[Bug mi/23703] -break-delete should allow to delete by file and line

simon.marchi at ericsson dot com sourceware-bugzilla@sourceware.org
Mon Sep 24 01:22:00 GMT 2018


https://sourceware.org/bugzilla/show_bug.cgi?id=23703

Simon Marchi <simon.marchi at ericsson dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |simon.marchi at ericsson dot com

--- Comment #2 from Simon Marchi <simon.marchi at ericsson dot com> ---
(In reply to Basile B. from comment #0)
> As it worked previously when using the "clear" TUI command when it still was
> allowed in MI mode.

Ok, I just tried it and I think I know why the command seems to not work
anymore.  Some commit, I think this one:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=a9e408182d2faaed5c2457b68ea3276c719a590f

changed how breakpoint locations are reported.  If you put a breakpoint on a
line that has no code associated to it, GDB will report that the breakpoint as
on the first following line that does have code associated to it.  So I tried
putting a breakpoint on line test.c:1 and it was actually installed on line
test.c:3.  Then, I tried "clear test.c:1", which didn't work, as I had to do
"clear test.c:3".

> Expected syntax is "-break-delete --source <fname> --line <number>", to be
> consistent with the "-break-insert" equivalent.
> 
> For example if my UI at some points send
> 
> 
>     -break-insert --source foo.c --line 42
> 
> I really expect to be able to do the opposite, i.e
> 
> 
>     -break-delete --source foo.c --line 42
> 
> Actually i even think that any syntax allowed with "-break-insert" MUST have
> its "-break-delete" counterpart.

I am really not convinced of that.  The current way of handling breakpoint
deletion, by breakpoint number, is easy to implement and unambiguous.  I don't
understand what would be the advantage of doing it by explicit
location/linespec over the existing way.

-- 
You are receiving this mail because:
You are on the CC list for the bug.


More information about the Gdb-prs mailing list