The ClamScan scan summary prints bytes scanned and bytes read in
multiples of 4096 (aka `CL_COUNT_PRECISION`), as is provided by the
`cl_scanfile()`, `cl_scandesc()`, `cl_scanfile_callback()`, and
`cl_scandesc_callback()` functions.
I believe this imprecision was the result of using an `unsigned long int`
which may be 64bit or 32bit, depending on platform. I believe the
intention was to be able to support scanning more than 4 GiB of data.
Since the new `cl_scan*_ex()` functions use a `uint64_t`, which
guarantees a 64bit integer and supports ~16,777,216 terabytes, I find no
reason not to report an accurate count.
For the legacy scan functions (above) I've kept the `CL_COUNT_PRECISION`
behavior to maintain backwards compatibility.
I have also improved the bytes scanned/read output to report GiB, MiB,
KiB, or B as appropriate. Previously, it always report "MB".
CLAM-1433
* Added loglevel parameter to logg()
* Fix logg and mprintf internals with new loglevels
* Update all logg calls to set loglevel
* Update all mprintf calls to set loglevel
* Fix hidden logg calls
* Executed clam-format
* fix scan of partial messages
* allow for tempfiles to be cleaned up based on age
* new clamd.conf option ScanPartialMessages
* sample cleanup script
* clamd/thrmgr.c: fix item_count
git-svn: trunk@4031
