Michael Mann
changed
bug 8326
| What |
Removed |
Added |
| Attachment #10038 is obsolete |
|
1
|
| Attachment #10039 is obsolete |
|
1
|
| Attachment #10040 is obsolete |
|
1
|
| Attachment #10042 is obsolete |
|
1
|
| Attachment #10043 is obsolete |
|
1
|
| Attachment #10043 Flags |
review_for_checkin?
|
|
Comment # 13
on bug 8326
from Michael Mann
Created attachment 10046 [details]
RTP Proxy v4
Updated v3 with "removed tree handling". All columns should be "tree
independent", so the COL_INFO stuff in the switch statement shouldn't be under
a tree either. In the interest of simplicity I just removed the tree check
entirely. You could add it back for individual cases in the switch statement
(AFTER the COL_INFO calls), but I don't think the performance benefit is
necessarily worth the effort.
Also, has this been fuzztested? I don't think "fail" cases have been properly
accounted for, like when tvb_find_guint8() returns -1 to 'offset' (which is a
guint32). See packet-bitcoin.c for examples as it has to handle the same
issue.
You are receiving this mail because:
- You are watching all bug changes.
