A TLS sendData() failing with SSL_ERROR_WANT_READ (mid-handshake
application data, or a post-handshake session ticket exchange) has
nothing to do with write-readiness, but resetSocket() kept re-arming
EPOLLWRNORM anyway - an already-writable TCP socket fires that
immediately, so the retry loop spun the CPU as fast as it could
until the connection was eventually dropped, without ever reading
the peer's first command. Confirmed via a 15GB+ debug log from one
stuck connection. sendNeedsRead() lets Socket now tell the
difference and wait for a real EPOLLIN instead of busy-retrying.
TCPSocket's SSL_CTX*-taking constructor - the one every accepted
TCPSession actually goes through (TCPSession(EPoll&, TCPServer*, ...)
passes server->ctx into it) - never initialized the `tlsInfo` member,
and TCPSocket.h gave it no default either. Every session's `tlsInfo`
was uninitialized garbage.
receiveData()/sendData() key their TLS-vs-plain dispatch on
`if(tlsInfo)`. Since that garbage value is non-zero often enough on a
real stack, sessions on a plain, non-TLS server were intermittently
routed through TLS::receiveData()/sendData() (SSL_read()/SSL_write())
against an also-uninitialized SSL*, segfaulting inside libssl. Confirmed
via dmesg: repeated "general protection fault ... in libssl.so.3",
signal 11, non-deterministic timing consistent with depending on
whatever garbage happened to be on the stack for a given session.
Fixed by dispatching on `ctx` instead of `tlsInfo`. `ctx` (TLS.h) now
defaults to NULL (previously also uninitialized) and is the value that
actually gets set for real in TLS::createContext() and correctly
inherited by every session via TLS(SSL_CTX*) - unlike `tlsInfo`, which
is only ever set on the listening TCPServer's own constructor and never
propagated to sessions at all. For the previously-working TLS-enabled
case this is a no-op: ctx and tlsInfo were always both set (or both
unset) together on the listening server, so the observable behavior
there is unchanged - only the previously-undefined session-level
behavior is fixed.
Verified live: restarted the APIServer test instance (a plain, non-TLS
TCPServer - the exact configuration that was crashing), ran 100
sequential requests with no failures, and confirmed no new libssl
segfaults in dmesg since the fix (three separate pre-fix crash
incidents are all still visible in the dmesg history for comparison).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The previous commit picked up CRLF line endings from a Windows-side
working copy used during development. Converting to plain LF to match
the rest of this Linux-native repo - content unchanged, this is purely
a line-ending normalization (932 insertions/932 deletions, 1:1).
TCPSocket inherits from both Socket and TLS, both declaring receiveData()
as unrelated sibling methods - only Socket::receiveData was virtual, so
TLS::receiveData (SSL_read) was never reached regardless of TLS config,
and even when reached its result was never forwarded to onDataReceived
(that call was commented out). Added an explicit TCPSocket::receiveData
override dispatching to TLS::receiveData or Socket::receiveData based on
tlsInfo, and completed TLS::receiveData to actually return/forward the
decrypted bytes.
Write path did not exist at all. Added Socket::sendData() as a new
virtual hook mirroring receiveData, routed Socket::writeSocket() through
it instead of a raw ::write(), added TLS::sendData() (SSL_write, same
error-handling style as the existing TLS methods), and TCPSocket::sendData()
dispatching the same way the read side does.
Also fixed: a queued write in writeSocket() was being popped off the fifo
as if sent even when shutDown was true and nothing was actually written;
removed a dead empty if(tlsInfo){} stub in onDataReceived, no longer
needed now dispatch happens one level up; replaced leftover debug
std::cout lines with proper Log() calls matching the rest of the file.
Verified via full local compile: all 20 .cpp files build clean, link
into libServerCore.a, and nm confirms the new/fixed symbols are correctly
defined in the right object files.
Brad Arant