/*

 * Copyright 1999-2007 Sun Microsystems, Inc.  All Rights Reserved.

 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.

 *

 * This code is free software; you can redistribute it and/or modify it

 * under the terms of the GNU General Public License version 2 only, as

 * published by the Free Software Foundation.

 *

 * This code is distributed in the hope that it will be useful, but WITHOUT

 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or

 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License

 * version 2 for more details (a copy is included in the LICENSE file that

 * accompanied this code).

 *

 * You should have received a copy of the GNU General Public License version

 * 2 along with this work; if not, write to the Free Software Foundation,

 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.

 *

 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,

 * CA 95054 USA or visit www.sun.com if you need additional information or

 * have any questions.

 *

 */

// do not include  precompiled  header file

# include "incls/_os_linux_x86.cpp.incl"

// put OS-includes here

# include <sys/types.h>

# include <sys/mman.h>

# include <pthread.h>

# include <signal.h>

# include <errno.h>

# include <dlfcn.h>

# include <stdlib.h>

# include <stdio.h>

# include <unistd.h>

# include <sys/resource.h>

# include <pthread.h>

# include <sys/stat.h>

# include <sys/time.h>

# include <sys/utsname.h>

# include <sys/socket.h>

# include <sys/wait.h>

# include <pwd.h>

# include <poll.h>

# include <ucontext.h>

# include <fpu_control.h>

#ifdef AMD64

#define REG_SP REG_RSP

#define REG_PC REG_RIP

#define REG_FP REG_RBP

#define SPELL_REG_SP "rsp"

#define SPELL_REG_FP "rbp"

#else

#define REG_SP REG_UESP

#define REG_PC REG_EIP

#define REG_FP REG_EBP

#define SPELL_REG_SP "esp"

#define SPELL_REG_FP "ebp"

#endif // AMD64

address os::current_stack_pointer() {

  register void *esp __asm__ (SPELL_REG_SP);

  return (address) esp;

}

char* os::non_memory_address_word() {

  // Must never look like an address returned by reserve_memory,

  // even in its subfields (as defined by the CPU immediate fields,

  // if the CPU splits constants across multiple instructions).

  return (char*) -1;

}

void os::initialize_thread() {

// Nothing to do.

}

address os::Linux::ucontext_get_pc(ucontext_t * uc) {

  return (address)uc->uc_mcontext.gregs[REG_PC];

}

intptr_t* os::Linux::ucontext_get_sp(ucontext_t * uc) {

  return (intptr_t*)uc->uc_mcontext.gregs[REG_SP];

}

intptr_t* os::Linux::ucontext_get_fp(ucontext_t * uc) {

  return (intptr_t*)uc->uc_mcontext.gregs[REG_FP];

}

// For Forte Analyzer AsyncGetCallTrace profiling support - thread

// is currently interrupted by SIGPROF.

// os::Solaris::fetch_frame_from_ucontext() tries to skip nested signal

// frames. Currently we don't do that on Linux, so it's the same as

// os::fetch_frame_from_context().

ExtendedPC os::Linux::fetch_frame_from_ucontext(Thread* thread,

  ucontext_t* uc, intptr_t** ret_sp, intptr_t** ret_fp) {

  assert(thread != NULL, "just checking");

  assert(ret_sp != NULL, "just checking");

  assert(ret_fp != NULL, "just checking");

  return os::fetch_frame_from_context(uc, ret_sp, ret_fp);

}

ExtendedPC os::fetch_frame_from_context(void* ucVoid,

                    intptr_t** ret_sp, intptr_t** ret_fp) {

  ExtendedPC  epc;

  ucontext_t* uc = (ucontext_t*)ucVoid;

  if (uc != NULL) {

    epc = ExtendedPC(os::Linux::ucontext_get_pc(uc));

    if (ret_sp) *ret_sp = os::Linux::ucontext_get_sp(uc);

    if (ret_fp) *ret_fp = os::Linux::ucontext_get_fp(uc);

  } else {

    // construct empty ExtendedPC for return value checking

    epc = ExtendedPC(NULL);

    if (ret_sp) *ret_sp = (intptr_t *)NULL;

    if (ret_fp) *ret_fp = (intptr_t *)NULL;

  }

  return epc;

}

frame os::fetch_frame_from_context(void* ucVoid) {

  intptr_t* sp;

  intptr_t* fp;

  ExtendedPC epc = fetch_frame_from_context(ucVoid, &sp, &fp);

  return frame(sp, fp, epc.pc());

}

// By default, gcc always save frame pointer (%ebp/%rbp) on stack. It may get

// turned off by -fomit-frame-pointer,

frame os::get_sender_for_C_frame(frame* fr) {

  return frame(fr->sender_sp(), fr->link(), fr->sender_pc());

}

intptr_t* _get_previous_fp() {

  register intptr_t **ebp __asm__ (SPELL_REG_FP);

  return (intptr_t*) *ebp;   // we want what it points to.

}

frame os::current_frame() {

  intptr_t* fp = _get_previous_fp();

  frame myframe((intptr_t*)os::current_stack_pointer(),

                (intptr_t*)fp,

                CAST_FROM_FN_PTR(address, os::current_frame));

  if (os::is_first_C_frame(&myframe)) {

    // stack is not walkable

    return frame(NULL, NULL, NULL);

  } else {

    return os::get_sender_for_C_frame(&myframe);

  }

}

// Utility functions

julong os::allocatable_physical_memory(julong size) {

#ifdef AMD64

  return size;

#else

  julong result = MIN2(size, (julong)3800*M);

   if (!is_allocatable(result)) {

     // See comments under solaris for alignment considerations

     julong reasonable_size = (julong)2*G - 2 * os::vm_page_size();

     result =  MIN2(size, reasonable_size);

   }

   return result;

#endif // AMD64

}

// From IA32 System Programming Guide

enum {

  trap_page_fault = 0xE

};

extern "C" void Fetch32PFI () ;

extern "C" void Fetch32Resume () ;

#ifdef AMD64

extern "C" void FetchNPFI () ;

extern "C" void FetchNResume () ;

#endif // AMD64

extern "C" int

JVM_handle_linux_signal(int sig,

                        siginfo_t* info,

                        void* ucVoid,

                        int abort_if_unrecognized) {

  ucontext_t* uc = (ucontext_t*) ucVoid;

  Thread* t = ThreadLocalStorage::get_thread_slow();

  SignalHandlerMark shm(t);

  // Note: it's not uncommon that JNI code uses signal/sigset to install

  // then restore certain signal handler (e.g. to temporarily block SIGPIPE,

  // or have a SIGILL handler when detecting CPU type). When that happens,

  // JVM_handle_linux_signal() might be invoked with junk info/ucVoid. To

  // avoid unnecessary crash when libjsig is not preloaded, try handle signals

  // that do not require siginfo/ucontext first.

  if (sig == SIGPIPE || sig == SIGXFSZ) {

    // allow chained handler to go first

    if (os::Linux::chained_handler(sig, info, ucVoid)) {

      return true;

    } else {

      if (PrintMiscellaneous && (WizardMode || Verbose)) {

        char buf[64];

        warning("Ignoring %s - see bugs 4229104 or 646499219",

                os::exception_name(sig, buf, sizeof(buf)));

      }

      return true;

    }

  }

  JavaThread* thread = NULL;

  VMThread* vmthread = NULL;

  if (os::Linux::signal_handlers_are_installed) {

    if (t != NULL ){

      if(t->is_Java_thread()) {

        thread = (JavaThread*)t;

      }

      else if(t->is_VM_thread()){

        vmthread = (VMThread *)t;

      }

    }

  }

/*

  NOTE: does not seem to work on linux.

  if (info == NULL || info->si_code <= 0 || info->si_code == SI_NOINFO) {

    // can't decode this kind of signal

    info = NULL;

  } else {

    assert(sig == info->si_signo, "bad siginfo");

  }

*/

  // decide if this trap can be handled by a stub

  address stub = NULL;

  address pc          = NULL;

  //%note os_trap_1

  if (info != NULL && uc != NULL && thread != NULL) {

    pc = (address) os::Linux::ucontext_get_pc(uc);

    if (pc == (address) Fetch32PFI) {

       uc->uc_mcontext.gregs[REG_PC] = intptr_t(Fetch32Resume) ;

       return 1 ;

    }

#ifdef AMD64

    if (pc == (address) FetchNPFI) {

       uc->uc_mcontext.gregs[REG_PC] = intptr_t (FetchNResume) ;

       return 1 ;

    }

#endif // AMD64

    // Handle ALL stack overflow variations here

    if (sig == SIGSEGV) {

      address addr = (address) info->si_addr;

      // check if fault address is within thread stack

      if (addr < thread->stack_base() &&

          addr >= thread->stack_base() - thread->stack_size()) {

        // stack overflow

        if (thread->in_stack_yellow_zone(addr)) {

          thread->disable_stack_yellow_zone();

          if (thread->thread_state() == _thread_in_Java) {

            // Throw a stack overflow exception.  Guard pages will be reenabled

            // while unwinding the stack.

            stub = SharedRuntime::continuation_for_implicit_exception(thread, pc, SharedRuntime::STACK_OVERFLOW);

          } else {

            // Thread was in the vm or native code.  Return and try to finish.

            return 1;

          }

        } else if (thread->in_stack_red_zone(addr)) {

          // Fatal red zone violation.  Disable the guard pages and fall through

          // to handle_unexpected_exception way down below.

          thread->disable_stack_red_zone();

          tty->print_raw_cr("An irrecoverable stack overflow has occurred.");

        } else {

          // Accessing stack address below sp may cause SEGV if current

          // thread has MAP_GROWSDOWN stack. This should only happen when

          // current thread was created by user code with MAP_GROWSDOWN flag

          // and then attached to VM. See notes in os_linux.cpp.

          if (thread->osthread()->expanding_stack() == 0) {

             thread->osthread()->set_expanding_stack();

             if (os::Linux::manually_expand_stack(thread, addr)) {

               thread->osthread()->clear_expanding_stack();

               return 1;

             }

             thread->osthread()->clear_expanding_stack();

          } else {

             fatal("recursive segv. expanding stack.");

          }

        }

      }

    }

    if (thread->thread_state() == _thread_in_Java) {

      // Java thread running in Java code => find exception handler if any

      // a fault inside compiled code, the interpreter, or a stub

      if (sig == SIGSEGV && os::is_poll_address((address)info->si_addr)) {

        stub = SharedRuntime::get_poll_stub(pc);

      } else if (sig == SIGBUS /* && info->si_code == BUS_OBJERR */) {

        // BugId 4454115: A read from a MappedByteBuffer can fault

        // here if the underlying file has been truncated.

        // Do not crash the VM in such a case.

        CodeBlob* cb = CodeCache::find_blob_unsafe(pc);

        nmethod* nm = cb->is_nmethod() ? (nmethod*)cb : NULL;

        if (nm != NULL && nm->has_unsafe_access()) {

          stub = StubRoutines::handler_for_unsafe_access();

        }

      }

      else

#ifdef AMD64

      if (sig == SIGFPE  &&

          (info->si_code == FPE_INTDIV || info->si_code == FPE_FLTDIV)) {

        stub =

          SharedRuntime::

          continuation_for_implicit_exception(thread,

                                              pc,

                                              SharedRuntime::

                                              IMPLICIT_DIVIDE_BY_ZERO);

#else

      if (sig == SIGFPE /* && info->si_code == FPE_INTDIV */) {

        // HACK: si_code does not work on linux 2.2.12-20!!!

        int op = pc[0];

        if (op == 0xDB) {

          // FIST

          // TODO: The encoding of D2I in i486.ad can cause an exception

          // prior to the fist instruction if there was an invalid operation

          // pending. We want to dismiss that exception. From the win_32

          // side it also seems that if it really was the fist causing

          // the exception that we do the d2i by hand with different

          // rounding. Seems kind of weird.

          // NOTE: that we take the exception at the NEXT floating point instruction.

          assert(pc[0] == 0xDB, "not a FIST opcode");

          assert(pc[1] == 0x14, "not a FIST opcode");

          assert(pc[2] == 0x24, "not a FIST opcode");

          return true;

        } else if (op == 0xF7) {

          // IDIV

          stub = SharedRuntime::continuation_for_implicit_exception(thread, pc, SharedRuntime::IMPLICIT_DIVIDE_BY_ZERO);

        } else {

          // TODO: handle more cases if we are using other x86 instructions

          //   that can generate SIGFPE signal on linux.

          tty->print_cr("unknown opcode 0x%X with SIGFPE.", op);

          fatal("please update this code.");

        }

#endif // AMD64

      } else if (sig == SIGSEGV &&

               !MacroAssembler::needs_explicit_null_check((intptr_t)info->si_addr)) {

          // Determination of interpreter/vtable stub/compiled code null exception

          stub = SharedRuntime::continuation_for_implicit_exception(thread, pc, SharedRuntime::IMPLICIT_NULL);

      }

    } else if (thread->thread_state() == _thread_in_vm &&

               sig == SIGBUS && /* info->si_code == BUS_OBJERR && */

               thread->doing_unsafe_access()) {

        stub = StubRoutines::handler_for_unsafe_access();

    }

    // jni_fast_Get<Primitive>Field can trap at certain pc's if a GC kicks in

    // and the heap gets shrunk before the field access.

    if ((sig == SIGSEGV) || (sig == SIGBUS)) {

      address addr = JNI_FastGetField::find_slowcase_pc(pc);

      if (addr != (address)-1) {

        stub = addr;

      }

    }

    // Check to see if we caught the safepoint code in the

    // process of write protecting the memory serialization page.

    // It write enables the page immediately after protecting it

    // so we can just return to retry the write.

    if ((sig == SIGSEGV) &&

        os::is_memory_serialize_page(thread, (address) info->si_addr)) {

      // Block current thread until the memory serialize page permission restored.

      os::block_on_serialize_page_trap();

      return true;

    }

  }

#ifndef AMD64

  // Execution protection violation

  //

  // This should be kept as the last step in the triage.  We don't

  // have a dedicated trap number for a no-execute fault, so be

  // conservative and allow other handlers the first shot.

  //

  // Note: We don't test that info->si_code == SEGV_ACCERR here.

  // this si_code is so generic that it is almost meaningless; and

  // the si_code for this condition may change in the future.

  // Furthermore, a false-positive should be harmless.

  if (UnguardOnExecutionViolation > 0 &&

      (sig == SIGSEGV || sig == SIGBUS) &&

      uc->uc_mcontext.gregs[REG_TRAPNO] == trap_page_fault) {

    int page_size = os::vm_page_size();

    address addr = (address) info->si_addr;

    address pc = os::Linux::ucontext_get_pc(uc);

    // Make sure the pc and the faulting address are sane.

    //

    // If an instruction spans a page boundary, and the page containing

    // the beginning of the instruction is executable but the following

    // page is not, the pc and the faulting address might be slightly

    // different - we still want to unguard the 2nd page in this case.

    //

    // 15 bytes seems to be a (very) safe value for max instruction size.

    bool pc_is_near_addr =

      (pointer_delta((void*) addr, (void*) pc, sizeof(char)) < 15);

    bool instr_spans_page_boundary =

      (align_size_down((intptr_t) pc ^ (intptr_t) addr,

                       (intptr_t) page_size) > 0);

    if (pc == addr || (pc_is_near_addr && instr_spans_page_boundary)) {

      static volatile address last_addr =

        (address) os::non_memory_address_word();

      // In conservative mode, don't unguard unless the address is in the VM

      if (addr != last_addr &&

          (UnguardOnExecutionViolation > 1 || os::address_is_in_vm(addr))) {

        // Unguard and retry

        address page_start =

          (address) align_size_down((intptr_t) addr, (intptr_t) page_size);

        bool res = os::unguard_memory((char*) page_start, page_size);

        if (PrintMiscellaneous && Verbose) {

          char buf[256];

          jio_snprintf(buf, sizeof(buf), "Execution protection violation "

                       "at " INTPTR_FORMAT

                       ", unguarding " INTPTR_FORMAT ": %s, errno=%d", addr,

                       page_start, (res ? "success" : "failed"), errno);

          tty->print_raw_cr(buf);

        }

        stub = pc;

        // Set last_addr so if we fault again at the same address, we don't end

        // up in an endless loop.

        //

        // There are two potential complications here.  Two threads trapping at

        // the same address at the same time could cause one of the threads to

        // think it already unguarded, and abort the VM.  Likely very rare.

        //

        // The other race involves two threads alternately trapping at

        // different addresses and failing to unguard the page, resulting in

        // an endless loop.  This condition is probably even more unlikely than

        // the first.

        //

        // Although both cases could be avoided by using locks or thread local

        // last_addr, these solutions are unnecessary complication: this

        // handler is a best-effort safety net, not a complete solution.  It is

        // disabled by default and should only be used as a workaround in case

        // we missed any no-execute-unsafe VM code.

        last_addr = addr;

      }

    }

  }

#endif // !AMD64

  if (stub != NULL) {

    // save all thread context in case we need to restore it

    if (thread != NULL) thread->set_saved_exception_pc(pc);

    uc->uc_mcontext.gregs[REG_PC] = (greg_t)stub;

    return true;

  }

  // signal-chaining

  if (os::Linux::chained_handler(sig, info, ucVoid)) {

     return true;

  }

  if (!abort_if_unrecognized) {

    // caller wants another chance, so give it to him

    return false;

  }

  if (pc == NULL && uc != NULL) {

    pc = os::Linux::ucontext_get_pc(uc);

  }

  // unmask current signal

  sigset_t newset;

  sigemptyset(&newset);

  sigaddset(&newset, sig);

  sigprocmask(SIG_UNBLOCK, &newset, NULL);

  VMError err(t, sig, pc, info, ucVoid);

  err.report_and_die();

  ShouldNotReachHere();

}

void os::Linux::init_thread_fpu_state(void) {

#ifndef AMD64

  // set fpu to 53 bit precision

  set_fpu_control_word(0x27f);

#endif // !AMD64

}

int os::Linux::get_fpu_control_word(void) {

#ifdef AMD64

  return 0;

#else

  int fpu_control;

  _FPU_GETCW(fpu_control);

  return fpu_control & 0xffff;

#endif // AMD64

}

void os::Linux::set_fpu_control_word(int fpu_control) {

#ifndef AMD64

  _FPU_SETCW(fpu_control);

#endif // !AMD64

}

// Check that the linux kernel version is 2.4 or higher since earlier

// versions do not support SSE without patches.

bool os::supports_sse() {

#ifdef AMD64