[PATCH] Improve i386 prologue analyzer

[Andrew Cagney]

>>Date: Fri, 06 Aug 2004 16:29:47 -0400
>>> From: Andrew Cagney <cagney@gnu.org>
>>> 
>>> Ah.  We need to push the MIPS fixes out fairly quickly.  Sounds like 
>>> this should be held back, perhaphs for 6.3 or a 6.2.2.
> 
> 
> Please don't hold this back!  Producing a decent backtrace is one of
> the most important features of a debugger (think about debugging a
> crash from a core dump); a debugger that cannot reliably do that
> cannot be trusted.
> 
> I really don't understand what's to ``hold back'' here: Mark already
> made a patch, that patch was reviewed by myself and by Michael, and I
> already reported that it fixes the original problems.  So it is beyond
> me why the MIPS fixes can be in GDB 6.2.1, while the i386 prolog
> analysis patch cannot.

Mark wrote:

> I'd really like this to get some exposure in HEAD before backporting
> it, but otherwise I do agree.

Remember, at present MIPS is dead in the water.  For how long should we 
delay pushing out a fix for that?

Andrew