1 /* 2 * Xen event channels (FIFO-based ABI) 3 * 4 * Copyright (C) 2013 Citrix Systems R&D ltd. 5 * 6 * This source code is free software; you can redistribute it and/or 7 * modify it under the terms of the GNU General Public License as 8 * published by the Free Software Foundation; either version 2 of the 9 * License, or (at your option) any later version. 10 * 11 * Or, when distributed separately from the Linux kernel or 12 * incorporated into other software packages, subject to the following 13 * license: 14 * 15 * Permission is hereby granted, free of charge, to any person obtaining a copy 16 * of this source file (the "Software"), to deal in the Software without 17 * restriction, including without limitation the rights to use, copy, modify, 18 * merge, publish, distribute, sublicense, and/or sell copies of the Software, 19 * and to permit persons to whom the Software is furnished to do so, subject to 20 * the following conditions: 21 * 22 * The above copyright notice and this permission notice shall be included in 23 * all copies or substantial portions of the Software. 24 * 25 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 26 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 27 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 28 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 29 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 30 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS 31 * IN THE SOFTWARE. 32 */ 33 34 #define pr_fmt(fmt) "xen:" KBUILD_MODNAME ": " fmt 35 36 #include <linux/linkage.h> 37 #include <linux/interrupt.h> 38 #include <linux/irq.h> 39 #include <linux/smp.h> 40 #include <linux/percpu.h> 41 #include <linux/cpu.h> 42 43 #include <asm/barrier.h> 44 #include <asm/sync_bitops.h> 45 #include <asm/xen/hypercall.h> 46 #include <asm/xen/hypervisor.h> 47 48 #include <xen/xen.h> 49 #include <xen/xen-ops.h> 50 #include <xen/events.h> 51 #include <xen/interface/xen.h> 52 #include <xen/interface/event_channel.h> 53 #include <xen/page.h> 54 55 #include "events_internal.h" 56 57 #define EVENT_WORDS_PER_PAGE (XEN_PAGE_SIZE / sizeof(event_word_t)) 58 #define MAX_EVENT_ARRAY_PAGES (EVTCHN_FIFO_NR_CHANNELS / EVENT_WORDS_PER_PAGE) 59 60 struct evtchn_fifo_queue { 61 uint32_t head[EVTCHN_FIFO_MAX_QUEUES]; 62 }; 63 64 static DEFINE_PER_CPU(struct evtchn_fifo_control_block *, cpu_control_block); 65 static DEFINE_PER_CPU(struct evtchn_fifo_queue, cpu_queue); 66 static event_word_t *event_array[MAX_EVENT_ARRAY_PAGES] __read_mostly; 67 static unsigned event_array_pages __read_mostly; 68 69 /* 70 * sync_set_bit() and friends must be unsigned long aligned. 71 */ 72 #if BITS_PER_LONG > 32 73 74 #define BM(w) (unsigned long *)((unsigned long)w & ~0x7UL) 75 #define EVTCHN_FIFO_BIT(b, w) \ 76 (((unsigned long)w & 0x4UL) ? (EVTCHN_FIFO_ ##b + 32) : EVTCHN_FIFO_ ##b) 77 78 #else 79 80 #define BM(w) ((unsigned long *)(w)) 81 #define EVTCHN_FIFO_BIT(b, w) EVTCHN_FIFO_ ##b 82 83 #endif 84 85 static inline event_word_t *event_word_from_port(evtchn_port_t port) 86 { 87 unsigned i = port / EVENT_WORDS_PER_PAGE; 88 89 return event_array[i] + port % EVENT_WORDS_PER_PAGE; 90 } 91 92 static unsigned evtchn_fifo_max_channels(void) 93 { 94 return EVTCHN_FIFO_NR_CHANNELS; 95 } 96 97 static unsigned evtchn_fifo_nr_channels(void) 98 { 99 return event_array_pages * EVENT_WORDS_PER_PAGE; 100 } 101 102 static int init_control_block(int cpu, 103 struct evtchn_fifo_control_block *control_block) 104 { 105 struct evtchn_fifo_queue *q = &per_cpu(cpu_queue, cpu); 106 struct evtchn_init_control init_control; 107 unsigned int i; 108 109 /* Reset the control block and the local HEADs. */ 110 clear_page(control_block); 111 for (i = 0; i < EVTCHN_FIFO_MAX_QUEUES; i++) 112 q->head[i] = 0; 113 114 init_control.control_gfn = virt_to_gfn(control_block); 115 init_control.offset = 0; 116 init_control.vcpu = xen_vcpu_nr(cpu); 117 118 return HYPERVISOR_event_channel_op(EVTCHNOP_init_control, &init_control); 119 } 120 121 static void free_unused_array_pages(void) 122 { 123 unsigned i; 124 125 for (i = event_array_pages; i < MAX_EVENT_ARRAY_PAGES; i++) { 126 if (!event_array[i]) 127 break; 128 free_page((unsigned long)event_array[i]); 129 event_array[i] = NULL; 130 } 131 } 132 133 static void init_array_page(event_word_t *array_page) 134 { 135 unsigned i; 136 137 for (i = 0; i < EVENT_WORDS_PER_PAGE; i++) 138 array_page[i] = 1 << EVTCHN_FIFO_MASKED; 139 } 140 141 static int evtchn_fifo_setup(evtchn_port_t port) 142 { 143 unsigned new_array_pages; 144 int ret; 145 146 new_array_pages = port / EVENT_WORDS_PER_PAGE + 1; 147 148 if (new_array_pages > MAX_EVENT_ARRAY_PAGES) 149 return -EINVAL; 150 151 while (event_array_pages < new_array_pages) { 152 void *array_page; 153 struct evtchn_expand_array expand_array; 154 155 /* Might already have a page if we've resumed. */ 156 array_page = event_array[event_array_pages]; 157 if (!array_page) { 158 array_page = (void *)__get_free_page(GFP_KERNEL); 159 if (array_page == NULL) { 160 ret = -ENOMEM; 161 goto error; 162 } 163 event_array[event_array_pages] = array_page; 164 } 165 166 /* Mask all events in this page before adding it. */ 167 init_array_page(array_page); 168 169 expand_array.array_gfn = virt_to_gfn(array_page); 170 171 ret = HYPERVISOR_event_channel_op(EVTCHNOP_expand_array, &expand_array); 172 if (ret < 0) 173 goto error; 174 175 event_array_pages++; 176 } 177 return 0; 178 179 error: 180 if (event_array_pages == 0) 181 panic("xen: unable to expand event array with initial page (%d)\n", ret); 182 else 183 pr_err("unable to expand event array (%d)\n", ret); 184 free_unused_array_pages(); 185 return ret; 186 } 187 188 static void evtchn_fifo_bind_to_cpu(evtchn_port_t evtchn, unsigned int cpu, 189 unsigned int old_cpu) 190 { 191 /* no-op */ 192 } 193 194 static void evtchn_fifo_clear_pending(evtchn_port_t port) 195 { 196 event_word_t *word = event_word_from_port(port); 197 sync_clear_bit(EVTCHN_FIFO_BIT(PENDING, word), BM(word)); 198 } 199 200 static void evtchn_fifo_set_pending(evtchn_port_t port) 201 { 202 event_word_t *word = event_word_from_port(port); 203 sync_set_bit(EVTCHN_FIFO_BIT(PENDING, word), BM(word)); 204 } 205 206 static bool evtchn_fifo_is_pending(evtchn_port_t port) 207 { 208 event_word_t *word = event_word_from_port(port); 209 return sync_test_bit(EVTCHN_FIFO_BIT(PENDING, word), BM(word)); 210 } 211 212 static void evtchn_fifo_mask(evtchn_port_t port) 213 { 214 event_word_t *word = event_word_from_port(port); 215 sync_set_bit(EVTCHN_FIFO_BIT(MASKED, word), BM(word)); 216 } 217 218 static bool evtchn_fifo_is_masked(evtchn_port_t port) 219 { 220 event_word_t *word = event_word_from_port(port); 221 return sync_test_bit(EVTCHN_FIFO_BIT(MASKED, word), BM(word)); 222 } 223 /* 224 * Clear MASKED if not PENDING, spinning if BUSY is set. 225 * Return true if mask was cleared. 226 */ 227 static bool clear_masked_cond(volatile event_word_t *word) 228 { 229 event_word_t new, old, w; 230 231 w = *word; 232 233 do { 234 if (!(w & (1 << EVTCHN_FIFO_MASKED))) 235 return true; 236 237 if (w & (1 << EVTCHN_FIFO_PENDING)) 238 return false; 239 240 old = w & ~(1 << EVTCHN_FIFO_BUSY); 241 new = old & ~(1 << EVTCHN_FIFO_MASKED); 242 w = sync_cmpxchg(word, old, new); 243 } while (w != old); 244 245 return true; 246 } 247 248 static void evtchn_fifo_unmask(evtchn_port_t port) 249 { 250 event_word_t *word = event_word_from_port(port); 251 252 BUG_ON(!irqs_disabled()); 253 254 if (!clear_masked_cond(word)) { 255 struct evtchn_unmask unmask = { .port = port }; 256 (void)HYPERVISOR_event_channel_op(EVTCHNOP_unmask, &unmask); 257 } 258 } 259 260 static uint32_t clear_linked(volatile event_word_t *word) 261 { 262 event_word_t new, old, w; 263 264 w = *word; 265 266 do { 267 old = w; 268 new = (w & ~((1 << EVTCHN_FIFO_LINKED) 269 | EVTCHN_FIFO_LINK_MASK)); 270 } while ((w = sync_cmpxchg(word, old, new)) != old); 271 272 return w & EVTCHN_FIFO_LINK_MASK; 273 } 274 275 static void consume_one_event(unsigned cpu, struct evtchn_loop_ctrl *ctrl, 276 struct evtchn_fifo_control_block *control_block, 277 unsigned priority, unsigned long *ready) 278 { 279 struct evtchn_fifo_queue *q = &per_cpu(cpu_queue, cpu); 280 uint32_t head; 281 evtchn_port_t port; 282 event_word_t *word; 283 284 head = q->head[priority]; 285 286 /* 287 * Reached the tail last time? Read the new HEAD from the 288 * control block. 289 */ 290 if (head == 0) { 291 virt_rmb(); /* Ensure word is up-to-date before reading head. */ 292 head = control_block->head[priority]; 293 } 294 295 port = head; 296 word = event_word_from_port(port); 297 head = clear_linked(word); 298 299 /* 300 * If the link is non-zero, there are more events in the 301 * queue, otherwise the queue is empty. 302 * 303 * If the queue is empty, clear this priority from our local 304 * copy of the ready word. 305 */ 306 if (head == 0) 307 clear_bit(priority, ready); 308 309 if (evtchn_fifo_is_pending(port) && !evtchn_fifo_is_masked(port)) { 310 if (unlikely(!ctrl)) 311 pr_warn("Dropping pending event for port %u\n", port); 312 else 313 handle_irq_for_port(port, ctrl); 314 } 315 316 q->head[priority] = head; 317 } 318 319 static void __evtchn_fifo_handle_events(unsigned cpu, 320 struct evtchn_loop_ctrl *ctrl) 321 { 322 struct evtchn_fifo_control_block *control_block; 323 unsigned long ready; 324 unsigned q; 325 326 control_block = per_cpu(cpu_control_block, cpu); 327 328 ready = xchg(&control_block->ready, 0); 329 330 while (ready) { 331 q = find_first_bit(&ready, EVTCHN_FIFO_MAX_QUEUES); 332 consume_one_event(cpu, ctrl, control_block, q, &ready); 333 ready |= xchg(&control_block->ready, 0); 334 } 335 } 336 337 static void evtchn_fifo_handle_events(unsigned cpu, 338 struct evtchn_loop_ctrl *ctrl) 339 { 340 __evtchn_fifo_handle_events(cpu, ctrl); 341 } 342 343 static void evtchn_fifo_resume(void) 344 { 345 unsigned cpu; 346 347 for_each_possible_cpu(cpu) { 348 void *control_block = per_cpu(cpu_control_block, cpu); 349 int ret; 350 351 if (!control_block) 352 continue; 353 354 /* 355 * If this CPU is offline, take the opportunity to 356 * free the control block while it is not being 357 * used. 358 */ 359 if (!cpu_online(cpu)) { 360 free_page((unsigned long)control_block); 361 per_cpu(cpu_control_block, cpu) = NULL; 362 continue; 363 } 364 365 ret = init_control_block(cpu, control_block); 366 BUG_ON(ret < 0); 367 } 368 369 /* 370 * The event array starts out as empty again and is extended 371 * as normal when events are bound. The existing pages will 372 * be reused. 373 */ 374 event_array_pages = 0; 375 } 376 377 static int evtchn_fifo_alloc_control_block(unsigned cpu) 378 { 379 void *control_block = NULL; 380 int ret = -ENOMEM; 381 382 control_block = (void *)__get_free_page(GFP_KERNEL); 383 if (control_block == NULL) 384 goto error; 385 386 ret = init_control_block(cpu, control_block); 387 if (ret < 0) 388 goto error; 389 390 per_cpu(cpu_control_block, cpu) = control_block; 391 392 return 0; 393 394 error: 395 free_page((unsigned long)control_block); 396 return ret; 397 } 398 399 static int evtchn_fifo_percpu_init(unsigned int cpu) 400 { 401 if (!per_cpu(cpu_control_block, cpu)) 402 return evtchn_fifo_alloc_control_block(cpu); 403 return 0; 404 } 405 406 static int evtchn_fifo_percpu_deinit(unsigned int cpu) 407 { 408 __evtchn_fifo_handle_events(cpu, NULL); 409 return 0; 410 } 411 412 static const struct evtchn_ops evtchn_ops_fifo = { 413 .max_channels = evtchn_fifo_max_channels, 414 .nr_channels = evtchn_fifo_nr_channels, 415 .setup = evtchn_fifo_setup, 416 .bind_to_cpu = evtchn_fifo_bind_to_cpu, 417 .clear_pending = evtchn_fifo_clear_pending, 418 .set_pending = evtchn_fifo_set_pending, 419 .is_pending = evtchn_fifo_is_pending, 420 .mask = evtchn_fifo_mask, 421 .unmask = evtchn_fifo_unmask, 422 .handle_events = evtchn_fifo_handle_events, 423 .resume = evtchn_fifo_resume, 424 .percpu_init = evtchn_fifo_percpu_init, 425 .percpu_deinit = evtchn_fifo_percpu_deinit, 426 }; 427 428 int __init xen_evtchn_fifo_init(void) 429 { 430 int cpu = smp_processor_id(); 431 int ret; 432 433 ret = evtchn_fifo_alloc_control_block(cpu); 434 if (ret < 0) 435 return ret; 436 437 pr_info("Using FIFO-based ABI\n"); 438 439 evtchn_ops = &evtchn_ops_fifo; 440 441 return ret; 442 } 443