linux_machdep.c revision 1.48.14.1
1 /* $NetBSD: linux_machdep.c,v 1.48.14.1 2017/04/21 16:53:41 bouyer Exp $ */ 2 3 /*- 4 * Copyright (c) 2005 Emmanuel Dreyfus, all rights reserved. 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 1. Redistributions of source code must retain the above copyright 10 * notice, this list of conditions and the following disclaimer. 11 * 2. Redistributions in binary form must reproduce the above copyright 12 * notice, this list of conditions and the following disclaimer in the 13 * documentation and/or other materials provided with the distribution. 14 * 3. All advertising materials mentioning features or use of this software 15 * must display the following acknowledgement: 16 * This product includes software developed by Emmanuel Dreyfus 17 * 4. The name of the author may not be used to endorse or promote 18 * products derived from this software without specific prior written 19 * permission. 20 * 21 * THIS SOFTWARE IS PROVIDED BY THE THE AUTHOR AND CONTRIBUTORS ``AS IS'' 22 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, 23 * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 24 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS 25 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 26 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 27 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 28 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 29 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 30 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 31 * POSSIBILITY OF SUCH DAMAGE. 32 */ 33 34 #include <sys/cdefs.h> 35 36 __KERNEL_RCSID(0, "$NetBSD: linux_machdep.c,v 1.48.14.1 2017/04/21 16:53:41 bouyer Exp $"); 37 38 #include <sys/param.h> 39 #include <sys/types.h> 40 #include <sys/systm.h> 41 #include <sys/signal.h> 42 #include <sys/exec.h> 43 #include <sys/proc.h> 44 #include <sys/ptrace.h> /* for process_read_fpregs() */ 45 #include <sys/ucontext.h> 46 #include <sys/conf.h> 47 48 #include <machine/reg.h> 49 #include <machine/pcb.h> 50 #include <machine/mcontext.h> 51 #include <machine/specialreg.h> 52 #include <machine/vmparam.h> 53 #include <machine/cpufunc.h> 54 #include <x86/include/sysarch.h> 55 56 /* 57 * To see whether wscons is configured (for virtual console ioctl calls). 58 */ 59 #if defined(_KERNEL_OPT) 60 #include "wsdisplay.h" 61 #endif 62 #if (NWSDISPLAY > 0) 63 #include <dev/wscons/wsconsio.h> 64 #include <dev/wscons/wsdisplay_usl_io.h> 65 #endif 66 67 68 #include <compat/linux/common/linux_signal.h> 69 #include <compat/linux/common/linux_errno.h> 70 #include <compat/linux/common/linux_exec.h> 71 #include <compat/linux/common/linux_ioctl.h> 72 #include <compat/linux/common/linux_prctl.h> 73 #include <compat/linux/common/linux_machdep.h> 74 #include <compat/linux/common/linux_ipc.h> 75 #include <compat/linux/common/linux_sem.h> 76 #include <compat/linux/linux_syscall.h> 77 #include <compat/linux/linux_syscallargs.h> 78 79 static void linux_buildcontext(struct lwp *, void *, void *); 80 81 void 82 linux_setregs(struct lwp *l, struct exec_package *epp, vaddr_t stack) 83 { 84 struct pcb *pcb = lwp_getpcb(l); 85 struct trapframe *tf; 86 87 #ifdef USER_LDT 88 pmap_ldt_cleanup(l); 89 #endif 90 91 fpu_save_area_clear(l, __NetBSD_NPXCW__); 92 pcb->pcb_flags = 0; 93 94 l->l_proc->p_flag &= ~PK_32; 95 96 tf = l->l_md.md_regs; 97 tf->tf_rax = 0; 98 tf->tf_rbx = 0; 99 tf->tf_rcx = epp->ep_entry; 100 tf->tf_rdx = 0; 101 tf->tf_rsi = 0; 102 tf->tf_rdi = 0; 103 tf->tf_rbp = 0; 104 tf->tf_rsp = stack; 105 tf->tf_r8 = 0; 106 tf->tf_r9 = 0; 107 tf->tf_r10 = 0; 108 tf->tf_r11 = 0; 109 tf->tf_r12 = 0; 110 tf->tf_r13 = 0; 111 tf->tf_r14 = 0; 112 tf->tf_r15 = 0; 113 tf->tf_rip = epp->ep_entry; 114 tf->tf_rflags = PSL_USERSET; 115 tf->tf_cs = GSEL(GUCODE_SEL, SEL_UPL); 116 tf->tf_ss = GSEL(GUDATA_SEL, SEL_UPL); 117 tf->tf_ds = GSEL(GUDATA_SEL, SEL_UPL); 118 tf->tf_es = 0; 119 cpu_fsgs_zero(l); 120 121 return; 122 } 123 124 void 125 linux_sendsig(const ksiginfo_t *ksi, const sigset_t *mask) 126 { 127 struct lwp *l = curlwp; 128 struct proc *p = l->l_proc; 129 struct pcb *pcb = lwp_getpcb(l); 130 struct sigacts *ps = p->p_sigacts; 131 int onstack, error; 132 int sig = ksi->ksi_signo; 133 struct linux_rt_sigframe *sfp, sigframe; 134 struct linux__fpstate *fpsp; 135 struct fpreg fpregs; 136 struct trapframe *tf = l->l_md.md_regs; 137 sig_t catcher = SIGACTION(p, sig).sa_handler; 138 linux_sigset_t lmask; 139 char *sp; 140 141 /* Do we need to jump onto the signal stack? */ 142 onstack = 143 (l->l_sigstk.ss_flags & (SS_DISABLE | SS_ONSTACK)) == 0 && 144 (SIGACTION(p, sig).sa_flags & SA_ONSTACK) != 0; 145 146 /* Allocate space for the signal handler context. */ 147 if (onstack) 148 sp = ((char *)l->l_sigstk.ss_sp + 149 l->l_sigstk.ss_size); 150 else 151 sp = (char *)tf->tf_rsp - 128; 152 153 /* Save FPU state */ 154 sp = (char *) (((long)sp - sizeof (*fpsp)) & ~0xfUL); 155 fpsp = (struct linux__fpstate *)sp; 156 157 /* 158 * Populate the rt_sigframe 159 */ 160 sp = (char *) 161 ((((long)sp - sizeof(struct linux_rt_sigframe)) & ~0xfUL) - 8); 162 sfp = (struct linux_rt_sigframe *)sp; 163 164 memset(&sigframe, 0, sizeof(sigframe)); 165 if (ps->sa_sigdesc[sig].sd_vers != 0) 166 sigframe.pretcode = 167 (char *)(u_long)ps->sa_sigdesc[sig].sd_tramp; 168 else 169 sigframe.pretcode = NULL; 170 171 /* 172 * The user context 173 */ 174 sigframe.uc.luc_flags = 0; 175 sigframe.uc.luc_link = NULL; 176 177 /* This is used regardless of SA_ONSTACK in Linux */ 178 sigframe.uc.luc_stack.ss_sp = l->l_sigstk.ss_sp; 179 sigframe.uc.luc_stack.ss_size = l->l_sigstk.ss_size; 180 sigframe.uc.luc_stack.ss_flags = 0; 181 if (l->l_sigstk.ss_flags & SS_ONSTACK) 182 sigframe.uc.luc_stack.ss_flags |= LINUX_SS_ONSTACK; 183 if (l->l_sigstk.ss_flags & SS_DISABLE) 184 sigframe.uc.luc_stack.ss_flags |= LINUX_SS_DISABLE; 185 186 sigframe.uc.luc_mcontext.r8 = tf->tf_r8; 187 sigframe.uc.luc_mcontext.r9 = tf->tf_r9; 188 sigframe.uc.luc_mcontext.r10 = tf->tf_r10; 189 sigframe.uc.luc_mcontext.r11 = tf->tf_r11; 190 sigframe.uc.luc_mcontext.r12 = tf->tf_r12; 191 sigframe.uc.luc_mcontext.r13 = tf->tf_r13; 192 sigframe.uc.luc_mcontext.r14 = tf->tf_r14; 193 sigframe.uc.luc_mcontext.r15 = tf->tf_r15; 194 sigframe.uc.luc_mcontext.rdi = tf->tf_rdi; 195 sigframe.uc.luc_mcontext.rsi = tf->tf_rsi; 196 sigframe.uc.luc_mcontext.rbp = tf->tf_rbp; 197 sigframe.uc.luc_mcontext.rbx = tf->tf_rbx; 198 sigframe.uc.luc_mcontext.rdx = tf->tf_rdx; 199 sigframe.uc.luc_mcontext.rax = tf->tf_rax; 200 sigframe.uc.luc_mcontext.rcx = tf->tf_rcx; 201 sigframe.uc.luc_mcontext.rsp = tf->tf_rsp; 202 sigframe.uc.luc_mcontext.rip = tf->tf_rip; 203 sigframe.uc.luc_mcontext.eflags = tf->tf_rflags; 204 sigframe.uc.luc_mcontext.cs = tf->tf_cs; 205 sigframe.uc.luc_mcontext.gs = tf->tf_gs; 206 sigframe.uc.luc_mcontext.fs = tf->tf_fs; 207 sigframe.uc.luc_mcontext.err = tf->tf_err; 208 sigframe.uc.luc_mcontext.trapno = tf->tf_trapno; 209 native_to_linux_sigset(&lmask, mask); 210 sigframe.uc.luc_mcontext.oldmask = lmask.sig[0]; 211 sigframe.uc.luc_mcontext.cr2 = (long)pcb->pcb_onfault; 212 sigframe.uc.luc_mcontext.fpstate = fpsp; 213 native_to_linux_sigset(&sigframe.uc.luc_sigmask, mask); 214 native_to_linux_siginfo(&sigframe.info, &ksi->ksi_info); 215 sendsig_reset(l, sig); 216 mutex_exit(p->p_lock); 217 error = 0; 218 219 /* 220 * Save FPU state, if any 221 */ 222 if (fpsp != NULL) { 223 size_t fp_size = sizeof fpregs; 224 /* The netbsd and linux structures both match the fxsave data */ 225 (void)process_read_fpregs(l, &fpregs, &fp_size); 226 error = copyout(&fpregs, fpsp, sizeof(*fpsp)); 227 } 228 229 if (error == 0) 230 error = copyout(&sigframe, sp, sizeof(sigframe)); 231 232 mutex_enter(p->p_lock); 233 234 if (error != 0) { 235 sigexit(l, SIGILL); 236 return; 237 } 238 239 if ((vaddr_t)catcher >= VM_MAXUSER_ADDRESS) { 240 sigexit(l, SIGILL); 241 return; 242 } 243 244 linux_buildcontext(l, catcher, sp); 245 tf->tf_rdi = sigframe.info.lsi_signo; 246 tf->tf_rax = 0; 247 tf->tf_rsi = (long)&sfp->info; 248 tf->tf_rdx = (long)&sfp->uc; 249 250 /* 251 * Remember we use signal stack 252 */ 253 if (onstack) 254 l->l_sigstk.ss_flags |= SS_ONSTACK; 255 return; 256 } 257 258 int 259 linux_sys_modify_ldt(struct lwp *l, const struct linux_sys_modify_ldt_args *v, register_t *retval) 260 { 261 printf("linux_sys_modify_ldt\n"); 262 return 0; 263 } 264 265 int 266 linux_sys_iopl(struct lwp *l, const struct linux_sys_iopl_args *v, register_t *retval) 267 { 268 return 0; 269 } 270 271 int 272 linux_sys_ioperm(struct lwp *l, const struct linux_sys_ioperm_args *v, register_t *retval) 273 { 274 return 0; 275 } 276 277 dev_t 278 linux_fakedev(dev_t dev, int raw) 279 { 280 281 extern const struct cdevsw ptc_cdevsw, pts_cdevsw; 282 const struct cdevsw *cd = cdevsw_lookup(dev); 283 284 if (raw) { 285 #if (NWSDISPLAY > 0) 286 extern const struct cdevsw wsdisplay_cdevsw; 287 if (cd == &wsdisplay_cdevsw) 288 return makedev(LINUX_CONS_MAJOR, (minor(dev) + 1)); 289 #endif 290 } 291 292 if (cd == &ptc_cdevsw) 293 return makedev(LINUX_PTC_MAJOR, minor(dev)); 294 if (cd == &pts_cdevsw) 295 return makedev(LINUX_PTS_MAJOR, minor(dev)); 296 297 return ((minor(dev) & 0xff) | ((major(dev) & 0xfff) << 8) 298 | (((unsigned long long int) (minor(dev) & ~0xff)) << 12) 299 | (((unsigned long long int) (major(dev) & ~0xfff)) << 32)); 300 } 301 302 int 303 linux_machdepioctl(struct lwp *l, const struct linux_sys_ioctl_args *v, register_t *retval) 304 { 305 return 0; 306 } 307 308 int 309 linux_sys_rt_sigreturn(struct lwp *l, const void *v, register_t *retval) 310 { 311 struct linux_ucontext *luctx; 312 struct trapframe *tf = l->l_md.md_regs; 313 struct linux_sigcontext *lsigctx; 314 struct linux_rt_sigframe frame, *fp; 315 ucontext_t uctx; 316 mcontext_t *mctx; 317 struct fxsave *fxarea; 318 int error; 319 320 fp = (struct linux_rt_sigframe *)(tf->tf_rsp - 8); 321 if ((error = copyin(fp, &frame, sizeof(frame))) != 0) { 322 mutex_enter(l->l_proc->p_lock); 323 sigexit(l, SIGILL); 324 return error; 325 } 326 luctx = &frame.uc; 327 lsigctx = &luctx->luc_mcontext; 328 329 memset(&uctx, 0, sizeof(uctx)); 330 mctx = (mcontext_t *)&uctx.uc_mcontext; 331 fxarea = (struct fxsave *)&mctx->__fpregs; 332 333 /* 334 * Set the flags. Linux always have CPU, stack and signal state, 335 * FPU is optional. uc_flags is not used to tell what we have. 336 */ 337 uctx.uc_flags = (_UC_SIGMASK|_UC_CPU|_UC_STACK|_UC_CLRSTACK); 338 if (lsigctx->fpstate != NULL) 339 uctx.uc_flags |= _UC_FPU; 340 uctx.uc_link = NULL; 341 342 /* 343 * Signal set 344 */ 345 linux_to_native_sigset(&uctx.uc_sigmask, &luctx->luc_sigmask); 346 347 /* 348 * CPU state 349 */ 350 mctx->__gregs[_REG_R8] = lsigctx->r8; 351 mctx->__gregs[_REG_R9] = lsigctx->r9; 352 mctx->__gregs[_REG_R10] = lsigctx->r10; 353 mctx->__gregs[_REG_R11] = lsigctx->r11; 354 mctx->__gregs[_REG_R12] = lsigctx->r12; 355 mctx->__gregs[_REG_R13] = lsigctx->r13; 356 mctx->__gregs[_REG_R14] = lsigctx->r14; 357 mctx->__gregs[_REG_R15] = lsigctx->r15; 358 mctx->__gregs[_REG_RDI] = lsigctx->rdi; 359 mctx->__gregs[_REG_RSI] = lsigctx->rsi; 360 mctx->__gregs[_REG_RBP] = lsigctx->rbp; 361 mctx->__gregs[_REG_RBX] = lsigctx->rbx; 362 mctx->__gregs[_REG_RAX] = lsigctx->rax; 363 mctx->__gregs[_REG_RDX] = lsigctx->rdx; 364 mctx->__gregs[_REG_RCX] = lsigctx->rcx; 365 mctx->__gregs[_REG_RIP] = lsigctx->rip; 366 mctx->__gregs[_REG_RFLAGS] = lsigctx->eflags; 367 mctx->__gregs[_REG_CS] = lsigctx->cs; 368 mctx->__gregs[_REG_GS] = lsigctx->gs; 369 mctx->__gregs[_REG_FS] = lsigctx->fs; 370 mctx->__gregs[_REG_ERR] = lsigctx->err; 371 mctx->__gregs[_REG_TRAPNO] = lsigctx->trapno; 372 mctx->__gregs[_REG_ES] = tf->tf_es; 373 mctx->__gregs[_REG_DS] = tf->tf_ds; 374 mctx->__gregs[_REG_RSP] = lsigctx->rsp; /* XXX */ 375 mctx->__gregs[_REG_SS] = tf->tf_ss; 376 377 /* 378 * FPU state 379 */ 380 if (lsigctx->fpstate != NULL) { 381 /* Both structures match the fxstate data */ 382 error = copyin(lsigctx->fpstate, fxarea, sizeof(*fxarea)); 383 if (error != 0) { 384 mutex_enter(l->l_proc->p_lock); 385 sigexit(l, SIGILL); 386 return error; 387 } 388 } 389 390 /* 391 * And the stack 392 */ 393 uctx.uc_stack.ss_flags = 0; 394 if (luctx->luc_stack.ss_flags & LINUX_SS_ONSTACK) 395 uctx.uc_stack.ss_flags |= SS_ONSTACK; 396 397 if (luctx->luc_stack.ss_flags & LINUX_SS_DISABLE) 398 uctx.uc_stack.ss_flags |= SS_DISABLE; 399 400 uctx.uc_stack.ss_sp = luctx->luc_stack.ss_sp; 401 uctx.uc_stack.ss_size = luctx->luc_stack.ss_size; 402 403 /* 404 * And let setucontext deal with that. 405 */ 406 mutex_enter(l->l_proc->p_lock); 407 error = setucontext(l, &uctx); 408 mutex_exit(l->l_proc->p_lock); 409 if (error) 410 return error; 411 412 return EJUSTRETURN; 413 } 414 415 int 416 linux_sys_arch_prctl(struct lwp *l, 417 const struct linux_sys_arch_prctl_args *uap, register_t *retval) 418 { 419 /* { 420 syscallarg(int) code; 421 syscallarg(unsigned long) addr; 422 } */ 423 void *addr = (void *)SCARG(uap, addr); 424 425 switch(SCARG(uap, code)) { 426 case LINUX_ARCH_SET_GS: 427 return x86_set_sdbase(addr, 'g', l, true); 428 429 case LINUX_ARCH_GET_GS: 430 return x86_get_sdbase(addr, 'g'); 431 432 case LINUX_ARCH_SET_FS: 433 return x86_set_sdbase(addr, 'f', l, true); 434 435 case LINUX_ARCH_GET_FS: 436 return x86_get_sdbase(addr, 'f'); 437 438 default: 439 #ifdef DEBUG_LINUX 440 printf("linux_sys_arch_prctl: unexpected code %d\n", 441 SCARG(uap, code)); 442 #endif 443 return EINVAL; 444 } 445 /* NOTREACHED */ 446 } 447 448 const int linux_vsyscall_to_syscall[] = { 449 LINUX_SYS_gettimeofday, 450 LINUX_SYS_time, 451 LINUX_SYS_nosys, /* nosys */ 452 LINUX_SYS_nosys, /* nosys */ 453 }; 454 455 int 456 linux_usertrap(struct lwp *l, vaddr_t trapaddr, void *arg) 457 { 458 struct trapframe *tf = arg; 459 uint64_t retaddr; 460 size_t vsyscallnr; 461 462 /* 463 * Check for a vsyscall. %rip must be the fault address, 464 * and the address must be in the Linux vsyscall area. 465 * Also, vsyscalls are only done at 1024-byte boundaries. 466 */ 467 468 if (__predict_true(trapaddr < LINUX_VSYSCALL_START)) 469 return 0; 470 471 if (trapaddr != tf->tf_rip) 472 return 0; 473 474 if ((tf->tf_rip & (LINUX_VSYSCALL_SIZE - 1)) != 0) 475 return 0; 476 477 vsyscallnr = (tf->tf_rip - LINUX_VSYSCALL_START) / LINUX_VSYSCALL_SIZE; 478 479 if (vsyscallnr > LINUX_VSYSCALL_MAXNR) 480 return 0; 481 482 /* 483 * Get the return address from the top of the stack, 484 * and fix up the return address. 485 * This assumes the faulting instruction was callq *reg, 486 * which is the only way that vsyscalls are ever entered. 487 */ 488 if (copyin((void *)tf->tf_rsp, &retaddr, sizeof retaddr) != 0) 489 return 0; 490 if ((vaddr_t)retaddr >= VM_MAXUSER_ADDRESS) 491 return 0; 492 tf->tf_rip = retaddr; 493 tf->tf_rax = linux_vsyscall_to_syscall[vsyscallnr]; 494 tf->tf_rsp += 8; /* "pop" the return address */ 495 496 #if 0 497 printf("usertrap: rip %p rsp %p retaddr %p vsys %d sys %d\n", 498 (void *)tf->tf_rip, (void *)tf->tf_rsp, (void *)retaddr, 499 vsyscallnr, (int)tf->tf_rax); 500 #endif 501 502 (*l->l_proc->p_md.md_syscall)(tf); 503 504 return 1; 505 } 506 507 static void 508 linux_buildcontext(struct lwp *l, void *catcher, void *f) 509 { 510 struct trapframe *tf = l->l_md.md_regs; 511 512 tf->tf_ds = GSEL(GUDATA_SEL, SEL_UPL); 513 tf->tf_rip = (u_int64_t)catcher; 514 tf->tf_cs = GSEL(GUCODE_SEL, SEL_UPL); 515 tf->tf_rflags &= ~PSL_CLEARSIG; 516 tf->tf_rsp = (u_int64_t)f; 517 tf->tf_ss = GSEL(GUDATA_SEL, SEL_UPL); 518 } 519
Indexes created Tue Sep 30 20:09:53 GMT 2025