Early resolution: some but limited potential

Article: Microservers must Specialize to Survive
By: Paul A. Clayton (paaronclayton.delete@this.gmail.com), February 19, 2013 4:52 pm
Room: Moderated Discussions
EBFE (x.delete@this.y.com) on February 19, 2013 1:03 am wrote:
[snip]
> By "early branch resolution" you mean that the branch simply doesn't predict and stall for the condition?

No, I was thinking of using early availability of a branch condition to potentially override any prediction. However, in some (quite possibly rare) cases, stalling may be a better option than prediction or dynamic predication. Fine-grained or simultaneous multithreading can reduce the cost of such a stall. If a high-latency load severely constrains the extractable parallelism, stalling an unpredictable branch could be more efficient than the alternatives and possibly even higher performance (depending on the cost of a misprediction).

(As I think I mentioned already, early resolution can also have benefits similar to good confidence estimation. Knowing that a path is correct [even if excluding exceptions] can facilitate more optimal allocation of resources [e.g., possibly freeing registers early] or accelerate the commitment of a memory transaction [even in a non-TM system one could conceive of forwarding store buffer values to a requester speculatively when the probability of the stores not actually occurring is tiny and revocation is possible].

> And its candidate usecase is limited to high-misprediction (or
> use prediction), not-in-loop-body (or use CFD) branches.
> And compiler has to shuffle code to fill a potentially large and unpredictable
> bubble, which IMO is unlikely to find that much work to fill.

I agree that the use cases seem to be relatively limited. However, some unpredictable branches are not friendly to predication (especially multi-way and indirect branches) and control flow decoupling seems too novel to be adopted as a reasonably safe design point.

(I think I already mentioned that branches that can be resolved early might also tend to be relatively predictable anyway, reducing the benefit of early resolution.)

I suspect that in general a reorganization of code which would hurt performance (or power efficiency) under perfect prediction would be undesirable, so the question would largely be about the choice of how much compile time should be allocated. The size of the bubble is somewhat less important if earlier resolution always notices mispredictions earlier. Assuming the compiler is targeting good performance on multiple likely implementations, a moderately large bubble assumption might be appropriate (processors that usually experience a larger bubble are likely to have better predictors and the benefit of early condition setting is less likely to be worth the compiler effort and possible performance harm).

David I. August et al.'s "Architectural Support Compiler-Synthesized Dynamic Branch Prediction Strategies: Rationale and Initial Results" (1997) has a graph (Figure 2, page 2) where the average among several benchmarks was 30% of branches (ranging from under 5% to over 90%--for word count!) could be resolved 6 cycles early (on a 4-wide machine). Of course, that was also assuming aggressive scheduling and predication (which would tend to add "useless" instructions and increase the flexibility of scheduling), so the potential for a less EPIC ISA might be substantially lower. However, I think that an aggressive compiler could do better than what was typically done for delayed branches (IIRC, filling two delay slots was difficult).

(A similar problem applies to load scheduling and prefetching. Just because the targeted processors have extensive OoO capability and fast caches with good hardware prefetching does not mean that loads should be scheduled immediately before use or that prefetching is necessarily completely useless. Admittedly, there may be more profitable efforts for the compiler writer than trying to do clever scheduling for a small to minuscule benefit. Perhaps in open source development there might be some who would enjoy such scheduling problems more than algorithmic transformation or data layout optimizations.)

I would also note that, despite what Linus Torvalds seems to think, not all processors are Great Big OoO processors. For the low end (say, 3 to 5 pipeline stages, scalar) multiple condition codes would mainly only make sense if certain condition values are relatively persistent (where such might save some energy, otherwise a single condition code is likely quite adequate to provide early branch resolution). A bit higher in performance (say, up to 9 pipeline stages, two-wide, in-order) multiple condition codes could have a performance benefit. Even higher performance (9+ pipeline stages, 2+ issue, OoO) might make multiple condition codes a slight disadvantage (considering the extra complexity in renaming and data flow arranging). At the highest performance, multiple condition codes might not make much difference.

Early condition code setting seems to be kind of like delayed branches without the need to architect the delay; it can benefit some lower end designs, is a bit painful in the middle, and in the noise at the high end.
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/28 12:47 PM
  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 03:42 AM
    Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 05:09 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 06:01 AM
        Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 12:35 PM
          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 01:42 PM
            Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 11:12 PM
              Article Online: Microservers must Specialize to SurviveMichael S2013/01/30 01:06 AM
              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 08:58 AM
              Article Online: Microservers must Specialize to SurviveDoug S2013/01/30 11:38 AM
                FRU type and granularity selectionPaul A. Clayton2013/01/30 01:29 PM
                  FRU type and granularity selectionrwessel2013/01/30 03:00 PM
                    FRU type and granularity selectionEtienne2013/01/31 02:26 AM
                Article Online: Microservers must Specialize to SurviveMaynard Handley2013/02/13 01:46 PM
                  Article Online: Microservers must Specialize to SurviveDavid Kanter2013/02/16 02:16 PM
                    Article Online: Microservers must Specialize to SurviveDoug S2013/02/17 11:24 AM
        Article Online: Microservers must Specialize to SurviveJohan2013/02/01 12:40 AM
          Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/02/01 02:06 AM
            Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/02/01 02:13 AM
    Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 05:49 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 06:19 AM
        Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 08:47 AM
          Article Online: Microservers must Specialize to Survivebakaneko2013/01/29 01:18 PM
            Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 05:14 PM
              Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 10:51 PM
                Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 07:34 AM
                  Article Online: Microservers must Specialize to Surviveanon2013/01/31 05:45 AM
                    Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 11:25 AM
                      Article Online: Microservers must Specialize to Surviveanon2013/01/31 11:47 AM
                        Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 12:36 PM
                          Article Online: Microservers must Specialize to Surviveanon2013/01/31 03:59 PM
                            Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 04:37 PM
                              Article Online: Microservers must Specialize to Surviveanon2013/01/31 05:47 PM
                                Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 06:16 PM
                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/31 06:36 PM
                                    Article Online: Microservers must Specialize to Surviveanon2013/02/01 12:47 AM
                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 03:26 AM
                                        Article Online: Microservers must Specialize to Surviveanon2013/02/01 04:56 AM
                                          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 05:35 AM
                                            Article Online: Microservers must Specialize to SurviveMichael S2013/02/01 06:07 AM
                                            Article Online: Microservers must Specialize to Surviveanon2013/02/01 07:08 AM
                                              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 07:49 AM
                                                Article Online: Microservers must Specialize to Surviveanon2013/02/01 08:36 AM
                                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/01 09:11 AM
                                                    Article Online: Microservers must Specialize to Surviveanon2013/02/01 05:57 PM
                                                      Article Online: Microservers must Specialize to Surviverwessel2013/02/02 12:42 AM
                                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 05:41 AM
                                                        Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 01:14 PM
                                                          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 01:50 PM
                                                            Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 03:05 PM
                                                              Article Online: Microservers must Specialize to SurviveMichael S2013/02/02 03:20 PM
                                                              Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 04:05 PM
                                                                ASP estimatesRichard Cownie2013/02/02 05:07 PM
                                                                  ASP estimatesRicardo B2013/02/02 05:48 PM
                                                                    ASP estimatesRichard Cownie2013/02/02 06:04 PM
                                                                      ASP estimatesanon2013/02/02 09:19 PM
                                                                        ASP estimatesRichard Cownie2013/02/03 05:06 AM
                                                                          ASP estimates...cache?Per Hesselgren2013/02/03 11:08 AM
                                                                      ASP estimatesMichael S2013/02/03 01:51 AM
                                                                        ASP estimatesRichard Cownie2013/02/03 05:10 AM
                                                                          ASP estimatesDaniel Bizo2013/02/04 02:45 PM
                                                                            ASP estimatesRichard Cownie2013/02/04 04:01 PM
                                                                              ASP estimatesJS2013/02/05 03:25 AM
                                                                              ASP estimatesDaniel Bizo2013/02/05 05:31 AM
                                                                                ASP estimatesRichard Cownie2013/02/05 05:46 AM
                                                                  ARM OEMs get free silicon?someone2013/02/03 12:44 PM
                                                                    ARM OEMs get free silicon?Richard Cownie2013/02/03 07:28 PM
                                                                      ARM OEMs get free silicon?Doug S2013/02/04 10:08 AM
                                                                        ARM OEMs get free silicon?Richard Cownie2013/02/04 11:17 AM
                                                                          ARM OEMs get free silicon?geeker2013/02/06 11:16 AM
                                                                            ARM OEMs get free silicon?Gabriele Svelto2013/02/07 03:18 AM
                                                                              ARM OEMs get free silicon?izotop2013/02/07 05:11 AM
                                                                            ARM OEMs get free silicon?Richard Cownie2013/02/07 05:48 AM
                                                                Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 05:31 PM
                                                                  Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 05:55 PM
                                                                    Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 06:25 PM
                                                                      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/02 06:50 PM
                                                                        Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 07:22 PM
                                                                        Article Online: Microservers must Specialize to SurviveBrendan2013/02/02 11:48 PM
                                                                        Article Online: Microservers must Specialize to SurviveMichael S2013/02/03 01:22 AM
                                                                    Article Online: Microservers must Specialize to Surviveanon2013/02/02 09:21 PM
                                                                      Article Online: Microservers must Specialize to SurviveMichael S2013/02/03 01:40 AM
                                                                        Article Online: Microservers must Specialize to Surviveanon2013/02/03 02:54 AM
                                                                      Overlooking the advantage (or not) of x86 in different marketsDoug S2013/02/03 12:54 PM
                                                                Article Online: Microservers must Specialize to Surviveanon2013/02/02 09:14 PM
                                                                  Article Online: Microservers must Specialize to Surviveanon2013/02/02 09:32 PM
                                                                    Article Online: Microservers must Specialize to SurviveRichard Cownie2013/02/03 05:24 AM
                                                        Article Online: Microservers must Specialize to Surviveanon2013/02/02 09:08 PM
                                          Article Online: Microservers must Specialize to SurviveMichael S2013/02/01 05:41 AM
                                            Article Online: Microservers must Specialize to Surviveanon2013/02/01 06:41 AM
                            Article Online: Microservers must Specialize to SurviveDoug S2013/02/01 03:41 PM
                              Article Online: Microservers must Specialize to Surviveanon2013/02/01 06:08 PM
                                Article Online: Microservers must Specialize to SurviveDoug S2013/02/02 01:38 PM
                                  Article Online: Microservers must Specialize to SurviveRicardo B2013/02/02 02:03 PM
              Article Online: Microservers must Specialize to Survivebakaneko2013/01/30 12:16 AM
        Article Online: Microservers must Specialize to Surviveanon2013/01/29 10:34 PM
    Article Online: Microservers must Specialize to Survivesomeone2013/01/29 07:58 AM
      Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/29 08:28 AM
      Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 08:42 AM
      Article Online: Microservers must Specialize to SurviveMark Roulo2013/01/29 05:08 PM
        Article Online: Microservers must Specialize to Surviverwessel2013/01/29 09:15 PM
          Article Online: Microservers must Specialize to SurviveRichard Cownie2013/01/30 01:39 PM
            Article Online: Microservers must Specialize to Surviverwessel2013/01/30 02:54 PM
          Or compare to *desktop* ASPsMark Roulo2013/01/30 02:51 PM
            Or compare to *desktop* ASPsrwessel2013/01/30 03:07 PM
            Or compare to *desktop* ASPsCeleron2013/01/30 04:49 PM
            Or compare to *desktop* ASPsDoug S2013/01/30 08:50 PM
              Or compare to *desktop* ASPsrwessel2013/01/30 10:04 PM
              when thoughput matters, ultra-cheap Intel CPUs are not attractiveMichael S2013/01/31 01:59 AM
                when thoughput matters, ultra-cheap Intel CPUs are not attractiveDoug S2013/01/31 11:00 AM
                  when thoughput matters, ultra-cheap Intel CPUs are not attractiveMichael S2013/01/31 01:04 PM
                    Price/perf curve on intel.Jouni Osmala2013/01/31 10:28 PM
                    when thoughput matters, ultra-cheap Intel CPUs are not attractiveDoug S2013/02/01 03:45 PM
                      when thoughput matters, ultra-cheap Intel CPUs are not attractiverwessel2013/02/02 12:54 AM
              Or compare to *desktop* ASPsGabriele Svelto2013/01/31 02:33 AM
    Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 09:36 AM
      Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 12:52 PM
        Article Online: Microservers must Specialize to SurviveARMandLeg2013/01/29 02:31 PM
        Article Online: Microservers must Specialize to SurviveNo2013/01/29 02:32 PM
          Article Online: Microservers must Specialize to SurviveMichael S2013/01/29 03:36 PM
            Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 06:25 PM
        Article Online: Microservers must Specialize to SurviveDoug S2013/01/29 06:20 PM
          Article Online: Microservers must Specialize to SurviveEduardoS2013/01/30 02:48 AM
  Article Online: Microservers must Specialize to SurvivePer Hesselgren2013/01/29 08:08 AM
  Article Online: Microservers must Specialize to Survivecarop2013/01/29 01:55 PM
    Article Online: Microservers must Specialize to SurviveDavid Kanter2013/01/29 10:33 PM
      Article Online: Microservers must Specialize to SurviveGabriele Svelto2013/01/31 02:51 AM
    Article Online: Microservers must Specialize to Survivebakaneko2013/01/30 12:32 AM
      Cache capacity benefits (babbling a bit)Paul A. Clayton2013/01/30 12:42 PM
        Cache size experiments may have been done on ARMsMark Roulo2013/01/30 01:14 PM
          Way and line locking and associativityPaul A. Clayton2013/01/30 02:08 PM
            Way and line locking and associativityMark Roulo2013/01/30 02:37 PM
        Cache capacity benefits may have been done in simulationPatrick Chase2013/01/30 06:25 PM
          Simulation has benefits but is SLOWPaul A. Clayton2013/01/30 08:08 PM
            Simulation has benefits but is SLOWPatrick Chase2013/01/30 09:36 PM
              Simulation has benefits but is SLOWRichard Cownie2013/02/01 11:32 AM
                Simulation has benefits but is SLOWPatrick Chase2013/02/01 12:06 PM
                Simulation has benefits but is SLOWnone2013/02/01 12:13 PM
      Article Online: Microservers must Specialize to Survivesoupdragon2013/01/31 11:02 AM
    Article Online: Microservers must Specialize to SurvivePatrick Chase2013/01/30 11:26 AM
  4x or 10x advantageS. Rao2013/01/30 04:35 PM
    4x or 10x advantagerwessel2013/01/30 05:32 PM
      4x or 10x advantageS. Rao2013/02/03 09:32 PM
        4x or 10x advantageDavid Kanter2013/02/03 10:09 PM
        4x or 10x advantagerwessel2013/02/03 11:08 PM
          x86_64Richard Cownie2013/02/04 04:52 PM
            x86_64Michael S2013/02/05 01:51 AM
              x86_64Richard Cownie2013/02/05 03:33 AM
                x86_64EduardoS2013/02/05 03:41 AM
                  x86_64sJ2013/02/05 04:52 AM
                  x86_64Richard Cownie2013/02/05 05:47 AM
                x86_64someone2013/02/05 08:22 AM
                  x86_64Richard Cownie2013/02/05 08:44 AM
              "Just recompile"Mark Roulo2013/02/05 07:40 AM
                "Just recompile"Michael S2013/02/05 09:35 AM
                  "Just recompile"Mark Roulo2013/02/05 09:55 AM
  Virtualization and custom interconnectsGabriele Svelto2013/01/31 04:05 AM
  x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/01 09:11 PM
    x86 penalty depends on performance and microarchitecturempx2013/02/02 03:31 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 08:27 AM
        Heterogenous multicore in serversPaul A. Clayton2013/02/02 09:25 AM
    x86 penalty depends on performance and microarchitectureanon2013/02/02 04:04 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 08:42 AM
        x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 08:58 AM
        x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 09:26 AM
        x86 penalty depends on performance and microarchitectureanon2013/02/02 04:49 PM
          x86 penalty depends on performance and microarchitecturePer Hesselgren2013/02/03 12:44 AM
            x86 penalty depends on performance and microarchitectureMichael S2013/02/03 01:59 AM
          x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 01:27 PM
            x86 penalty depends on performance and microarchitectureanon2013/02/03 01:58 PM
              x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 02:35 PM
                x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 02:39 PM
                x86 penalty depends on performance and microarchitectureanon2013/02/03 05:08 PM
                  x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/03 07:05 PM
                    x86 penalty depends on performance and microarchitectureanon2013/02/03 08:52 PM
                      x86 penalty depends on performance and microarchitectureDavid Kanter2013/02/03 10:23 PM
                        U shape curve on performance...Jouni Osmala2013/02/04 12:07 AM
                          U shape curve on performance...none2013/02/04 02:03 AM
                            U shape curve on performance...Michael S2013/02/04 06:55 AM
                              U shape curve on performance...Patrick Chase2013/02/04 09:14 AM
                                U shape curve on performance...Michael S2013/02/04 01:43 PM
                                  U shape curve on performance...Patrick Chase2013/02/04 03:06 PM
                            Microcode can be usefulPaul A. Clayton2013/02/04 07:08 AM
                          U shape curve on performance...Michael S2013/02/04 06:48 AM
                          U shape curve on performance...Patrick Chase2013/02/04 09:05 AM
                            It only works if ... ??Paul A. Clayton2013/02/04 11:20 AM
                              It only works if ... ??Patrick Chase2013/02/04 11:48 AM
                                It only works if ... ??Patrick Chase2013/02/04 12:02 PM
                                It only works if ... ??Patrick Chase2013/02/04 12:08 PM
                                  It only works if ... ??Patrick Chase2013/02/04 12:15 PM
                                  It only works if ... ??anon2013/02/04 04:45 PM
                            I don't follow...Megol2013/02/05 09:16 AM
                      x86 penalty depends on performance and microarchitecturerwessel2013/02/03 11:28 PM
            spec95 vs Spec2kMichael S2013/02/04 01:27 AM
              spec95 vs Spec2kPer Hesselgren2013/02/04 03:42 AM
              spec95 vs Spec2kbakaneko2013/02/04 03:48 AM
                correct linksMichael S2013/02/04 05:10 AM
              spec95 vs Spec2kPatrick Chase2013/02/04 10:47 AM
                spec95 vs Spec2kPatrick Chase2013/02/04 10:55 AM
                  spec95 vs Spec2kMichael S2013/02/04 12:39 PM
                    spec95 vs Spec2kPatrick Chase2013/02/04 01:07 PM
                    spec95 vs Spec2knone2013/02/04 03:28 PM
                      spec95 vs Spec2kMichael S2013/02/05 01:35 AM
                spec95 vs Spec2kMichael S2013/02/04 12:50 PM
                spec95 vs Spec2k (hair-splitting)Kira2013/02/04 02:26 PM
    x86 penalty depends on performance and microarchitecturenone2013/02/02 04:33 AM
      x86 penalty depends on performance and microarchitecturePatrick Chase2013/02/02 09:11 AM
        x86 penalty depends on performance and microarchitecturenone2013/02/02 09:43 AM
          Cortex A9 ROBPatrick Chase2013/02/02 12:13 PM
            Cortex A9 ROBPatrick Chase2013/02/02 01:01 PM
              Cortex A9 ROBnone2013/02/04 03:12 PM
                Cortex A9 ROBPatrick Chase2013/02/04 03:43 PM
                  Cortex A9 ROBnone2013/02/04 03:58 PM
    ARM diversity might be an advantagePaul A. Clayton2013/02/02 10:10 AM
      ARM diversity might be an advantagePatrick Chase2013/02/03 03:29 PM
        ARMv8 cleanlinessanon2013/02/03 05:11 PM
          ARMv8 cleanlinessPatrick Chase2013/02/03 06:40 PM
            Does AArch64 not support table node page sizes?Paul A. Clayton2013/02/03 08:40 PM
              Does AArch64 not support table node page sizes?Patrick Chase2013/02/03 09:14 PM
                Does AArch64 not support table node page sizes?none2013/02/04 01:55 AM
              LPAE supports 2MiB and 1 GiB blocksPaul A. Clayton2013/02/03 09:52 PM
            ARMv8 cleanlinessanon2013/02/03 09:06 PM
              ARMv8 cleanlinessrwessel2013/02/03 11:37 PM
                ARMv8 cleanlinessMichael S2013/02/04 12:50 AM
              ARMv8 cleanlinessEtienne2013/02/04 07:01 AM
                Similar to Thumb If-Then instructionPaul A. Clayton2013/02/04 07:43 AM
                  Similar to Thumb If-Then instructionEtienne2013/02/05 06:22 AM
                ARMv8 cleanlinessJouni Osmala2013/02/04 10:15 PM
                  ARMv8 cleanlinessEtienne2013/02/05 05:57 AM
                    ARMv8 cleanlinessJouni Osmala2013/02/05 06:59 AM
                      ARMv8 cleanlinessEtienne2013/02/05 07:55 AM
              ARMv8 cleanlinessMaynard Handley2013/02/13 02:53 PM
                ARMv8 cleanlinessGabriele Svelto2013/02/14 12:37 AM
                ARMv8 cleanlinessEtienne2013/02/14 05:00 AM
                  ARMv8 cleanlinessrwessel2013/02/14 03:03 PM
                ARMv8 cleanlinessSeni2013/02/14 06:45 AM
                  ARMv8 cleanlinessrwessel2013/02/14 03:14 PM
                    ARMv8 cleanlinessLinus Torvalds2013/02/14 03:44 PM
                      ARMv8 cleanlinessGabriele Svelto2013/02/15 01:28 AM
                        ARMv8 cleanlinessEduardoS2013/02/15 09:45 AM
                        ARMv8 cleanlinessMax2013/02/15 12:24 PM
                          ARMv8 cleanlinessWilco2013/02/15 04:46 PM
                            ARMv8 cleanlinessLinus Torvalds2013/02/15 05:34 PM
                              ARMv8 cleanlinessWilco2013/02/17 12:47 PM
                                ARMv8 cleanlinessLinus Torvalds2013/02/17 03:08 PM
                                  ARMv8 cleanlinessWilco2013/02/18 04:14 PM
                                    ARMv8 cleanlinessanon2013/02/19 12:56 AM
                                      ARMv8 cleanlinessWilco2013/02/19 01:20 PM
                                    ARMv8 cleanlinessLinus Torvalds2013/02/19 09:13 AM
                                      ARMv8 cleanlinessWilco2013/02/19 01:10 PM
                                        ARMv8 cleanlinessLinus Torvalds2013/02/19 05:16 PM
                                          ARMv8 cleanlinessLinus Torvalds2013/02/19 05:41 PM
                                            ARMv8 cleanlinessWilco2013/02/25 05:05 AM
                                              ARMv8 cleanlinessLinus Torvalds2013/02/25 05:22 PM
                                                ARMv8 cleanlinessnone2013/02/26 12:31 AM
                                                  ARMv8 cleanlinessMichael S2013/02/26 04:05 AM
                                                ARMv8 cleanlinessDavid Hess2013/02/26 10:01 PM
                                                  ARMv8 cleanlinessLinus Torvalds2013/02/27 10:22 AM
                                                    ARMv8 cleanlinessDavid Hess2013/03/01 12:58 PM
                                                      What is wrong with ARM?Paul A. Clayton2013/03/01 05:56 PM
                                                        What is wrong with ARM?David Hess2013/03/01 06:49 PM
                                                          What is wrong with ARM?Michael S2013/03/02 04:11 PM
                                                            What is wrong with ARM?David Hess2013/03/03 10:32 PM
                                                              What is wrong with ARM?Michael S2013/03/04 01:05 AM
                                                        What is wrong with ARM?David Hess2013/03/07 05:30 PM
                                                          Thank you very much!Paul A. Clayton2013/03/07 06:59 PM
                                                            Thank you very much!David Hess2013/03/07 10:22 PM
                                                              Thank you very much!gallier22013/03/08 12:43 AM
                                                                Thank you very much!David Hess2013/03/08 10:24 AM
                                                                  Thank you very much!Ungo2013/03/08 03:37 PM
                                                                    Thank you very much!David Hess2013/03/08 07:47 PM
                                                            Thank you very much!Ricardo B2013/03/08 04:07 AM
                                                              ARM lackings a historical accident?Paul A. Clayton2013/03/08 07:54 AM
                                                                ARM lackings a historical accident?Ricardo B2013/03/08 02:33 PM
                                                          What is wrong with ARM?Michael S2013/03/08 05:25 AM
                                                            What is wrong with ARM?David Hess2013/03/08 11:37 AM
                                                            What is wrong with ARM?dmsc2013/03/12 02:47 PM
                                                          What is wrong with ARM?dmsc2013/03/12 02:43 PM
                                                      ARMv8 cleanlinessMichael S2013/03/02 04:06 PM
                                                    ParityAndi Kleen2013/03/03 12:47 PM
                                                      ParityLinus Torvalds2013/03/03 02:06 PM
                                                        ParityAndi Kleen2013/03/03 06:15 PM
                                            Except the borrow bitMatthew D2013/02/25 11:07 PM
                                              Except the borrow bitrwessel2013/02/25 11:42 PM
                                              Except the borrow bitLinus Torvalds2013/02/26 10:13 AM
                                                are ZSCO flags really fundamental?Michael S2013/02/27 03:24 PM
                                                  are ZSCO flags really fundamental?Linus Torvalds2013/02/27 04:41 PM
                                                    are ZSCO flags really fundamental?Michael S2013/02/28 12:39 AM
                                                    are ZSCO flags really fundamental?2013/03/01 09:23 AM
                                                  are ZSCO flags really fundamental?EduardoS2013/02/27 10:07 PM
                                                    are ZSCO flags really fundamental?Michael S2013/02/28 03:58 AM
                                                  are ZSCO flags really fundamental?rwessel2013/02/27 11:29 PM
                                                    are ZSCO flags really fundamental?rwessel2013/02/27 11:32 PM
                                                    Bit numbering on big endian machinesKonrad Schwarz2013/02/28 12:49 AM
                                                      Bit numbering on big endian machinesrwessel2013/02/28 01:32 AM
                                          ARMv8 cleanlinessWilco2013/02/25 04:54 AM
                                            ARMv8 cleanlinessEduardoS2013/02/25 05:24 AM
                                              ARMv8 cleanlinessWilco2013/02/25 06:36 AM
                                                ARMv8 cleanlinessEduardoS2013/02/25 12:06 PM
                                                  ARMv8 cleanlinessWilco2013/02/25 04:24 PM
                                                    ARMv8 cleanlinessEduardoS2013/02/25 04:55 PM
                                                    Utility of data loss exceptionsPaul A. Clayton2013/02/25 07:18 PM
                                                      Utility of data loss exceptionsrwessel2013/02/26 12:00 AM
                                                        Excessively expensive mode changingPaul A. Clayton2013/02/26 05:15 AM
                                                      Utility of data loss exceptionsEduardoS2013/02/26 01:55 AM
                                                        Utility of data loss exceptionsTREZA2013/02/26 03:53 AM
                                                          Utility of data loss exceptionsEduardoS2013/02/26 04:37 AM
                                                      Utility of data loss exceptionsEtienne2013/02/26 05:01 AM
                                            1b vs. 4b not a huge difference?Paul A. Clayton2013/02/25 08:57 AM
                                  Possible benefit of multi-bit CCPaul A. Clayton2013/02/23 07:09 AM
                                ARMv8 cleanlinessrwessel2013/02/17 11:01 PM
                                  ARMv8 cleanlinessWilco2013/02/18 04:33 PM
                                    ARMv8 cleanlinessEduardoS2013/02/18 07:18 PM
                                      ARMv8 cleanlinessWilco2013/02/25 05:19 AM
                      ARMv8 cleanlinessTREZA2013/02/15 03:21 AM
                      ARMv8 cleanliness2013/02/17 09:49 AM
                        ARMv8 cleanlinessLinus Torvalds2013/02/17 10:54 AM
                          ARMv8 cleanlinessLinus Torvalds2013/02/17 10:58 AM
                          ARMv8 cleanlinessnone2013/02/17 03:36 PM
                        ARMv8 cleanlinessTREZA2013/02/17 10:59 AM
                          ARMv8 cleanlinessWilco2013/02/17 11:30 AM
                            and-compare and or-comparePaul A. Clayton2013/02/17 04:28 PM
                              early branch resolutionanon2013/02/18 01:05 AM
                                My ignorance about compilers is vasterPaul A. Clayton2013/02/18 12:30 PM
                                early branch resolutionanon2013/02/18 04:12 PM
                              early branch resolution?EBFE2013/02/19 12:03 AM
                                Early resolution: some but limited potentialPaul A. Clayton2013/02/19 04:52 PM
                                  Early resolution: some but limited potentialhobold2013/02/20 01:02 PM
                                    PPC750 was an inspirationPaul A. Clayton2013/02/23 07:09 AM
                          ARMv8 cleanlinessTREZA2013/02/17 12:53 PM
                            ARMv8 cleanlinessExophase2013/02/17 03:49 PM
                        ARMv8 cleanlinessWilco2013/02/17 11:07 AM
                        ARMv8 cleanlinessbakaneko2013/02/17 11:43 AM
                        ARMv8 cleanlinessJouni Osmala2013/02/17 12:10 PM
                          -1 comparison true seems betterPaul A. Clayton2013/02/17 05:07 PM
                  ARMv8 cleanlinessEtienne2013/07/17 01:58 AM
      ARM server benchmark- why Apache?Per Hesselgren2013/02/04 11:13 PM
        ARM server benchmark- why Apache?anon2013/02/04 11:38 PM
          ARM server benchmark- why Apache?Per Hesselgren2013/02/05 12:25 AM
            ARM server benchmark- why Apache?Michael S2013/02/05 01:13 AM
          ARM server benchmark- why Apache?Michael S2013/02/05 12:51 AM
          I think that was debunked alreadyDavid Kanter2013/02/05 06:36 PM
        ARM server benchmark- why Apache?bakaneko2013/02/06 01:37 AM
          ARM server benchmark- why Apache?Michael S2013/02/06 02:43 AM
            Marketing to PHBs?Paul A. Clayton2013/02/06 05:40 AM
            ARM server benchmark- why Apache?bakaneko2013/02/06 06:08 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?