When investigating PR 40464 I have come across a declaration with DECL_HAS_VALUE_EXPR_P and DECL_VALUE_EXPR set (which indeed turned out to be the root of the problem). However, according to a comment in tree.h: "once this field has been set, the decl itself may not legitimately appear in the function."
Richard Guenther has written a patch to verify there are none such declarations in the function statements and found none on an x86_64 (and so did I). However, when I run the test on a hppa I got over two hundred failures in the c testsuite alone. The patch is at http://gcc.gnu.org/ml/gcc-patches/2009-08/msg00419.html Steps to reproduce: 1) apply the patch above, 2) compile gcc (probably irrespective of configuration but see below) and 3) run the c testsuite. For the sake of completeness, I have configured gcc with: /home/jamborm/gcc/trunk/configure --prefix=/home/jamborm/gcc/inst/trunk/ --enable-checking=yes --enable-languages=c,c++ --disable-bootstrap --with-mpfr=/opt/cfarm/mpfr-2.3.2/ --disable-multilib -- Summary: hppa has DECL_VALUE_EXPR decls appearing in the function Product: gcc Version: 4.5.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: middle-end AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: jamborm at gcc dot gnu dot org GCC host triplet: hppa-linux-gnu GCC target triplet: hppa-linux-gnu http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41250
