From mboxrd@z Thu Jan 1 00:00:00 1970 From: Joel Brobecker To: gdb-patches@sources.redhat.com Subject: Re: [RFA/RFC] New command: ``start'' Date: Fri, 21 May 2004 01:31:00 -0000 Message-id: <20040521013133.GH10684@gnat.com> References: <20040518222740.GZ10684@gnat.com> <20040518224131.GA6026@nevyn.them.org> <20040519153615.GD10684@gnat.com> <20040519154155.GA8567@nevyn.them.org> <20040520010145.GQ10684@gnat.com> <20040520134600.GA11705@nevyn.them.org> <20040520160259.GV10684@gnat.com> <20040520171404.GA3880@nevyn.them.org> <20040520221214.GC10684@gnat.com> <20040521002645.GA17522@nevyn.them.org> X-SW-Source: 2004-05/msg00624.html > This looks pretty good. I have one question: > > > + /* Check that there is a program to debug. Some languages such as Ada > > + need to search inside the program symbols for the location where to > > + put the temporary breakpoint before starting. */ > > + if (!have_full_symbols () && !have_partial_symbols ()) > > + error ("No symbol table loaded. Use the \"file\" command."); > > Shouldn't you accept have_minimal_symbols here? Hmmmm, that's a very sharp remark. Looking at our current implementation in begin_command, we indeed only rely on minimal symbols, which makes sense (we only need the address, we know how to read it afterwards). So a check against have_minimal_symbols is indeed more appropriate. Thanks for catching this. Here is an updated version. 2004-05-20 Joel Brobecker * infcmd.c (kill_if_already_running): New function, extracted from run_command(). (run_command): Replace extracted code by call to kill_if_already_running(). (start_command): New function. (_initialize_infcmd): Add "start" command. -- Joel Attachment: infcmd.c.diff Description: Text document