Question on internal_get_tmp_var

[Richard Henderson]

On Wed, Jun 23, 2004 at 06:27:13PM -0400, Richard Kenner wrote:
> I'm not quite sure I understand what "1)" in front of the function means
> or why it claims it can't be used for SAVE_EXPR, but it is indeed used
> for SAVE_EXPR.

The comment actually applies to get_formal_tmp_var, not to
internal_get_tmp_var.  Tsk, tsk, Jason.

> Also, why isn't TREE_READONLY set on such a variable?

TREE_READONLY for scalar values doesn't help anything.  I wouldn't
have expected this to get used for aggregates very often.

> Finally, is there any good reason why DECL_INITIAL couldn't be set to the
> initializing value?

DECL_INITIAL isn't gimple.  Notice that we *clear* DECL_INITIAL
in gimplify_decl_stmt.


r~