Andrew Pinski wrote:
> On Jan 9, 2005, at 11:51 AM, Tobias Schlüter wrote:
>> While investigating if PR 19334 was a problem in the Fortran frontend, I
>> needed the tree dumps to include information if an integer constant was
>> of the actual type I expected it to be. To this end I added code to
>> tree-pretty-print.c which adds a type suffix to an integer constant in a
>> way modeled after the Fortran way of doing this kind of things.
> I would like to second this patch, since we can find more type mismatch
> with this patch than before. Maybe only adding the suffixes if we are in
> detail mode, or adding an extra mode for this.
I have built a complete tree yesterday, and I'm currently doing testsuite runs
for this patch, and another patch for which this was required, so unlike I
said in my original mail, I will soon have tested this patch.