[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On 09/01/18 13:21, Julian Hall wrote: > FYI.. partially on topic as Linux is mentioned.. I don't know where you got that body of text from (it's not in the included link) but it's not only *not* [OT], it's also comically and almost completely factually incorrect. Who wrote that crap? This specific issue is only affecting a narrow band of users with specific Athlon or Sempron branded CPUs, not just all AMD units. A rather critical omission. It can also be trivially fixed with a rollback. The writer doesn't understand what "bricking" means, amongst other things - a temporary and completely recoverable fault is not bricking for god's sake. >> Windows 10 doesn't allow users to turn off the update either Factually incorrect. >> CPUs haven't got any faster for ten years Haha what. Seriously? >> this is why you occasionally get a tiny glitch while running games Oh my sides! Is this person for real? >> The Meltdown and Spectre exploits target this predictive behaviour by getting the CPU to execute arbitrary code in RAM Hmm. Not really a technical expert are they? >> exploit is easy to set up with JavaScript So easy that there are no known exploits outside of the original academic PoCs? >> There is no fix other than replacing the CPU with a redesigned one Well, other than ALL THE OTHER FIXES currently being rolled out by multiple vendors... >> what the patch does is to mitigate the effects of the exploit by disabling the speculative execution Oh my goodness, they just keep digging and digging and digging. The patches do absolutely no such thing of course. >> Microsoft famously develops on all-Intel racks By "famously" they presumably mean "some half arsed internet rumour I once heard and have lazily decided to spread about regardless. >> if you're running Windows on an AMD system, you're better off not patching it Great, they have now crossed the line from just being hopelessly stupid and incorrect about everything to now disseminating dangerous advice that they are woefully under-qualified to be giving. This person needs to shut the hell up. To be fair, this is a big fat mess at the moment and none of the involved parties have been exactly covering themselves with glory - Microsoft's casual approach to QA has indeed bitten some of their unfortunate AMD Athlon/Sempron wielding users with an extra dose of misery on top of everything else. However, early reports indicate that this is probably an issue with some of the low level documentation that AMD themselves provided to Microsoft regarding the effected chips. Of course Microsoft QA still should have caught it but let's not pretend that Redmond don't just outsource the majority of their testing to their unfortunate user base these days. The signal to noise ratio on the whole Spectre/Meltdown situation is untenable at the moment so I seriously advise just ignoring all the sensationalist and uninformed crap flying around online and wait for calmer, saner professional heads to chime in with rock solid information over the coming few days. Our very own Martijn Grooten for example actually works in the security industry and hopefully may pop in with a clear and concise summary from the frontline at some point (hint hint Martijn!) For now, half-arsed nonsense really isn't helping anyone. El Reg as ever have timely and comparatively unsensationalised reporting as updates unfold. Phoronix have *actual* benchmarking on Linux + BSD systems pre and post patching. 95% of the panic-stricken mainstream tech news at the moment is just uninformed clickbaity crap and full of inaccuracies, ungrounded accusations and sheer speculation. Which come to think of it, is basically business as usual I suppose :[ Cheers -- The Mailing List for the Devon & Cornwall LUG https://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq