While this issue is visible in mysqli_poll() functions, the cause
lays deeper in the stream to socket casting API. On Win x64 the
SOCKET datatype is a 64 or 32 bit unsigned, while on Linux/Unix-like
it's 32 bit signed integer. The game of casting 32 bit var to/from
64 bit pointer back and forth is the best way to break it.
Further more, while socket and file descriptors are always integers
on Linux, those are different things using different APIs on Windows.
Even though using integer instead of SOCKET might work on Windows, this
issue might need to be revamped more carefully later. By this time
this patch is tested well with phpt and apps and shows no regressions,
neither in mysqli_poll() nor in any other parts.
crypto method (SSLv3, SSLv2 etc.) by calling
stream_context_set_option($ctx, "ssl", "crypto_method", $crypto_method)
where $crypto_method can be one of STREAM_CRYPTO_METHOD_SSLv2_CLIENT,
STREAM_CRYPTO_METHOD_SSLv3_CLIENT, STREAM_CRYPTO_METHOD_SSLv23_CLIENT
or STREAM_CRYPTO_METHOD_TLS_CLIENT. SSLv23 remains the default crypto
method.
This change makes it possible to fopen() SSL URLs that are only
provided using SSL v3.
Per requests from users, I refactored the read / write methods and pulled out some of the common code between the new refactored method and php_openssl_enable_crypto(). Personally, I think that too much factoring can reduce readability, but it was specifically asked for.
The CRIME attack vector exploits TLS compression. This patch adds a stream context option
allowing servers to disable TLS compression for versions of OpenSSL >= 1.0.0 (which first
introduced the SSL_OP_NO_COMPRESSION option). A summary rundown of the CRIME attack can
be found at https://community.qualys.com/blogs/securitylabs/2012/09/14/crime-information-leakage-attack-against-ssltls
Thanks to @DaveRandom for pointing out the relevant section of code.
See http://www.openssl.org/~bodo/tls-cbc.txt
The biggest reason for this mode being in SSL_OP_ALL was older versions
of IE (2002) talking to servers using OpenSSL.
Can hopefully get this into 5.4.
do not support it).
I haven't written a test due to the need for such a test to have a HTTPS server
available which mishandles SessionTicket requests; it's likely that server
administrators will gradually fix this either intentionally or through OpenSSL
upgrades. That said, if there's a great clamoring for a test, I'll work one up.
