From 82a2edfbcf9e89d869c4f60d82c69bc4851aa4e3 Mon Sep 17 00:00:00 2001 From: Jim Kingdon Date: Thu, 3 Mar 1994 01:00:57 +0000 Subject: * breakpoint.c, breakpoint.h, c-valprint.c, ch-valprint.c, cp-valprint.c, eval.c, expprint.c, findvar.c, language.c, objfiles.h, infcmd.c, printcmd.c, stack.c, typeprint.c, valarith.c, valops.c, valprint.c, value.h, values.c: Replace value with value_ptr. This is for the ptx compiler. * objfiles.h, target.h: Don't declare a "sec_ptr" field using a "sec_ptr" typedef. * symm-nat.c: Add a bunch of stuff for symmetry's ptrace stuff. #if 0 i386_float_info. * symm-tdep.c (round): Remove. Also remove sgttyb. * symm-tdep.c: Remove lots of stuff which duplicates stuff from i386-tdep.c. Remove register_addr and ptx_coff_regno_to_gdb. * i386-tdep.c (i386_frame_find_saved_regs): Put in I386_REGNO_TO_SYMMETRY check in case it is needed for Dynix someday. * config/i386/nm-symmetry.h: Change KERNEL_U_ADDR. Move stuff from PTRACE_READ_REGS, PTRACE_WRITE_REGS macros to symm-nat.c. Define CHILD_WAIT and declare child_wait(). * config/i386/tm-symmetry.h: Remove call function stuff; stuff in tm-i386v.h is apparently OK. * config/i386/xm-symmetry.h [_SEQUENT_]: Define HAVE_TERMIOS not HAVE_TERMIO. Define MEM_FNS_DECLARED, NEED_POSIX_SETPGID, and USE_O_NOCTTY. --- gdb/symm-nat.c | 503 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 492 insertions(+), 11 deletions(-) (limited to 'gdb/symm-nat.c') diff --git a/gdb/symm-nat.c b/gdb/symm-nat.c index 752263d..f910d7d 100644 --- a/gdb/symm-nat.c +++ b/gdb/symm-nat.c @@ -1,5 +1,5 @@ /* Sequent Symmetry host interface, for GDB when running under Unix. - Copyright 1986, 1987, 1989, 1991, 1992 Free Software Foundation, Inc. + Copyright 1986, 1987, 1989, 1991, 1992, 1994 Free Software Foundation, Inc. This file is part of GDB. @@ -24,13 +24,20 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include "frame.h" #include "inferior.h" #include "symtab.h" +#include "target.h" +/* FIXME: What is the _INKERNEL define for? */ +#define _INKERNEL #include +#undef _INKERNEL +#include #include #include +#include #include #include #include +#include #include "gdbcore.h" #include #include @@ -61,7 +68,7 @@ int regno; regs.pr_fpa.fpa_regs[i] = *(int *)®isters[REGISTER_BYTE(FP1_REGNUM+i)]; } - PTRACE_WRITE_REGS (inferior_pid, (PTRACE_ARG3_TYPE) ®s); + mptrace (XPT_WREGS, inferior_pid, (PTRACE_ARG3_TYPE) ®s, 0); } void @@ -74,7 +81,7 @@ fetch_inferior_registers (regno) registers_fetched (); - PTRACE_READ_REGS (inferior_pid, (PTRACE_ARG3_TYPE) ®s); + mptrace (XPT_RREGS, (pid), (regaddr), 0); *(int *)®isters[REGISTER_BYTE(EAX_REGNUM)] = regs.pr_eax; *(int *)®isters[REGISTER_BYTE(EBX_REGNUM)] = regs.pr_ebx; *(int *)®isters[REGISTER_BYTE(ECX_REGNUM)] = regs.pr_ecx; @@ -146,7 +153,7 @@ struct pt_regset ep; for (i = 9; i >= 0; i--) printf_unfiltered ("%02x", ep.pr_fpu.fpu_stack[fpreg][i]); - i387_to_double (ep.pr_fpu.fpu_stack[fpreg], (char *)&val); + i387_to_double ((char *)ep.pr_fpu.fpu_stack[fpreg], (char *)&val); printf_unfiltered (" %g\n", val); } if (ep.pr_fpu.fpu_rsvd1) @@ -203,24 +210,46 @@ unsigned int pcr; if (pcr_tmp & FPA_PCR_20MHZ) printf_unfiltered(" 20MHZ"); if (pcr_tmp & FPA_PCR_CC_Z) printf_unfiltered(" Z"); if (pcr_tmp & FPA_PCR_CC_C2) printf_unfiltered(" C2"); + + /* Dynix defines FPA_PCR_CC_C0 to 0x100 and ptx defines + FPA_PCR_CC_C1 to 0x100. Use whichever is defined and assume + the OS knows what it is doing. */ +#ifdef FPA_PCR_CC_C1 if (pcr_tmp & FPA_PCR_CC_C1) printf_unfiltered(" C1"); - switch (pcr_tmp) { - case FPA_PCR_CC_Z: +#endif +#ifdef FPA_PCR_CC_C0 + if (pcr_tmp & FPA_PCR_CC_C1) printf_unfiltered(" C0"); +#endif + + switch (pcr_tmp) + { + case FPA_PCR_CC_Z: printf_unfiltered(" (Equal)"); break; - case FPA_PCR_CC_C1: +#ifdef FPA_PCR_CC_C1 + case FPA_PCR_CC_C1: +#endif +#ifdef FPA_PCR_CC_C0 + case FPA_PCR_CC_C0: +#endif printf_unfiltered(" (Less than)"); break; - case 0: + case 0: printf_unfiltered(" (Greater than)"); break; - case FPA_PCR_CC_Z | FPA_PCR_CC_C1 | FPA_PCR_CC_C2: + case FPA_PCR_CC_Z | +#ifdef FPA_PCR_CC_C1 + FPA_PCR_CC_C1 +#else + FPA_PCR_CC_C0 +#endif + | FPA_PCR_CC_C2: printf_unfiltered(" (Unordered)"); break; - default: + default: printf_unfiltered(" (Undefined)"); break; - } + } printf_unfiltered("\n"); pcr_tmp = pcr & FPA_PCR_AE; printf_unfiltered("\tAE= %#x", pcr_tmp); @@ -279,6 +308,7 @@ struct pt_regset ep; } } +#if 0 /* disabled because it doesn't go through the target vector. */ i386_float_info () { char ubuf[UPAGES*NBPG]; @@ -306,3 +336,454 @@ i386_float_info () print_fpu_status(regset); print_fpa_status(regset); } +#endif + +static volatile int got_sigchld; + +/*ARGSUSED*/ +/* This will eventually be more interesting. */ +void +sigchld_handler(signo) + int signo; +{ + got_sigchld++; +} + +/* + * Signals for which the default action does not cause the process + * to die. See for where this came from (alas, we + * can't use those macros directly) + */ +#ifndef sigmask +#define sigmask(s) (1 << ((s) - 1)) +#endif +#define SIGNALS_DFL_SAFE sigmask(SIGSTOP) | sigmask(SIGTSTP) | \ + sigmask(SIGTTIN) | sigmask(SIGTTOU) | sigmask(SIGCHLD) | \ + sigmask(SIGCONT) | sigmask(SIGWINCH) | sigmask(SIGPWR) | \ + sigmask(SIGURG) | sigmask(SIGPOLL) + + +/* + * Thanks to XPT_MPDEBUGGER, we have to mange child_wait(). + */ +int +child_wait(pid, status) + int pid; + struct target_waitstatus *status; +{ + int save_errno, rv, xvaloff, saoff, sa_hand; + struct pt_stop pt; + struct user u; + sigset_t set; + /* Host signal number for a signal which the inferior terminates with, or + 0 if it hasn't terminated due to a signal. */ + static int death_by_signal = 0; +#ifdef SVR4_SHARED_LIBS /* use this to distinguish ptx 2 vs ptx 4 */ + prstatus_t pstatus; +#endif + + do { + if (attach_flag) + set_sigint_trap(); /* Causes SIGINT to be passed on to the + attached process. */ + save_errno = errno; + + got_sigchld = 0; + + sigemptyset(&set); + + while (got_sigchld == 0) { + sigsuspend(&set); + } + + if (attach_flag) + clear_sigint_trap(); + + rv = mptrace(XPT_STOPSTAT, 0, (char *)&pt, 0); + if (-1 == rv) { + printf("XPT_STOPSTAT: errno %d\n", errno); /* DEBUG */ + continue; + } + + pid = pt.ps_pid; + + if (pid != inferior_pid) { + /* NOTE: the mystery fork in csh/tcsh needs to be ignored. + * We should not return new children for the initial run + * of a process until it has done the exec. + */ + /* inferior probably forked; send it on its way */ + rv = mptrace(XPT_UNDEBUG, pid, 0, 0); + if (-1 == rv) { + printf("child_wait: XPT_UNDEBUG: pid %d: %s\n", pid, + safe_strerror(errno)); + } + continue; + } + /* FIXME: Do we deal with fork notification correctly? */ + switch (pt.ps_reason) { + case PTS_FORK: + /* multi proc: treat like PTS_EXEC */ + /* + * Pretend this didn't happen, since gdb isn't set up + * to deal with stops on fork. + */ + rv = ptrace(PT_CONTSIG, pid, 1, 0); + if (-1 == rv) { + printf("PTS_FORK: PT_CONTSIG: error %d\n", errno); + } + continue; + case PTS_EXEC: + /* + * Pretend this is a SIGTRAP. + */ + status->kind = TARGET_WAITKIND_STOPPED; + status->value.sig = TARGET_SIGNAL_TRAP; + break; + case PTS_EXIT: + /* + * Note: we stop before the exit actually occurs. Extract + * the exit code from the uarea. If we're stopped in the + * exit() system call, the exit code will be in + * u.u_ap[0]. An exit due to an uncaught signal will have + * something else in here, see the comment in the default: + * case, below. Finally,let the process exit. + */ + if (death_by_signal) + { + status->kind = TARGET_WAITKIND_SIGNALED; + status->value.sig = target_signal_from_host (death_by_signal); + death_by_signal = 0; + break; + } + xvaloff = (unsigned long)&u.u_ap[0] - (unsigned long)&u; + errno = 0; + rv = ptrace(PT_RUSER, pid, (char *)xvaloff, 0); + status->kind = TARGET_WAITKIND_EXITED; + status->value.integer = rv; + /* + * addr & data to mptrace() don't matter here, since + * the process is already dead. + */ + rv = mptrace(XPT_UNDEBUG, pid, 0, 0); + if (-1 == rv) { + printf("child_wait: PTS_EXIT: XPT_UNDEBUG: pid %d error %d\n", pid, + errno); + } + break; + case PTS_WATCHPT_HIT: + fatal("PTS_WATCHPT_HIT\n"); + break; + default: + /* stopped by signal */ + status->kind = TARGET_WAITKIND_STOPPED; + status->value.sig = target_signal_from_host (pt.ps_reason); + death_by_signal = 0; + + if (0 == (SIGNALS_DFL_SAFE & sigmask(pt.ps_reason))) { + break; + } + /* else default action of signal is to die */ +#ifdef SVR4_SHARED_LIBS + rv = ptrace(PT_GET_PRSTATUS, pid, (char *)&pstatus, 0); + if (-1 == rv) + error("child_wait: signal %d PT_GET_PRSTATUS: %s\n", + pt.ps_reason, safe_strerror(errno)); + if (pstatus.pr_cursig != pt.ps_reason) { + printf("pstatus signal %d, pt signal %d\n", + pstatus.pr_cursig, pt.ps_reason); + } + sa_hand = (int)pstatus.pr_action.sa_handler; +#else + saoff = (unsigned long)&u.u_sa[0] - (unsigned long)&u; + saoff += sizeof(struct sigaction) * (pt.ps_reason - 1); + errno = 0; + sa_hand = ptrace(PT_RUSER, pid, (char *)saoff, 0); + if (errno) + error("child_wait: signal %d: RUSER: %s\n", + pt.ps_reason, safe_strerror(errno)); +#endif + if ((int)SIG_DFL == sa_hand) { + /* we will be dying */ + death_by_signal = pt.ps_reason; + } + break; + } + + } while (pid != inferior_pid); /* Some other child died or stopped */ + + return pid; +} + + + +/* This function simply calls ptrace with the given arguments. + It exists so that all calls to ptrace are isolated in this + machine-dependent file. */ +int +call_ptrace (request, pid, addr, data) + int request, pid; + PTRACE_ARG3_TYPE addr; + int data; +{ + return ptrace (request, pid, addr, data); +} + +int +call_mptrace(request, pid, addr, data) + int request, pid; + PTRACE_ARG3_TYPE addr; + int data; +{ + return mptrace(request, pid, addr, data); +} + +#if defined (DEBUG_PTRACE) +/* For the rest of the file, use an extra level of indirection */ +/* This lets us breakpoint usefully on call_ptrace. */ +#define ptrace call_ptrace +#define mptrace call_mptrace +#endif + +void +kill_inferior () +{ + if (inferior_pid == 0) + return; + /* + * Don't use PT_KILL, since the child will stop again with a PTS_EXIT. + * Just hit him with SIGKILL (so he stops) and detach. + */ + kill (inferior_pid, SIGKILL); + detach(SIGKILL); + target_mourn_inferior (); +} + +/* Resume execution of the inferior process. + If STEP is nonzero, single-step it. + If SIGNAL is nonzero, give it that signal. */ + +void +child_resume (pid, step, signal) + int pid; + int step; + int signal; +{ + errno = 0; + + if (pid == -1) + pid = inferior_pid; + + /* An address of (PTRACE_ARG3_TYPE)1 tells ptrace to continue from where + it was. (If GDB wanted it to start some other way, we have already + written a new PC value to the child.) + + If this system does not support PT_SSTEP, a higher level function will + have called single_step() to transmute the step request into a + continue request (by setting breakpoints on all possible successor + instructions), so we don't have to worry about that here. */ + + if (step) + ptrace (PT_SSTEP, pid, (PTRACE_ARG3_TYPE) 1, signal); + else + ptrace (PT_CONTSIG, pid, (PTRACE_ARG3_TYPE) 1, signal); + + if (errno) + perror_with_name ("ptrace"); +} + +#ifdef ATTACH_DETACH +/* Start debugging the process whose number is PID. */ +int +attach (pid) + int pid; +{ + sigset_t set; + int rv; + + rv = mptrace(XPT_DEBUG, pid, 0, 0); + if (-1 == rv) { + error("mptrace(XPT_DEBUG): %s", safe_strerror(errno)); + } + rv = mptrace(XPT_SIGNAL, pid, 0, SIGSTOP); + if (-1 == rv) { + error("mptrace(XPT_SIGNAL): %s", safe_strerror(errno)); + } + attach_flag = 1; + return pid; +} + +void +detach (signo) + int signo; +{ + int rv; + + rv = mptrace(XPT_UNDEBUG, inferior_pid, 1, signo); + if (-1 == rv) { + error("mptrace(XPT_UNDEBUG): %s", safe_strerror(errno)); + } + attach_flag = 0; +} + +#endif /* ATTACH_DETACH */ + +/* Default the type of the ptrace transfer to int. */ +#ifndef PTRACE_XFER_TYPE +#define PTRACE_XFER_TYPE int +#endif + + +/* NOTE! I tried using PTRACE_READDATA, etc., to read and write memory + in the NEW_SUN_PTRACE case. + It ought to be straightforward. But it appears that writing did + not write the data that I specified. I cannot understand where + it got the data that it actually did write. */ + +/* Copy LEN bytes to or from inferior's memory starting at MEMADDR + to debugger memory starting at MYADDR. Copy to inferior if + WRITE is nonzero. + + Returns the length copied, which is either the LEN argument or zero. + This xfer function does not do partial moves, since child_ops + doesn't allow memory operations to cross below us in the target stack + anyway. */ + +int +child_xfer_memory (memaddr, myaddr, len, write, target) + CORE_ADDR memaddr; + char *myaddr; + int len; + int write; + struct target_ops *target; /* ignored */ +{ + register int i; + /* Round starting address down to longword boundary. */ + register CORE_ADDR addr = memaddr & - sizeof (PTRACE_XFER_TYPE); + /* Round ending address up; get number of longwords that makes. */ + register int count + = (((memaddr + len) - addr) + sizeof (PTRACE_XFER_TYPE) - 1) + / sizeof (PTRACE_XFER_TYPE); + /* Allocate buffer of that many longwords. */ + register PTRACE_XFER_TYPE *buffer + = (PTRACE_XFER_TYPE *) alloca (count * sizeof (PTRACE_XFER_TYPE)); + + if (write) + { + /* Fill start and end extra bytes of buffer with existing memory data. */ + + if (addr != memaddr || len < (int) sizeof (PTRACE_XFER_TYPE)) { + /* Need part of initial word -- fetch it. */ + buffer[0] = ptrace (PT_RTEXT, inferior_pid, (PTRACE_ARG3_TYPE) addr, + 0); + } + + if (count > 1) /* FIXME, avoid if even boundary */ + { + buffer[count - 1] + = ptrace (PT_RTEXT, inferior_pid, + ((PTRACE_ARG3_TYPE) + (addr + (count - 1) * sizeof (PTRACE_XFER_TYPE))), + 0); + } + + /* Copy data to be written over corresponding part of buffer */ + + memcpy ((char *) buffer + (memaddr & (sizeof (PTRACE_XFER_TYPE) - 1)), + myaddr, + len); + + /* Write the entire buffer. */ + + for (i = 0; i < count; i++, addr += sizeof (PTRACE_XFER_TYPE)) + { + errno = 0; + ptrace (PT_WRITE_D, inferior_pid, (PTRACE_ARG3_TYPE) addr, + buffer[i]); + if (errno) + { + /* Using the appropriate one (I or D) is necessary for + Gould NP1, at least. */ + errno = 0; + ptrace (PT_WTEXT, inferior_pid, (PTRACE_ARG3_TYPE) addr, + buffer[i]); + } + if (errno) + return 0; + } + } + else + { + /* Read all the longwords */ + for (i = 0; i < count; i++, addr += sizeof (PTRACE_XFER_TYPE)) + { + errno = 0; + buffer[i] = ptrace (PT_RTEXT, inferior_pid, + (PTRACE_ARG3_TYPE) addr, 0); + if (errno) + return 0; + QUIT; + } + + /* Copy appropriate bytes out of the buffer. */ + memcpy (myaddr, + (char *) buffer + (memaddr & (sizeof (PTRACE_XFER_TYPE) - 1)), + len); + } + return len; +} + + +void +_initialize_symm_nat () +{ +/* + * the MPDEBUGGER is necessary for process tree debugging and attach + * to work, but it alters the behavior of debugged processes, so other + * things (at least child_wait()) will have to change to accomodate + * that. + * + * Note that attach is not implemented in dynix 3, and not in ptx + * until version 2.1 of the OS. + */ + int rv; + sigset_t set; + struct sigaction sact; + + rv = mptrace(XPT_MPDEBUGGER, 0, 0, 0); + if (-1 == rv) { + fatal("_initialize_symm_nat(): mptrace(XPT_MPDEBUGGER): %s", + safe_strerror(errno)); + } + + /* + * Under MPDEBUGGER, we get SIGCLHD when a traced process does + * anything of interest. + */ + + /* + * Block SIGCHLD. We leave it blocked all the time, and then + * call sigsuspend() in child_wait() to wait for the child + * to do something. None of these ought to fail, but check anyway. + */ + sigemptyset(&set); + rv = sigaddset(&set, SIGCHLD); + if (-1 == rv) { + fatal("_initialize_symm_nat(): sigaddset(SIGCHLD): %s", + safe_strerror(errno)); + } + rv = sigprocmask(SIG_BLOCK, &set, (sigset_t *)NULL); + if (-1 == rv) { + fatal("_initialize_symm_nat(): sigprocmask(SIG_BLOCK): %s", + safe_strerror(errno)); + } + + sact.sa_handler = sigchld_handler; + sigemptyset(&sact.sa_mask); + sact.sa_flags = SA_NOCLDWAIT; /* keep the zombies away */ + rv = sigaction(SIGCHLD, &sact, (struct sigaction *)NULL); + if (-1 == rv) { + fatal("_initialize_symm_nat(): sigaction(SIGCHLD): %s", + safe_strerror(errno)); + } +} -- cgit v1.1