From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Kettenis To: msnyder@cygnus.com Cc: gdb-patches@sources.redhat.com Subject: Re: [PATCH]: Make Linux use the new unified x86 watchpoint support Date: Tue, 27 Mar 2001 00:46:00 -0000 Message-id: <200103270845.f2R8jHr21772@debye.wins.uva.nl> References: <200103212123.f2LLNUh30080@delius.kettenis.local> <3ABFF768.B557AD29@cygnus.com> X-SW-Source: 2001-03/msg00485.html Date: Mon, 26 Mar 2001 18:14:00 -0800 From: Michael Snyder Mark Kettenis wrote: > > FYI, I checked this in. HJ can finally be happy now (although things > probably won't work correctly for multithreaded programs). Mark, this breaks remote i386 targets debugged from linux hosts. STOPPED_BY_WATCHPOINT is unconditionally defined to a function in i386-nat.c, but we may not be debugging a native target. STOPPED_BY_WATCHPOINT always has been unconditionally defined. The only difference I can see is that with the old stuff is that I now call perror_with_name if the ptrace call fails. If I'm right, the attached patch should fix your problems. I cannot test it right now (no access to a Linux/x86 box), but if it works for you, feel free to check it in. More details in the reply to your other message. Mark Index: ChangeLog from Mark Kettenis * i386-linux-nat.c (i386_linux_dr_get): Return 0 if ptrace call fails instead of calling perror_with_name. This should fix debugging remote i386 targets with a native Linux/x86 GDB. Add FIXME for this hack. Index: i386-linux-nat.c =================================================================== RCS file: /cvs/src/src/gdb/i386-linux-nat.c,v retrieving revision 1.24 diff -u -r1.24 i386-linux-nat.c --- i386-linux-nat.c 2001/03/21 21:22:48 1.24 +++ i386-linux-nat.c 2001/03/27 08:40:06 @@ -712,11 +712,20 @@ one thread. */ tid = PIDGET (inferior_pid); + /* FIXME: kettenis/2001-03-27: Calling perror_with_name if the + ptrace call fails breaks debugging remote targets. The correct + way to fix this is to add the hardware breakpoint and watchpoint + stuff to the target vectore. For now, just return zero if the + ptrace call fails. */ errno = 0; value = ptrace (PT_READ_U, tid, offsetof (struct user, u_debugreg[regnum]), 0); if (errno != 0) +#if 0 perror_with_name ("Couldn't read debug register"); +#else + return 0; +#endif return value; }