This is the mail archive of the gdb-patches@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Remote debugging without a binary (regression)


Hi Gary,

On 02/18/2016 10:30 AM, Gary Benson wrote:
Hi Luis,

Luis Machado wrote:
The following patch guards such a call to prevent those errors
from disrupting a potential debugging session, and display only
a warning.

After looking into this it seems your patch is the right way to
go.  There is a separate bug in exec_file_locate_attach in that
in some cases the sysroot is ignored, but fixing that will not
remove the need for what you are proposing.

My only nit is that I would prefer the TRY-CATCH block smaller,
covering just exec_file_locate_attach, like so:

   /* If no main executable is currently open then attempt to
      open the file that was executed to create this inferior.  */
   if (try_open_exec && get_exec_file (0) == NULL)
     {
       /* exec_file_locate_attach may throw an error...
       TRY
         {
           exec_file_locate_attach (pid, 1);
         }
       CATCH...

Other than that I am ok with this change.

Thanks for tracking this down.

No problem. I'll come up with a new version of the patch and, while at it, i'll try Pedro's suggestion for the testcase.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]