 Nick Mathewson
|
496e414e52
Basic RAM poisoning and magic-checking to notice connection and circuit
|
22 years ago |
 Roger Dingledine
|
3456adec31
adding the reattach-after-timeout feature wasn't so easy after all.
|
22 years ago |
|
Nick Mathewson
|
bd14023afd
Propagate yes/no/maybe a little farther forward.
|
22 years ago |
|
Roger Dingledine
|
afdaff63ef
turn some knobs, add more debugging
|
22 years ago |
|
Roger Dingledine
|
3516ae1a26
add an ap_conn connecting state; report connecting delay
|
22 years ago |
|
Roger Dingledine
|
53061b3778
fix endian issue: rh.integrity was getting sent wrong
|
22 years ago |
|
Roger Dingledine
|
8085235e5f
make loglevel info less noisy
|
22 years ago |
|
Roger Dingledine
|
f3fdbadfdf
randomize the initial circ_id and stream_id, so an adversary who
|
22 years ago |
|
Roger Dingledine
|
71739b486a
back out the insert-padding-if-conflict code
|
22 years ago |
|
Roger Dingledine
|
43fcb70bad
checkpoint: revamp relay cell packaging and handling
|
22 years ago |
|
Roger Dingledine
|
a5e6ec244f
rename circ_id_t to uint16_t for code clarity
|
22 years ago |
|
Roger Dingledine
|
45a66c6955
make a relay_header_t struct and pack/unpack funcs
|
22 years ago |
|
Roger Dingledine
|
6a19e64066
remove trailing whitespace
|
22 years ago |
|
Roger Dingledine
|
5ecd6b6bad
make fetch_from_buf_http malloc its strings rather
|
22 years ago |
|
Roger Dingledine
|
aba237e3e2
end-to-end integrity checking now works
|
22 years ago |
|
Roger Dingledine
|
21cc01299b
infrastructure for integrity-checks in relay cells
|
22 years ago |
|
Roger Dingledine
|
4e50f79b11
resolve an edge case in get_unique_circ_id_by_conn
|
22 years ago |
|
Roger Dingledine
|
8712a30e91
move cell size to 512 bytes
|
22 years ago |
|
Roger Dingledine
|
961ecf7abf
add H(K|1) to the onionskin reply
|
22 years ago |
|
Nick Mathewson
|
4885e90490
Make compile warning-free on cygwin
|
22 years ago |
|
Roger Dingledine
|
08534a6271
start tracking the 'It appears I've already sent the end' warning
|
22 years ago |
|
Roger Dingledine
|
05c05d6444
more friendly warning when there are too many dns workers
|
22 years ago |
|
Roger Dingledine
|
7a3a3ae1df
bugfix: a circuit that immediately failed still counts as a failed circuit
|
22 years ago |
|
Roger Dingledine
|
9c66e2bf9a
if >=2 circs are being built that handle a given stream,
|
22 years ago |
|
Roger Dingledine
|
38ce03d0e7
close the circuit when we get a truncated cell
|
22 years ago |
|
Roger Dingledine
|
d172fdd300
fix vicious bug in connection_ap_attach_pending that caused it to
|
22 years ago |
|
Roger Dingledine
|
975bb68010
simplify: options.OnionRouter==1 iff options.ORPort>0
|
22 years ago |
|
Roger Dingledine
|
f5994416c4
bugfix and cleanups
|
22 years ago |
|
Nick Mathewson
|
fd07872a72
Separate failure-count tracking from circuit-launching.
|
22 years ago |
|
Roger Dingledine
|
25ac8857f3
bugfix: sometimes we closed a circ while cpuworker was cranking,
|
22 years ago |
