DNA shearing (also know as DNA fragmentation), Chromatin shearing and, less commonly, RNA shearing are required for a number of sample prep applications, for instance prior to sequencing.
The most commonly used homogenizers for DNA shearing and chromatin shearing are ultrasonics. There is a lot of literature showing the effectiveness of ultrasonic homogenization for shearing DNA and chromatin, and this is the recommended method unless there is a specific reason that ultrasonic homogenization cannot be used in your application.
Many high-pressure homogenizers are also capable of shearing DNA, however most high-pressure homogenizers require larger volumes than are present in most DNA fragmentation applications. The size of the fragment is generally inversely related to the pressure drop caused by the homogenizer - the larger the pressure drop, the smaller the fragments.
There is evidence that most other methods can also shear DNA, however are not as effective and may not be able to achieve as small of fragments. Rotor-stators are able to shear DNA, and it is reported that they can achieve relatively small fragments if ran at high speeds for extended periods of time. High-power bead mills may also achieve meaningful fragmentation when run for extended periods of time, but ironically some bead mills are also very good devices for obtaining relatively intact DNA.
Mortar and pestle methods are generally incapable of shearing DNA to fragment sizes which would be meaningful in downstream applications requiring sheared DNA. Paddle blenders, which are designed not to lyse cells, are not suitable.
DNA and chromatin shearing often involve small samples, and because of this there can be very significant foaming as well as overheating if direct probe sonication is used. Instead, it is recommended that indirect sonication is used. This can be achieved using a cup horn (example) or similar probe accessory. Cup horns also have the ancillary benefit of increasing throughput and reducing cross-contamination risk.
Longer runs at higher powers will achieve smaller fragments. This is a fairly reproducible process, but some variation in DNA fragment lengths is normal as DNA fragmentation during ultrasonication is not entirely random (reference). Expect that the run times necessary to achieve small fragments will be far greater than those to simply achieve cell lysis.
Don't forget to treat your samples with RNase! Otherwise you may see "fragments" at about 100 - 150 bp that are not actually DNA, but RNA.