Apparently @musllibc's memcpy/memset are most-hardened against rowhammer use. :-) See page 5: http://seclab.cs.sunysb.edu/seclab/pubs/host16.pdf …
@hrnrj ECC does not prevent rowhammer, but gives you a fairly good chance that arbitrary-code-execution will be reduced to system-crash.
-
-
@RichFelker@hrnrj ECC corrects some errors. If conf'd properly that's advances warning, and if conf'd properly crashes when irrecoverable. -
@jfbastien@hrnrj AFAIK the only real, complete mitigation for rowhammer is drastically lowering the dram refresh interval. -
@RichFelker@hrnrj right, you can also mitigate w/ increased refresh rate. Mention above is good anyways, RAM can break w/o rowhammer! -
@jfbastien@hrnrj Do you know any practical guide to refresh rate based mitigation, including info on _how_ to adjust it when BIOS sucks? -
@RichFelker@hrnrj that's dark magic. Some have the chance to talk to OEMs and experiment at scale, I'm but a poor compiler engineer! -
@jfbastien@hrnrj The current situation is horrible. RH has been around how long, and there's NO INFO on how ordinary people can defend. -
@RichFelker@hrnrj I entirely agree! Defenses and mitigations aren't well documented, neither are the vulnerabilities.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.