From: Guy Harris <gharris@xxxxxxxxx>
Subject: Re: [Ethereal-dev] Help please - CVS ethereal blows up everytime
Date: Thu, 4 Apr 2002 02:15:49 -0800
> If it does, fix that dissector not to use the return value of an
> "proto_tree_add_" routine as a "proto_tree *"; yes, "proto_tree" and
> "proto_item" happen to be typedeffed to the same value, but that doesn't
> mean they're the same sort of item - dissectors should assign an ett_
> value to all subtrees, and use "proto_item_add_subtree()" to convert a
> "proto_item *" to a "proto_tree *".
Ok, I have fixed packet-iscsi.c as you suggested and it has stopped
blowing up (diff attached).
This must be a new compatibility issue as the iSCSI dissector has not
changed in this respect. It seems to be affecting other dissectors as
well. For example, if I just capture some stuff off the local ethernet
it crashes and gdb says:
#4 0x081f37cf in alloc_field_info (tree=0x84abe5c, hfindex=139100576,
tvb=0x81f1d7c, start=139100576, length=0x846f0a4) at proto.c:1624
#5 0x081f2519 in proto_tree_add_none_format (tree=0x84abe5c, hfindex=6080,
tvb=0x846f0a4, start=8, length=4, format=0x1 <Address 0x1 out of bounds>)
at proto.c:688
#6 0x08188451 in windowAttributes (tvb=0x846f0a4) at packet-x11.c:1447
#7 0x08188880 in dissect_x11_request_loop (tvb=0x846f0a4, root=0x84ac104)
at packet-x11.c:1548
#8 0x0818c77f in dissect_x11 (tvb=0x846f0a4, pinfo=0x847bce0, tree=0x84abbc8)
at packet-x11.c:2894
I guess we can expect lots of breakage until everyone fixes up their
dissectors.
Mark
Index: packet-iscsi.c
===================================================================
RCS file: /cvsroot/ethereal/packet-iscsi.c,v
retrieving revision 1.28
diff -u -3 -p -r1.28 packet-iscsi.c
--- packet-iscsi.c 2002/04/04 10:20:24 1.28
+++ packet-iscsi.c 2002/04/04 23:04:48
@@ -181,6 +181,7 @@ static int hf_iscsi_BegRun = -1;
static int hf_iscsi_RunLength = -1;
/* Initialize the subtree pointers */
+static gint ett_iscsi = -1;
static gint ett_iscsi_KeyValues = -1;
static gint ett_iscsi_CDB = -1;
static gint ett_iscsi_Flags = -1;
@@ -719,7 +720,7 @@ static void
dissect_iscsi_pdu(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, guint offset, guint8 opcode, const char *opcode_str, guint32 data_segment_len) {
guint original_offset = offset;
- proto_item *ti = NULL;
+ proto_tree *ti = NULL;
char *scsi_command_name = NULL;
guint8 scsi_status = 0;
guint cdb_offset = offset + 32; /* offset of CDB from start of PDU */
@@ -861,11 +862,12 @@ dissect_iscsi_pdu(tvbuff_t *tvb, packet_
/* In the interest of speed, if "tree" is NULL, don't do any
work not necessary to generate protocol tree items. */
if (tree) {
-
+ proto_item *tp;
/* create display subtree for the protocol */
- ti = proto_tree_add_protocol_format(tree, proto_iscsi, tvb,
+ tp = proto_tree_add_protocol_format(tree, proto_iscsi, tvb,
offset, -1, "iSCSI (%s)",
(char *)opcode_str);
+ ti = proto_item_add_subtree(tp, ett_iscsi);
proto_tree_add_uint(ti, hf_iscsi_Opcode, tvb,
offset + 0, 1, opcode);
@@ -2040,6 +2042,7 @@ proto_register_iscsi(void)
/* Setup protocol subtree array */
static gint *ett[] = {
+ &ett_iscsi,
&ett_iscsi_KeyValues,
&ett_iscsi_CDB,
&ett_iscsi_Flags,
This message has been 'sanitized'. This means that potentially
dangerous content has been rewritten or removed. The following
log describes which actions were taken.
Sanitizer (start="1017961684"):
Replaced MIME boundary: >>--Next_Part--<<
with: >>MIMEStream=_0+254952_23195930223765_936149043<<
Writer (pos="1559"):
Total modifications so far: 1
Part (pos="1605"):
SanitizeFile (filename="unnamed.txt", mimetype="Text/Plain"):
Match (rule="2"):
Enforced policy: accept
Part (pos="3291"):
SanitizeFile (filename="foo.diff", mimetype="Text/Plain"):
Match (rule="default"):
Enforced policy: accept
Anomy 0.0.0 : Sanitizer.pm
$Id: Sanitizer.pm,v 1.32 2001/10/11 19:27:15 bre Exp $
