Re: Code size regressions, mainline+branch

[Jan Hubicka <jh at suse dot cz>]

> 	I assume that you are referring to x86 executable size.  Can you
> narrow the size regression any further?  Was binutils updated?

Andreas didn't made any changes on setup of machine and the cfg-branch merger
provoked same behaviour on the branch as well, so it is not configuration
change.  Still I am feeling puzzled by what happends, hope to find time to
debug it sometime later this week.

Note that the size is non-stripped, so possibly only debug info got
bloated but I didn't find anything that may be responsible for that in
the diffs as well.

Honza
> 
> 	Nothing in the first list of patch should have any effect unless
> something strange is happening with unrolling sibling calls.  The second
> list of patches also includes a sibcall change.
> 
> 	The expand_expr change correctly makes the values assigned from
> the string sign-extended.  Is the x86 port somehow choosing a different,
> longer x86 instruction because it does not think the signed char fits in a
> byte?
> 
> David