Re: [PATCH 11/12] Test case

[Keith Seitz <keiths at redhat dot com>]

On 04/18/2014 04:42 AM, Yao Qi wrote:
> On 02/14/2014 04:44 PM, Yao Qi wrote:
> >   - Remove hard-coded breakpoint number, but it depends on this patch
> >
> >    [PATCH] Accept convenience variable in commands -break-passcount and -break-commands
> >    https://sourceware.org/ml/gdb-patches/2014-01/msg00936.html
>
> This patch above isn't acceptable as its current shape, so I revert
> this patch back to use hard-coded breakpoint number.

Yeah, I apologize for that. Something tells me we should consider 
extending the mi_create_breakpoint API to not only return the breakpoint 
regexp but also the number of the breakpoint that was set. That would 
make things like this quite a bit easier in the future.

I really only have three (related) nits about this patch:

In gdb.trace/mi-available-children-only-cxx.exp:

+mi_gdb_test "-break-insert -a marker" "\\^done.*" \
+    "trace marker"
+

and in gdb.trace/mi-available-children-only.exp:

+mi_gdb_test "-break-insert -a marker1" "\\^done.*" \
+    "trace marker1"

[snip]

+mi_gdb_test "-break-insert -a marker2" "\\^done.*" \
+    "trace marker2"

Please use mi_create_breakpoint unless there is some feature that is not 
supported by that API.

Otherwise, this looks okay. I recommend a maintainer approve with these 
changes.

Keith