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] Cannot set pending bp if condition set explicitly


>>>>> "Pedro" == Pedro Alves <palves@redhat.com> writes:

Pedro> An alternative would be to restrict pending breakpoints conditions
Pedro> to globals only.  Not sure whether that would fly.

Yeah, I'd rather not.

For example, it would mean that a breakpoint that examines function
arguments would work fine if the function was in the main program; but
if you refactored the code to move it to a shared library, it would not
work.

Also it seems to me that you'd have to add extra semantics to decide
what to do in the re-run case, where a conditional breakpoint is set in
a .so that is loaded at some point during the run.  Here, won't the
breakpoint convert itself to pending, and then back when the .so is
loaded?  Would this then mean that the expression is interpreted
differently?

Tom


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