This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] PR fortran/9972
- From: David Edelsohn <dje at watson dot ibm dot com>
- To: Zack Weinberg <zack at codesourcery dot com>
- Cc: Jan Hubicka <jh at suse dot cz>, gcc-patches at gcc dot gnu dot org, Toon Moene <toon at moene dot indiv dot nluug dot nl>
- Date: Tue, 13 May 2003 14:01:35 -0400
- Subject: Re: [PATCH] PR fortran/9972
>>>>> Zack Weinberg writes:
Zack> Further, I don't understand this, which seems to be the key component
Zack> of the change. Could you please explain why
Zack> (flag_inline_functions || flag_inline_trees == 0)
Zack> is the proper test?
Honza added the flag_inline_functions test to query if RTL
inliners are enabled. However, most languages are converting
flag_inline_functions into flag_inline_trees. The current test
essentially queries: "Is RTL inlining enabled". My test adds: "Is Tree
inlining disabled". We might only need the latter.
Without flag_inline_functions enabled and WITHOUT optimization,
the compilation deferral is performing dead function elimination. The
Fortran front-end is relying on that behavior because it did not emit the
global data symbols referenced by the dead function code.