This new field counts all the times Redis is configured with AOF enabled and
fsync policy 'everysec', but the previous fsync performed by the
background thread was not able to complete within two seconds, forcing
Redis to perform a write against the AOF file while the fsync is still
in progress (likely a blocking operation).

Added a top-function comment to rioWriteHashIteratorCursor() to better specify what the function does. Not immediately clear from the name.

Implementation of the internals that make possible to terminate clients overcoming configured output buffer (soft and hard) limits.

Track the length of the client pending output buffers (still to transfer) in a new field in the client structure.

AOF fileds in the global server state, and define names, renamed with more consistent names. More work to do.

Fix: when aof_write_rewrite is true don't append on the AOF buffer but accumulate the differences for the rewrite.

AOF fixes in the context of replicaiton (when AOF is used by slave) and CONFIG SET appendonly yes/no.

Fixed issues with expire introduced with latest millisecond resolution feature. Many time_t were not converted to long long, and one time() call was not replaced with mstime().

high resolution expires API modified to use separated commands. AOF transation to PEXPIREAT of all the expire-style commands fixed.

Initial support for key expire times with millisecond resolution. RDB version is now 3, new opcoded added for high resolution times. Redis is still able to correctly load RDB version 2. Tests passing but still a work in progress. API to specify milliseconds expires still missing, but the precision of normal expires is now already improved and working.

Clear the AOF rewrite scheduled flag once an AOF rewrite is triggered. Fix for issue #161, probably fixing 159 as well.

Thanks to @ardsrk.

rioInitWithFile nad rioInitWithBuffer functions now take a rio structure pointer to avoid copying a structure to return value to the caller.

Emit a log message when AOF fsync is still in progress but we are forced to write from the main thread since two seconds already elapsed.

Added aof_backgronud_fsync() function, and use it in the bacground rewrite done handler when the fsync policy is everysec.

use bio.c instead of libeio for closing the AOF file in background. Some comment added and other minor changes.