Hi,
While writing a debugger [http://www.secretmango.com/jimb/Whit...e/ptrace.html]
After attaching to the process using ptrace(PTRACE_ATTACH....), I am unable to set my breakpoints in the target tracee , it crashes whenever we set one

But when we use ptrace SingleStep after we return from waitpid, it all works fine
I want to know why must we call ptrace Single step (causing the debugger to break on waitpid with SIGTRAP) right after the tracee has stopped (due to attaching).

I cant find any material on this.

if(WIFSTOPPED(status)){.....

user uregs1={};
ptrace(PTRACE_GETREGS,rpid,NULL,&uregs1);
printf("rip0=%lx\n",uregs1.regs.rip);fflush(stdout);

ptrace(PTRACE_SINGLESTEP, tid, 0, 0); //let it continue one step
tid=waitpid(-1, &status, __WALL);

ptrace(PTRACE_GETREGS,rpid,NULL,&uregs1);
printf("rip0=%lx\n",uregs1.regs.rip);fflush(stdout);

both print the same RIP (after executing one instruction)...this probably tells me that the first RIP value was incorrect (and now ptrace with single step is required to get the right value), probably incorrect due to the pipeline nature of a processor