Re: [patch 2/5] allow pending tracepoint

[Pedro Alves <pedro at codesourcery dot com>]

On Tuesday 15 November 2011 07:43:27, Yao Qi wrote:
> Existing breakpoint infrastructure in gdb makes easier to do pending
> tracepoint.  We just create right type (tracepoint) for pending state.

> Note that we don't check/validate SALs for pending fast trace point,
> because inferior may not be ready to access at that moment.

Hmm, confusing comment.  You don't validate the SALs for pending
tracepoints because a pending tracepoint has only one dummy sal,
and it makes no sense to validate a dummy sal.

> -      struct breakpoint *b;
> +      struct breakpoint *b = NULL;

Unnecessary initialization.

Okay.

-- 
Pedro Alves