The sendfile() implementation for evbuffer_add_file is potentially more
efficient, but it has a problem: you can only use it to send bytes over
a socket using sendfile(). If you are writing bytes via SSL_send() or
via a filter, or if you need to be able to inspect your buffer, it
doesn't work.
As an easy fix, this patch disables the sendfile-based implementation of
evbuffer_add_file on an evbuffer unless the user sets a new
EVBUFFER_FLAG_DRAINS_TO_FD flag on that evbuffer, indicating that the
evbuffer will not be inspected, but only written out via
evbuffer_write(), evbuffer_write_atmost(), or drained with stuff like
evbuffer_drain() or evbuffer_add_buffer(). This flag is off by
default, except for evbuffers used for output on bufferevent_socket.
In the future, it could be interesting to make a best-effort file
segment implementation that tries to send via sendfile, but mmaps on
demand. That's too much complexity for a stable release series, though.
Some hosts require you to define certain options to get a large off_t
instead of a small one, to get useful ftell and fseek calls instead of
ones that can only support 2GB files, and so on. This patch makes
Libevent support those platforms by:
* Defining the right options when we build, and
* Changing our API so that it does not depend on the platform's
definition of off_t.
Based on discusion with Michael Herf
https://sourceforge.net/tracker/index.php?func=detail&aid=3078187&group_id=50884&atid=461324
The problem is that bufferevent_disable() doesn't disable EV_WRITE
when 'connecting' flag is set. However from evhttp_connection_reset()
we want to disable EV_WRITE for sure (we are closing the socket next).
So we add bufferevent_disable_hard(), which acts like
bufferevent_disable(), but resets 'connecting' flag before the call to
the actual handler.
TODO: bufferevent_disable_hard() shouldn't be public, remove it from
event2/bufferevent.h.
If the EVHTTP_URI_NONCONFORMANT flag is passed in (which it is when
parsing URIs we get over the wire), then we relax our checks a lot.
Specifically, we do nothing to check for correct characters in the
path, query, and fragment parts of such a URI.
We could do much more here: we could relax our hostname requirements,
deal with spaces differently/better, trap some errors but not others,
etc. But this should solve the worst user-agent compatibility issues
for now; the other issues can wait for a later release.
Previously, debug logs were turned on if you built with -DUSE_DEBUG
and off otherwise. This make builds with -DUSE_DEBUG hideously slow
and other builds unable to get debug logs.
This is based off a patch by Ralph Castain from October. It tries a
little harder to avoid needless function calls, it doesn't require
stdbool, and makes the controlling parameter a mask rather than a
boolean so that we can later support enabling only the debugging
messages for the parts of Libevent you're trying to debug.
