JVM for amd64 RC1?

Mike Larkin mlarkin at azathoth.net
Thu Feb 5 09:39:46 UTC 2004


I've been running Tomcat on x86-64 using Blackdown 1.4.2 amd64 build all 
day and it's working fine. I tested a few webapps that are fairly complex.

I also tried Blackdown's 1.4.2 for i386 and it also gave me no problems, 
although I didn't test that as much.

The one problem I did notice is that JNI fails on the amd64 build if you 
use 32 bit libraries (have to use 32 bit JVM, but it works fine if you do).

What version of Tomcat are you using? Anything odd about your server.xml 
config?

PS - it seems to work fine under both SuSE 9.0 and FC1, but most of the 
testing I've done has been under SuSE.

-ml

Barry Roberts wrote:

>On Fri, Jan 30, 2004 at 04:40:10PM -0500, Nathan Bryant wrote:
>  
>
>>This is the wrong list. You should post the JVM/Hotspot error to the 
>>java-linux list somewhere on Blackdown.org. It might be a good idea not 
>>to use ZIP format.
>>
>>    
>>
>
>So far no response on that list, and the plain text was too big for this list.
>
>  
>
>>The 32 bit jdk should, in theory, work on Fedora but you will need to 
>>install the i686 glibc rpm's. If you selected mozilla during 
>>installation, the 32-bit libraries should be present, otherwise you can 
>>get them from the distribution or the external updates respository that 
>>has been discussed here recently.
>>    
>>
>
>Ok, so much for the theory.  I've tried the latest 32-bit jvms from
>Sun, blackdown.org, and IBM.  With the 32-bit glibc and libstdc++,
>they install, and will run Tomcat.  Sometimes.  For a few seconds.
>Then I hit the server a couple of times and get something like the
>error below.
>
>So if anybody is succesfully using Java, especially tomcat or any
>servlet container on an x86-64, I would love to hear how it was done
>(what jvm, distro, etc.), since this is about what I was getting on
>SuSE 8.1.
>
>Thanks,
>Barry Roberts
>
>Jan 31 04:15:35 optitwo kernel: Oops: 0010
>Jan 31 04:15:35 optitwo kernel: CPU 1 
>Jan 31 04:15:35 optitwo kernel: Pid: 18527, comm: java Not tainted
>Jan 31 04:15:35 optitwo kernel: RIP: 0010:[<00000001801c7fce>]
>Jan 31 04:15:35 optitwo kernel: RSP: 0000:0000010078583e48  EFLAGS: 00010212
>Jan 31 04:15:35 optitwo kernel: RAX: 0000000000000000 RBX: 00000000cb5e2d40 RCX: 0000000000000000
>Jan 31 04:15:35 optitwo kernel: RDX: 0000000000000000 RSI: 0000010078583f38 RDI: 00000000cb5e2dd0
>Jan 31 04:15:35 optitwo kernel: RBP: 0000000000000000 R08: 0000000008052b7c R09: 00000000c8988920
>Jan 31 04:15:35 optitwo kernel: R10: 0000000000000000 R11: 00000000c898878c R12: 0000010078583f58
>Jan 31 04:15:35 optitwo kernel: R13: 0000010005412728 R14: 000000000000000c R15: 0000010078582a58
>Jan 31 04:15:35 optitwo kernel: FS:  0000002a95565060(0000) GS:ffffffff80572880(005b) knlGS:00000000cb5e4bb0
>Jan 31 04:15:35 optitwo kernel: CS:  0010 DS: 002b ES: 002b CR0: 000000008005003b
>Jan 31 04:15:35 optitwo kernel: CR2: 00000001801c7fce CR3: 0000000037ffe000 CR4: 00000000000006e0
>Jan 31 04:15:35 optitwo kernel: Process java (pid: 18527, stackpage=10078583000)
>Jan 31 04:15:35 optitwo kernel: Stack: 0000010078583e48 0000000000000000 00000001801c7fce 0000000100000000 
>Jan 31 04:15:35 optitwo kernel:        0000000000000000 55b37e4400000000 0000000100000000 08052b7c00000000 
>Jan 31 04:15:35 optitwo kernel:        8013daaa00000000 00000000ffffffff 3b9ab29000000000 c898878c00000000 
>Jan 31 04:15:35 optitwo kernel: Call Trace: [<ffffffff801101ce>]{do_signal+158} [<ffffffff8013daaa>]{compat_sys_futex+202} 
>Jan 31 04:15:35 optitwo kernel:        [<ffffffff8011061b>]{intret_signal+45} 
>Jan 31 04:15:35 optitwo kernel: 
>Jan 31 04:15:35 optitwo kernel: Code:  Bad RIP value.
>Jan 31 04:15:35 optitwo kernel: RIP [<00000001801c7fce>] RSP <0000010078583e48>
>Jan 31 04:15:35 optitwo kernel: CR2: 00000001801c7fce
>
>
>  
>





More information about the fedora-test-list mailing list