need someone to check a changeset

[Ben Abbott]

I've been trying to clear out some of the bug reports.

There is one regarding axes limits for logscale that is accompanied by a 
changeset that looks ok to me, 

As trivial as this changeset is, my c\c++ skill are so poor that I'd like 
someone to take a quick look before I push it.

        https://savannah.gnu.org/bugs/index.php?30872

For convenience, the changeset is attached.

The change fixes the roundoff to zero for the min value of the axis limits that 
occurs for the simple script below.

        m = 19:27 
        semilogy (m, exp (-m.^2)) 

Ben

changeset.patch
Description: Binary data