Re: odd (or not so odd?) segmentation fault in 5.8.0

Message-ID: <RtQa+gzkg2kF092yn@efn.org>

p4raw-id: //depot/perl@18889

diff --git a/mg.c b/mg.c
index 941338b..98e4c09 100644 (file)
--- a/mg.c
+++ b/mg.c
@@ -133,6 +133,12 @@ Perl_mg_get(pTHX_ SV *sv)
 
        if (!(mg->mg_flags & MGf_GSKIP) && vtbl && vtbl->svt_get) {
            CALL_FPTR(vtbl->svt_get)(aTHX_ sv, mg);
+
+           /* guard against sv having been freed */
+           if (SvTYPE(sv) == SVTYPEMASK) {
+               Perl_croak(aTHX_ "Tied variable freed while still in use");
+           }
+
            /* Don't restore the flags for this entry if it was deleted. */
            if (mg->mg_flags & MGf_GSKIP)
                (SSPTR(mgs_ix, MGS *))->mgs_flags = 0;

diff --git a/pod/perldiag.pod b/pod/perldiag.pod
index 3881288..cb314e9 100644 (file)
--- a/pod/perldiag.pod
+++ b/pod/perldiag.pod
@@ -3672,6 +3672,12 @@ target of the change to
 
 (F) The entry point function of threads->create() failed for some reason.
 
+=item Tied variable freed while still in use
+
+(F) An access method for a tied variable (e.g. FETCH) did something to
+free the variable.  Since continuing the current operation is likely
+to result in a coredump, Perl is bailing out instead.
+
 =item times not implemented
 
 (F) Your version of the C library apparently doesn't do times().  I

diff --git a/t/op/tie.t b/t/op/tie.t
index 6e73cee..49c189e 100755 (executable)
--- a/t/op/tie.t
+++ b/t/op/tie.t
@@ -286,3 +286,12 @@ EXPECT
 7
 8
 0
+########
+#
+# FETCH freeing tie'd SV
+sub TIESCALAR { bless [] }
+sub FETCH { *a = \1; 1 }
+tie $a, 'main';
+print $a;
+EXPECT
+Tied variable freed while still in use at - line 6.