Re: [RFA]: Modified Watchthreads Patch

[Jeff Johnston <jjohnstn@redhat.com>]

Eli Zaretskii wrote:
>>Date: Thu, 09 Dec 2004 18:36:13 -0500
>>From: Jeff Johnston <jjohnstn@redhat.com>
>>
>>The following is a modified version of my thread watchpoint patch from 
>>October/November.  It removes the code I had used to switch between lwp ptids 
>>and thread ptids now that Daniel's lwp patch is in place.  It uses the former 
>>version of my observer that is linux-specific and is activated in attach_thread 
>>in linux-thread-db.c.  Eli, I renamed the observer as asked to indicate this.
> 
> 
> Thanks.
> 
> 
>>         * breakpoint.c (insert_watchpoints_for_new_thread): New function.
>>         (print_it_typical): Do not issue an error for bp_thread_event
>>         if a subsequent event is on the chain.
>>         * breakpoint.h (insert_watchpoints_for_new_thread): New prototype.
> 
> 
> Hmm... the new function insert_watchpoints_for_new_thread is called
> only by ia64_linux_new_thread.  Is there any policy for functions that
> are only used by a single port?  Do we care that all the other GDB
> builds will get a useless function compiled into them?  Should we
> perhaps #ifdef it away conditioned on some symbol?
> 
> 
>>+@deftypefun void linux_new_thread (ptid_t @var{ptid})
>>+A new linux thread described by @var{ptid} has been officially attached
>>+to by gdb.
>>+@end deftypefun
> 
> 
> What does it mean ``officially attached''?  Can a thread be attached
> to ``unofficially''?
>

I'm referring to the act of gdb recognizing the thread.  The function itself is 
called attach_thread but it has a #ifdef governing whether a low-level ATTACH is 
required or not.  Gdb now recognizes it has "attached" to the thread whether a 
physical attach is needed or not.  I can drop the "officially" qualifier if it 
is confusing.

-- Jeff J.