aboutsummaryrefslogtreecommitdiff
path: root/gcc/flow.c
diff options
context:
space:
mode:
authorBrendan Kehoe <brendan@lisa.cygnus.com>1997-10-31 23:17:35 +0000
committerBrendan Kehoe <brendan@gcc.gnu.org>1997-10-31 18:17:35 -0500
commitcf9d67e3e8aa596271353f56c49dbe6ebbe3f7ad (patch)
treedff5b59bd20c5f553511ec9e6abb17639fc45d92 /gcc/flow.c
parent27ca375af668068978d277ae1d1b92149f121c6a (diff)
downloadgcc-cf9d67e3e8aa596271353f56c49dbe6ebbe3f7ad.zip
gcc-cf9d67e3e8aa596271353f56c49dbe6ebbe3f7ad.tar.gz
gcc-cf9d67e3e8aa596271353f56c49dbe6ebbe3f7ad.tar.bz2
except.c (push_eh_info): Pass the number of fields - 1 down, not the exact number of fields.
* except.c (push_eh_info): Pass the number of fields - 1 down, not the exact number of fields. cuz in finish_builtin_type, its comment sez LEN is the number of elements in FIELDS minus one, or put another way, it is the maximum subscript used in FIELDS. and its code does for (i = 0; i < len; i++) { layout_type (TREE_TYPE (fields[i])); DECL_FIELD_CONTEXT (fields[i]) = type; TREE_CHAIN (fields[i]) = fields[i+1]; } DECL_FIELD_CONTEXT (fields[i]) = type; DECL_CLASS_CONTEXT (fields[i]) = type; thus expecting the final ones to be fields[4], not fields[5] (which is the actual size from 0, not the last field member) From-SVN: r16257
Diffstat (limited to 'gcc/flow.c')
0 files changed, 0 insertions, 0 deletions