tools/firewire: nosy-dump: work around segfault in decode_fcp
If I run "nosy-dump --view=transaction" with my camcorder on battery
instead of mains, it segfaults very quickly because of !t->request.
Perhaps this is because of increased likelyhood of incomplete
transactions (ack_busy when host writes to camcorder's FCP_Request)
and a bug deeper in nosy-dump's transaction housekeeping. This is a
quick workaround to get me going.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
diff --git a/tools/firewire/nosy-dump.c b/tools/firewire/nosy-dump.c
index 0bad0a4..1d4dd54 100644
--- a/tools/firewire/nosy-dump.c
+++ b/tools/firewire/nosy-dump.c
@@ -230,6 +230,11 @@
struct subaction *sa;
int i;
+ if (!t->request) {
+ printf("BUG in handle_transaction\n");
+ return;
+ }
+
for (i = 0; i < array_length(protocol_decoders); i++)
if (protocol_decoders[i].decode(t))
break;