I decided about three weeks ago to go ahead and try 64bit Arch, after about a year of using the i686 version. The issues that held me back appeared to have been resolved, so using some free space on my hard drive I installed, set up a basic system, and proceeded to test things out. Everyhting worked perfectly, including flash and java.
Great. Time to back up my data, repartition (because I had been meaning to change my partition scheme anyway) and Install for real. This was about a week ago. This time, java, or more precisely the java plugin, did not work properly evn though it did when I was testing things out. Hmm, must have been something different about my setup, I figured, so I start trying to figure out what I did differently, including a few reinstalls in the past week on the off chance that the order I installed things might play a role.
No such luck. After extensive searching I can't seem to find any answers to my problem, so I'm hoping someone can point me in the right direction. I am fairly certain that this is not a common issue, and inclined to beleive it is something related to my setup.
OK, here are my symtoms. Upon going to a website (in this case Pogo.com, one that has to work to get my wife to even consider a Windowless home) that uses java applets, java seems to freeze. Specifically trying to launch one of the games (using firefox by the way, in case that helps) launches a new window as usual. This window will show the typical java loading screen. Then it will clear the window and start the next loading phase (try one of the Pogo games if you are not sure what I mean, this is standard behavior at least on their site) and nothing appears to happen. If I leave the new firefox window open long enough (several minutes) it will eventually load, but then trying to play the game is impossible as everything is extremely slow.
So, any thoughts on how to troubleshoot this and where to look? I am usually pretty good about solving my problems, but this has me stumped since it worked on the first install but not the next, using the same firefox and jre packages. I could switch to openjdk (which seemed to work, but required kludges to get sound since I'm using oss, and also caused random browser shutdowns) but would prefer to find a solution to the Sun Java problem.
Last edited by Kevind71 (2009-04-24 18:51:37)
whats ur system look like
I'm not really sure what this tells you, but here is the cpuinfo:
processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 107 model name : AMD Athlon(tm) 64 X2 Dual Core Processor 4400+ stepping : 1 cpu MHz : 2294.089 cache size : 512 KB physical id : 0 siblings : 2 core id : 0 cpu cores : 2 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow rep_good pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch bogomips : 4590.35 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc 100mhzsteps processor : 1 vendor_id : AuthenticAMD cpu family : 15 model : 107 model name : AMD Athlon(tm) 64 X2 Dual Core Processor 4400+ stepping : 1 cpu MHz : 2294.089 cache size : 512 KB physical id : 0 siblings : 2 core id : 1 cpu cores : 2 apicid : 1 initial apicid : 1 fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow rep_good pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch bogomips : 4590.70 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: ts fid vid ttp tm stc 100mhzsteps
And here is the meminfo:
MemTotal: 3090516 kB MemFree: 2726648 kB Buffers: 9560 kB Cached: 142400 kB SwapCached: 0 kB Active: 189072 kB Inactive: 122248 kB Active(anon): 161692 kB Inactive(anon): 0 kB Active(file): 27380 kB Inactive(file): 122248 kB Unevictable: 0 kB Mlocked: 0 kB SwapTotal: 4000176 kB SwapFree: 4000176 kB Dirty: 116 kB Writeback: 0 kB AnonPages: 159360 kB Mapped: 52620 kB Slab: 17268 kB SReclaimable: 11364 kB SUnreclaim: 5904 kB PageTables: 3984 kB NFS_Unstable: 0 kB Bounce: 0 kB WritebackTmp: 0 kB CommitLimit: 5545432 kB Committed_AS: 468336 kB VmallocTotal: 34359738367 kB VmallocUsed: 302400 kB VmallocChunk: 34359433723 kB DirectMap4k: 12160 kB DirectMap2M: 3133440 kB
Also, seeing that new versions of both firefox and jre were in the repos, I have upgraded to them, but that didn't change anything.
it looks like ur system is new enough so its not a hardware issue.
the only other thing i can think of is if ur not using dial-up or running huge torrents while ur trying to play...
Nope, nothing like that. At the moment it is a clean install, with just enough to be able to run firefox, plus jre and the flash plugin. Using LXDE, using the nvidia driver, in case that makes a difference. Like I said I had it working fine, just testing things out, then did a clean install and haven't been able to get it working since.
Between my first 64 bit install and the time I repartitioned and reinstalled, there were updates to the kernel, nvidia, and X. I have considered the possibility that one of those is somehow the issue, but it seems like more people would have a problem in that case.
I was hoping that someone could point me to settings I should check (anything in xorg.conf possibly? Could this be a rendering problem?) or log files that might shed some light on it. Any thoughts would be greatly appreciated.
Unfortunately I am not 100% certain what fixed it, but it definately was related to nvidia. I did yet another clean install, this time being careful to closely follow the wiki instructions about xorg.conf and nvidia. As a result I added some oftions to my xorg.conf that weren't there before, specifically
Option "RenderAccel" "true" Option "AGPFastWrite" "true" Option "NoLogo" "true" Option "EnablePageFlip" "true"
in the device section.
The other thing I did differently was running 'depmod -a' as mentioned in the beginners guide. These are the only things I did differently this time, so one of them solved the issue.
It may be a little late, but this sounds a lot like a poor interaction between the JVM and hugepaged in recent Linux kernels.
An easy way to tell is to try (as root):
echo madvise > /sys/kernel/mm/transparent_hugepage/defrag
which will disable the hugepage auto defragging. This should immediately resolve the problem. It's easiest to just run this on each boot. You can also search the web for -XX:+UseLargePages for some information on how to configure things so that the JVM can make use of huge pages, but it requires setting aside memory just for the JVM, which may not be desired.