Undefined behavior

By: David Kanter (dkanter.delete@this.realworldtech.com), January 31, 2018 11:18 am
Room: Moderated Discussions
hobold (hobold.delete@this.vectorizer.org) on January 31, 2018 9:33 am wrote:
> anon.1 (abc.delete@this.def.com) on January 27, 2018 2:55 pm wrote:
>
> [...]
> >Meltdown does not violate the architectural
> > specification: a user process cannot access kernel pages without triggering a GP fault.
>
> Upon reading the above sentence I kinda lost my marbles. I had to take a break and ponder if the
> world really was going crazy or if it was just me (the proverbially more likely alternative).
>
> But now I better understand why Meltdown is one of these "once in a blue moon"
> cases where I simply cannot quietly stand on the sidelines. So back to topic.
>
> Memory protection is not about passing control to exception handlers. Memory protection
> is about preventing unauthorized access to memory data. Or, phrased more closely aligned
> to the quoted wording above: "a user process cannot access kernel data, period".

That's really the responsibility of both HW and SW. Not all OSes have memory protection (or want it). E.g., many microcontrollers don't have virtual memory.

> Triggering some fault is merely an implementation detail. It is certainly convenient
> to trigger privilege violation exceptions, because that way we can consistently handle
> other cases of ... unserviceable memory requests by passing control to the OS.

From a system architect's perspective that might be true. But given the way modern ISAs work, it's not true. Exception handling, fault handling, etc. are all architecturally defined. That is a contract between HW and SW, and this is about the grey areas not covered by that contract.

I think what you are getting at is that in an ideal world, it would be a system responsibility and not divided between HW and SW. In fact, the implementation could vary (e.g., some systems might use mostly HW memory isolation, some might use SW only, and some might be a hybrid). I like that idea, but that's not the systems we have today.

> But we could implement memory protection without faults: unauthorized reads always return zero, and unauthorized
> writes silently fail and do not change the contents of the destination address. That would have its own
> set of downsides, but would maintain barriers between protection domains. The concept of triggering faults
> is not directly related to the concept of protecting data in memory from unauthorized access.

How would you distinguish an unauthorized read from a read of a recently zeroed page? A page that was zeroed by another core, or an I/O device?

> > What Intel inadvertently
> > allows is undefined behavior: x86 (or ARM, SPARC, POWER,
> > PA-RISC, Alpha, VAX...) does not forbid speculative
> > accesses of memory nor does it qualify this with privilege level checks.

> Back when I was reading ISA specifications (specifically a few revisions of the PowerPC ISA,
> in 32bit and 64bit variants), those documents were intentionally worded to be complete. Complete
> in the sense that all possible behaviours of an instruction were listed exhaustively in the specification.
> In the few cases where the architects wanted to leave some leeway for the architects, they explicitly
> listed a special behaviour that they called "boundedly undefined".

I sincerely doubt any ISA is fully defined. In fact, PPC suffers from the same problems as x86 with respect to Spectre and Meltdown (https://www.ibm.com/blogs/psirt/potential-impact-processors-power-family/). One could argue its worse because of the weak consistency model.

> The definition of boundedly undefined was something like: the same effect that any legal
> sequence of valid instructions could have. As I google for that now, I find this:
>
> ... "undefined results are bounded in that a spurious change from user to supervisor
> state is not allowed, and the level of privilege exercised by the program in relation
> to memory access and other system resources cannot be exceeded."
>
> In other words, even where the PowerPC architects left things undefined,
> they explicitly maintained the security of the machine.

Except for the fact that PowerPC implementations are vulnerable to Spectre and Meltdown. So they didn't. Just like Intel and Apple, IBM's architects aren't idiots. They missed a subtle interplay of undefined behavior.

> > AMD implemented the undefined behavior
> > by disallowing speculation across privilege levels entirely.
>
> That's not what AMD did in Zen. AMD's Zen simply honours memory protection by
> not returning memory contents from an unauthorized load operation. Not unlike
> the hypothetical example above of an alternative memory protection behaviour.

Zen makes a different (and clearly better!) choice about how to handle undefined behavior. It's still undefined behavior.

> Intel's mistake with Meltdown is not in overlooking a side channel. Intel's mistake with Meltdown lies in passing
> on data to subsequent instructions, through the primary channel, even when that load was not permitted to read
> from the given address. Intel's mistake with Meltdown lies in not honouring memory protection.

Not really. If there were no side channels, it wouldn't matter that a register somehow accessed data it shouldn't.

Also, I think you tend to look at this from a black and white perspective, whereas I see security as more of a probabilistic concern. No system is fully secure, it's all a matter of degrees and what you are willing to sacrifice for that security (e.g., if you really care, stay off networks!).

David
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Intel recommends not using recent microcode updatesanonymous22018/01/22 12:01 PM
  Intel recommends not using recent microcode updateshobold2018/01/22 02:54 PM
    I imagine most of the effort is going into addressing it in silicon asap (NT)john2018/01/22 03:17 PM
    Intel recommends not using recent microcode updatesjuanrga2018/01/22 05:04 PM
    how about... less bitching and more appreciation?anonymou52018/01/22 07:05 PM
      it's not the engineers I'm bitching abouthobold2018/01/23 01:13 AM
        it's not the engineers I'm bitching aboutjuanrga2018/01/23 03:55 AM
          it's not the engineers I'm bitching abouthobold2018/01/23 04:16 AM
            it's not the engineers I'm bitching aboutanon.12018/01/23 01:48 PM
              it's not the engineers I'm bitching aboutanon.12018/01/23 02:20 PM
                it's not the engineers I'm bitching aboutDavid Hess2018/01/23 06:53 PM
                  it's not the engineers I'm bitching aboutTravis2018/01/24 12:03 AM
                  it's not the engineers I'm bitching aboutanon.12018/01/24 12:11 AM
                    it's not the engineers I'm bitching aboutDavid Kanter2018/01/24 12:22 AM
                    it's not the engineers I'm bitching aboutrwessel2018/01/24 01:38 AM
                    More implementation than performance advantageRay Terry2018/02/04 06:45 PM
                      More implementation than performance advantagedmcq2018/02/11 02:22 AM
                        More implementation than performance advantageRay Terry2018/02/17 09:14 AM
                          More implementation than performance advantagedmcq2018/02/17 09:33 AM
                          More implementation than performance advantageanon2018/02/17 06:07 PM
                            More implementation than performance advantageRay Terry2018/02/18 07:40 AM
                              More implementation than performance advantageanon2018/02/18 08:55 PM
                                More implementation than performance advantageTravis2018/02/20 12:57 PM
                                  More implementation than performance advantageRay Terry2018/02/20 03:53 PM
                                    More implementation than performance advantageLinus Torvalds2018/02/20 06:43 PM
                                      More implementation than performance advantageanon2018/02/21 03:13 AM
                                      More implementation than performance advantageRay Terry2018/02/21 04:03 AM
                                        More implementation than performance advantageLinus Torvalds2018/02/21 08:59 AM
                                    More implementation than performance advantageanon2018/02/21 03:18 AM
                                      More implementation than performance advantageRay Terry2018/02/21 08:32 AM
                                        More implementation than performance advantageTravis2018/02/21 09:35 AM
                                      More implementation than performance advantageTravis2018/02/21 09:29 AM
                                        you can mostly ignore the parent postTravis2018/02/21 09:36 AM
                                      More implementation than performance advantageDavid Hess2018/02/23 11:33 AM
                                        More implementation than performance advantageTravis2018/02/23 02:57 PM
                                          More implementation than performance advantageDavid Hess2018/02/23 06:12 PM
                                        More implementation than performance advantageanon2018/02/23 09:01 PM
                                    More implementation than performance advantageTravis2018/02/21 09:43 AM
                                      More implementation than performance advantageRay Terry2018/02/21 04:44 PM
                                        More implementation than performance advantageTravis2018/02/21 05:16 PM
                                          More implementation than performance advantageLinus Torvalds2018/02/21 06:31 PM
                                            More implementation than performance advantageTravis2018/02/22 10:48 AM
                                              More implementation than performance advantageTravis2018/02/22 11:29 AM
                                                More implementation than performance advantageLinus Torvalds2018/02/22 11:42 AM
                                                  More implementation than performance advantageTravis2018/02/22 05:38 PM
                                                    More implementation than performance advantageLinus Torvalds2018/02/22 08:14 PM
                                                      More implementation than performance advantageTravis2018/02/22 10:20 PM
                                                        More implementation than performance advantageLinus Torvalds2018/02/23 08:55 AM
                                                          More implementation than performance advantageLinus Torvalds2018/02/23 09:43 AM
                                                            AMD's SYSCALLYuhong Bao2018/02/23 01:29 PM
                                                              AMD's SYSCALLLinus Torvalds2018/02/23 02:05 PM
                                                                AMD's SYSCALLYuhong Bao2018/02/23 02:26 PM
                                                                  AMD's SYSCALLLinus Torvalds2018/02/23 02:40 PM
                                                                AMD's SYSCALLYuhong Bao2018/02/23 02:34 PM
                                                                AMD's SYSCALLYuhong Bao2018/02/23 02:45 PM
                                                                AMD's SYSCALLBrendan2018/02/23 07:03 PM
                                                                  AMD's SYSCALLYuhong Bao2018/02/23 07:10 PM
                                                                  AMD's SYSCALLLinus Torvalds2018/02/23 08:40 PM
                                                                    AMD's SYSCALLBrendan2018/02/24 03:47 AM
                                                                    AMD's SYSCALLYuhong Bao2018/02/24 11:02 AM
                                                                      AMD's SYSCALLLinus Torvalds2018/02/24 03:45 PM
                                                                        AMD's SYSCALLYuhong Bao2018/02/24 06:31 PM
                                                                  AMD's SYSCALLanonymou52018/02/23 09:03 PM
                                                                    AMD's SYSCALLBrendan2018/02/24 03:49 AM
                                              More implementation than performance advantageLinus Torvalds2018/02/22 11:33 AM
                                              More implementation than performance advantageanon2018/02/22 12:23 PM
                                                More implementation than performance advantageanon2018/02/22 06:45 PM
                                                  More implementation than performance advantageTravis2018/02/22 10:15 PM
                                        serialization of syscall/sysretTravis2018/02/23 06:10 PM
                                          serialization of syscall/sysretanonymou52018/02/23 08:58 PM
                                            serialization of syscall/sysretTravis2018/02/23 10:43 PM
                it's not the engineers I'm bitching abouthobold2018/01/24 09:27 AM
                  it's not the engineers I'm bitching aboutanon.12018/01/26 12:26 AM
                    it's not the engineers I'm bitching abouthobold2018/01/26 06:41 AM
                      it's not the engineers I'm bitching aboutanon.12018/01/27 10:42 AM
                        it's not the engineers I'm bitching aboutMichael S2018/01/27 11:31 AM
                        it's not the engineers I'm bitching abouthobold2018/01/27 01:16 PM
                          it's not the engineers I'm bitching aboutanon.12018/01/27 02:55 PM
                            memory protection and sanityhobold2018/01/31 09:33 AM
                              Undefined behaviorDavid Kanter2018/01/31 11:18 AM
                                Undefined behaviorhobold2018/01/31 12:36 PM
                                  Undefined behavioranon2018/01/31 04:00 PM
                                    erring on the side of securityhobold2018/01/31 11:38 PM
                                      There is no primary channel before the retire stageHeikki Kultala2018/01/31 11:43 PM
                                        retire is about keeping recordshobold2018/02/01 01:07 AM
                                          retire is about keeping recordsanon2018/02/01 02:32 AM
                                            Well saidRay Terry2018/02/04 10:27 PM
                                            retire is about keeping recordshobold2018/02/08 07:46 AM
                                              retire is about keeping recordsdmcq2018/02/08 08:12 AM
                                              retire is NOT about keeping recordsHeikki Kultala2018/02/08 08:33 AM
                                                retire is NOT about keeping recordshobold2018/02/08 11:55 AM
                                                  retire is NOT about keeping recordsHeikki Kultala2018/02/08 01:01 PM
                                              retire is about keeping recordsanon2018/02/08 02:45 PM
                                              retire is about keeping recordsanon2018/02/08 08:43 PM
                                      erring on the side of securitypgerassi2018/02/01 12:43 AM
                                      erring on the side of securityanon2018/02/01 02:38 AM
                                      erring on the side of securityanon2018/02/01 07:11 AM
                                        erring on the side of securityhobold2018/02/08 08:24 AM
                                          erring on the side of securityanon2018/02/08 02:12 PM
                                  Undefined behavioranon2018/01/31 06:32 PM
                                    memory protection is fully definedhobold2018/01/31 10:48 PM
                                      Doubtfulquestioning critic2018/01/31 11:33 PM
                                        insert irrelevant title herehobold2018/02/01 12:49 AM
                                      memory protection is fully definedanon2018/02/01 02:17 AM
                                        memory protection is fully definedanon.12018/02/02 12:44 AM
                                          memory protection is fully definedhobold2018/02/08 08:43 AM
                                            memory protection is fully definedanon2018/02/08 08:59 PM
                                        memory protection is fully definedhobold2018/02/08 08:34 AM
                                          memory protection is fully definedHeikki Kultala2018/02/08 08:58 AM
                                            memory protection is fully definedMichael_S2018/02/08 10:52 AM
                                          memory protection is fully definedanon2018/02/08 09:01 PM
                                            memory protection is fully definedgallier22018/02/09 01:57 AM
                                              memory protection is fully defineddmcq2018/02/09 02:53 AM
                                              memory protection is fully definedanon2018/02/09 04:48 AM
                                                memory protection is fully definedgallier22018/02/09 05:53 AM
                                                  memory protection is fully definedMichael S2018/02/09 06:03 AM
                                                    memory protection is fully definedAdrian2018/02/10 01:51 AM
                                                      memory protection is fully defineddmcq2018/02/10 02:51 AM
                                                      memory protection is fully definedanon2018/02/10 03:29 AM
                                                        memory protection is fully definedAdrian2018/02/10 04:45 AM
                                                          memory protection is fully definedanon2018/02/10 06:57 AM
                                                            memory protection is fully definedBrett2018/02/10 02:19 PM
                                                              memory protection is fully definedrwessel2018/02/10 02:41 PM
                                                                memory protection is fully definedBrett2018/02/11 02:45 AM
                                                                  memory protection is fully definedanon2018/02/11 03:49 AM
                                                                    memory protection is fully defineddmcq2018/02/11 10:34 AM
                                                                      memory protection is fully definedanonymou52018/02/11 12:42 PM
                                                                      memory protection is fully definedBrett2018/02/11 12:57 PM
                                                                        memory protection is fully definedMichael S2018/02/11 01:20 PM
                                                                          memory protection is fully definedBrett2018/02/11 01:49 PM
                                                                            memory protection is fully definedanon2018/02/11 03:01 PM
                                                          memory protection is fully defineddmcq2018/02/10 07:34 AM
                                                          memory protection is fully definedrwessel2018/02/10 09:53 AM
                                                        memory protection is fully definedanon2018/02/10 04:55 AM
                                                      memory protection is fully definedanon2018/02/10 04:40 AM
                                                        memory protection is fully definedhobold2018/02/12 06:18 AM
                                                          memory protection is fully definedKlimax2018/02/12 12:34 PM
                                                  Load instruction performs corrrectlycommenting bystander2018/02/09 07:57 AM
                                                    Load instruction performs corrrectlyhobold2018/02/12 07:55 AM
                                                      Load instruction performs corrrectlyKlimax2018/02/12 12:35 PM
                                                        Load instruction performs corrrectlyhobold2018/02/12 08:24 PM
                                                          Load instruction performs corrrectlyrwessel2018/02/12 09:53 PM
                                                          Load instruction performs corrrectlyanon2018/02/13 03:27 AM
                                                            Load instruction performs corrrectlyMichael_S2018/02/13 04:17 AM
                                                            Load instruction performs corrrectlyMichael_S2018/02/13 04:22 AM
                                                          Load instruction performs corrrectlyDavid Hess2018/02/13 05:43 AM
                                                            Load instruction performs corrrectlyhobold2018/02/14 05:48 AM
                                                              Answer: unmapped VM page (NT)Heikki Kultala2018/02/14 09:31 AM
                                                          Load instruction performs corrrectlySimon Farnsworth2018/02/13 07:14 AM
                                                      Load instruction performs corrrectlyanony2018/02/13 04:46 PM
                                                        spec? what spec?hobold2018/02/14 06:22 AM
                                                          spec? what spec?dmcq2018/02/14 07:32 AM
                                                          spec? what spec?Simon Farnsworth2018/02/14 09:39 AM
                                                            Historical AsideTerry A Gray2018/02/16 06:29 PM
                                                              Historical AsideUngo2018/02/16 10:10 PM
                                                                Historical Asidedmcq2018/02/17 02:58 AM
                                                              Historical Asidedmcq2018/02/17 02:37 AM
                                                                Historical AsideMaynard Handley2018/02/17 12:50 PM
                                                                  Historical Asidedmcq2018/02/18 06:16 AM
                                                                    Historical AsideMaynard Handley2018/02/18 11:10 AM
                                                              Historical AsideSimon Farnsworth2018/02/17 03:34 AM
                                                          ISA docs are part of specDavid Kanter2018/02/14 02:55 PM
                                                            ISA docs are part of specdmcq2018/02/14 03:14 PM
                                                              ISA docs are part of specMichael S2018/02/14 03:46 PM
                                                                ISA docs are part of specrwessel2018/02/14 04:44 PM
                                                                ISA docs are part of specAaron Spink2018/02/14 06:10 PM
                                                              ISA docs are part of specnone2018/02/16 06:26 AM
                                                            ISA docs are part of specAnon2018/02/16 01:36 AM
                                                              ISA docs are part of specwumpus2018/02/16 03:03 AM
                                                                ISA docs are part of specdmcq2018/02/16 04:23 AM
                                                  memory protection is fully definedanon2018/02/09 08:10 AM
                                                    memory protection is fully definedgallier22018/02/09 08:19 AM
                                                      memory protection is fully definedhobold2018/02/12 06:46 AM
                                                        memory protection is fully definedanon2018/02/12 04:51 PM
                                                  memory protection is fully definedSimon Farnsworth2018/02/09 10:14 AM
                                                    memory protection is fully definedhobold2018/02/12 06:52 AM
                                                      memory protection is fully definedSimon Farnsworth2018/02/12 08:21 AM
                                                        memory protection is fully defineddmcq2018/02/12 09:00 AM
                                                          memory protection is fully definedhobold2018/02/12 10:41 AM
                                                            memory protection is fully definedSimon Farnsworth2018/02/12 12:45 PM
                                                              Meltdown on ARMvvid2018/02/12 04:45 PM
                                                                Meltdown on ARMSimon Farnsworth2018/02/13 04:43 AM
                                                              memory protection is fully definedhobold2018/02/12 08:33 PM
                                                                memory protection is fully definedSimon Farnsworth2018/02/13 04:54 AM
                                                                  memory protection is fully definedhobold2018/02/14 06:51 AM
                                                                    memory protection is fully definedSimon Farnsworth2018/02/14 09:31 AM
                                                                      protection speculation has zero web hitshobold2018/02/17 12:20 PM
                                                                        NitpickingPaul A. Clayton2018/02/17 04:49 PM
                                                                          Tu quoque, Paulus?hobold2018/02/20 06:38 AM
                                                                            Tu quoque, Paulus?Ungo2018/02/21 03:13 AM
                                                                              dudehobold2018/02/21 09:09 PM
                                                            memory protection is fully definedanon2018/02/12 04:53 PM
                                                              memory protection is fully definedhobold2018/02/12 08:52 PM
                                                                memory protection is fully definedanon2018/02/12 09:21 PM
                                          memory protection is fully definedcommenting bystander2018/02/08 10:16 PM
                                      memory protection is fully definedJukka Larja2018/02/01 07:53 AM
                                        memory protection is fully definednone2018/02/01 09:41 AM
                                          memory protection is fully defineddmcq2018/02/02 04:37 AM
                                        Ceterum Censeo ...hobold2018/02/08 08:48 AM
                                          Ceterum Censeo ...anonymou52018/02/08 09:31 AM
                                            Ceterum Censeo ...Michael_S2018/02/08 10:45 AM
                                              Ceterum Censeo ...hobold2018/02/08 11:46 AM
                                                Ceterum Censeo ...Alberto2018/02/09 01:00 AM
                                            Ceterum Censeo ...gallier22018/02/09 02:01 AM
                                              Ceterum Censeo ...anon2018/02/09 05:14 AM
                                                Ceterum Censeo ...dmcq2018/02/10 03:17 AM
                                          Ceterum Censeo ...Jukka Larja2018/02/09 08:37 AM
                                            Ceterum Censeo ...hobold2018/02/12 07:06 AM
                                              Ceterum Censeo ...Klimax2018/02/12 12:31 PM
                                  Undefined behaviorJukka Larja2018/02/01 08:08 AM
                    it's not the engineers I'm bitching aboutRay Terry2018/02/04 10:46 PM
            it's not the engineers I'm bitching aboutjuanrga2018/01/23 04:18 PM
              it's not the engineers I'm bitching aboutanon.12018/01/23 11:58 PM
    Intel recommends not using recent microcode updatesMegol2018/01/23 02:48 AM
      Intel recommends not using recent microcode updateshobold2018/01/23 04:30 AM
        Intel recommends not using recent microcode updatesMegol2018/01/24 02:13 AM
          Intel recommends not using recent microcode updateshobold2018/01/24 09:50 AM
            Intel recommends not using recent microcode updatesKlimax2018/01/24 10:45 AM
              Intel recommends not using recent microcode updatesBrett2018/01/24 11:38 AM
              Intel recommends not using recent microcode updateshobold2018/01/24 12:10 PM
                Intel recommends not using recent microcode updatesKlimax2018/01/26 05:25 AM
                  Intel recommends not using recent microcode updateshobold2018/01/26 05:49 AM
                    Intel recommends not using recent microcode updatesDavid Kanter2018/01/26 11:42 PM
                      Intel recommends not using recent microcode updateshobold2018/01/27 02:02 AM
                        Try to avoid partisan postingDavid Kanter2018/01/27 09:57 AM
                          Try to avoid partisan postinghobold2018/01/27 12:51 PM
                      Intel recommends not using recent microcode updatesknweiss2018/01/27 03:22 AM
                        Intel recommends not using recent microcode updatesanon2018/01/27 05:56 AM
                      Intel recommends not using recent microcode updatesanon2018/01/27 06:01 AM
                        Intel recommends not using recent microcode updatesDavid Hess2018/01/27 11:42 AM
                          Intel recommends not using recent microcode updatesanon2018/01/28 02:07 AM
                    Intel recommends not using recent microcode updatesanon2018/01/27 06:28 AM
                Some background for Meltdown/Spectrejuanrga2018/01/27 05:37 AM
                  Some background for Meltdown/Spectrehobold2018/01/27 08:44 AM
                    Some background for Meltdown/SpectreAlberto2018/01/27 09:00 AM
                    Some background for Meltdown/Spectrejuanrga2018/01/27 11:35 AM
                      Some background for Meltdown/SpectreLinus Torvalds2018/01/27 12:09 PM
                        Some background for Meltdown/SpectreYuhong Bao2018/01/27 06:55 PM
                        Some background for Meltdown/SpectreDoug S2018/01/28 01:48 AM
                          Some background for Meltdown/SpectreAlberto2018/01/28 02:01 AM
                        Some background for Meltdown/Spectrejuanrga2018/01/28 06:49 AM
                      Some background for Meltdown/Spectrehobold2018/01/27 01:50 PM
                        Some background for Meltdown/SpectreAlberto2018/01/28 03:31 AM
                        Some background for Meltdown/Spectrejuanrga2018/01/28 06:39 AM
                          Spectre V2 and AMD.Heikki Kultala2018/01/29 03:21 AM
                            That's not a full fixRay Terry2018/01/29 03:43 AM
                            Spectre V2 and AMD.juanrga2018/01/30 02:37 PM
                              Spectre V2 and AMD.Brett2018/01/31 12:04 AM
                                Spectre V2 and AMD.juanrga2018/01/31 03:33 AM
                                  Spectre V2 and AMD.hobel2018/01/31 03:54 AM
                                    Spectre V2 and AMD.Doug S2018/01/31 08:41 AM
                  Some background for Meltdown/Spectrejuanrga2018/03/10 12:53 PM
            Intel recommends not using recent microcode updatesMichael S2018/01/24 12:38 PM
              Intel recommends not using recent microcode updateshobold2018/01/24 12:59 PM
              Intel recommends not using recent microcode updatesGabriele Svelto2018/01/25 01:56 AM
                Intel recommends not using recent microcode updatesMichael_S2018/01/25 02:20 AM
                  Intel recommends not using recent microcode updatessomeone2018/01/25 02:46 AM
                    Intel recommends not using recent microcode updatesMegol2018/01/25 03:35 AM
                      Intel recommends not using recent microcode updatesMichael_S2018/01/25 04:09 AM
                  Intel recommends not using recent microcode updatesGabriele Svelto2018/01/26 12:21 AM
                    Intel recommends not using recent microcode updatesMichael S2018/01/26 03:07 AM
                      Intel recommends not using recent microcode updatesanon2018/01/26 05:06 AM
                      Intel recommends not using recent microcode updatesGabriele Svelto2018/01/26 07:56 AM
                        Intel recommends not using recent microcode updatesMichael S2018/01/27 11:59 AM
              Intel recommends not using recent microcode updatesWalt Berger2018/01/31 06:05 PM
            Is it material?David Kanter2018/01/24 03:35 PM
              Is it material?hobold2018/01/25 01:18 AM
                Is it material?Klimax2018/01/26 05:30 AM
                  Is it material?hobold2018/01/26 06:18 AM
              Is it material?OMG2018/01/30 05:27 AM
                Is it material?Doug S2018/01/30 10:48 AM
                  Is it material?Alberto2018/01/30 01:41 PM
                    Is it material?Doug S2018/01/30 03:14 PM
                      Is it material?anon2018/01/31 03:03 AM
                        Is it material?Alberto2018/01/31 03:43 AM
                          Is it material?anon2018/01/31 04:38 AM
                      Is it material?Alberto2018/01/31 03:49 AM
                      Is it material?joema2018/01/31 09:42 AM
            Intel recommends not using recent microcode updatesGabriele Svelto2018/01/25 01:54 AM
  Intel recommends not using recent microcode updatesDavid Hess2018/01/23 07:09 AM
    reboot expectationsanonymous22018/01/23 09:04 AM
      reboot expectationsFoo_2018/01/23 01:20 PM
        reboot expectationsDavid Hess2018/01/23 06:56 PM
          reboot expectations -Intel UpdateMS2018/01/24 09:29 AM
  Intel recommends not using recent microcode updatesMS2018/01/23 11:52 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell avocado?