Announcement

Collapse
No announcement yet.

The Huge Disaster Within The Linux 2.6.35 Kernel

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    Originally posted by Michael View Post
    That though wouldn't go towards addressing the fundamental problem that this article is about: how such a glaringly severe regression can be pulled into the tree in the first place and then live there for days. Improving the status quo is what this article is intended to be about more than this bug per se.
    I think what you mean is if the kernel developers paid you to run this service they'd get this information directly rather than you writing an article on it.

    Comment


    • #47
      Originally posted by FireBurn View Post
      I think what you mean is if the kernel developers paid you to run this service they'd get this information directly rather than you writing an article on it.
      Errm no, I am not looking for the kernel developers to pay me. As I've said before I am happy to run such trackers for free software projects such as the kernel as long as they are being used well.
      Michael Larabel
      http://www.michaellarabel.com/

      Comment


      • #48
        Originally posted by Michael View Post
        That though wouldn't go towards addressing the fundamental problem that this article is about: how such a glaringly severe regression can be pulled into the tree in the first place and then live there for days. Improving the status quo is what this article is intended to be about more than this bug per se.
        Uh. Because it still works? Linus's development model involves massively pulling commits he likes and then spending several RC releases hammering away at bugs and regressions.

        Hell, you yourself have reported on this. How is this suddenly a problem for you?

        Comment


        • #49
          Originally posted by bulletxt View Post
          People like you deserve to have a toilet kernel, and then I'de love to see you saying: "hey michael, you do 24/h benchmarks and didn't say anything to us!!!"

          That's what all of you complaining of Michael's work deserve.
          +1

          "We're getting bloated, yes it's a problem," Torvalds said, "I'd love to say we have a plan. I mean, sometimes it's a bit sad and we're definitely not the streamlined hyper-efficient kernel that I had envisioned 15 years ago. The kernel is huge and bloated, and our icache footprint is scary. I mean, there is no question about that. And whenever we add a new feature, it only gets worse."

          Comment


          • #50
            Originally posted by dyna View Post
            All the info needed to create a bug report is here so almost any reader could do it. So why haven't YOU done it yet? (or me for that matter).
            Exactly.

            Thanks for the benchmarks, Michael.

            Comment


            • #51
              I have two problems with the article,

              1. It's over-inflammatory - what is Phoronix trying to be? The tech-equivalent of The Sun?
              2. The bug wasn't reported

              As plenty of people have mentioned, this is still the merge window, it's possible that a further merge will rectify the issue. If this was happening at RC1 or 2 then fair enough, but even then "disaster" is way over the top.

              Is it really that hard to shoot a mail over the the LKML saying "someone might wanna take a look at this performance regression that our benchmarks picked up"? It's just the right thing to do, especially if your entire business is based on Linux like Phoronix's is.

              Comment


              • #52
                Originally posted by Michael View Post
                Errm no, I am not looking for the kernel developers to pay me. As I've said before I am happy to run such trackers for free software projects such as the kernel as long as they are being used well.
                Is your point that none of the developers noticed the regression on Phoromatic which you host for free?

                If your tool is noticing regressions based on specific commits why not get Phoromatic to e-mail directly the committer with an automated warning?

                Hopefully the developer will either get it fixed, reverted or explain why a regression was necessary. The latter would provide extra information for your articles so you wont need to pad them out with so many graphs.

                As numerous people have said the merge window is for getting all the new code in (which in it's self is a huge achievement)

                The specific commit causing this regression might be aggravated by other new code say in VFS that wouldn't be seen in the developers local testing

                Stop writing like a Rupert Murdoch tabloid rag, you've produced many great articles in the past, why have you been resorting to trash writing over the last few months?

                Comment


                • #53
                  Well, I don't understand all the fuss. Phoronix is a site that REPORTS news related to linux. The journalists job is only to report what they see, they don't interfere with what's going on. That's one of the rules of journalism AFAIK. So as a journalist, Michael sees a problem on the kernel development and reports it. End of story. As an individual he may file a bug report or not, but that's not our business. He's not supposed to save the world from world war 3 or erradicate starvation. On the same article he also points out that the kernel devs can use the freely availabe services of the phoronix test suite to better deal with the kind of problems reported. What's the problem in that?

                  Comment


                  • #54
                    Well I'll admit I didn't read the article at first (the headline sounded too sensationalist for me to take it seriously - but that's a comment more about me than anything else), after having read the article, I would point out that it was clearly stated that the kernel is not ready for mainstream use.
                    As for waiting - well, I'll admit I don't know what "organically caught and addressed" refers to (unless skynet is also maintaining the linux kernel) but I can understand wait a little to see if something has been committed that "just works", and if the developer will fix it to work properly later. Breathing down the necks of developers for every single commit might not help them much.
                    I will however state that the article does still sound a bit sensationalist, and as a result comes across the wrong way (rather obvious from the comments posted).
                    It still will be interesting to know how performance goes as development on 2.6.35 continues.

                    Comment


                    • #55
                      Originally posted by devius View Post
                      Well, I don't understand all the fuss. Phoronix is a site that REPORTS news related to linux. The journalists job is only to report what they see, they don't interfere with what's going on. That's one of the rules of journalism AFAIK. So as a journalist, Michael sees a problem on the kernel development and reports it. End of story. As an individual he may file a bug report or not, but that's not our business. He's not supposed to save the world from world war 3 or erradicate starvation. On the same article he also points out that the kernel devs can use the freely availabe services of the phoronix test suite to better deal with the kind of problems reported. What's the problem in that?
                      Perhaps they just love acting like trolls. Perhaps they're just fanatics (40% regression reports in the new Linux kernel? never! just report the fact to the devs and shut up pleez.)

                      OTOH if this benchmark was about Windows' kernel, I suspect the same persons would be a lot more condescending

                      Comment


                      • #56
                        Originally posted by zoomblab View Post
                        Says it all.
                        wrong. That there is no mail from Michael on lkml - THAT says it all.

                        There always have been performance regressions been caught by some benchmarks - people reported it on lkml, they got fixed. Without alarming 'oh my god we are all going to die!!!!' articles on some site.

                        Comment


                        • #57
                          Originally posted by zoomblab View Post
                          Thing is, bugs like this shouldn't be allowed to enter into the main repository. Again, proper practices and test procedures...
                          wrong. And wrong. Linux kernel development does not work that way. Just look up Greg's videos about it. You can not keep up that high level of development if you do it the old style.

                          That makes testing more important.

                          And even more important is reporting your findings on lkml.

                          Phoronix failed that very hard.

                          Compare phoronix with the many posts of performance regressions on lkml - and look at the outcome:
                          stuff gets reported, stuff got fixed.

                          This might be hard to crasp for lazy 'I don't test and I don't report' people.

                          Comment


                          • #58
                            This is a troll bait article I would expect on digg from a person who knows nothing about Linux or kernel development. This sadly brings phoronix down to a new low.

                            If phoronix editors had any decency they would retract this entire article and apologize to their readers.

                            Comment


                            • #59
                              Originally posted by Michael View Post
                              That though wouldn't go towards addressing the fundamental problem that this article is about: how such a glaringly severe regression can be pulled into the tree in the first place and then live there for days. Improving the status quo is what this article is intended to be about more than this bug per se.
                              a) it is the merge window. Most devs are busy merging their stuff and unbreak build problems caused by this.

                              b) if someone does not report his findings, how should they know?

                              Other people report regressions on lkml - and they get fixed. Did you?

                              Comment


                              • #60
                                Originally posted by energyman View Post
                                Other people report regressions on lkml - and they get fixed. Did you?
                                Question: did YOU? "Other people," eh?

                                Comment

                                Working...
                                X