Post AY58WvaO2PXDcXFsEi by SEGFAULT@oldbytes.space
 (DIR) More posts by SEGFAULT@oldbytes.space
 (DIR) Post #AY57xzVXSsepeGKPxo by erincandescent@queer.af
       2023-07-26T09:13:22Z
       
       1 likes, 0 repeats
       
       From inferences, it seems like AMD accidentally mailed the Zenbleed patch to LKML ahead of time, which is why all the microcode updates aren't ready on announcement day.If you're running a Zen 2 CPU that's not an EPYC, update your kernel for the workaround to apply
       
 (DIR) Post #AY57y1zkDY0XMWTZK4 by erincandescent@queer.af
       2023-07-26T09:13:52Z
       
       0 likes, 0 repeats
       
       In other news,microcode  : 0x830107a✅ Zenbleed microcode patch applied to our server
       
 (DIR) Post #AY584TElqO6dafZhdg by lanodan@queer.hacktivis.me
       2023-07-26T09:17:44.908719Z
       
       0 likes, 0 repeats
       
       @erincandescent So they planned to embargo a massive leak for months?
       
 (DIR) Post #AY58JdgxFxBMMHzU4e by erincandescent@queer.af
       2023-07-26T09:19:19Z
       
       1 likes, 0 repeats
       
       See specifically the Project Zero document, Timeline...2023-07-20 AMD unexpectedly publish patches, earlier than an agreed embargo date.2023-07-21 As the fix is now public, we propose privately notifying major distributions that they should begin preparing updated firmware packages.2023-07-24 Public disclosure.
       
 (DIR) Post #AY58SbmrhuH93d23Tk by erincandescent@queer.af
       2023-07-26T09:20:10Z
       
       0 likes, 0 repeats
       
       @lanodan Not the first time, won't be the last; preparing CPU workarounds and microcode patches is a slow process
       
 (DIR) Post #AY58Scj0DkmPxwEU64 by lanodan@queer.hacktivis.me
       2023-07-26T09:22:07.474564Z
       
       0 likes, 0 repeats
       
       @erincandescent Even for just a chicken-bit? (I'm maybe too much of a software person…)
       
 (DIR) Post #AY58WvaO2PXDcXFsEi by SEGFAULT@oldbytes.space
       2023-07-26T09:18:41Z
       
       0 likes, 0 repeats
       
       @erincandescent more performance regressions by chance?
       
 (DIR) Post #AY58WwRYqi4MHS8L7Q by erincandescent@queer.af
       2023-07-26T09:21:03Z
       
       1 likes, 0 repeats
       
       @SEGFAULT The microcode patch has little-to-no perf impact to my knowledge(the chicken bit has "probably slightly higher" impact on AVX workloads)
       
 (DIR) Post #AY59ZZYAWTFC6J2hEG by erincandescent@queer.af
       2023-07-26T09:24:19Z
       
       0 likes, 0 repeats
       
       @lanodan The chicken bit workaround is out for all Zen 2 CPUsPeople are complaining about the delay on the microcode patches
       
 (DIR) Post #AY59ZaE00whkC2c6Qi by lanodan@queer.hacktivis.me
       2023-07-26T09:34:33.398574Z
       
       0 likes, 0 repeats
       
       @erincandescent Well at least I'm glad we got the workaround early, even though it's almost a zero-day in practice with traviso'd code.I'd say delay for a proper fix is fine *if* there's a workaround in the meantime.
       
 (DIR) Post #AY5PTopmNmoB24YfVg by lanodan@queer.hacktivis.me
       2023-07-26T12:32:45.964418Z
       
       0 likes, 0 repeats
       
       @becomethewaifu @erincandescent The current microcode update is only for Epyc servers.
       
 (DIR) Post #AY6Cl3AvDC98BfIG6i by rotopenguin@mastodon.social
       2023-07-26T14:34:16Z
       
       0 likes, 0 repeats
       
       @ariadne @chjara @erincandescent @lanodan We should stick to simple microcontrollers that are much easier to verifyhttps://mastodon.social/@whitequark/110770366839001808
       
 (DIR) Post #AY6Cl487f5V99GzXNo by lanodan@queer.hacktivis.me
       2023-07-26T21:44:58.587439Z
       
       0 likes, 0 repeats
       
       @rotopenguin @ariadne @chjara @erincandescent Which is why things like smartcards and HSM are used for a lot of security purposes.But microcontrollers for your main CPU… not going to happen.