Re: [PATCH] gdb: change thread_info::name to unique_xmalloc_ptr, add helper function

[Tom Tromey <tom@tromey.com>]

>>>>> "Simon" == Simon Marchi via Gdb-patches <gdb-patches@sourceware.org> writes:

Simon> This started out as changing thread_info::name to a unique_xmalloc_ptr.
Simon> That showed that almost all users of that field had the same logic to
Simon> get a thread's name: use thread_info::name if non-nullptr, else ask the
Simon> target.  Factor out this logic in a new thread_name free function.  Make
Simon> the field private (rename to m_name) and add some accessors.

Thank you.

Simon> +  /* The user-give name of the thread.

Missing an "n" on "given".

Simon> +/* Return THREAD's name.
Simon> +
Simon> +   If THREAD has a user-given name, return it.  Otherwise, query the thread's
Simon> +   target to get the name.  May return nullptr.  */
Simon> +extern const char *thread_name (thread_info *thread);

It seems to me that this should probably be the only caller of
target_thread_name.  So, what about putting the extra logic into
target_thread_name instead of having a new function?

One idea here may be to keep target_thread_name as a simple wrapper for
the target method.  But in that case, I think target_thread_name's
declaration should be updated to say something like "don't call this,
use thread_name instead".

Other than this issue, this all looks great to me.

Tom