[34] | 1 | /****************************************************************************** |
---|
| 2 | * xen.h |
---|
| 3 | * |
---|
| 4 | * Guest OS interface to Xen. |
---|
| 5 | * |
---|
| 6 | * Permission is hereby granted, free of charge, to any person obtaining a copy |
---|
| 7 | * of this software and associated documentation files (the "Software"), to |
---|
| 8 | * deal in the Software without restriction, including without limitation the |
---|
| 9 | * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or |
---|
| 10 | * sell copies of the Software, and to permit persons to whom the Software is |
---|
| 11 | * furnished to do so, subject to the following conditions: |
---|
| 12 | * |
---|
| 13 | * The above copyright notice and this permission notice shall be included in |
---|
| 14 | * all copies or substantial portions of the Software. |
---|
| 15 | * |
---|
| 16 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
---|
| 17 | * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
---|
| 18 | * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE |
---|
| 19 | * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER |
---|
| 20 | * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING |
---|
| 21 | * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER |
---|
| 22 | * DEALINGS IN THE SOFTWARE. |
---|
| 23 | * |
---|
| 24 | * Copyright (c) 2004, K A Fraser |
---|
| 25 | */ |
---|
| 26 | |
---|
| 27 | #ifndef __XEN_PUBLIC_XEN_H__ |
---|
| 28 | #define __XEN_PUBLIC_XEN_H__ |
---|
| 29 | |
---|
| 30 | #include "xen-compat.h" |
---|
| 31 | |
---|
| 32 | #if defined(__i386__) || defined(__x86_64__) |
---|
| 33 | #include "arch-x86/xen.h" |
---|
| 34 | #elif defined(__ia64__) |
---|
| 35 | #include "arch-ia64.h" |
---|
| 36 | #elif defined(__powerpc__) |
---|
| 37 | #include "arch-powerpc.h" |
---|
| 38 | #else |
---|
| 39 | #error "Unsupported architecture" |
---|
| 40 | #endif |
---|
| 41 | |
---|
| 42 | /* |
---|
| 43 | * HYPERCALLS |
---|
| 44 | */ |
---|
| 45 | |
---|
| 46 | #define __HYPERVISOR_set_trap_table 0 |
---|
| 47 | #define __HYPERVISOR_mmu_update 1 |
---|
| 48 | #define __HYPERVISOR_set_gdt 2 |
---|
| 49 | #define __HYPERVISOR_stack_switch 3 |
---|
| 50 | #define __HYPERVISOR_set_callbacks 4 |
---|
| 51 | #define __HYPERVISOR_fpu_taskswitch 5 |
---|
| 52 | #define __HYPERVISOR_sched_op_compat 6 /* compat since 0x00030101 */ |
---|
| 53 | #define __HYPERVISOR_platform_op 7 |
---|
| 54 | #define __HYPERVISOR_set_debugreg 8 |
---|
| 55 | #define __HYPERVISOR_get_debugreg 9 |
---|
| 56 | #define __HYPERVISOR_update_descriptor 10 |
---|
| 57 | #define __HYPERVISOR_memory_op 12 |
---|
| 58 | #define __HYPERVISOR_multicall 13 |
---|
| 59 | #define __HYPERVISOR_update_va_mapping 14 |
---|
| 60 | #define __HYPERVISOR_set_timer_op 15 |
---|
| 61 | #define __HYPERVISOR_event_channel_op_compat 16 /* compat since 0x00030202 */ |
---|
| 62 | #define __HYPERVISOR_xen_version 17 |
---|
| 63 | #define __HYPERVISOR_console_io 18 |
---|
| 64 | #define __HYPERVISOR_physdev_op_compat 19 /* compat since 0x00030202 */ |
---|
| 65 | #define __HYPERVISOR_grant_table_op 20 |
---|
| 66 | #define __HYPERVISOR_vm_assist 21 |
---|
| 67 | #define __HYPERVISOR_update_va_mapping_otherdomain 22 |
---|
| 68 | #define __HYPERVISOR_iret 23 /* x86 only */ |
---|
| 69 | #define __HYPERVISOR_vcpu_op 24 |
---|
| 70 | #define __HYPERVISOR_set_segment_base 25 /* x86/64 only */ |
---|
| 71 | #define __HYPERVISOR_mmuext_op 26 |
---|
| 72 | #define __HYPERVISOR_acm_op 27 |
---|
| 73 | #define __HYPERVISOR_nmi_op 28 |
---|
| 74 | #define __HYPERVISOR_sched_op 29 |
---|
| 75 | #define __HYPERVISOR_callback_op 30 |
---|
| 76 | #define __HYPERVISOR_xenoprof_op 31 |
---|
| 77 | #define __HYPERVISOR_event_channel_op 32 |
---|
| 78 | #define __HYPERVISOR_physdev_op 33 |
---|
| 79 | #define __HYPERVISOR_hvm_op 34 |
---|
| 80 | #define __HYPERVISOR_sysctl 35 |
---|
| 81 | #define __HYPERVISOR_domctl 36 |
---|
| 82 | #define __HYPERVISOR_kexec_op 37 |
---|
| 83 | |
---|
| 84 | /* Architecture-specific hypercall definitions. */ |
---|
| 85 | #define __HYPERVISOR_arch_0 48 |
---|
| 86 | #define __HYPERVISOR_arch_1 49 |
---|
| 87 | #define __HYPERVISOR_arch_2 50 |
---|
| 88 | #define __HYPERVISOR_arch_3 51 |
---|
| 89 | #define __HYPERVISOR_arch_4 52 |
---|
| 90 | #define __HYPERVISOR_arch_5 53 |
---|
| 91 | #define __HYPERVISOR_arch_6 54 |
---|
| 92 | #define __HYPERVISOR_arch_7 55 |
---|
| 93 | |
---|
| 94 | /* |
---|
| 95 | * HYPERCALL COMPATIBILITY. |
---|
| 96 | */ |
---|
| 97 | |
---|
| 98 | /* New sched_op hypercall introduced in 0x00030101. */ |
---|
| 99 | #if __XEN_INTERFACE_VERSION__ < 0x00030101 |
---|
| 100 | #undef __HYPERVISOR_sched_op |
---|
| 101 | #define __HYPERVISOR_sched_op __HYPERVISOR_sched_op_compat |
---|
| 102 | #endif |
---|
| 103 | |
---|
| 104 | /* New event-channel and physdev hypercalls introduced in 0x00030202. */ |
---|
| 105 | #if __XEN_INTERFACE_VERSION__ < 0x00030202 |
---|
| 106 | #undef __HYPERVISOR_event_channel_op |
---|
| 107 | #define __HYPERVISOR_event_channel_op __HYPERVISOR_event_channel_op_compat |
---|
| 108 | #undef __HYPERVISOR_physdev_op |
---|
| 109 | #define __HYPERVISOR_physdev_op __HYPERVISOR_physdev_op_compat |
---|
| 110 | #endif |
---|
| 111 | |
---|
| 112 | /* New platform_op hypercall introduced in 0x00030204. */ |
---|
| 113 | #if __XEN_INTERFACE_VERSION__ < 0x00030204 |
---|
| 114 | #define __HYPERVISOR_dom0_op __HYPERVISOR_platform_op |
---|
| 115 | #endif |
---|
| 116 | |
---|
| 117 | /* |
---|
| 118 | * VIRTUAL INTERRUPTS |
---|
| 119 | * |
---|
| 120 | * Virtual interrupts that a guest OS may receive from Xen. |
---|
| 121 | * |
---|
| 122 | * In the side comments, 'V.' denotes a per-VCPU VIRQ while 'G.' denotes a |
---|
| 123 | * global VIRQ. The former can be bound once per VCPU and cannot be re-bound. |
---|
| 124 | * The latter can be allocated only once per guest: they must initially be |
---|
| 125 | * allocated to VCPU0 but can subsequently be re-bound. |
---|
| 126 | */ |
---|
| 127 | #define VIRQ_TIMER 0 /* V. Timebase update, and/or requested timeout. */ |
---|
| 128 | #define VIRQ_DEBUG 1 /* V. Request guest to dump debug info. */ |
---|
| 129 | #define VIRQ_CONSOLE 2 /* G. (DOM0) Bytes received on emergency console. */ |
---|
| 130 | #define VIRQ_DOM_EXC 3 /* G. (DOM0) Exceptional event for some domain. */ |
---|
| 131 | #define VIRQ_TBUF 4 /* G. (DOM0) Trace buffer has records available. */ |
---|
| 132 | #define VIRQ_DEBUGGER 6 /* G. (DOM0) A domain has paused for debugging. */ |
---|
| 133 | #define VIRQ_XENOPROF 7 /* V. XenOprofile interrupt: new sample available */ |
---|
| 134 | #define VIRQ_CON_RING 8 /* G. (DOM0) Bytes received on console */ |
---|
| 135 | |
---|
| 136 | /* Architecture-specific VIRQ definitions. */ |
---|
| 137 | #define VIRQ_ARCH_0 16 |
---|
| 138 | #define VIRQ_ARCH_1 17 |
---|
| 139 | #define VIRQ_ARCH_2 18 |
---|
| 140 | #define VIRQ_ARCH_3 19 |
---|
| 141 | #define VIRQ_ARCH_4 20 |
---|
| 142 | #define VIRQ_ARCH_5 21 |
---|
| 143 | #define VIRQ_ARCH_6 22 |
---|
| 144 | #define VIRQ_ARCH_7 23 |
---|
| 145 | |
---|
| 146 | #define NR_VIRQS 24 |
---|
| 147 | |
---|
| 148 | /* |
---|
| 149 | * MMU-UPDATE REQUESTS |
---|
| 150 | * |
---|
| 151 | * HYPERVISOR_mmu_update() accepts a list of (ptr, val) pairs. |
---|
| 152 | * A foreigndom (FD) can be specified (or DOMID_SELF for none). |
---|
| 153 | * Where the FD has some effect, it is described below. |
---|
| 154 | * ptr[1:0] specifies the appropriate MMU_* command. |
---|
| 155 | * |
---|
| 156 | * ptr[1:0] == MMU_NORMAL_PT_UPDATE: |
---|
| 157 | * Updates an entry in a page table. If updating an L1 table, and the new |
---|
| 158 | * table entry is valid/present, the mapped frame must belong to the FD, if |
---|
| 159 | * an FD has been specified. If attempting to map an I/O page then the |
---|
| 160 | * caller assumes the privilege of the FD. |
---|
| 161 | * FD == DOMID_IO: Permit /only/ I/O mappings, at the priv level of the caller. |
---|
| 162 | * FD == DOMID_XEN: Map restricted areas of Xen's heap space. |
---|
| 163 | * ptr[:2] -- Machine address of the page-table entry to modify. |
---|
| 164 | * val -- Value to write. |
---|
| 165 | * |
---|
| 166 | * ptr[1:0] == MMU_MACHPHYS_UPDATE: |
---|
| 167 | * Updates an entry in the machine->pseudo-physical mapping table. |
---|
| 168 | * ptr[:2] -- Machine address within the frame whose mapping to modify. |
---|
| 169 | * The frame must belong to the FD, if one is specified. |
---|
| 170 | * val -- Value to write into the mapping entry. |
---|
| 171 | */ |
---|
| 172 | #define MMU_NORMAL_PT_UPDATE 0 /* checked '*ptr = val'. ptr is MA. */ |
---|
| 173 | #define MMU_MACHPHYS_UPDATE 1 /* ptr = MA of frame to modify entry for */ |
---|
| 174 | |
---|
| 175 | /* |
---|
| 176 | * MMU EXTENDED OPERATIONS |
---|
| 177 | * |
---|
| 178 | * HYPERVISOR_mmuext_op() accepts a list of mmuext_op structures. |
---|
| 179 | * A foreigndom (FD) can be specified (or DOMID_SELF for none). |
---|
| 180 | * Where the FD has some effect, it is described below. |
---|
| 181 | * |
---|
| 182 | * cmd: MMUEXT_(UN)PIN_*_TABLE |
---|
| 183 | * mfn: Machine frame number to be (un)pinned as a p.t. page. |
---|
| 184 | * The frame must belong to the FD, if one is specified. |
---|
| 185 | * |
---|
| 186 | * cmd: MMUEXT_NEW_BASEPTR |
---|
| 187 | * mfn: Machine frame number of new page-table base to install in MMU. |
---|
| 188 | * |
---|
| 189 | * cmd: MMUEXT_NEW_USER_BASEPTR [x86/64 only] |
---|
| 190 | * mfn: Machine frame number of new page-table base to install in MMU |
---|
| 191 | * when in user space. |
---|
| 192 | * |
---|
| 193 | * cmd: MMUEXT_TLB_FLUSH_LOCAL |
---|
| 194 | * No additional arguments. Flushes local TLB. |
---|
| 195 | * |
---|
| 196 | * cmd: MMUEXT_INVLPG_LOCAL |
---|
| 197 | * linear_addr: Linear address to be flushed from the local TLB. |
---|
| 198 | * |
---|
| 199 | * cmd: MMUEXT_TLB_FLUSH_MULTI |
---|
| 200 | * vcpumask: Pointer to bitmap of VCPUs to be flushed. |
---|
| 201 | * |
---|
| 202 | * cmd: MMUEXT_INVLPG_MULTI |
---|
| 203 | * linear_addr: Linear address to be flushed. |
---|
| 204 | * vcpumask: Pointer to bitmap of VCPUs to be flushed. |
---|
| 205 | * |
---|
| 206 | * cmd: MMUEXT_TLB_FLUSH_ALL |
---|
| 207 | * No additional arguments. Flushes all VCPUs' TLBs. |
---|
| 208 | * |
---|
| 209 | * cmd: MMUEXT_INVLPG_ALL |
---|
| 210 | * linear_addr: Linear address to be flushed from all VCPUs' TLBs. |
---|
| 211 | * |
---|
| 212 | * cmd: MMUEXT_FLUSH_CACHE |
---|
| 213 | * No additional arguments. Writes back and flushes cache contents. |
---|
| 214 | * |
---|
| 215 | * cmd: MMUEXT_SET_LDT |
---|
| 216 | * linear_addr: Linear address of LDT base (NB. must be page-aligned). |
---|
| 217 | * nr_ents: Number of entries in LDT. |
---|
| 218 | */ |
---|
| 219 | #define MMUEXT_PIN_L1_TABLE 0 |
---|
| 220 | #define MMUEXT_PIN_L2_TABLE 1 |
---|
| 221 | #define MMUEXT_PIN_L3_TABLE 2 |
---|
| 222 | #define MMUEXT_PIN_L4_TABLE 3 |
---|
| 223 | #define MMUEXT_UNPIN_TABLE 4 |
---|
| 224 | #define MMUEXT_NEW_BASEPTR 5 |
---|
| 225 | #define MMUEXT_TLB_FLUSH_LOCAL 6 |
---|
| 226 | #define MMUEXT_INVLPG_LOCAL 7 |
---|
| 227 | #define MMUEXT_TLB_FLUSH_MULTI 8 |
---|
| 228 | #define MMUEXT_INVLPG_MULTI 9 |
---|
| 229 | #define MMUEXT_TLB_FLUSH_ALL 10 |
---|
| 230 | #define MMUEXT_INVLPG_ALL 11 |
---|
| 231 | #define MMUEXT_FLUSH_CACHE 12 |
---|
| 232 | #define MMUEXT_SET_LDT 13 |
---|
| 233 | #define MMUEXT_NEW_USER_BASEPTR 15 |
---|
| 234 | |
---|
| 235 | #ifndef __ASSEMBLY__ |
---|
| 236 | struct mmuext_op { |
---|
| 237 | unsigned int cmd; |
---|
| 238 | union { |
---|
| 239 | /* [UN]PIN_TABLE, NEW_BASEPTR, NEW_USER_BASEPTR */ |
---|
| 240 | xen_pfn_t mfn; |
---|
| 241 | /* INVLPG_LOCAL, INVLPG_ALL, SET_LDT */ |
---|
| 242 | unsigned long linear_addr; |
---|
| 243 | } arg1; |
---|
| 244 | union { |
---|
| 245 | /* SET_LDT */ |
---|
| 246 | unsigned int nr_ents; |
---|
| 247 | /* TLB_FLUSH_MULTI, INVLPG_MULTI */ |
---|
| 248 | XEN_GUEST_HANDLE_00030205(void) vcpumask; |
---|
| 249 | } arg2; |
---|
| 250 | }; |
---|
| 251 | typedef struct mmuext_op mmuext_op_t; |
---|
| 252 | DEFINE_XEN_GUEST_HANDLE(mmuext_op_t); |
---|
| 253 | #endif |
---|
| 254 | |
---|
| 255 | /* These are passed as 'flags' to update_va_mapping. They can be ORed. */ |
---|
| 256 | /* When specifying UVMF_MULTI, also OR in a pointer to a CPU bitmap. */ |
---|
| 257 | /* UVMF_LOCAL is merely UVMF_MULTI with a NULL bitmap pointer. */ |
---|
| 258 | #define UVMF_NONE (0UL<<0) /* No flushing at all. */ |
---|
| 259 | #define UVMF_TLB_FLUSH (1UL<<0) /* Flush entire TLB(s). */ |
---|
| 260 | #define UVMF_INVLPG (2UL<<0) /* Flush only one entry. */ |
---|
| 261 | #define UVMF_FLUSHTYPE_MASK (3UL<<0) |
---|
| 262 | #define UVMF_MULTI (0UL<<2) /* Flush subset of TLBs. */ |
---|
| 263 | #define UVMF_LOCAL (0UL<<2) /* Flush local TLB. */ |
---|
| 264 | #define UVMF_ALL (1UL<<2) /* Flush all TLBs. */ |
---|
| 265 | |
---|
| 266 | /* |
---|
| 267 | * Commands to HYPERVISOR_console_io(). |
---|
| 268 | */ |
---|
| 269 | #define CONSOLEIO_write 0 |
---|
| 270 | #define CONSOLEIO_read 1 |
---|
| 271 | |
---|
| 272 | /* |
---|
| 273 | * Commands to HYPERVISOR_vm_assist(). |
---|
| 274 | */ |
---|
| 275 | #define VMASST_CMD_enable 0 |
---|
| 276 | #define VMASST_CMD_disable 1 |
---|
| 277 | |
---|
| 278 | /* x86/32 guests: simulate full 4GB segment limits. */ |
---|
| 279 | #define VMASST_TYPE_4gb_segments 0 |
---|
| 280 | |
---|
| 281 | /* x86/32 guests: trap (vector 15) whenever above vmassist is used. */ |
---|
| 282 | #define VMASST_TYPE_4gb_segments_notify 1 |
---|
| 283 | |
---|
| 284 | /* |
---|
| 285 | * x86 guests: support writes to bottom-level PTEs. |
---|
| 286 | * NB1. Page-directory entries cannot be written. |
---|
| 287 | * NB2. Guest must continue to remove all writable mappings of PTEs. |
---|
| 288 | */ |
---|
| 289 | #define VMASST_TYPE_writable_pagetables 2 |
---|
| 290 | |
---|
| 291 | /* x86/PAE guests: support PDPTs above 4GB. */ |
---|
| 292 | #define VMASST_TYPE_pae_extended_cr3 3 |
---|
| 293 | |
---|
| 294 | #define MAX_VMASST_TYPE 3 |
---|
| 295 | |
---|
| 296 | #ifndef __ASSEMBLY__ |
---|
| 297 | |
---|
| 298 | typedef uint16_t domid_t; |
---|
| 299 | |
---|
| 300 | /* Domain ids >= DOMID_FIRST_RESERVED cannot be used for ordinary domains. */ |
---|
| 301 | #define DOMID_FIRST_RESERVED (0x7FF0U) |
---|
| 302 | |
---|
| 303 | /* DOMID_SELF is used in certain contexts to refer to oneself. */ |
---|
| 304 | #define DOMID_SELF (0x7FF0U) |
---|
| 305 | |
---|
| 306 | /* |
---|
| 307 | * DOMID_IO is used to restrict page-table updates to mapping I/O memory. |
---|
| 308 | * Although no Foreign Domain need be specified to map I/O pages, DOMID_IO |
---|
| 309 | * is useful to ensure that no mappings to the OS's own heap are accidentally |
---|
| 310 | * installed. (e.g., in Linux this could cause havoc as reference counts |
---|
| 311 | * aren't adjusted on the I/O-mapping code path). |
---|
| 312 | * This only makes sense in MMUEXT_SET_FOREIGNDOM, but in that context can |
---|
| 313 | * be specified by any calling domain. |
---|
| 314 | */ |
---|
| 315 | #define DOMID_IO (0x7FF1U) |
---|
| 316 | |
---|
| 317 | /* |
---|
| 318 | * DOMID_XEN is used to allow privileged domains to map restricted parts of |
---|
| 319 | * Xen's heap space (e.g., the machine_to_phys table). |
---|
| 320 | * This only makes sense in MMUEXT_SET_FOREIGNDOM, and is only permitted if |
---|
| 321 | * the caller is privileged. |
---|
| 322 | */ |
---|
| 323 | #define DOMID_XEN (0x7FF2U) |
---|
| 324 | |
---|
| 325 | /* |
---|
| 326 | * Send an array of these to HYPERVISOR_mmu_update(). |
---|
| 327 | * NB. The fields are natural pointer/address size for this architecture. |
---|
| 328 | */ |
---|
| 329 | struct mmu_update { |
---|
| 330 | uint64_t ptr; /* Machine address of PTE. */ |
---|
| 331 | uint64_t val; /* New contents of PTE. */ |
---|
| 332 | }; |
---|
| 333 | typedef struct mmu_update mmu_update_t; |
---|
| 334 | DEFINE_XEN_GUEST_HANDLE(mmu_update_t); |
---|
| 335 | |
---|
| 336 | /* |
---|
| 337 | * Send an array of these to HYPERVISOR_multicall(). |
---|
| 338 | * NB. The fields are natural register size for this architecture. |
---|
| 339 | */ |
---|
| 340 | struct multicall_entry { |
---|
| 341 | unsigned long op, result; |
---|
| 342 | unsigned long args[6]; |
---|
| 343 | }; |
---|
| 344 | typedef struct multicall_entry multicall_entry_t; |
---|
| 345 | DEFINE_XEN_GUEST_HANDLE(multicall_entry_t); |
---|
| 346 | |
---|
| 347 | /* |
---|
| 348 | * Event channel endpoints per domain: |
---|
| 349 | * 1024 if a long is 32 bits; 4096 if a long is 64 bits. |
---|
| 350 | */ |
---|
| 351 | #define NR_EVENT_CHANNELS (sizeof(unsigned long) * sizeof(unsigned long) * 64) |
---|
| 352 | |
---|
| 353 | struct vcpu_time_info { |
---|
| 354 | /* |
---|
| 355 | * Updates to the following values are preceded and followed by an |
---|
| 356 | * increment of 'version'. The guest can therefore detect updates by |
---|
| 357 | * looking for changes to 'version'. If the least-significant bit of |
---|
| 358 | * the version number is set then an update is in progress and the guest |
---|
| 359 | * must wait to read a consistent set of values. |
---|
| 360 | * The correct way to interact with the version number is similar to |
---|
| 361 | * Linux's seqlock: see the implementations of read_seqbegin/read_seqretry. |
---|
| 362 | */ |
---|
| 363 | uint32_t version; |
---|
| 364 | uint32_t pad0; |
---|
| 365 | uint64_t tsc_timestamp; /* TSC at last update of time vals. */ |
---|
| 366 | uint64_t system_time; /* Time, in nanosecs, since boot. */ |
---|
| 367 | /* |
---|
| 368 | * Current system time: |
---|
| 369 | * system_time + |
---|
| 370 | * ((((tsc - tsc_timestamp) << tsc_shift) * tsc_to_system_mul) >> 32) |
---|
| 371 | * CPU frequency (Hz): |
---|
| 372 | * ((10^9 << 32) / tsc_to_system_mul) >> tsc_shift |
---|
| 373 | */ |
---|
| 374 | uint32_t tsc_to_system_mul; |
---|
| 375 | int8_t tsc_shift; |
---|
| 376 | int8_t pad1[3]; |
---|
| 377 | }; /* 32 bytes */ |
---|
| 378 | typedef struct vcpu_time_info vcpu_time_info_t; |
---|
| 379 | |
---|
| 380 | struct vcpu_info { |
---|
| 381 | /* |
---|
| 382 | * 'evtchn_upcall_pending' is written non-zero by Xen to indicate |
---|
| 383 | * a pending notification for a particular VCPU. It is then cleared |
---|
| 384 | * by the guest OS /before/ checking for pending work, thus avoiding |
---|
| 385 | * a set-and-check race. Note that the mask is only accessed by Xen |
---|
| 386 | * on the CPU that is currently hosting the VCPU. This means that the |
---|
| 387 | * pending and mask flags can be updated by the guest without special |
---|
| 388 | * synchronisation (i.e., no need for the x86 LOCK prefix). |
---|
| 389 | * This may seem suboptimal because if the pending flag is set by |
---|
| 390 | * a different CPU then an IPI may be scheduled even when the mask |
---|
| 391 | * is set. However, note: |
---|
| 392 | * 1. The task of 'interrupt holdoff' is covered by the per-event- |
---|
| 393 | * channel mask bits. A 'noisy' event that is continually being |
---|
| 394 | * triggered can be masked at source at this very precise |
---|
| 395 | * granularity. |
---|
| 396 | * 2. The main purpose of the per-VCPU mask is therefore to restrict |
---|
| 397 | * reentrant execution: whether for concurrency control, or to |
---|
| 398 | * prevent unbounded stack usage. Whatever the purpose, we expect |
---|
| 399 | * that the mask will be asserted only for short periods at a time, |
---|
| 400 | * and so the likelihood of a 'spurious' IPI is suitably small. |
---|
| 401 | * The mask is read before making an event upcall to the guest: a |
---|
| 402 | * non-zero mask therefore guarantees that the VCPU will not receive |
---|
| 403 | * an upcall activation. The mask is cleared when the VCPU requests |
---|
| 404 | * to block: this avoids wakeup-waiting races. |
---|
| 405 | */ |
---|
| 406 | uint8_t evtchn_upcall_pending; |
---|
| 407 | uint8_t evtchn_upcall_mask; |
---|
| 408 | unsigned long evtchn_pending_sel; |
---|
| 409 | struct arch_vcpu_info arch; |
---|
| 410 | struct vcpu_time_info time; |
---|
| 411 | }; /* 64 bytes (x86) */ |
---|
| 412 | #ifndef __XEN__ |
---|
| 413 | typedef struct vcpu_info vcpu_info_t; |
---|
| 414 | #endif |
---|
| 415 | |
---|
| 416 | /* |
---|
| 417 | * Xen/kernel shared data -- pointer provided in start_info. |
---|
| 418 | * |
---|
| 419 | * This structure is defined to be both smaller than a page, and the |
---|
| 420 | * only data on the shared page, but may vary in actual size even within |
---|
| 421 | * compatible Xen versions; guests should not rely on the size |
---|
| 422 | * of this structure remaining constant. |
---|
| 423 | */ |
---|
| 424 | struct shared_info { |
---|
| 425 | struct vcpu_info vcpu_info[MAX_VIRT_CPUS]; |
---|
| 426 | |
---|
| 427 | /* |
---|
| 428 | * A domain can create "event channels" on which it can send and receive |
---|
| 429 | * asynchronous event notifications. There are three classes of event that |
---|
| 430 | * are delivered by this mechanism: |
---|
| 431 | * 1. Bi-directional inter- and intra-domain connections. Domains must |
---|
| 432 | * arrange out-of-band to set up a connection (usually by allocating |
---|
| 433 | * an unbound 'listener' port and avertising that via a storage service |
---|
| 434 | * such as xenstore). |
---|
| 435 | * 2. Physical interrupts. A domain with suitable hardware-access |
---|
| 436 | * privileges can bind an event-channel port to a physical interrupt |
---|
| 437 | * source. |
---|
| 438 | * 3. Virtual interrupts ('events'). A domain can bind an event-channel |
---|
| 439 | * port to a virtual interrupt source, such as the virtual-timer |
---|
| 440 | * device or the emergency console. |
---|
| 441 | * |
---|
| 442 | * Event channels are addressed by a "port index". Each channel is |
---|
| 443 | * associated with two bits of information: |
---|
| 444 | * 1. PENDING -- notifies the domain that there is a pending notification |
---|
| 445 | * to be processed. This bit is cleared by the guest. |
---|
| 446 | * 2. MASK -- if this bit is clear then a 0->1 transition of PENDING |
---|
| 447 | * will cause an asynchronous upcall to be scheduled. This bit is only |
---|
| 448 | * updated by the guest. It is read-only within Xen. If a channel |
---|
| 449 | * becomes pending while the channel is masked then the 'edge' is lost |
---|
| 450 | * (i.e., when the channel is unmasked, the guest must manually handle |
---|
| 451 | * pending notifications as no upcall will be scheduled by Xen). |
---|
| 452 | * |
---|
| 453 | * To expedite scanning of pending notifications, any 0->1 pending |
---|
| 454 | * transition on an unmasked channel causes a corresponding bit in a |
---|
| 455 | * per-vcpu selector word to be set. Each bit in the selector covers a |
---|
| 456 | * 'C long' in the PENDING bitfield array. |
---|
| 457 | */ |
---|
| 458 | unsigned long evtchn_pending[sizeof(unsigned long) * 8]; |
---|
| 459 | unsigned long evtchn_mask[sizeof(unsigned long) * 8]; |
---|
| 460 | |
---|
| 461 | /* |
---|
| 462 | * Wallclock time: updated only by control software. Guests should base |
---|
| 463 | * their gettimeofday() syscall on this wallclock-base value. |
---|
| 464 | */ |
---|
| 465 | uint32_t wc_version; /* Version counter: see vcpu_time_info_t. */ |
---|
| 466 | uint32_t wc_sec; /* Secs 00:00:00 UTC, Jan 1, 1970. */ |
---|
| 467 | uint32_t wc_nsec; /* Nsecs 00:00:00 UTC, Jan 1, 1970. */ |
---|
| 468 | |
---|
| 469 | struct arch_shared_info arch; |
---|
| 470 | |
---|
| 471 | }; |
---|
| 472 | #ifndef __XEN__ |
---|
| 473 | typedef struct shared_info shared_info_t; |
---|
| 474 | #endif |
---|
| 475 | |
---|
| 476 | /* |
---|
| 477 | * Start-of-day memory layout: |
---|
| 478 | * 1. The domain is started within contiguous virtual-memory region. |
---|
| 479 | * 2. The contiguous region ends on an aligned 4MB boundary. |
---|
| 480 | * 3. This the order of bootstrap elements in the initial virtual region: |
---|
| 481 | * a. relocated kernel image |
---|
| 482 | * b. initial ram disk [mod_start, mod_len] |
---|
| 483 | * c. list of allocated page frames [mfn_list, nr_pages] |
---|
| 484 | * d. start_info_t structure [register ESI (x86)] |
---|
| 485 | * e. bootstrap page tables [pt_base, CR3 (x86)] |
---|
| 486 | * f. bootstrap stack [register ESP (x86)] |
---|
| 487 | * 4. Bootstrap elements are packed together, but each is 4kB-aligned. |
---|
| 488 | * 5. The initial ram disk may be omitted. |
---|
| 489 | * 6. The list of page frames forms a contiguous 'pseudo-physical' memory |
---|
| 490 | * layout for the domain. In particular, the bootstrap virtual-memory |
---|
| 491 | * region is a 1:1 mapping to the first section of the pseudo-physical map. |
---|
| 492 | * 7. All bootstrap elements are mapped read-writable for the guest OS. The |
---|
| 493 | * only exception is the bootstrap page table, which is mapped read-only. |
---|
| 494 | * 8. There is guaranteed to be at least 512kB padding after the final |
---|
| 495 | * bootstrap element. If necessary, the bootstrap virtual region is |
---|
| 496 | * extended by an extra 4MB to ensure this. |
---|
| 497 | */ |
---|
| 498 | |
---|
| 499 | #define MAX_GUEST_CMDLINE 1024 |
---|
| 500 | struct start_info { |
---|
| 501 | /* THE FOLLOWING ARE FILLED IN BOTH ON INITIAL BOOT AND ON RESUME. */ |
---|
| 502 | char magic[32]; /* "xen-<version>-<platform>". */ |
---|
| 503 | unsigned long nr_pages; /* Total pages allocated to this domain. */ |
---|
| 504 | unsigned long shared_info; /* MACHINE address of shared info struct. */ |
---|
| 505 | uint32_t flags; /* SIF_xxx flags. */ |
---|
| 506 | xen_pfn_t store_mfn; /* MACHINE page number of shared page. */ |
---|
| 507 | uint32_t store_evtchn; /* Event channel for store communication. */ |
---|
| 508 | union { |
---|
| 509 | struct { |
---|
| 510 | xen_pfn_t mfn; /* MACHINE page number of console page. */ |
---|
| 511 | uint32_t evtchn; /* Event channel for console page. */ |
---|
| 512 | } domU; |
---|
| 513 | struct { |
---|
| 514 | uint32_t info_off; /* Offset of console_info struct. */ |
---|
| 515 | uint32_t info_size; /* Size of console_info struct from start.*/ |
---|
| 516 | } dom0; |
---|
| 517 | } console; |
---|
| 518 | /* THE FOLLOWING ARE ONLY FILLED IN ON INITIAL BOOT (NOT RESUME). */ |
---|
| 519 | unsigned long pt_base; /* VIRTUAL address of page directory. */ |
---|
| 520 | unsigned long nr_pt_frames; /* Number of bootstrap p.t. frames. */ |
---|
| 521 | unsigned long mfn_list; /* VIRTUAL address of page-frame list. */ |
---|
| 522 | unsigned long mod_start; /* VIRTUAL address of pre-loaded module. */ |
---|
| 523 | unsigned long mod_len; /* Size (bytes) of pre-loaded module. */ |
---|
| 524 | int8_t cmd_line[MAX_GUEST_CMDLINE]; |
---|
| 525 | }; |
---|
| 526 | typedef struct start_info start_info_t; |
---|
| 527 | |
---|
| 528 | /* New console union for dom0 introduced in 0x00030203. */ |
---|
| 529 | #if __XEN_INTERFACE_VERSION__ < 0x00030203 |
---|
| 530 | #define console_mfn console.domU.mfn |
---|
| 531 | #define console_evtchn console.domU.evtchn |
---|
| 532 | #endif |
---|
| 533 | |
---|
| 534 | /* These flags are passed in the 'flags' field of start_info_t. */ |
---|
| 535 | #define SIF_PRIVILEGED (1<<0) /* Is the domain privileged? */ |
---|
| 536 | #define SIF_INITDOMAIN (1<<1) /* Is this the initial control domain? */ |
---|
| 537 | |
---|
| 538 | typedef struct dom0_vga_console_info { |
---|
| 539 | uint8_t video_type; /* DOM0_VGA_CONSOLE_??? */ |
---|
| 540 | #define XEN_VGATYPE_TEXT_MODE_3 0x03 |
---|
| 541 | #define XEN_VGATYPE_VESA_LFB 0x23 |
---|
| 542 | |
---|
| 543 | union { |
---|
| 544 | struct { |
---|
| 545 | /* Font height, in pixels. */ |
---|
| 546 | uint16_t font_height; |
---|
| 547 | /* Cursor location (column, row). */ |
---|
| 548 | uint16_t cursor_x, cursor_y; |
---|
| 549 | /* Number of rows and columns (dimensions in characters). */ |
---|
| 550 | uint16_t rows, columns; |
---|
| 551 | } text_mode_3; |
---|
| 552 | |
---|
| 553 | struct { |
---|
| 554 | /* Width and height, in pixels. */ |
---|
| 555 | uint16_t width, height; |
---|
| 556 | /* Bytes per scan line. */ |
---|
| 557 | uint16_t bytes_per_line; |
---|
| 558 | /* Bits per pixel. */ |
---|
| 559 | uint16_t bits_per_pixel; |
---|
| 560 | /* LFB physical address, and size (in units of 64kB). */ |
---|
| 561 | uint32_t lfb_base; |
---|
| 562 | uint32_t lfb_size; |
---|
| 563 | /* RGB mask offsets and sizes, as defined by VBE 1.2+ */ |
---|
| 564 | uint8_t red_pos, red_size; |
---|
| 565 | uint8_t green_pos, green_size; |
---|
| 566 | uint8_t blue_pos, blue_size; |
---|
| 567 | uint8_t rsvd_pos, rsvd_size; |
---|
| 568 | } vesa_lfb; |
---|
| 569 | } u; |
---|
| 570 | } dom0_vga_console_info_t; |
---|
| 571 | |
---|
| 572 | typedef uint8_t xen_domain_handle_t[16]; |
---|
| 573 | |
---|
| 574 | /* Turn a plain number into a C unsigned long constant. */ |
---|
| 575 | #define __mk_unsigned_long(x) x ## UL |
---|
| 576 | #define mk_unsigned_long(x) __mk_unsigned_long(x) |
---|
| 577 | |
---|
| 578 | DEFINE_XEN_GUEST_HANDLE(uint8_t); |
---|
| 579 | DEFINE_XEN_GUEST_HANDLE(uint16_t); |
---|
| 580 | DEFINE_XEN_GUEST_HANDLE(uint32_t); |
---|
| 581 | DEFINE_XEN_GUEST_HANDLE(uint64_t); |
---|
| 582 | |
---|
| 583 | #else /* __ASSEMBLY__ */ |
---|
| 584 | |
---|
| 585 | /* In assembly code we cannot use C numeric constant suffixes. */ |
---|
| 586 | #define mk_unsigned_long(x) x |
---|
| 587 | |
---|
| 588 | #endif /* !__ASSEMBLY__ */ |
---|
| 589 | |
---|
| 590 | /* Default definitions for macros used by domctl/sysctl. */ |
---|
| 591 | #if defined(__XEN__) || defined(__XEN_TOOLS__) |
---|
| 592 | #ifndef uint64_aligned_t |
---|
| 593 | #define uint64_aligned_t uint64_t |
---|
| 594 | #endif |
---|
| 595 | #ifndef XEN_GUEST_HANDLE_64 |
---|
| 596 | #define XEN_GUEST_HANDLE_64(name) XEN_GUEST_HANDLE(name) |
---|
| 597 | #endif |
---|
| 598 | #endif |
---|
| 599 | |
---|
| 600 | #endif /* __XEN_PUBLIC_XEN_H__ */ |
---|
| 601 | |
---|
| 602 | /* |
---|
| 603 | * Local variables: |
---|
| 604 | * mode: C |
---|
| 605 | * c-set-style: "BSD" |
---|
| 606 | * c-basic-offset: 4 |
---|
| 607 | * tab-width: 4 |
---|
| 608 | * indent-tabs-mode: nil |
---|
| 609 | * End: |
---|
| 610 | */ |
---|