Re: [Haifux] Oops your system with plain malloc()
Hello Eli, On Thu, 17 May 2012 21:31:16 +0300 Eli Billauer e...@billauer.co.il wrote: On 05/17/2012 08:56 PM, Shlomi Fish wrote: Oh wait! You're using kernel-2.6.35.4 - not kernel 3.5.x (which does not exist yet). That's a really old version. Is it an up-to-date distribution kernel? If so - you can report it to your distributor. If not - you should upgrade. Of course I should upgrade. That way I'll exchange a problem which has zero impact on my system (currently) with one that will possibly cause real trouble. That's the good thing about upgrading all the time: You have a lot of quality time with you computer, fixing whatever broke when you upgraded to fix the previous problem. I may consider upgrading to a kernel 2.6.35 which has gone through a long phase of bug fixes only but I understand there is no such around at the moment. What I call a vintage kernel. This post by Gabor is of relevance here: http://szabgab.com/what-does--if-it-aint-broke-dont-fix-it--really-mean.html Namely, an upgrade may always introduce some broken things, but if you delay it, then the gap will be larger, and will likely break more things. Furthermore, I have been on development distributions (first Mandriva Cooker and now Mageia Cauldron) and while there were breakages in the past, most of the time, the upgrades went pretty well, and I didn't spend a lot of time in fixing broken things. Regards, Shlomi Fish -- - Shlomi Fish http://www.shlomifish.org/ My Favourite FOSS - http://www.shlomifish.org/open-source/favourite/ We don’t know his cellphone number, and even if we did, we would tell you that we didn’t know it. Please reply to list if it's a mailing list post - http://shlom.in/reply . ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Oops your system with plain malloc()
On 05/17/2012 08:56 PM, Shlomi Fish wrote: Oh wait! You're using kernel-2.6.35.4 - not kernel 3.5.x (which does not exist yet). That's a really old version. Is it an up-to-date distribution kernel? If so - you can report it to your distributor. If not - you should upgrade. Of course I should upgrade. That way I'll exchange a problem which has zero impact on my system (currently) with one that will possibly cause real trouble. That's the good thing about upgrading all the time: You have a lot of quality time with you computer, fixing whatever broke when you upgraded to fix the previous problem. I may consider upgrading to a kernel 2.6.35 which has gone through a long phase of bug fixes only but I understand there is no such around at the moment. What I call a vintage kernel. Eli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Oops your system with plain malloc()
Hi Eli, On Thu, 17 May 2012 20:48:59 +0300 Shlomi Fish shlo...@shlomifish.org wrote: Hi Eli, On Thu, 17 May 2012 19:34:12 +0300 Eli Billauer e...@billauer.co.il wrote: Hi all, I just wrote a post in my blog about Linux (2.6.35 on x86_64) oopsing and hanging when allocating huge amounts of RAM. There are a couple of plain C programs in the post that did the job pretty well. You may want to try them yourselves. Or maybe suggest reasons why this happens at all. This is not an issue for me right now, because no real program does what those test snippets do. And still, there's something worrying about being able to crash the computer completely that easily. http://billauer.co.il/blog/?p=2837 Oh wait! You're using kernel-2.6.35.4 - not kernel 3.5.x (which does not exist yet). That's a really old version. Is it an up-to-date distribution kernel? If so - you can report it to your distributor. If not - you should upgrade. Relevant signature this time. Regards, Shlomi Fish -- - Shlomi Fish http://www.shlomifish.org/ Rethinking CPAN - http://shlom.in/rethinking-cpan Chuck Norris reads all messages posted to LKML (= the Linux Kernel Mailing List), understands them all, and he kills all gnomes he sees in sight. Please reply to list if it's a mailing list post - http://shlom.in/reply . ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Oops your system with plain malloc()
Hi Eli, On Thu, 17 May 2012 19:34:12 +0300 Eli Billauer e...@billauer.co.il wrote: Hi all, I just wrote a post in my blog about Linux (2.6.35 on x86_64) oopsing and hanging when allocating huge amounts of RAM. There are a couple of plain C programs in the post that did the job pretty well. You may want to try them yourselves. Or maybe suggest reasons why this happens at all. This is not an issue for me right now, because no real program does what those test snippets do. And still, there's something worrying about being able to crash the computer completely that easily. http://billauer.co.il/blog/?p=2837 Thanks for the interesting post. On my system (specs below) with 8GB of RAM and kernel Linux telaviv1.shlomifish.org 3.3.6-desktop-1.mga2 #1 SMP Sat May 12 20:01:24 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux memeater does not cause the system to crash. In fact, everything appears to be fine after running memeater. I'll try it with a vanilla 3.5.x kernel next and report it here. Here are my specs: An Intel Core i3 CPU (x86-64). 8 GB of RAM. Intel Corporation Sandy Bridge Integrated Graphics Controller (rev 09) A 2 TB hard-disk. A 19״ LCD Screen by ViewSonic. Intel Corporation Cougar Point High Definition Audio Controller. Intel Corporation 82579V Gigabit Network Connection. Regards, Shlomi Fish -- - Shlomi Fish http://www.shlomifish.org/ What Makes Software Apps High Quality - http://shlom.in/sw-quality Knuth is not God! Unless you confuse him with Dijkstra. Please reply to list if it's a mailing list post - http://shlom.in/reply . ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Oops your system with plain malloc()
On Thu, May 17, 2012 at 09:31:16PM +0300, Eli Billauer wrote: On 05/17/2012 08:56 PM, Shlomi Fish wrote: Oh wait! You're using kernel-2.6.35.4 - not kernel 3.5.x (which does not exist yet). That's a really old version. Is it an up-to-date distribution kernel? If so - you can report it to your distributor. If not - you should upgrade. Of course I should upgrade. That way I'll exchange a problem which has zero impact on my system (currently) with one that will possibly cause real trouble. That's the good thing about upgrading all the time: You have a lot of quality time with you computer, fixing whatever broke when you upgraded to fix the previous problem. I may consider upgrading to a kernel 2.6.35 which has gone through a long phase of bug fixes only but I understand there is no such around at the moment. What I call a vintage kernel. You obviously have a beefy system with memory to spare. You obviously don't really care rebooting it. The logical conclusion: build some kernels and test it. Grab a kernel tree, see if this still crashes v2.6.35, and if so, git bisect is your friend. -- Tzafrir Cohen | tzaf...@jabber.org | VIM is http://tzafrir.org.il || a Mutt's tzaf...@cohens.org.il || best tzaf...@debian.org|| friend ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Oops your system with plain malloc()
On 05/17/2012 10:23 PM, Tzafrir Cohen wrote: I may consider upgrading to a kernel 2.6.35 which has gone through a long phase of "bug fixes only" but I understand there is no such around at the moment. What I call a vintage kernel. You obviously have a beefy system with memory to spare. You obviously don't really care rebooting it. The logical conclusion: build some kernels and test it. Grab a kernel tree, see if this still crashes v2.6.35, and if so, git bisect is your friend. Beefy it is, yes. Which is why it hasn't reached its full memory usage ever during the two years it has been running. Rebooting? Yes, I do care about that. Neither do I want to spend time playing around with this. Actually, I could play around with a virtual machine. I suppose the problem would appear likewise. I wrote to this list in case someone else wants to check his or her computer up. It does look like this was fixed in later kernels, but it just proves my point that no matter which kernel you pick, there's always something. Except, maybe, those kernels that have been maintained for a long while for stability. Eli -- Web: http://www.billauer.co.il ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
