Message ID | yddy47euyvj.fsf@CeBiTec.Uni-Bielefeld.DE |
---|---|
State | New |
Headers | show |
On Fri, May 13, 2016 at 10:48 AM, Rainer Orth <ro@cebitec.uni-bielefeld.de> wrote: > When running the full testsuite with -mcmodel=medium, several tests FAIL > with a NULL pointer dereference as described in the PR. > > Since these calls ultimately originate from get_named_section(NULL_TREE, > ...) calls, the obvious fix seems the right thing to do. > > Bootstrapped without regressions on x86_64-pc-linux-gnu, followed by an > additional testsuite run with -mcmodel=medium. The only changes present > are those reported in PR target/71097, and I've verified that they also > exist on the gcc-6 branch, i.e. without my patch. > > Ok for mainline? > > Rainer > > > 2016-05-12 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE> > > PR target/71080 > * config/i386/i386.c (ix86_in_large_data_p): Guard against NULL exp. OK. Thanks, Uros.
# HG changeset patch # Parent df465f483742f325ab7a60c2058cd31681949212 Fix SEGV in ix86_in_large_data_p (PR target/71080) diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c --- a/gcc/config/i386/i386.c +++ b/gcc/config/i386/i386.c @@ -6823,6 +6823,9 @@ ix86_in_large_data_p (tree exp) if (ix86_cmodel != CM_MEDIUM && ix86_cmodel != CM_MEDIUM_PIC) return false; + if (exp == NULL_TREE) + return false; + /* Functions are never large data. */ if (TREE_CODE (exp) == FUNCTION_DECL) return false;