Bug: Non-fatal assertion failed in process_win32_read_from_handle
On win32, latest Tor 0.4.6.10 when starting with torrc consisting of:
ClientTransportPlugin trebuchet exec /path/to/snowflake-client.exe
Bridge trebuchet 127.0.0.1:123
Logs the following:
Jan 1 00:00:00.000 [notice] Bootstrapped 0% (starting): Starting
Jan 1 00:00:00.000 [notice] Starting with guard context "default"
Jan 1 00:00:00.000 [warn] Client managed proxy encountered a method error. (trebuchet no such method)
Jan 1 00:00:00.000 [warn] Managed proxy at '/path/to/snowflake-client.exe' failed the configuration protocol and will be destroyed.
Jan 1 00:00:00.000 [warn] tor_bug_occurred_(): Bug: process_win32.c:891: process_win32_read_from_handle: Non-fatal assertion !(handle->reached_eof) failed. (on Tor 0.4.6.10 22fd351cf582aa2b)
Jan 1 00:00:00.000 [warn] Bug: Tor 0.4.6.10 (git-22fd351cf582aa2b): Non-fatal assertion !(handle->reached_eof) failed in process_win32_read_from_handle at process_win32.c:891. (Stack trace not available) (on Tor 0.4.6.10 22fd351cf582aa2b)
Jan 1 00:00:00.000 [warn] Pluggable Transport process terminated with status code 0
Jan 1 00:00:00.000 [notice] Bootstrapped 5% (conn): Connecting to a relay
Interestingly, bug doesn't happen if I put obfs4proxy instead of snowflake-client, so something additional triggers it in Snowflake (version from latest TBB). But I think tor mustn't log Bug's no matter how broken PT is.
Edited by David Goulet