This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: SSA usage question
- From: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- To: dnovillo at redhat dot com
- Cc: gcc at gcc dot gnu dot org
- Date: Mon, 7 Jun 04 07:45:28 EDT
- Subject: Re: SSA usage question
[ It has been impossible for me to recreate the thread of this
discussion. Kenner, please fix your mail program! ]
That's on my list. I have to be sure I have the right sources to
start from first.
GIMPLE is not supposed to be powerful. It's supposed to be really
simplistic to cater to the optimizers. You are supposed to expose
everything in excruciating detail.
That's fine for scalars, but for aggregates, the whole point is that the
structure of the access is what contains the information. If you break it
down, you lose that information.
With scalars, if you have an extra copy, so what.
With aggregates, the copy might be prohibitively expensive, involve variable
size temporaries (which the gimplifier doesn't support) or not be permitted
by the language (e.g., limited types in Ada).