 | list_for_each_entry_safe(indirect_page, n, &rinfo->indirect_pages, lru) { |
bfe11d6d RPM |
1259 | list_del(&indirect_page->lru); |
1260 | __free_page(indirect_page); | |
1261 | } | |
1262 | } | |
1263 | ||
73716df7 BL |
1264 | /* Remove all persistent grants. */ |
1265 | if (!list_empty(&rinfo->grants)) { | |
1266 | list_for_each_entry_safe(persistent_gnt, n, | |
1267 | &rinfo->grants, node) { | |
1268 | list_del(&persistent_gnt->node); | |
1269 | if (persistent_gnt->gref != GRANT_INVALID_REF) { | |
1270 | gnttab_end_foreign_access(persistent_gnt->gref, | |
1271 | 0, 0UL); | |
1272 | rinfo->persistent_gnts_c--; | |
1273 | } | |
1274 | if (info->feature_persistent) | |
1275 | __free_page(persistent_gnt->page); | |
1276 | kfree(persistent_gnt); | |
1277 | } | |
1278 | } | |
1279 | BUG_ON(rinfo->persistent_gnts_c != 0); | |
1280 | ||
86839c56 | 1281 | for (i = 0; i < BLK_RING_SIZE(info); i++) { |
402b27f9 RPM |
1282 | /* |
1283 | * Clear persistent grants present in requests already | |
1284 | * on the shared ring | |
1285 | */ | |
81f35161 | 1286 | if (!rinfo->shadow[i].request) |
402b27f9 RPM |
1287 | goto free_shadow; |
1288 | ||
81f35161 BL |
1289 | segs = rinfo->shadow[i].req.operation == BLKIF_OP_INDIRECT ? |
1290 | rinfo->shadow[i].req.u.indirect.nr_segments : | |
1291 | rinfo->shadow[i].req.u.rw.nr_segments; | |
402b27f9 | 1292 | for (j = 0; j < segs; j++) { |
81f35161 | 1293 | persistent_gnt = rinfo->shadow[i].grants_used[j]; |
402b27f9 | 1294 | gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL); |
bfe11d6d | 1295 | if (info->feature_persistent) |
a7a6df22 | 1296 | __free_page(persistent_gnt->page); |
402b27f9 RPM |
1297 | kfree(persistent_gnt); |
1298 | } | |
1299 | ||
81f35161 | 1300 | if (rinfo->shadow[i].req.operation != BLKIF_OP_INDIRECT) |
402b27f9 RPM |
1301 | /* |
1302 | * If this is not an indirect operation don't try to | |
1303 | * free indirect segments | |
1304 | */ | |
1305 | goto free_shadow; | |
1306 | ||
1307 | for (j = 0; j < INDIRECT_GREFS(segs); j++) { | |
81f35161 | 1308 | persistent_gnt = rinfo->shadow[i].indirect_grants[j]; |
402b27f9 | 1309 | gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL); |
a7a6df22 | 1310 | __free_page(persistent_gnt->page); |
402b27f9 RPM |
1311 | kfree(persistent_gnt); |
1312 | } | |
1313 | ||
1314 | free_shadow: | |
81f35161 BL |
1315 | kfree(rinfo->shadow[i].grants_used); |
1316 | rinfo->shadow[i].grants_used = NULL; | |
1317 | kfree(rinfo->shadow[i].indirect_grants); | |
1318 | rinfo->shadow[i].indirect_grants = NULL; | |
1319 | kfree(rinfo->shadow[i].sg); | |
1320 | rinfo->shadow[i].sg = NULL; | |
402b27f9 RPM |
1321 | } |
1322 | ||
9f27ee59 | 1323 | /* No more gnttab callback work. */ |
81f35161 | 1324 | gnttab_cancel_free_callback(&rinfo->callback); |
9f27ee59 JF |
1325 | |
1326 | /* Flush gnttab callback work. Must be done with no locks held. */ | |
81f35161 | 1327 | flush_work(&rinfo->work); |
9f27ee59 JF |
1328 | |
1329 | /* Free resources associated with old device channel. */ | |
86839c56 | 1330 | for (i = 0; i < info->nr_ring_pages; i++) { |
81f35161 BL |
1331 | if (rinfo->ring_ref[i] != GRANT_INVALID_REF) { |
1332 | gnttab_end_foreign_access(rinfo->ring_ref[i], 0, 0); | |
1333 | rinfo->ring_ref[i] = GRANT_INVALID_REF; | |
86839c56 | 1334 | } |
9f27ee59 | 1335 | } |
6c647b0e | 1336 | free_pages((unsigned long)rinfo->ring.sring, get_order(info->nr_ring_pages * XEN_PAGE_SIZE)); |
81f35161 | 1337 | rinfo->ring.sring = NULL; |
86839c56 | 1338 | |
81f35161 BL |
1339 | if (rinfo->irq) |
1340 | unbind_from_irqhandler(rinfo->irq, rinfo); | |
1341 | rinfo->evtchn = rinfo->irq = 0; | |
3df0e505 | 1342 | } |
9f27ee59 | 1343 | |
3df0e505 BL |
1344 | static void blkif_free(struct blkfront_info *info, int suspend) |
1345 | { | |
3df0e505 BL |
1346 | unsigned int i; |
1347 | ||
1348 | /* Prevent new requests being issued until we fix things up. */ | |
3df0e505 BL |
1349 | info->connected = suspend ? |
1350 | BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED; | |
1351 | /* No more blkif_request(). */ | |
1352 | if (info->rq) | |
1353 | blk_mq_stop_hw_queues(info->rq); | |
1354 | ||
3df0e505 BL |
1355 | for (i = 0; i < info->nr_rings; i++) |
1356 | blkif_free_ring(&info->rinfo[i]); | |
1357 | ||
1358 | kfree(info->rinfo); | |
1359 | info->rinfo = NULL; | |
1360 | info->nr_rings = 0; | |
9f27ee59 JF |
1361 | } |
1362 | ||
c004a6fe JG |
1363 | struct copy_from_grant { |
1364 | const struct blk_shadow *s; | |
1365 | unsigned int grant_idx; | |
1366 | unsigned int bvec_offset; | |
1367 | char *bvec_data; | |
1368 | }; | |
1369 | ||
1370 | static void blkif_copy_from_grant(unsigned long gfn, unsigned int offset, | |
1371 | unsigned int len, void *data) | |
1372 | { | |
1373 | struct copy_from_grant *info = data; | |
1374 | char *shared_data; | |
1375 | /* Convenient aliases */ | |
1376 | const struct blk_shadow *s = info->s; | |
1377 | ||
1378 | shared_data = kmap_atomic(s->grants_used[info->grant_idx]->page); | |
1379 | ||
1380 | memcpy(info->bvec_data + info->bvec_offset, | |
1381 | shared_data + offset, len); | |
1382 | ||
1383 | info->bvec_offset += len; | |
1384 | info->grant_idx++; | |
1385 | ||
1386 | kunmap_atomic(shared_data); | |
1387 | } | |
1388 | ||
6cc56833 JG |
1389 | static enum blk_req_status blkif_rsp_to_req_status(int rsp) |
1390 | { | |
1391 | switch (rsp) | |
1392 | { | |
1393 | case BLKIF_RSP_OKAY: | |
1394 | return REQ_DONE; | |
1395 | case BLKIF_RSP_EOPNOTSUPP: | |
1396 | return REQ_EOPNOTSUPP; | |
1397 | case BLKIF_RSP_ERROR: | |
1398 | /* Fallthrough. */ | |
1399 | default: | |
1400 | return REQ_ERROR; | |
1401 | } | |
1402 | } | |
1403 | ||
1404 | /* | |
1405 | * Get the final status of the block request based on two ring response | |
1406 | */ | |
1407 | static int blkif_get_final_status(enum blk_req_status s1, | |
1408 | enum blk_req_status s2) | |
1409 | { | |
1410 | BUG_ON(s1 == REQ_WAITING); | |
1411 | BUG_ON(s2 == REQ_WAITING); | |
1412 | ||
1413 | if (s1 == REQ_ERROR || s2 == REQ_ERROR) | |
1414 | return BLKIF_RSP_ERROR; | |
1415 | else if (s1 == REQ_EOPNOTSUPP || s2 == REQ_EOPNOTSUPP) | |
1416 | return BLKIF_RSP_EOPNOTSUPP; | |
1417 | return BLKIF_RSP_OKAY; | |
1418 | } | |
1419 | ||
1420 | static bool blkif_completion(unsigned long *id, | |
1421 | struct blkfront_ring_info *rinfo, | |
0a8704a5 | 1422 | struct blkif_response *bret) |
9f27ee59 | 1423 | { |
d62f6918 | 1424 | int i = 0; |
b7649158 | 1425 | struct scatterlist *sg; |
c004a6fe | 1426 | int num_sg, num_grant; |
81f35161 | 1427 | struct blkfront_info *info = rinfo->dev_info; |
6cc56833 | 1428 | struct blk_shadow *s = &rinfo->shadow[*id]; |
c004a6fe | 1429 | struct copy_from_grant data = { |
c004a6fe JG |
1430 | .grant_idx = 0, |
1431 | }; | |
402b27f9 | 1432 | |
c004a6fe | 1433 | num_grant = s->req.operation == BLKIF_OP_INDIRECT ? |
402b27f9 | 1434 | s->req.u.indirect.nr_segments : s->req.u.rw.nr_segments; |
6cc56833 JG |
1435 | |
1436 | /* The I/O request may be split in two. */ | |
1437 | if (unlikely(s->associated_id != NO_ASSOCIATED_ID)) { | |
1438 | struct blk_shadow *s2 = &rinfo->shadow[s->associated_id]; | |
1439 | ||
1440 | /* Keep the status of the current response in shadow. */ | |
1441 | s->status = blkif_rsp_to_req_status(bret->status); | |
1442 | ||
1443 | /* Wait the second response if not yet here. */ | |
1444 | if (s2->status == REQ_WAITING) | |
1445 | return 0; | |
1446 | ||
1447 | bret->status = blkif_get_final_status(s->status, | |
1448 | s2->status); | |
1449 | ||
1450 | /* | |
1451 | * All the grants is stored in the first shadow in order | |
1452 | * to make the completion code simpler. | |
1453 | */ | |
1454 | num_grant += s2->req.u.rw.nr_segments; | |
1455 | ||
1456 | /* | |
1457 | * The two responses may not come in order. Only the | |
1458 | * first request will store the scatter-gather list. | |
1459 | */ | |
1460 | if (s2->num_sg != 0) { | |
1461 | /* Update "id" with the ID of the first response. */ | |
1462 | *id = s->associated_id; | |
1463 | s = s2; | |
1464 | } | |
1465 | ||
1466 | /* | |
1467 | * We don't need anymore the second request, so recycling | |
1468 | * it now. | |
1469 | */ | |
1470 | if (add_id_to_freelist(rinfo, s->associated_id)) | |
1471 | WARN(1, "%s: can't recycle the second part (id = %ld) of the request\n", | |
1472 | info->gd->disk_name, s->associated_id); | |
1473 | } | |
1474 | ||
1475 | data.s = s; | |
c004a6fe | 1476 | num_sg = s->num_sg; |
0a8704a5 | 1477 | |
bfe11d6d | 1478 | if (bret->operation == BLKIF_OP_READ && info->feature_persistent) { |
c004a6fe | 1479 | for_each_sg(s->sg, sg, num_sg, i) { |
b7649158 | 1480 | BUG_ON(sg->offset + sg->length > PAGE_SIZE); |
c004a6fe JG |
1481 | |
1482 | data.bvec_offset = sg->offset; | |
1483 | data.bvec_data = kmap_atomic(sg_page(sg)); | |
1484 | ||
1485 | gnttab_foreach_grant_in_range(sg_page(sg), | |
1486 | sg->offset, | |
1487 | sg->length, | |
1488 | blkif_copy_from_grant, | |
1489 | &data); | |
1490 | ||
1491 | kunmap_atomic(data.bvec_data); | |
0a8704a5 RPM |
1492 | } |
1493 | } | |
1494 | /* Add the persistent grant into the list of free grants */ | |
c004a6fe | 1495 | for (i = 0; i < num_grant; i++) { |
fbe363c4 RPM |
1496 | if (gnttab_query_foreign_access(s->grants_used[i]->gref)) { |
1497 | /* | |
1498 | * If the grant is still mapped by the backend (the | |
1499 | * backend has chosen to make this grant persistent) | |
1500 | * we add it at the head of the list, so it will be | |
1501 | * reused first. | |
1502 | */ | |
bfe11d6d RPM |
1503 | if (!info->feature_persistent) |
1504 | pr_alert_ratelimited("backed has not unmapped grant: %u\n", | |
1505 | s->grants_used[i]->gref); | |
73716df7 BL |
1506 | list_add(&s->grants_used[i]->node, &rinfo->grants); |
1507 | rinfo->persistent_gnts_c++; | |
fbe363c4 RPM |
1508 | } else { |
1509 | /* | |
1510 | * If the grant is not mapped by the backend we end the | |
1511 | * foreign access and add it to the tail of the list, | |
1512 | * so it will not be picked again unless we run out of | |
1513 | * persistent grants. | |
1514 | */ | |
1515 | gnttab_end_foreign_access(s->grants_used[i]->gref, 0, 0UL); | |
1516 | s->grants_used[i]->gref = GRANT_INVALID_REF; | |
73716df7 | 1517 | list_add_tail(&s->grants_used[i]->node, &rinfo->grants); |
fbe363c4 | 1518 | } |
0a8704a5 | 1519 | } |
402b27f9 | 1520 | if (s->req.operation == BLKIF_OP_INDIRECT) { |
c004a6fe | 1521 | for (i = 0; i < INDIRECT_GREFS(num_grant); i++) { |
fbe363c4 | 1522 | if (gnttab_query_foreign_access(s->indirect_grants[i]->gref)) { |
bfe11d6d RPM |
1523 | if (!info->feature_persistent) |
1524 | pr_alert_ratelimited("backed has not unmapped grant: %u\n", | |
1525 | s->indirect_grants[i]->gref); | |
73716df7 BL |
1526 | list_add(&s->indirect_grants[i]->node, &rinfo->grants); |
1527 | rinfo->persistent_gnts_c++; | |
fbe363c4 | 1528 | } else { |
bfe11d6d RPM |
1529 | struct page *indirect_page; |
1530 | ||
fbe363c4 | 1531 | gnttab_end_foreign_access(s->indirect_grants[i]->gref, 0, 0UL); |
bfe11d6d RPM |
1532 | /* |
1533 | * Add the used indirect page back to the list of | |
1534 | * available pages for indirect grefs. | |
1535 | */ | |
7b076750 | 1536 | if (!info->feature_persistent) { |
a7a6df22 | 1537 | indirect_page = s->indirect_grants[i]->page; |
81f35161 | 1538 | list_add(&indirect_page->lru, &rinfo->indirect_pages); |
7b076750 | 1539 | } |
fbe363c4 | 1540 | s->indirect_grants[i]->gref = GRANT_INVALID_REF; |
73716df7 | 1541 | list_add_tail(&s->indirect_grants[i]->node, &rinfo->grants); |
fbe363c4 | 1542 | } |
402b27f9 RPM |
1543 | } |
1544 | } | |
6cc56833 JG |
1545 | |
1546 | return 1; | |
9f27ee59 JF |
1547 | } |
1548 | ||
1549 | static irqreturn_t blkif_interrupt(int irq, void *dev_id) | |
1550 | { | |
1551 | struct request *req; | |
1552 | struct blkif_response *bret; | |
1553 | RING_IDX i, rp; | |
1554 | unsigned long flags; | |
81f35161 BL |
1555 | struct blkfront_ring_info *rinfo = (struct blkfront_ring_info *)dev_id; |
1556 | struct blkfront_info *info = rinfo->dev_info; | |
9f27ee59 | 1557 | |
11659569 | 1558 | if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) |
9f27ee59 | 1559 | return IRQ_HANDLED; |
9f27ee59 | 1560 | |
11659569 | 1561 | spin_lock_irqsave(&rinfo->ring_lock, flags); |
9f27ee59 | 1562 | again: |
81f35161 | 1563 | rp = rinfo->ring.sring->rsp_prod; |
9f27ee59 JF |
1564 | rmb(); /* Ensure we see queued responses up to 'rp'. */ |
1565 | ||
81f35161 | 1566 | for (i = rinfo->ring.rsp_cons; i != rp; i++) { |
9f27ee59 | 1567 | unsigned long id; |
9f27ee59 | 1568 | |
81f35161 | 1569 | bret = RING_GET_RESPONSE(&rinfo->ring, i); |
9f27ee59 | 1570 | id = bret->id; |
6878c32e KRW |
1571 | /* |
1572 | * The backend has messed up and given us an id that we would | |
1573 | * never have given to it (we stamp it up to BLK_RING_SIZE - | |
1574 | * look in get_id_from_freelist. | |
1575 | */ | |
86839c56 | 1576 | if (id >= BLK_RING_SIZE(info)) { |
6878c32e KRW |
1577 | WARN(1, "%s: response to %s has incorrect id (%ld)\n", |
1578 | info->gd->disk_name, op_name(bret->operation), id); | |
1579 | /* We can't safely get the 'struct request' as | |
1580 | * the id is busted. */ | |
1581 | continue; | |
1582 | } | |
81f35161 | 1583 | req = rinfo->shadow[id].request; |
9f27ee59 | 1584 | |
6cc56833 JG |
1585 | if (bret->operation != BLKIF_OP_DISCARD) { |
1586 | /* | |
1587 | * We may need to wait for an extra response if the | |
1588 | * I/O request is split in 2 | |
1589 | */ | |
1590 | if (!blkif_completion(&id, rinfo, bret)) | |
1591 | continue; | |
1592 | } | |
9f27ee59 | 1593 | |
81f35161 | 1594 | if (add_id_to_freelist(rinfo, id)) { |
6878c32e KRW |
1595 | WARN(1, "%s: response to %s (id %ld) couldn't be recycled!\n", |
1596 | info->gd->disk_name, op_name(bret->operation), id); | |
1597 | continue; | |
1598 | } | |
9f27ee59 | 1599 | |
2a842aca CH |
1600 | if (bret->status == BLKIF_RSP_OKAY) |
1601 | blkif_req(req)->error = BLK_STS_OK; | |
1602 | else | |
1603 | blkif_req(req)->error = BLK_STS_IOERR; | |
1604 | ||
9f27ee59 | 1605 | switch (bret->operation) { |
ed30bf31 LD |
1606 | case BLKIF_OP_DISCARD: |
1607 | if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) { | |
1608 | struct request_queue *rq = info->rq; | |
6878c32e KRW |
1609 | printk(KERN_WARNING "blkfront: %s: %s op failed\n", |
1610 | info->gd->disk_name, op_name(bret->operation)); | |
2a842aca | 1611 | blkif_req(req)->error = BLK_STS_NOTSUPP; |
ed30bf31 | 1612 | info->feature_discard = 0; |
5ea42986 | 1613 | info->feature_secdiscard = 0; |
ed30bf31 | 1614 | queue_flag_clear(QUEUE_FLAG_DISCARD, rq); |
288dab8a | 1615 | queue_flag_clear(QUEUE_FLAG_SECERASE, rq); |
ed30bf31 | 1616 | } |
ed30bf31 | 1617 | break; |
edf6ef59 | 1618 | case BLKIF_OP_FLUSH_DISKCACHE: |
9f27ee59 JF |
1619 | case BLKIF_OP_WRITE_BARRIER: |
1620 | if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) { | |
6878c32e KRW |
1621 | printk(KERN_WARNING "blkfront: %s: %s op failed\n", |
1622 | info->gd->disk_name, op_name(bret->operation)); | |
31c4ccc3 | 1623 | blkif_req(req)->error = BLK_STS_NOTSUPP; |
dcb8baec JF |
1624 | } |
1625 | if (unlikely(bret->status == BLKIF_RSP_ERROR && | |
81f35161 | 1626 | rinfo->shadow[id].req.u.rw.nr_segments == 0)) { |
6878c32e KRW |
1627 | printk(KERN_WARNING "blkfront: %s: empty %s op failed\n", |
1628 | info->gd->disk_name, op_name(bret->operation)); | |
2a842aca | 1629 | blkif_req(req)->error = BLK_STS_NOTSUPP; |
dcb8baec | 1630 | } |
2609587c | 1631 | if (unlikely(blkif_req(req)->error)) { |
2a842aca CH |
1632 | if (blkif_req(req)->error == BLK_STS_NOTSUPP) |
1633 | blkif_req(req)->error = BLK_STS_OK; | |
a418090a | 1634 | info->feature_fua = 0; |
4913efe4 TH |
1635 | info->feature_flush = 0; |
1636 | xlvbd_flush(info); | |
9f27ee59 JF |
1637 | } |
1638 | /* fall through */ | |
1639 | case BLKIF_OP_READ: | |
1640 | case BLKIF_OP_WRITE: | |
1641 | if (unlikely(bret->status != BLKIF_RSP_OKAY)) | |
1642 | dev_dbg(&info->xbdev->dev, "Bad return from blkdev data " | |
1643 | "request: %x\n", bret->status); | |
1644 | ||
9f27ee59 JF |
1645 | break; |
1646 | default: | |
1647 | BUG(); | |
1648 | } | |
2609587c | 1649 | |
08e0029a | 1650 | blk_mq_complete_request(req); |
9f27ee59 JF |
1651 | } |
1652 | ||
81f35161 | 1653 | rinfo->ring.rsp_cons = i; |
9f27ee59 | 1654 | |
81f35161 | 1655 | if (i != rinfo->ring.req_prod_pvt) { |
9f27ee59 | 1656 | int more_to_do; |
81f35161 | 1657 | RING_FINAL_CHECK_FOR_RESPONSES(&rinfo->ring, more_to_do); |
9f27ee59 JF |
1658 | if (more_to_do) |
1659 | goto again; | |
1660 | } else | |
81f35161 | 1661 | rinfo->ring.sring->rsp_event = i + 1; |
9f27ee59 | 1662 | |
11659569 | 1663 | kick_pending_request_queues_locked(rinfo); |
9f27ee59 | 1664 | |
11659569 | 1665 | spin_unlock_irqrestore(&rinfo->ring_lock, flags); |
9f27ee59 JF |
1666 | |
1667 | return IRQ_HANDLED; | |
1668 | } | |
1669 | ||
1670 | ||
1671 | static int setup_blkring(struct xenbus_device *dev, | |
81f35161 | 1672 | struct blkfront_ring_info *rinfo) |
9f27ee59 JF |
1673 | { |
1674 | struct blkif_sring *sring; | |
86839c56 | 1675 | int err, i; |
81f35161 | 1676 | struct blkfront_info *info = rinfo->dev_info; |
c004a6fe | 1677 | unsigned long ring_size = info->nr_ring_pages * XEN_PAGE_SIZE; |
9cce2914 | 1678 | grant_ref_t gref[XENBUS_MAX_RING_GRANTS]; |
9f27ee59 | 1679 | |
86839c56 | 1680 | for (i = 0; i < info->nr_ring_pages; i++) |
81f35161 | 1681 | rinfo->ring_ref[i] = GRANT_INVALID_REF; |
9f27ee59 | 1682 | |
86839c56 BL |
1683 | sring = (struct blkif_sring *)__get_free_pages(GFP_NOIO | __GFP_HIGH, |
1684 | get_order(ring_size)); | |
9f27ee59 JF |
1685 | if (!sring) { |
1686 | xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring"); | |
1687 | return -ENOMEM; | |
1688 | } | |
1689 | SHARED_RING_INIT(sring); | |
81f35161 | 1690 | FRONT_RING_INIT(&rinfo->ring, sring, ring_size); |
9e973e64 | 1691 | |
81f35161 | 1692 | err = xenbus_grant_ring(dev, rinfo->ring.sring, info->nr_ring_pages, gref); |
9f27ee59 | 1693 | if (err < 0) { |
86839c56 | 1694 | free_pages((unsigned long)sring, get_order(ring_size)); |
81f35161 | 1695 | rinfo->ring.sring = NULL; |
9f27ee59 JF |
1696 | goto fail; |
1697 | } | |
86839c56 | 1698 | for (i = 0; i < info->nr_ring_pages; i++) |
81f35161 | 1699 | rinfo->ring_ref[i] = gref[i]; |
9f27ee59 | 1700 | |
81f35161 | 1701 | err = xenbus_alloc_evtchn(dev, &rinfo->evtchn); |
9f27ee59 JF |
1702 | if (err) |
1703 | goto fail; | |
1704 | ||
81f35161 BL |
1705 | err = bind_evtchn_to_irqhandler(rinfo->evtchn, blkif_interrupt, 0, |
1706 | "blkif", rinfo); | |
9f27ee59 JF |
1707 | if (err <= 0) { |
1708 | xenbus_dev_fatal(dev, err, | |
1709 | "bind_evtchn_to_irqhandler failed"); | |
1710 | goto fail; | |
1711 | } | |
81f35161 | 1712 | rinfo->irq = err; |
9f27ee59 JF |
1713 | |
1714 | return 0; | |
1715 | fail: | |
1716 | blkif_free(info, 0); | |
1717 | return err; | |
1718 | } | |
1719 | ||
28d949bc BL |
1720 | /* |
1721 | * Write out per-ring/queue nodes including ring-ref and event-channel, and each | |
1722 | * ring buffer may have multi pages depending on ->nr_ring_pages. | |
1723 | */ | |
1724 | static int write_per_ring_nodes(struct xenbus_transaction xbt, | |
1725 | struct blkfront_ring_info *rinfo, const char *dir) | |
1726 | { | |
1727 | int err; | |
1728 | unsigned int i; | |
1729 | const char *message = NULL; | |
1730 | struct blkfront_info *info = rinfo->dev_info; | |
1731 | ||
1732 | if (info->nr_ring_pages == 1) { | |
1733 | err = xenbus_printf(xbt, dir, "ring-ref", "%u", rinfo->ring_ref[0]); | |
1734 | if (err) { | |
1735 | message = "writing ring-ref"; | |
1736 | goto abort_transaction; | |
1737 | } | |
1738 | } else { | |
1739 | for (i = 0; i < info->nr_ring_pages; i++) { | |
1740 | char ring_ref_name[RINGREF_NAME_LEN]; | |
1741 | ||
1742 | snprintf(ring_ref_name, RINGREF_NAME_LEN, "ring-ref%u", i); | |
1743 | err = xenbus_printf(xbt, dir, ring_ref_name, | |
1744 | "%u", rinfo->ring_ref[i]); | |
1745 | if (err) { | |
1746 | message = "writing ring-ref"; | |
1747 | goto abort_transaction; | |
1748 | } | |
1749 | } | |
1750 | } | |
1751 | ||
1752 | err = xenbus_printf(xbt, dir, "event-channel", "%u", rinfo->evtchn); | |
1753 | if (err) { | |
1754 | message = "writing event-channel"; | |
1755 | goto abort_transaction; | |
1756 | } | |
1757 | ||
1758 | return 0; | |
1759 | ||
1760 | abort_transaction: | |
1761 | xenbus_transaction_end(xbt, 1); | |
1762 | if (message) | |
1763 | xenbus_dev_fatal(info->xbdev, err, "%s", message); | |
1764 | ||
1765 | return err; | |
1766 | } | |
9f27ee59 JF |
1767 | |
1768 | /* Common code used when first setting up, and when resuming. */ | |
203fd61f | 1769 | static int talk_to_blkback(struct xenbus_device *dev, |
9f27ee59 JF |
1770 | struct blkfront_info *info) |
1771 | { | |
1772 | const char *message = NULL; | |
1773 | struct xenbus_transaction xbt; | |
28d949bc | 1774 | int err; |
f27dc1ac JG |
1775 | unsigned int i, max_page_order; |
1776 | unsigned int ring_page_order; | |
86839c56 | 1777 | |
7b2b3579 BD |
1778 | if (!info) |
1779 | return -ENODEV; | |
1780 | ||
f27dc1ac JG |
1781 | max_page_order = xenbus_read_unsigned(info->xbdev->otherend, |
1782 | "max-ring-page-order", 0); | |
1783 | ring_page_order = min(xen_blkif_max_ring_order, max_page_order); | |
1784 | info->nr_ring_pages = 1 << ring_page_order; | |
9f27ee59 | 1785 | |
7b2b3579 BD |
1786 | err = negotiate_mq(info); |
1787 | if (err) | |
1788 | goto destroy_blkring; | |
1789 | ||
3df0e505 | 1790 | for (i = 0; i < info->nr_rings; i++) { |
28d949bc BL |
1791 | struct blkfront_ring_info *rinfo = &info->rinfo[i]; |
1792 | ||
3df0e505 BL |
1793 | /* Create shared ring, alloc event channel. */ |
1794 | err = setup_blkring(dev, rinfo); | |
1795 | if (err) | |
1796 | goto destroy_blkring; | |
1797 | } | |
9f27ee59 JF |
1798 | |
1799 | again: | |
1800 | err = xenbus_transaction_start(&xbt); | |
1801 | if (err) { | |
1802 | xenbus_dev_fatal(dev, err, "starting transaction"); | |
1803 | goto destroy_blkring; | |
1804 | } | |
1805 | ||
28d949bc BL |
1806 | if (info->nr_ring_pages > 1) { |
1807 | err = xenbus_printf(xbt, dev->nodename, "ring-page-order", "%u", | |
1808 | ring_page_order); | |
1809 | if (err) { | |
1810 | message = "writing ring-page-order"; | |
1811 | goto abort_transaction; | |
1812 | } | |
1813 | } | |
3df0e505 | 1814 | |
28d949bc BL |
1815 | /* We already got the number of queues/rings in _probe */ |
1816 | if (info->nr_rings == 1) { | |
1817 | err = write_per_ring_nodes(xbt, &info->rinfo[0], dev->nodename); | |
1818 | if (err) | |
1819 | goto destroy_blkring; | |
1820 | } else { | |
1821 | char *path; | |
1822 | size_t pathsize; | |
3df0e505 | 1823 | |
28d949bc BL |
1824 | err = xenbus_printf(xbt, dev->nodename, "multi-queue-num-queues", "%u", |
1825 | info->nr_rings); | |
3df0e505 | 1826 | if (err) { |
28d949bc | 1827 | message = "writing multi-queue-num-queues"; |
3df0e505 BL |
1828 | goto abort_transaction; |
1829 | } | |
28d949bc BL |
1830 | |
1831 | pathsize = strlen(dev->nodename) + QUEUE_NAME_LEN; | |
1832 | path = kmalloc(pathsize, GFP_KERNEL); | |
1833 | if (!path) { | |
1834 | err = -ENOMEM; | |
1835 | message = "ENOMEM while writing ring references"; | |
1836 | goto abort_transaction; | |
1837 | } | |
1838 | ||
1839 | for (i = 0; i < info->nr_rings; i++) { | |
1840 | memset(path, 0, pathsize); | |
1841 | snprintf(path, pathsize, "%s/queue-%u", dev->nodename, i); | |
1842 | err = write_per_ring_nodes(xbt, &info->rinfo[i], path); | |
1843 | if (err) { | |
1844 | kfree(path); | |
1845 | goto destroy_blkring; | |
1846 | } | |
1847 | } | |
1848 | kfree(path); | |
9f27ee59 | 1849 | } |
3e334239 MA |
1850 | err = xenbus_printf(xbt, dev->nodename, "protocol", "%s", |
1851 | XEN_IO_PROTO_ABI_NATIVE); | |
1852 | if (err) { | |
1853 | message = "writing protocol"; | |
1854 | goto abort_transaction; | |
1855 | } | |
0a8704a5 | 1856 | err = xenbus_printf(xbt, dev->nodename, |
cb5bd4d1 | 1857 | "feature-persistent", "%u", 1); |
0a8704a5 RPM |
1858 | if (err) |
1859 | dev_warn(&dev->dev, | |
1860 | "writing persistent grants feature to xenbus"); | |
9f27ee59 JF |
1861 | |
1862 | err = xenbus_transaction_end(xbt, 0); | |
1863 | if (err) { | |
1864 | if (err == -EAGAIN) | |
1865 | goto again; | |
1866 | xenbus_dev_fatal(dev, err, "completing transaction"); | |
1867 | goto destroy_blkring; | |
1868 | } | |
1869 | ||
3df0e505 BL |
1870 | for (i = 0; i < info->nr_rings; i++) { |
1871 | unsigned int j; | |
28d949bc | 1872 | struct blkfront_ring_info *rinfo = &info->rinfo[i]; |
3df0e505 BL |
1873 | |
1874 | for (j = 0; j < BLK_RING_SIZE(info); j++) | |
1875 | rinfo->shadow[j].req.u.rw.id = j + 1; | |
1876 | rinfo->shadow[BLK_RING_SIZE(info)-1].req.u.rw.id = 0x0fffffff; | |
1877 | } | |
9f27ee59 JF |
1878 | xenbus_switch_state(dev, XenbusStateInitialised); |
1879 | ||
1880 | return 0; | |
1881 | ||
1882 | abort_transaction: | |
1883 | xenbus_transaction_end(xbt, 1); | |
1884 | if (message) | |
1885 | xenbus_dev_fatal(dev, err, "%s", message); | |
1886 | destroy_blkring: | |
1887 | blkif_free(info, 0); | |
3df0e505 | 1888 | |
c31ecf6c KRW |
1889 | kfree(info); |
1890 | dev_set_drvdata(&dev->dev, NULL); | |
1891 | ||
9f27ee59 JF |
1892 | return err; |
1893 | } | |
1894 | ||
3db70a85 BL |
1895 | static int negotiate_mq(struct blkfront_info *info) |
1896 | { | |
f27dc1ac | 1897 | unsigned int backend_max_queues; |
3db70a85 BL |
1898 | unsigned int i; |
1899 | ||
1900 | BUG_ON(info->nr_rings); | |
1901 | ||
1902 | /* Check if backend supports multiple queues. */ | |
f27dc1ac JG |
1903 | backend_max_queues = xenbus_read_unsigned(info->xbdev->otherend, |
1904 | "multi-queue-max-queues", 1); | |
3db70a85 BL |
1905 | info->nr_rings = min(backend_max_queues, xen_blkif_max_queues); |
1906 | /* We need at least one ring. */ | |
1907 | if (!info->nr_rings) | |
1908 | info->nr_rings = 1; | |
1909 | ||
1910 | info->rinfo = kzalloc(sizeof(struct blkfront_ring_info) * info->nr_rings, GFP_KERNEL); | |
1911 | if (!info->rinfo) { | |
1912 | xenbus_dev_fatal(info->xbdev, -ENOMEM, "allocating ring_info structure"); | |
814a3ad2 | 1913 | info->nr_rings = 0; |
3db70a85 BL |
1914 | return -ENOMEM; |
1915 | } | |
1916 | ||
1917 | for (i = 0; i < info->nr_rings; i++) { | |
1918 | struct blkfront_ring_info *rinfo; | |
1919 | ||
1920 | rinfo = &info->rinfo[i]; | |
1921 | INIT_LIST_HEAD(&rinfo->indirect_pages); | |
1922 | INIT_LIST_HEAD(&rinfo->grants); | |
1923 | rinfo->dev_info = info; | |
1924 | INIT_WORK(&rinfo->work, blkif_restart_queue); | |
1925 | spin_lock_init(&rinfo->ring_lock); | |
1926 | } | |
1927 | return 0; | |
1928 | } | |
9f27ee59 JF |
1929 | /** |
1930 | * Entry point to this code when a new device is created. Allocate the basic | |
1931 | * structures and the ring buffer for communication with the backend, and | |
1932 | * inform the backend of the appropriate details for those. Switch to | |
1933 | * Initialised state. | |
1934 | */ | |
1935 | static int blkfront_probe(struct xenbus_device *dev, | |
1936 | const struct xenbus_device_id *id) | |
1937 | { | |
86839c56 | 1938 | int err, vdevice; |
9f27ee59 JF |
1939 | struct blkfront_info *info; |
1940 | ||
1941 | /* FIXME: Use dynamic device id if this is not set. */ | |
1942 | err = xenbus_scanf(XBT_NIL, dev->nodename, | |
1943 | "virtual-device", "%i", &vdevice); | |
1944 | if (err != 1) { | |
9246b5f0 CL |
1945 | /* go looking in the extended area instead */ |
1946 | err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext", | |
1947 | "%i", &vdevice); | |
1948 | if (err != 1) { | |
1949 | xenbus_dev_fatal(dev, err, "reading virtual-device"); | |
1950 | return err; | |
1951 | } | |
9f27ee59 JF |
1952 | } |
1953 | ||
b98a409b SS |
1954 | if (xen_hvm_domain()) { |
1955 | char *type; | |
1956 | int len; | |
1957 | /* no unplug has been done: do not hook devices != xen vbds */ | |
51c71a3b | 1958 | if (xen_has_pv_and_legacy_disk_devices()) { |
b98a409b SS |
1959 | int major; |
1960 | ||
1961 | if (!VDEV_IS_EXTENDED(vdevice)) | |
1962 | major = BLKIF_MAJOR(vdevice); | |
1963 | else | |
1964 | major = XENVBD_MAJOR; | |
1965 | ||
1966 | if (major != XENVBD_MAJOR) { | |
1967 | printk(KERN_INFO | |
1968 | "%s: HVM does not support vbd %d as xen block device\n", | |
02f1f217 | 1969 | __func__, vdevice); |
b98a409b SS |
1970 | return -ENODEV; |
1971 | } | |
1972 | } | |
1973 | /* do not create a PV cdrom device if we are an HVM guest */ | |
1974 | type = xenbus_read(XBT_NIL, dev->nodename, "device-type", &len); | |
1975 | if (IS_ERR(type)) | |
1976 | return -ENODEV; | |
1977 | if (strncmp(type, "cdrom", 5) == 0) { | |
1978 | kfree(type); | |
c1c5413a SS |
1979 | return -ENODEV; |
1980 | } | |
b98a409b | 1981 | kfree(type); |
c1c5413a | 1982 | } |
9f27ee59 JF |
1983 | info = kzalloc(sizeof(*info), GFP_KERNEL); |
1984 | if (!info) { | |
1985 | xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure"); | |
1986 | return -ENOMEM; | |
1987 | } | |
1988 | ||
28d949bc | 1989 | info->xbdev = dev; |
81f35161 | 1990 | |
b70f5fa0 | 1991 | mutex_init(&info->mutex); |
9f27ee59 JF |
1992 | info->vdevice = vdevice; |
1993 | info->connected = BLKIF_STATE_DISCONNECTED; | |
9f27ee59 | 1994 | |
9f27ee59 JF |
1995 | /* Front end dir is a number, which is used as the id. */ |
1996 | info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0); | |
a1b4b12b | 1997 | dev_set_drvdata(&dev->dev, info); |
9f27ee59 | 1998 | |
9f27ee59 JF |
1999 | return 0; |
2000 | } | |
2001 | ||
9f27ee59 JF |
2002 | static int blkif_recover(struct blkfront_info *info) |
2003 | { | |
4559fa55 | 2004 | unsigned int r_index; |
402b27f9 | 2005 | struct request *req, *n; |
402b27f9 | 2006 | int rc; |
4559fa55 N |
2007 | struct bio *bio; |
2008 | unsigned int segs; | |
402b27f9 | 2009 | |
3df0e505 | 2010 | blkfront_gather_backend_features(info); |
172335ad BL |
2011 | /* Reset limits changed by blk_mq_update_nr_hw_queues(). */ |
2012 | blkif_set_queue_limits(info); | |
402b27f9 | 2013 | segs = info->max_indirect_segments ? : BLKIF_MAX_SEGMENTS_PER_REQUEST; |
6c647b0e | 2014 | blk_queue_max_segments(info->rq, segs / GRANTS_PER_PSEG); |
9f27ee59 | 2015 | |
3df0e505 | 2016 | for (r_index = 0; r_index < info->nr_rings; r_index++) { |
7b427a59 | 2017 | struct blkfront_ring_info *rinfo = &info->rinfo[r_index]; |
3df0e505 BL |
2018 | |
2019 | rc = blkfront_setup_indirect(rinfo); | |
7b427a59 | 2020 | if (rc) |
3df0e505 | 2021 | return rc; |
3df0e505 | 2022 | } |
9f27ee59 JF |
2023 | xenbus_switch_state(info->xbdev, XenbusStateConnected); |
2024 | ||
9f27ee59 JF |
2025 | /* Now safe for us to use the shared ring */ |
2026 | info->connected = BLKIF_STATE_CONNECTED; | |
2027 | ||
3df0e505 BL |
2028 | for (r_index = 0; r_index < info->nr_rings; r_index++) { |
2029 | struct blkfront_ring_info *rinfo; | |
2030 | ||
2031 | rinfo = &info->rinfo[r_index]; | |
2032 | /* Kick any other new requests queued since we resumed */ | |
2033 | kick_pending_request_queues(rinfo); | |
2034 | } | |
9f27ee59 | 2035 | |
7b427a59 | 2036 | list_for_each_entry_safe(req, n, &info->requests, queuelist) { |
402b27f9 RPM |
2037 | /* Requeue pending requests (flush or discard) */ |
2038 | list_del_init(&req->queuelist); | |
2039 | BUG_ON(req->nr_phys_segments > segs); | |
2b053aca | 2040 | blk_mq_requeue_request(req, false); |
402b27f9 | 2041 | } |
52d7f1b5 | 2042 | blk_mq_start_stopped_hw_queues(info->rq, true); |
907c3eb1 | 2043 | blk_mq_kick_requeue_list(info->rq); |
9f27ee59 | 2044 | |
7b427a59 | 2045 | while ((bio = bio_list_pop(&info->bio_list)) != NULL) { |
402b27f9 | 2046 | /* Traverse the list of pending bios and re-queue them */ |
4e49ea4a | 2047 | submit_bio(bio); |
402b27f9 RPM |
2048 | } |
2049 | ||
9f27ee59 JF |
2050 | return 0; |
2051 | } | |
2052 | ||
2053 | /** | |
2054 | * We are reconnecting to the backend, due to a suspend/resume, or a backend | |
2055 | * driver restart. We tear down our blkif structure and recreate it, but | |
2056 | * leave the device-layer structures intact so that this is transparent to the | |
2057 | * rest of the kernel. | |
2058 | */ | |
2059 | static int blkfront_resume(struct xenbus_device *dev) | |
2060 | { | |
a1b4b12b | 2061 | struct blkfront_info *info = dev_get_drvdata(&dev->dev); |
3db70a85 | 2062 | int err = 0; |
7b427a59 | 2063 | unsigned int i, j; |
9f27ee59 JF |
2064 | |
2065 | dev_dbg(&dev->dev, "blkfront_resume: %s\n", dev->nodename); | |
2066 | ||
7b427a59 BL |
2067 | bio_list_init(&info->bio_list); |
2068 | INIT_LIST_HEAD(&info->requests); | |
2069 | for (i = 0; i < info->nr_rings; i++) { | |
2070 | struct blkfront_ring_info *rinfo = &info->rinfo[i]; | |
2071 | struct bio_list merge_bio; | |
2072 | struct blk_shadow *shadow = rinfo->shadow; | |
2073 | ||
2074 | for (j = 0; j < BLK_RING_SIZE(info); j++) { | |
2075 | /* Not in use? */ | |
2076 | if (!shadow[j].request) | |
2077 | continue; | |
2078 | ||
2079 | /* | |
2080 | * Get the bios in the request so we can re-queue them. | |
2081 | */ | |
b15bd8cb MK |
2082 | if (req_op(shadow[j].request) == REQ_OP_FLUSH || |
2083 | req_op(shadow[j].request) == REQ_OP_DISCARD || | |
2084 | req_op(shadow[j].request) == REQ_OP_SECURE_ERASE || | |
3fc9d690 | 2085 | shadow[j].request->cmd_flags & REQ_FUA) { |
7b427a59 BL |
2086 | /* |
2087 | * Flush operations don't contain bios, so | |
2088 | * we need to requeue the whole request | |
3fc9d690 LT |
2089 | * |
2090 | * XXX: but this doesn't make any sense for a | |
2091 | * write with the FUA flag set.. | |
7b427a59 BL |
2092 | */ |
2093 | list_add(&shadow[j].request->queuelist, &info->requests); | |
2094 | continue; | |
2095 | } | |
2096 | merge_bio.head = shadow[j].request->bio; | |
2097 | merge_bio.tail = shadow[j].request->biotail; | |
2098 | bio_list_merge(&info->bio_list, &merge_bio); | |
2099 | shadow[j].request->bio = NULL; | |
2a842aca | 2100 | blk_mq_end_request(shadow[j].request, BLK_STS_OK); |
7b427a59 BL |
2101 | } |
2102 | } | |
2103 | ||
9f27ee59 JF |
2104 | blkif_free(info, info->connected == BLKIF_STATE_CONNECTED); |
2105 | ||
203fd61f | 2106 | err = talk_to_blkback(dev, info); |
2a6f71ad BL |
2107 | if (!err) |
2108 | blk_mq_update_nr_hw_queues(&info->tag_set, info->nr_rings); | |
402b27f9 RPM |
2109 | |
2110 | /* | |
2111 | * We have to wait for the backend to switch to | |
2112 | * connected state, since we want to read which | |
2113 | * features it supports. | |
2114 | */ | |
9f27ee59 JF |
2115 | |
2116 | return err; | |
2117 | } | |
2118 | ||
6f03a7ff | 2119 | static void blkfront_closing(struct blkfront_info *info) |
b70f5fa0 DS |
2120 | { |
2121 | struct xenbus_device *xbdev = info->xbdev; | |
2122 | struct block_device *bdev = NULL; | |
2123 | ||
2124 | mutex_lock(&info->mutex); | |
2125 | ||
2126 | if (xbdev->state == XenbusStateClosing) { | |
2127 | mutex_unlock(&info->mutex); | |
2128 | return; | |
2129 | } | |
2130 | ||
2131 | if (info->gd) | |
2132 | bdev = bdget_disk(info->gd, 0); | |
2133 | ||
2134 | mutex_unlock(&info->mutex); | |
2135 | ||
2136 | if (!bdev) { | |
2137 | xenbus_frontend_closed(xbdev); | |
2138 | return; | |
2139 | } | |
2140 | ||
2141 | mutex_lock(&bdev->bd_mutex); | |
2142 | ||
7b32d104 | 2143 | if (bdev->bd_openers) { |
b70f5fa0 DS |
2144 | xenbus_dev_error(xbdev, -EBUSY, |
2145 | "Device in use; refusing to close"); | |
2146 | xenbus_switch_state(xbdev, XenbusStateClosing); | |
2147 | } else { | |
2148 | xlvbd_release_gendisk(info); | |
2149 | xenbus_frontend_closed(xbdev); | |
2150 | } | |
2151 | ||
2152 | mutex_unlock(&bdev->bd_mutex); | |
2153 | bdput(bdev); | |
2154 | } | |
9f27ee59 | 2155 | |
ed30bf31 LD |
2156 | static void blkfront_setup_discard(struct blkfront_info *info) |
2157 | { | |
2158 | int err; | |
ed30bf31 LD |
2159 | unsigned int discard_granularity; |
2160 | unsigned int discard_alignment; | |
2161 | ||
1c8cad6c OH |
2162 | info->feature_discard = 1; |
2163 | err = xenbus_gather(XBT_NIL, info->xbdev->otherend, | |
2164 | "discard-granularity", "%u", &discard_granularity, | |
2165 | "discard-alignment", "%u", &discard_alignment, | |
2166 | NULL); | |
2167 | if (!err) { | |
2168 | info->discard_granularity = discard_granularity; | |
2169 | info->discard_alignment = discard_alignment; | |
2170 | } | |
f27dc1ac JG |
2171 | info->feature_secdiscard = |
2172 | !!xenbus_read_unsigned(info->xbdev->otherend, "discard-secure", | |
2173 | 0); | |
ed30bf31 LD |
2174 | } |
2175 | ||
81f35161 | 2176 | static int blkfront_setup_indirect(struct blkfront_ring_info *rinfo) |
402b27f9 | 2177 | { |
c004a6fe | 2178 | unsigned int psegs, grants; |
402b27f9 | 2179 | int err, i; |
81f35161 | 2180 | struct blkfront_info *info = rinfo->dev_info; |
402b27f9 | 2181 | |
6cc56833 JG |
2182 | if (info->max_indirect_segments == 0) { |
2183 | if (!HAS_EXTRA_REQ) | |
2184 | grants = BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
2185 | else { | |
2186 | /* | |
2187 | * When an extra req is required, the maximum | |
2188 | * grants supported is related to the size of the | |
2189 | * Linux block segment. | |
2190 | */ | |
2191 | grants = GRANTS_PER_PSEG; | |
2192 | } | |
2193 | } | |
d50babbe | 2194 | else |
c004a6fe | 2195 | grants = info->max_indirect_segments; |
3b4f1884 | 2196 | psegs = DIV_ROUND_UP(grants, GRANTS_PER_PSEG); |
402b27f9 | 2197 | |
81f35161 | 2198 | err = fill_grant_buffer(rinfo, |
c004a6fe | 2199 | (grants + INDIRECT_GREFS(grants)) * BLK_RING_SIZE(info)); |
402b27f9 RPM |
2200 | if (err) |
2201 | goto out_of_memory; | |
2202 | ||
bfe11d6d RPM |
2203 | if (!info->feature_persistent && info->max_indirect_segments) { |
2204 | /* | |
2205 | * We are using indirect descriptors but not persistent | |
2206 | * grants, we need to allocate a set of pages that can be | |
2207 | * used for mapping indirect grefs | |
2208 | */ | |
c004a6fe | 2209 | int num = INDIRECT_GREFS(grants) * BLK_RING_SIZE(info); |
bfe11d6d | 2210 | |
81f35161 | 2211 | BUG_ON(!list_empty(&rinfo->indirect_pages)); |
bfe11d6d RPM |
2212 | for (i = 0; i < num; i++) { |
2213 | struct page *indirect_page = alloc_page(GFP_NOIO); | |
2214 | if (!indirect_page) | |
2215 | goto out_of_memory; | |
81f35161 | 2216 | list_add(&indirect_page->lru, &rinfo->indirect_pages); |
bfe11d6d RPM |
2217 | } |
2218 | } | |
2219 | ||
86839c56 | 2220 | for (i = 0; i < BLK_RING_SIZE(info); i++) { |
81f35161 BL |
2221 | rinfo->shadow[i].grants_used = kzalloc( |
2222 | sizeof(rinfo->shadow[i].grants_used[0]) * grants, | |
402b27f9 | 2223 | GFP_NOIO); |
81f35161 | 2224 | rinfo->shadow[i].sg = kzalloc(sizeof(rinfo->shadow[i].sg[0]) * psegs, GFP_NOIO); |
402b27f9 | 2225 | if (info->max_indirect_segments) |
81f35161 BL |
2226 | rinfo->shadow[i].indirect_grants = kzalloc( |
2227 | sizeof(rinfo->shadow[i].indirect_grants[0]) * | |
c004a6fe | 2228 | INDIRECT_GREFS(grants), |
402b27f9 | 2229 | GFP_NOIO); |
81f35161 BL |
2230 | if ((rinfo->shadow[i].grants_used == NULL) || |
2231 | (rinfo->shadow[i].sg == NULL) || | |
402b27f9 | 2232 | (info->max_indirect_segments && |
81f35161 | 2233 | (rinfo->shadow[i].indirect_grants == NULL))) |
402b27f9 | 2234 | goto out_of_memory; |
81f35161 | 2235 | sg_init_table(rinfo->shadow[i].sg, psegs); |
402b27f9 RPM |
2236 | } |
2237 | ||
2238 | ||
2239 | return 0; | |
2240 | ||
2241 | out_of_memory: | |
86839c56 | 2242 | for (i = 0; i < BLK_RING_SIZE(info); i++) { |
81f35161 BL |
2243 | kfree(rinfo->shadow[i].grants_used); |
2244 | rinfo->shadow[i].grants_used = NULL; | |
2245 | kfree(rinfo->shadow[i].sg); | |
2246 | rinfo->shadow[i].sg = NULL; | |
2247 | kfree(rinfo->shadow[i].indirect_grants); | |
2248 | rinfo->shadow[i].indirect_grants = NULL; | |
402b27f9 | 2249 | } |
81f35161 | 2250 | if (!list_empty(&rinfo->indirect_pages)) { |
bfe11d6d | 2251 | struct page *indirect_page, *n; |
81f35161 | 2252 | list_for_each_entry_safe(indirect_page, n, &rinfo->indirect_pages, lru) { |
bfe11d6d RPM |
2253 | list_del(&indirect_page->lru); |
2254 | __free_page(indirect_page); | |
2255 | } | |
2256 | } | |
402b27f9 RPM |
2257 | return -ENOMEM; |
2258 | } | |
2259 | ||
d50babbe BL |
2260 | /* |
2261 | * Gather all backend feature-* | |
2262 | */ | |
3df0e505 | 2263 | static void blkfront_gather_backend_features(struct blkfront_info *info) |
d50babbe | 2264 | { |
d50babbe BL |
2265 | unsigned int indirect_segments; |
2266 | ||
2267 | info->feature_flush = 0; | |
a418090a | 2268 | info->feature_fua = 0; |
d50babbe | 2269 | |
d50babbe BL |
2270 | /* |
2271 | * If there's no "feature-barrier" defined, then it means | |
2272 | * we're dealing with a very old backend which writes | |
2273 | * synchronously; nothing to do. | |
2274 | * | |
2275 | * If there are barriers, then we use flush. | |
2276 | */ | |
f27dc1ac | 2277 | if (xenbus_read_unsigned(info->xbdev->otherend, "feature-barrier", 0)) { |
a418090a MC |
2278 | info->feature_flush = 1; |
2279 | info->feature_fua = 1; | |
2280 | } | |
2281 | ||
d50babbe BL |
2282 | /* |
2283 | * And if there is "feature-flush-cache" use that above | |
2284 | * barriers. | |
2285 | */ | |
f27dc1ac JG |
2286 | if (xenbus_read_unsigned(info->xbdev->otherend, "feature-flush-cache", |
2287 | 0)) { | |
a418090a MC |
2288 | info->feature_flush = 1; |
2289 | info->feature_fua = 0; | |
2290 | } | |
d50babbe | 2291 | |
f27dc1ac | 2292 | if (xenbus_read_unsigned(info->xbdev->otherend, "feature-discard", 0)) |
d50babbe BL |
2293 | blkfront_setup_discard(info); |
2294 | ||
f27dc1ac | 2295 | info->feature_persistent = |
b32728ff JB |
2296 | !!xenbus_read_unsigned(info->xbdev->otherend, |
2297 | "feature-persistent", 0); | |
d50babbe | 2298 | |
f27dc1ac JG |
2299 | indirect_segments = xenbus_read_unsigned(info->xbdev->otherend, |
2300 | "feature-max-indirect-segments", 0); | |
3b4f1884 JB |
2301 | if (indirect_segments > xen_blkif_max_segments) |
2302 | indirect_segments = xen_blkif_max_segments; | |
2303 | if (indirect_segments <= BLKIF_MAX_SEGMENTS_PER_REQUEST) | |
2304 | indirect_segments = 0; | |
2305 | info->max_indirect_segments = indirect_segments; | |
d50babbe BL |
2306 | } |
2307 | ||
9f27ee59 JF |
2308 | /* |
2309 | * Invoked when the backend is finally 'ready' (and has told produced | |
2310 | * the details about the physical device - #sectors, size, etc). | |
2311 | */ | |
2312 | static void blkfront_connect(struct blkfront_info *info) | |
2313 | { | |
2314 | unsigned long long sectors; | |
2315 | unsigned long sector_size; | |
7c4d7d71 | 2316 | unsigned int physical_sector_size; |
9f27ee59 | 2317 | unsigned int binfo; |
89515d02 | 2318 | char *envp[] = { "RESIZE=1", NULL }; |
3df0e505 | 2319 | int err, i; |
9f27ee59 | 2320 | |
1fa73be6 S |
2321 | switch (info->connected) { |
2322 | case BLKIF_STATE_CONNECTED: | |
2323 | /* | |
2324 | * Potentially, the back-end may be signalling | |
2325 | * a capacity change; update the capacity. | |
2326 | */ | |
2327 | err = xenbus_scanf(XBT_NIL, info->xbdev->otherend, | |
2328 | "sectors", "%Lu", §ors); | |
2329 | if (XENBUS_EXIST_ERR(err)) | |
2330 | return; | |
2331 | printk(KERN_INFO "Setting capacity to %Lu\n", | |
2332 | sectors); | |
2333 | set_capacity(info->gd, sectors); | |
2def141e | 2334 | revalidate_disk(info->gd); |
89515d02 MO |
2335 | kobject_uevent_env(&disk_to_dev(info->gd)->kobj, |
2336 | KOBJ_CHANGE, envp); | |
1fa73be6 | 2337 | |
402b27f9 | 2338 | return; |
1fa73be6 | 2339 | case BLKIF_STATE_SUSPENDED: |
402b27f9 RPM |
2340 | /* |
2341 | * If we are recovering from suspension, we need to wait | |
2342 | * for the backend to announce it's features before | |
2343 | * reconnecting, at least we need to know if the backend | |
2344 | * supports indirect descriptors, and how many. | |
2345 | */ | |
2346 | blkif_recover(info); | |
9f27ee59 JF |
2347 | return; |
2348 | ||
b4dddb49 JF |
2349 | default: |
2350 | break; | |
1fa73be6 | 2351 | } |
9f27ee59 JF |
2352 | |
2353 | dev_dbg(&info->xbdev->dev, "%s:%s.\n", | |
2354 | __func__, info->xbdev->otherend); | |
2355 | ||
2356 | err = xenbus_gather(XBT_NIL, info->xbdev->otherend, | |
2357 | "sectors", "%llu", §ors, | |
2358 | "info", "%u", &binfo, | |
2359 | "sector-size", "%lu", §or_size, | |
2360 | NULL); | |
2361 | if (err) { | |
2362 | xenbus_dev_fatal(info->xbdev, err, | |
2363 | "reading backend fields at %s", | |
2364 | info->xbdev->otherend); | |
2365 | return; | |
2366 | } | |
2367 | ||
7c4d7d71 SB |
2368 | /* |
2369 | * physcial-sector-size is a newer field, so old backends may not | |
2370 | * provide this. Assume physical sector size to be the same as | |
2371 | * sector_size in that case. | |
2372 | */ | |
f27dc1ac JG |
2373 | physical_sector_size = xenbus_read_unsigned(info->xbdev->otherend, |
2374 | "physical-sector-size", | |
2375 | sector_size); | |
3df0e505 BL |
2376 | blkfront_gather_backend_features(info); |
2377 | for (i = 0; i < info->nr_rings; i++) { | |
2378 | err = blkfront_setup_indirect(&info->rinfo[i]); | |
2379 | if (err) { | |
2380 | xenbus_dev_fatal(info->xbdev, err, "setup_indirect at %s", | |
2381 | info->xbdev->otherend); | |
2382 | blkif_free(info, 0); | |
2383 | break; | |
2384 | } | |
402b27f9 RPM |
2385 | } |
2386 | ||
7c4d7d71 SB |
2387 | err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size, |
2388 | physical_sector_size); | |
9f27ee59 JF |
2389 | if (err) { |
2390 | xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s", | |
2391 | info->xbdev->otherend); | |
4e876c2b | 2392 | goto fail; |
9f27ee59 JF |
2393 | } |
2394 | ||
2395 | xenbus_switch_state(info->xbdev, XenbusStateConnected); | |
2396 | ||
2397 | /* Kick pending requests. */ | |
9f27ee59 | 2398 | info->connected = BLKIF_STATE_CONNECTED; |
3df0e505 BL |
2399 | for (i = 0; i < info->nr_rings; i++) |
2400 | kick_pending_request_queues(&info->rinfo[i]); | |
9f27ee59 | 2401 | |
0d52c756 | 2402 | device_add_disk(&info->xbdev->dev, info->gd); |
1d78d705 CL |
2403 | |
2404 | info->is_ready = 1; | |
4e876c2b BL |
2405 | return; |
2406 | ||
2407 | fail: | |
2408 | blkif_free(info, 0); | |
2409 | return; | |
9f27ee59 JF |
2410 | } |
2411 | ||
9f27ee59 JF |
2412 | /** |
2413 | * Callback received when the backend's state changes. | |
2414 | */ | |
203fd61f | 2415 | static void blkback_changed(struct xenbus_device *dev, |
9f27ee59 JF |
2416 | enum xenbus_state backend_state) |
2417 | { | |
a1b4b12b | 2418 | struct blkfront_info *info = dev_get_drvdata(&dev->dev); |
9f27ee59 | 2419 | |
203fd61f | 2420 | dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state); |
9f27ee59 JF |
2421 | |
2422 | switch (backend_state) { | |
9f27ee59 | 2423 | case XenbusStateInitWait: |
a9b54bb9 BL |
2424 | if (dev->state != XenbusStateInitialising) |
2425 | break; | |
c31ecf6c | 2426 | if (talk_to_blkback(dev, info)) |
8ab0144a | 2427 | break; |
8ab0144a | 2428 | case XenbusStateInitialising: |
9f27ee59 | 2429 | case XenbusStateInitialised: |
b78c9512 NI |
2430 | case XenbusStateReconfiguring: |
2431 | case XenbusStateReconfigured: | |
9f27ee59 | 2432 | case XenbusStateUnknown: |
9f27ee59 JF |
2433 | break; |
2434 | ||
2435 | case XenbusStateConnected: | |
efd15352 BL |
2436 | /* |
2437 | * talk_to_blkback sets state to XenbusStateInitialised | |
2438 | * and blkfront_connect sets it to XenbusStateConnected | |
2439 | * (if connection went OK). | |
2440 | * | |
2441 | * If the backend (or toolstack) decides to poke at backend | |
2442 | * state (and re-trigger the watch by setting the state repeatedly | |
2443 | * to XenbusStateConnected (4)) we need to deal with this. | |
2444 | * This is allowed as this is used to communicate to the guest | |
2445 | * that the size of disk has changed! | |
2446 | */ | |
2447 | if ((dev->state != XenbusStateInitialised) && | |
2448 | (dev->state != XenbusStateConnected)) { | |
c31ecf6c KRW |
2449 | if (talk_to_blkback(dev, info)) |
2450 | break; | |
2451 | } | |
efd15352 | 2452 | |
9f27ee59 JF |
2453 | blkfront_connect(info); |
2454 | break; | |
2455 | ||
36613717 DV |
2456 | case XenbusStateClosed: |
2457 | if (dev->state == XenbusStateClosed) | |
2458 | break; | |
ccc22257 | 2459 | /* fall through */ |
9f27ee59 | 2460 | case XenbusStateClosing: |
a54c8f0f CA |
2461 | if (info) |
2462 | blkfront_closing(info); | |
9f27ee59 JF |
2463 | break; |
2464 | } | |
2465 | } | |
2466 | ||
fa1bd359 | 2467 | static int blkfront_remove(struct xenbus_device *xbdev) |
9f27ee59 | 2468 | { |
fa1bd359 DS |
2469 | struct blkfront_info *info = dev_get_drvdata(&xbdev->dev); |
2470 | struct block_device *bdev = NULL; | |
2471 | struct gendisk *disk; | |
9f27ee59 | 2472 | |
fa1bd359 | 2473 | dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename); |
9f27ee59 | 2474 | |
282e6259 VL |
2475 | if (!info) |
2476 | return 0; | |
2477 | ||
9f27ee59 JF |
2478 | blkif_free(info, 0); |
2479 | ||
fa1bd359 DS |
2480 | mutex_lock(&info->mutex); |
2481 | ||
2482 | disk = info->gd; | |
2483 | if (disk) | |
2484 | bdev = bdget_disk(disk, 0); | |
2485 | ||
2486 | info->xbdev = NULL; | |
2487 | mutex_unlock(&info->mutex); | |
2488 | ||
2489 | if (!bdev) { | |
2490 | kfree(info); | |
2491 | return 0; | |
2492 | } | |
2493 | ||
2494 | /* | |
2495 | * The xbdev was removed before we reached the Closed | |
2496 | * state. See if it's safe to remove the disk. If the bdev | |
2497 | * isn't closed yet, we let release take care of it. | |
2498 | */ | |
2499 | ||
2500 | mutex_lock(&bdev->bd_mutex); | |
2501 | info = disk->private_data; | |
2502 | ||
d54142c7 DS |
2503 | dev_warn(disk_to_dev(disk), |
2504 | "%s was hot-unplugged, %d stale handles\n", | |
2505 | xbdev->nodename, bdev->bd_openers); | |
2506 | ||
7b32d104 | 2507 | if (info && !bdev->bd_openers) { |
fa1bd359 DS |
2508 | xlvbd_release_gendisk(info); |
2509 | disk->private_data = NULL; | |
0e345826 | 2510 | kfree(info); |
fa1bd359 DS |
2511 | } |
2512 | ||
2513 | mutex_unlock(&bdev->bd_mutex); | |
2514 | bdput(bdev); | |
9f27ee59 JF |
2515 | |
2516 | return 0; | |
2517 | } | |
2518 | ||
1d78d705 CL |
2519 | static int blkfront_is_ready(struct xenbus_device *dev) |
2520 | { | |
a1b4b12b | 2521 | struct blkfront_info *info = dev_get_drvdata(&dev->dev); |
1d78d705 | 2522 | |
5d7ed20e | 2523 | return info->is_ready && info->xbdev; |
1d78d705 CL |
2524 | } |
2525 | ||
a63c848b | 2526 | static int blkif_open(struct block_device *bdev, fmode_t mode) |
9f27ee59 | 2527 | { |
13961743 DS |
2528 | struct gendisk *disk = bdev->bd_disk; |
2529 | struct blkfront_info *info; | |
2530 | int err = 0; | |
6e9624b8 | 2531 | |
2a48fc0a | 2532 | mutex_lock(&blkfront_mutex); |
6e9624b8 | 2533 | |
13961743 DS |
2534 | info = disk->private_data; |
2535 | if (!info) { | |
2536 | /* xbdev gone */ | |
2537 | err = -ERESTARTSYS; | |
2538 | goto out; | |
2539 | } | |
2540 | ||
2541 | mutex_lock(&info->mutex); | |
2542 | ||
2543 | if (!info->gd) | |
2544 | /* xbdev is closed */ | |
2545 | err = -ERESTARTSYS; | |
2546 | ||
2547 | mutex_unlock(&info->mutex); | |
2548 | ||
13961743 | 2549 | out: |
2a48fc0a | 2550 | mutex_unlock(&blkfront_mutex); |
13961743 | 2551 | return err; |
9f27ee59 JF |
2552 | } |
2553 | ||
db2a144b | 2554 | static void blkif_release(struct gendisk *disk, fmode_t mode) |
9f27ee59 | 2555 | { |
a63c848b | 2556 | struct blkfront_info *info = disk->private_data; |
7fd152f4 DS |
2557 | struct block_device *bdev; |
2558 | struct xenbus_device *xbdev; | |
2559 | ||
2a48fc0a | 2560 | mutex_lock(&blkfront_mutex); |
7fd152f4 DS |
2561 | |
2562 | bdev = bdget_disk(disk, 0); | |
7fd152f4 | 2563 | |
2f089cb8 FP |
2564 | if (!bdev) { |
2565 | WARN(1, "Block device %s yanked out from us!\n", disk->disk_name); | |
2566 | goto out_mutex; | |
2567 | } | |
acfca3c6 DS |
2568 | if (bdev->bd_openers) |
2569 | goto out; | |
2570 | ||
7fd152f4 DS |
2571 | /* |
2572 | * Check if we have been instructed to close. We will have | |
2573 | * deferred this request, because the bdev was still open. | |
2574 | */ | |
2575 | ||
2576 | mutex_lock(&info->mutex); | |
2577 | xbdev = info->xbdev; | |
2578 | ||
2579 | if (xbdev && xbdev->state == XenbusStateClosing) { | |
2580 | /* pending switch to state closed */ | |
d54142c7 | 2581 | dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n"); |
7fd152f4 DS |
2582 | xlvbd_release_gendisk(info); |
2583 | xenbus_frontend_closed(info->xbdev); | |
2584 | } | |
2585 | ||
2586 | mutex_unlock(&info->mutex); | |
2587 | ||
2588 | if (!xbdev) { | |
2589 | /* sudden device removal */ | |
d54142c7 | 2590 | dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n"); |
7fd152f4 DS |
2591 | xlvbd_release_gendisk(info); |
2592 | disk->private_data = NULL; | |
2593 | kfree(info); | |
9f27ee59 | 2594 | } |
7fd152f4 | 2595 | |
a4cc14ec | 2596 | out: |
dad5cf65 | 2597 | bdput(bdev); |
2f089cb8 | 2598 | out_mutex: |
2a48fc0a | 2599 | mutex_unlock(&blkfront_mutex); |
9f27ee59 JF |
2600 | } |
2601 | ||
83d5cde4 | 2602 | static const struct block_device_operations xlvbd_block_fops = |
9f27ee59 JF |
2603 | { |
2604 | .owner = THIS_MODULE, | |
a63c848b AV |
2605 | .open = blkif_open, |
2606 | .release = blkif_release, | |
597592d9 | 2607 | .getgeo = blkif_getgeo, |
8a6cfeb6 | 2608 | .ioctl = blkif_ioctl, |
9f27ee59 JF |
2609 | }; |
2610 | ||
2611 | ||
ec9c42ec | 2612 | static const struct xenbus_device_id blkfront_ids[] = { |
9f27ee59 JF |
2613 | { "vbd" }, |
2614 | { "" } | |
2615 | }; | |
2616 | ||
95afae48 DV |
2617 | static struct xenbus_driver blkfront_driver = { |
2618 | .ids = blkfront_ids, | |
9f27ee59 JF |
2619 | .probe = blkfront_probe, |
2620 | .remove = blkfront_remove, | |
2621 | .resume = blkfront_resume, | |
203fd61f | 2622 | .otherend_changed = blkback_changed, |
1d78d705 | 2623 | .is_ready = blkfront_is_ready, |
95afae48 | 2624 | }; |
9f27ee59 JF |
2625 | |
2626 | static int __init xlblk_init(void) | |
2627 | { | |
469738e6 | 2628 | int ret; |
28d949bc | 2629 | int nr_cpus = num_online_cpus(); |
469738e6 | 2630 | |
6e833587 | 2631 | if (!xen_domain()) |
9f27ee59 JF |
2632 | return -ENODEV; |
2633 | ||
3b4f1884 JB |
2634 | if (xen_blkif_max_segments < BLKIF_MAX_SEGMENTS_PER_REQUEST) |
2635 | xen_blkif_max_segments = BLKIF_MAX_SEGMENTS_PER_REQUEST; | |
2636 | ||
9cce2914 | 2637 | if (xen_blkif_max_ring_order > XENBUS_MAX_RING_GRANT_ORDER) { |
86839c56 | 2638 | pr_info("Invalid max_ring_order (%d), will use default max: %d.\n", |
9cce2914 | 2639 | xen_blkif_max_ring_order, XENBUS_MAX_RING_GRANT_ORDER); |
45fc8264 | 2640 | xen_blkif_max_ring_order = XENBUS_MAX_RING_GRANT_ORDER; |
86839c56 BL |
2641 | } |
2642 | ||
28d949bc BL |
2643 | if (xen_blkif_max_queues > nr_cpus) { |
2644 | pr_info("Invalid max_queues (%d), will use default max: %d.\n", | |
2645 | xen_blkif_max_queues, nr_cpus); | |
2646 | xen_blkif_max_queues = nr_cpus; | |
2647 | } | |
2648 | ||
51c71a3b | 2649 | if (!xen_has_pv_disk_devices()) |
b9136d20 IM |
2650 | return -ENODEV; |
2651 | ||
9f27ee59 JF |
2652 | if (register_blkdev(XENVBD_MAJOR, DEV_NAME)) { |
2653 | printk(KERN_WARNING "xen_blk: can't get major %d with name %s\n", | |
2654 | XENVBD_MAJOR, DEV_NAME); | |
2655 | return -ENODEV; | |
2656 | } | |
2657 | ||
73db144b | 2658 | ret = xenbus_register_frontend(&blkfront_driver); |
469738e6 LE |
2659 | if (ret) { |
2660 | unregister_blkdev(XENVBD_MAJOR, DEV_NAME); | |
2661 | return ret; | |
2662 | } | |
2663 | ||
2664 | return 0; | |
9f27ee59 JF |
2665 | } |
2666 | module_init(xlblk_init); | |
2667 | ||
2668 | ||
5a60d0cd | 2669 | static void __exit xlblk_exit(void) |
9f27ee59 | 2670 | { |
8605067f JB |
2671 | xenbus_unregister_driver(&blkfront_driver); |
2672 | unregister_blkdev(XENVBD_MAJOR, DEV_NAME); | |
2673 | kfree(minors); | |
9f27ee59 JF |
2674 | } |
2675 | module_exit(xlblk_exit); | |
2676 | ||
2677 | MODULE_DESCRIPTION("Xen virtual block device frontend"); | |
2678 | MODULE_LICENSE("GPL"); | |
2679 | MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR); | |
d2f0c52b | 2680 | MODULE_ALIAS("xen:vbd"); |
4f93f09b | 2681 | MODULE_ALIAS("xenblk"); |