Comment # 2
on bug 8515
from Guy Harris
Note that there are a number of ways in which the P1 dissector code can be
entered:
o through anybody who gets the dissector handle for the "p1" dissector;
o through anything that calls dissect_p1_mts_apdu();
o through any of the exported dissectors, where they're called from *other*
dissectors, such as the P7 dissector;
o through the "p1.extension" dissector table (which could mean re-entering
the P1 dissector code from the P1 dissector code);
so there are a number of places where the state information - whether stored as
global data or in an on-the-stack or allocated data structure - would need to
be initialized, and there is at least one place where they should *not* be
re-initialized, namely if the P1 dissector ends up processing an extension and
calling one of its own handlers for the "p1.extension" dissector table.
You are receiving this mail because:
- You are watching all bug changes.