asciilifeform: 'With the page table splitting patches merged, it becomes necessary for the kernel to flush these caches every time the kernel begins executing, and every time user code resumes executing. For some workloads, the effective total loss of the TLB lead around every system call leads to highly visible slowdowns: @grsecurity measured a simple case where Linux “du -s” suffered a 50% slowdown on a☟︎☟︎
asciilifeform: in other heathen lulz, http://archive.is/QmV5G >> 'there is presently an embargoed security bug impacting apparently all contemporary CPU architectures that implement virtual memory, requiring hardware changes to fully resolve'
asciilifeform: 'I did publish, separately in sci.crypt, a nostalgic piece about a possible ancestor to this system. You can't very well accuse me of claiming novelty. There is a lot of deviousness and a bit of intelligence in what I proposed, but the novelty is in the way I implement the idea, and in the beneficial side effects.'
asciilifeform: just as sci.physics was for 'phree energy' and 'ditch j00-einsteinism' etc
asciilifeform: i.e. 'take the fatal boojum and slide it under the carpet. bang! perfection!'
asciilifeform: 'There is the unavoidable problem of K exhaustion to be dealt with. The more streams we need, the faster we use up the K. But considering the ability of the system to provide a high level of confidence, there is nothing to prevent us from cheating a bit. A number of strategies for the rejuvenation of old K material can be left to the imagination of the clients...'
asciilifeform: typical of 'boy inventors', we see following pattern:
asciilifeform: upstack: the hilarity of http://btcbase.org/log/2017-12-31#1761777 is that it is provably equivalent to otp. but author appears to not have even heard of xor. ( to asciilifeform's anticredit, he as a boy knew of xor, and still came up with exact same rubbish )☝︎
