Another kernel panic message with leopard and fusion 1.x

[joelindien]

hello , since i've upgraded on leopard, i can't use anymore vmware fusion. The osx kernel panics.

ethernet seems involved to this crash. Before the black ring of the death appears, the console window, print a lot of error messages concerning "vlance" and i use pxe network boot over nfs on one of my vm machines ( it works succesfully on tiger )

i tried vmware on several installation type :

1.1b on leopard upgrade

1.1rc1 on leopard upgrade ( 1.1b desinstalled before )

1.0 on leopard fresh install

1.1rc1 on leopard fresh install ( uninstall 1.0 before )

Here below the result message :

-

---

panic(cpu 1 caller 0x003A3B2F): "soreceive corrupted so_rcv: m 0 cc 1514"@/Sourc

eCache/xnu/xnu-1228/bsd/kern/uipc_socket.c:2063

Backtrace, Format - Frame : Return Address (4 potential args on stack)

0x521d3d48 : 0x12b0e1 (0x4555b4 0x521d3d7c 0x133238 0x0)

0x521d3d98 : 0x3a3b2f (0x48d5b0 0x0 0x5ea 0x1939e07)

0x521d3e28 : 0x38d658 (0x7ca6e58 0x0 0x521d3eac 0x0)

0x521d3e58 : 0x3888be (0x434f150 0x521d3eac 0x0 0x45ef444)

0x521d3f08 : 0x388afe (0x45ef444 0x434f150 0x15e1d812 0x0)

0x521d3f78 : 0x3da8b0 (0x3f949d0 0x45ef340 0x45ef384 0x3e74ec0)

0x521d3fc8 : 0x19ea34 (0x46ee2a0 0x0 0x1a10b5 0x4ae46b0)

No mapping exists for frame pointer

Backtrace terminated-invalid frame pointer 0xb01846d8

BSD process name corresponding to current thread: vmware-vmx

Mac OS version:

9A581

Kernel version:

Darwin Kernel Version 9.0.0: Tue Oct 9 21:35:55 PDT 2007; root:xnu-1228~1/RELEA

SE_I386

System model name: Macmini2,1 (Mac-F4208EAA)

-

---

Until the problem be solved, rollback on tiger

thanks

[rcardona2k]

Permissions problems are common if you have not setup fstab correctly in Ubuntu. Please review this thread to make sure you've got things setup:

[Spike747]

Not the problem I'm seeing. First, I'm not using fstab, I'm issuing a fully-qualified mount command by hand (well, by script ). Second, I'm talking about completely trashed, bogus permissions and ownership, things like a single object file out of 30 with permissions -xxs and owned by daemon while all others are properly set to -rw-rw-r and owned by me. Then when I delete all of the objects and compile again a different object file has a similar problem with different details, like all permissions disabled and a different bogus owner. All of this in a VM that has worked perfectly under Tiger for months (and still does: I made a backup of my Tiger install that I can boot from before I upgraded), and was converted from a Parallels VM that worked perfectly for months, which was itself converted from a VirtualPC VM that worked perfectly on my old PowerBook G4. And the Parallels VM still works fine under Leopard, so it doesn't seem to be an nfsd problem on the host.

[rcardona2k]

Thanks for the details! Just checking that this isn't simple misinformation. Personally I use Samba/CIFS to share files from OS X, yes even on Linux with an smb client. The Windows sharing implementation is the most robust, and mature I've seen. Intuition says there's something Apple has changed or broken in Leopard to create the inconsistent permissions you're seeing. Hopefully it will be addressed soon.

[Spike747]

I now see the same problem (but much less frequently) under Parallels. Makes me think there's an nfsd problem in Leopard.

I've never been keen on SMB (let's just say my experience hasn't matched yours), but it looks like I'm headed there next.

[HPReg]

Folks,

We have finally been able to reproduce the problem in-house. We are investigating the problem as I write this.

Thanks for your numerous reports and your patience.

[kees_de_kapper]

I do not have so much to add to this thread, but the problem still remains in build number 72241 (Version 1.1.1).

This is my crash dump:

____________

panic(cpu 0 caller 0x004014A4): "IOPageableMapForAddress: null"@/SourceCache/xnu/xnu-1228.0.2/iokit/Kernel/IOLib.cpp:724

Backtrace, Format - Frame : Return Address (4 potential args on stack)

0x343ebbd8 : 0x12b0e1 (0x455670 0x343ebc0c 0x133238 0x0)

0x343ebc28 : 0x4014a4 (0x4963c0 0x496420 0x343ebc58 0x400953)

0x343ebc58 : 0x4cf72ee7 (0x3d55180 0x28 0x343ebc98 0x4cf76922)

0x343ebc78 : 0x4cf764dd (0x3d55184 0x0 0x24 0x4cf96d4c)

0x343ebc98 : 0x4cf77ca6 (0x4c698c0 0x1d 0x1 0x3435e000)

0x343ebd28 : 0x4cf71faf (0x5fc6004 0x343ebed0 0x343ebd58 0x1a8a6a)

0x343ebd78 : 0x200a2c (0x11000000 0xc00c561b 0x343ebed0 0x3)

0x343ebdb8 : 0x1f3e66 (0x343ebde8 0x246 0x343ebe18 0x1d803e)

0x343ebe18 : 0x1ea123 (0x53c3a70 0xc00c561b 0x343ebed0 0x3)

0x343ebe78 : 0x362e18 (0x45635e0 0xc00c561b 0x343ebed0 0x343ebf50)

0x343ebe98 : 0x389778 (0x45635e0 0xc00c561b 0x343ebed0 0x343ebf50)

0x343ebf78 : 0x3da847 (0x4576250 0x5e03dc0 0x5e03e04 0xffffffff)

0x343ebfc8 : 0x19ea34 (0x4dceb80 0x1 0x10 0x4dceb80)

No mapping exists for frame pointer

Backtrace terminated-invalid frame pointer 0xb0288e08

Kernel loadable modules in backtrace (with dependencies):

com.vmware.kext.vmx86(1.1.1fc2)@0x4cf70000->0x4cfd3fff

BSD process name corresponding to current thread: vmware-vmx

Mac OS version:

9B18

Kernel version:

Darwin Kernel Version 9.1.0: Wed Oct 31 17:46:22 PDT 2007; root:xnu-1228.0.2~1/RELEASE_I386

System model name: MacBook2,1 (Mac-F4208CA9)

____________

Good luck with the solution!

[admin]

I do not have so much to add to this thread, but the problem still remains in build number 72241 (Version 1.1.1).

panic(cpu 0 caller 0x004014A4): "IOPageableMapForAddress: null"@/SourceCache/xnu/xnu-1228.0.2/iokit/Kernel/IOLib.cpp:724

This is actually a totally different panic than the one the original poster of the thread observed -- you can tell this because your panic refers to a different file and line. Your panic is probably the same issue as , which has not yet been resolved.