[PATCH v3 4/4] gdb/mi: add a '--force-condition' flag to the '-break-insert' cmd

[Simon Marchi]

> diff --git a/gdb/testsuite/gdb.mi/mi-break.exp b/gdb/testsuite/gdb.mi/mi-break.exp
> index e46a6cc0ec6..5f1c6810389 100644
> --- a/gdb/testsuite/gdb.mi/mi-break.exp
> +++ b/gdb/testsuite/gdb.mi/mi-break.exp
> @@ -408,6 +408,30 @@ proc_with_prefix test_explicit_breakpoints {} {
>  	".*Source filename requires function, label, or line offset.*"
>  }
>  
> +# Test forcing an invalid condition.
> +
> +proc_with_prefix test_forced_conditions {} {
> +    set warning ".*warning: failed to validate condition .* disabling.*"
> +
> +    # Define a plain breakpoint first, and a condition later.
> +    mi_create_breakpoint "callme" "define a bp" ""
> +    mi_gdb_test "-break-condition --force \$bpnum bad" \
> +        "${warning}\\^done" \
> +        "invalid condition is forced"

This literally sends

  -break-condition --force $bpnum bad

But I don't understand what $bpnum refers to, can you explain?  I'm a bit lost.

> +
> +    set loc [mi_make_breakpoint_loc -enabled "N"]
> +    set args [list -cond "bad" -locations "\\\[$loc\\\]"]
> +    set bp [eval mi_make_breakpoint_multi $args]
> +
> +    mi_gdb_test "-break-insert -c bad --force-condition callme" \
> +	"${warning}\\^done,$bp" \
> +	"breakpoint with forced condition"

👍 (are we allowed emojis here?)

> +
> +    mi_gdb_test "-dprintf-insert -c bad --force-condition callme 123" \
> +        "${warning}\\^done,$bp" \
> +        "dprintf with forced condition"

It's really just a nit because it doesn't affect the test, but a dprintf
would normally take a format string, so I guess you meant:

  -dprintf-insert -c bad --force-condition callme "%d" 123

Or something like that.

So the patch pretty much LGTM, except the --force vs -force issue
mentioned in my other message.

Simon