My take was it depends on host details.  For example, I didn't track down
the Windows way of getting this data, but I wouldn't be surprised if it
turned into a method call to the system (which could be cached, but that's
an impl detail that a method call would allow us to hide).

I thought it made it more flexible on unknown details for current/future
platforms to have it be a function call.


On Fri, Jul 11, 2014 at 8:52 AM, Ed Maste <[email protected]> wrote:

> On 11 July 2014 11:04, Todd Fiala <[email protected]> wrote:
> > Hi all,
> >
> > This patch looks to provide a non-#ifdef manner to help choose better
> > debugger-defined thread names when operating on platforms that don't
> support
> > long thread names.  We have code that attempts to pick a smarter short
> name
> > when a longer thread name is too long to set, but that can fail too when
> the
> > shortened variant is still too long.
>
> I'm happy with this change, although it could be done as a
> compile-time constant instead, no?
>



-- 
-Todd
_______________________________________________
lldb-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits

Reply via email to