Mac Pro keeps crashing when using with KVM. where to start my investigation?

[AlexUK]

hi guys.

i have recently ordered ATEN CS-1762 KVM switch (2 ports model) to control my Mac and PC with a single keyboard and mouse. This KVM supports USB and DVI. the problem i have is Mac crashing when i switch from PC. I contacted mfg lately and they are unaware of this issue due to not having Mac Pro to test with this appliance. They have tested it with Mac Mini, G4 and G5 though and all was OK. based on this they put a description that this box is compatible with Mac. Anyway, being relatively new Mac user i ask for your help.

here are my steps and some ATEN's thoughts:

1. accroding to ATEN tech support, i should power up my second computer after the first one was started and is running. firstly i start the Mac Pro, 5 mins after that i power up my PC.

2. When both computers are up and running (Mac is in KVM focus), i switch to PC and all seems OK. I have noticed though that if NumLock is activated during the PC session, I won't be able to switch back to Mac. However, even though NumLock is OFF i still may have troubles when switching back to Mac. So it is not the rule, but maybe the clue to start with..

NOTES:

1. if i start PC first, Mac won't start at all. The power led will be on, but there will be no picture on the monitor and it won't accept any keyboard/mouse inputs as well as it will be network inaccessible.

2. I tried looking into Mac's logs but need your help in getting this clear.
here is the log. as you will see it stops at 13:04 and then continues at 13:08 after reboot.
i've seen some parallel's VMs but i don't have Parallels installed. i used to have it, but then it never happened to work so i deleted it from the machine. can this be the case it is still somewhere?

pls guys, any thoughts are very welcome!!!

PS: i can cause Mac to crash the same way if necessary or if log looks inclomplete. the thing forgot to mention is: when Mac crashes and i reboot it, it will not necessarily boot up completely. i've noticed that i would have to reset the KVM switch prior to Mac's successful start up. so maybe i copied the wrong log (not the one immediately after the crash). pls advise here

Alex

EDIT: specs for machines are:
Mac Pro dual 2.6GHz, 1GB (+1GB aftermarket added half a year back, works perfectly), HDD 260GB, Nvidia 7300 256Mb.
PC - Supermicro dual xeon 1.6GHz, 4GB Ram, 4 x 36GB SCSI, Nvidia 5200 256 MB with Windows 2000 Server.




Apr 18 13:03:38 localhost kernel[0]: hi mem tramps at 0xffe00000
Apr 18 13:03:38 localhost kernel[0]: PAE enabled
Apr 18 13:03:38 localhost kernel[0]: 64 bit mode enabled
Apr 18 13:03:38 localhost kernel[0]: standard timeslicing quantum is 10000 us
Apr 18 13:03:38 localhost kernel[0]: vm_page_bootstrap: 512109 free pages
Apr 18 13:03:38 localhost kernel[0]: mig_table_max_displ = 71
Apr 18 13:03:38 localhost kernel[0]: Enabling XMM register save/restore and SSE/SSE2 opcodes
Apr 18 13:03:38 localhost kernel[0]: 77 prelinked modules
Apr 18 13:03:38 localhost kernel[0]: ACPI CA 20060421
Apr 18 13:03:38 localhost kernel[0]: AppleIntelCPUPowerManagement: ready
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=0 LocalApicId=0 Enabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=1 LocalApicId=1 Enabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=2 LocalApicId=7 Enabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=3 LocalApicId=6 Enabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=4 LocalApicId=0 Disabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=5 LocalApicId=0 Disabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=6 LocalApicId=0 Disabled
Apr 18 13:03:38 localhost kernel[0]: AppleACPICPU: ProcessorApicId=7 LocalApicId=0 Disabled
Apr 18 13:03:38 localhost kernel[0]: Copyright (c) 1982, 1986, 1989, 1991, 1993
Apr 18 13:03:38 localhost kernel[0]: The Regents of the University of California. All rights reserved.
Apr 18 13:03:38 localhost kernel[0]: using 10485 buffer headers and 4096 cluster IO buffer headers
Apr 18 13:03:38 localhost kernel[0]: Enabling XMM register save/restore and SSE/SSE2 opcodes
Apr 18 13:03:38 localhost kernel[0]: Started CPU 01
Apr 18 13:03:38 localhost kernel[0]: Enabling XMM register save/restore and SSE/SSE2 opcodes
Apr 18 13:03:38 localhost kernel[0]: Started CPU 02
Apr 18 13:03:38 localhost kernel[0]: Enabling XMM register save/restore and SSE/SSE2 opcodes
Apr 18 13:03:38 localhost kernel[0]: IOAPIC: Version 0x20 Vectors 64:87
Apr 18 13:03:38 localhost kernel[0]: Started CPU 03
Apr 18 13:03:38 localhost kernel[0]: ACPI: System State [S0 S3 S4 S5] (S3)
Apr 18 13:03:38 localhost kernel[0]: Security auditing service present
Apr 18 13:03:38 localhost kernel[0]: BSM auditing present
Apr 18 13:03:38 localhost kernel[0]: disabled
Apr 18 13:03:38 localhost kernel[0]: rooting via boot-uuid from /chosen: BE5A00F3-0C3D-4125-9DE4-3F2193D3DAA8
Apr 18 13:03:38 localhost kernel[0]: Waiting on <dict ID="0"><key>IOProviderClass</key><string ID="1">IOResources</string><key>IOResourceMatch</key><string ID="2">boot-uuid-media</string></dict>
Apr 18 13:03:38 localhost kernel[0]: USB caused wake event (EHCI)
Apr 18 13:03:38 localhost kernel[0]: FireWire (OHCI) TI ID 8025 built-in now active, GUID 0016cbfffe6cfb5c; max speed s800.
Apr 18 13:03:38 localhost kernel[0]: Got boot device = IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/SATA@1F,2/AppleAHCI/PRT0@0/IOAHCIDevice@0/AppleAHCIDiskDriver/IOAHCIBlockStorageDevice/IOBlockStorageDriver/ST3160812AS P Media/IOGUIDPartitionScheme/Customer@2
Apr 18 13:03:38 localhost kernel[0]: BSD root: disk0s2, major 14, minor 2
Apr 18 13:03:38 localhost kernel[0]: jnl: replay_journal: from: 3560448 to: 7934976 (joffset 0x48b341000)
Apr 18 13:03:38 localhost kernel[0]: HFS: Removed 9 orphaned unlinked files
Apr 18 13:03:38 localhost kernel[0]: Jettisoning kernel linker.
Apr 18 13:03:38 localhost kernel[0]: Resetting IOCatalogue.
Apr 18 13:03:38 localhost kernel[0]: PXS1: family specific matching fails
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 1
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 2
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 2
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 2
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 2
Apr 18 13:03:38 localhost kernel[0]: Matching service count = 2
Apr 18 13:03:38 localhost kernel[0]: NVDANV40HAL loaded and registered.
Apr 18 13:03:38 localhost kernel[0]: PXS1: family specific matching fails
Apr 18 13:03:38 localhost kernel[0]: Previous Shutdown Cause: 3
Apr 18 13:03:38 localhost kernel[0]: IPv6 packet filtering initialized, default to accept, logging disabled
Apr 18 13:03:38 localhost mDNSResponder-108.4 (Jan 8 2007 20: 39:46)[30]: starting
Apr 18 13:03:38 localhost memberd[38]: memberd starting up
Apr 18 13:03:38 localhost DirectoryService[42]: Launched version 2.1 (v353.6)
Apr 18 13:03:39 localhost lookupd[44]: lookupd (version 369.6) starting - Wed Apr 18 13:03:39 2007
Apr 18 13:03:39 localhost kernel[0]: AppleIntel8254XEthernet: Ethernet address 00:17:f2:01:63:16
Apr 18 13:03:39 localhost configd[34]: No AirPort Driver found.
Apr 18 13:03:39 localhost kernel[0]: AppleIntel8254XEthernet: Ethernet address 00:17:f2:01:63:17
Apr 18 13:03:39 localhost diskarbitrationd[37]: disk0s2 hfs FDAB59BA-2B84-3AD5-AF66-4BD8CFA20261 Macintosh HD /
Apr 18 13:03:40 localhost lookupd[55]: lookupd (version 369.6) starting - Wed Apr 18 13:03:40 2007
Apr 18 13:03:42 localhost kernel[0]: Intel8254x -- Link Up -- 00:17:f2:01:63:16 -- called by interruptOccurred() --
Apr 18 13:03:42 localhost /System/Library/CoreServices/loginwindow.app/Contents/MacOS/loginwindow: Login Window Application Started
Apr 18 13:03:42 localhost kernel[0]: -- Auto-Neg Advertise Reg (04d) = 0xde1, Link Partner Ability Reg (05d) = 0x41e1, Gig Advertise Reg (09d) = 0xe00, Gig Link Partner Ability Reg (10d) = 0x4000
Apr 18 13:03:42 localhost mDNSResponder: Adding browse domain local.
Apr 18 13:03:43 localhost loginwindow[62]: Login Window Started Security Agent
Apr 18 13:03:46 localhost kernel[0]: Intel8254x -- Link down -- 00:17:f2:01:63:17 -- called by enableHardware() --
Apr 18 13:03:46 localhost kernel[0]: -- Auto-Neg Advertise Reg (04d) = 0xde1, Link Partner Ability Reg (05d) = 0x0, Gig Advertise Reg (09d) = 0xe00, Gig Link Partner Ability Reg (10d) = 0x4000
Apr 18 13:03:46 Alexs-Computer configd[34]: setting hostname to "Alexs-Computer.local"
Apr 18 13:03:46 Alexs-Computer TabletDriver[71]: RunApplication called while eRelaunching
Apr 18 13:03:52 Alexs-Computer configd[34]: executing /System/Library/SystemConfiguration/Kicker.bundle/Contents/Resources/enable-network
Apr 18 13:03:52 Alexs-Computer configd[34]: posting notification com.apple.system.config.network_change
Apr 18 13:03:52 Alexs-Computer lookupd[82]: lookupd (version 369.6) starting - Wed Apr 18 13:03:52 2007
Apr 18 13:03:52 Alexs-Computer ntpdate[94]: can't find host time.euro.apple.com\n
Apr 18 13:03:52 Alexs-Computer ntpdate[94]: no servers can be used, exiting
Apr 18 13:03:55 Alexs-Computer configd[34]: target=enable-network: disabled
Apr 18 13:04:15 Alexs-Computer configd[34]: posting notification com.apple.system.config.network_change
Apr 18 13:04:15 Alexs-Computer lookupd[148]: lookupd (version 369.6) starting - Wed Apr 18 13:04:15 2007
Apr 18 13:04:15 Alexs-Computer SystemStarter[41]: "/Library/StartupItems/Tablet" failed security check: permissions
Apr 18 13:04:17 Alexs-Computer kernel[0]: Parallels Workstation Hypervisor started.
Apr 18 13:04:17 Alexs-Computer kernel[0]: Parallels Workstation VMM Main driver started.
Apr 18 13:04:18 Alexs-Computer kernel[0]: com_parallels_kext_Pvsvnic: Ethernet address 00:01:23:45:67:89
Apr 18 13:04:19 Alexs-Computer SystemStarter[41]: Parallels VirtualMachine drivers (149) did not complete successfully
Apr 18 13:08:25 localhost kernel[0]: hi mem tramps at 0xffe00000
Apr 18 13:08:25 localhost kernel[0]: PAE enabled
Apr 18 13:08:25 localhost kernel[0]: 64 bit mode enabled


here couple more logs right at the moment of crash (found in paic.log). hopefully then can shed som elight on the troubles:

pls advise


*********

Wed Apr 18 13:04:16 2007
panic(cpu 0 caller 0x001A429B): Unresolved kernel trap (CPU 0, Type 14=page fault), registers:
CR0: 0x8001003b, CR2: 0x00004000, CR3: 0x01019000, CR4: 0x000006e0
EAX: 0x23a3d578, EBX: 0x00000000, ECX: 0x00001000, EDX: 0x00004000
CR2: 0x00004000, EBP: 0x2544bc18, ESI: 0x23a3d578, EDI: 0x00004000
EFL: 0x00010206, EIP: 0x0019637c, CS: 0x00000008, DS: 0x00000010

Backtrace, Format - Frame : Return Address (4 potential args on stack)
0x2544b9d8 : 0x128d08 (0x3cb134 0x2544b9fc 0x131de5 0x0)
0x2544ba18 : 0x1a429b (0x3d0e4c 0x0 0xe 0x3d0670)
0x2544bb28 : 0x19ada4 (0x2544bb40 0x23a39578 0x2544bb98 0x84ae50)
0x2544bc18 : 0x84b148 (0x23a39000 0x2 0x2544bc68 0x1a1738)
0x2544bc38 : 0x8393d4 (0x2544bddc 0x0 0x2544bc68 0x4b0000)
0x2544bd18 : 0x6ca33a (0x23a39000 0x2544bdaa 0x2544bdaa 0x84)
0x2544bd78 : 0x6be697 (0x3c34000 0x2544bdaa 0x2544bdaa 0x84)
0x2544bdf8 : 0x6a72e3 (0x3c34000 0x1 0x2544be48 0x1a1738)
0x2544be18 : 0x6aee4a (0x553f8000 0x0 0x3ff96a27 0x3d87)
0x2544be78 : 0x546b55 (0x3c34000 0x2544beac 0x1 0x1a26ce)
0x2544bec8 : 0x54a1d0 (0x3a21e00 0x3 0x3f60ddf8 0x3d87)
0x2544bee8 : 0x54bc8d (0x3a21e00 0x7abf 0x5cbd220 0x5cbd160)
0x2544bf28 : 0x39c536 (0x3a21e00 0x3a40280 0x8 0x2)
0x2544bf78 : 0x13d7d9 (0x3a40280 0x3a87e78 0x0 0x0)
0x2544bfc8 : 0x19ac1c (0x0 0x0 0x19e0b5 0x397fcec) Backtrace terminated-invalid frame pointer 0x0
Kernel loadable modules in backtrace (with dependencies):
com.apple.NVDAResman(4.5.2)@0x6c9000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
dependency: com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
dependency: com.apple.iokit.IONDRVSupport(1.4.5)@0x6ba000
com.apple.iokit.IONDRVSupport(1.4.5)@0x6ba000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
dependency: com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
com.apple.iokit.IOHIDFamily(1.4.10)@0x531000

Kernel version:
Darwin Kernel Version 8.9.1: Thu Feb 22 20:55:00 PST 2007; root:xnu-792.18.15~1/RELEASE_I386


*********


Wed Apr 18 17:06:26 2007
panic(cpu 1 caller 0x001A429B): Unresolved kernel trap (CPU 1, Type 14=page fault), registers:
CR0: 0x8001003b, CR2: 0x00004000, CR3: 0x01019000, CR4: 0x000006e0
EAX: 0x23a4d578, EBX: 0x00000000, ECX: 0x00001000, EDX: 0x00004000
CR2: 0x00004000, EBP: 0x25443c18, ESI: 0x23a4d578, EDI: 0x00004000
EFL: 0x00010206, EIP: 0x0019637c, CS: 0x00000008, DS: 0x04080010

Backtrace, Format - Frame : Return Address (4 potential args on stack)
0x254439d8 : 0x128d08 (0x3cb134 0x254439fc 0x131de5 0x0)
0x25443a18 : 0x1a429b (0x3d0e4c 0x1 0xe 0x3d0670)
0x25443b28 : 0x19ada4 (0x25443b40 0x23a49578 0x1 0x4065004)
0x25443c18 : 0x84b148 (0x23a49000 0x2 0x25443c48 0x713bbf)
0x25443c38 : 0x8393d4 (0x25443ddc 0x0 0x25443c68 0x7116c2)
0x25443d18 : 0x6ca33a (0x23a49000 0x25443daa 0x25443daa 0x84)
0x25443d78 : 0x6be697 (0x3c61c00 0x25443daa 0x25443daa 0x84)
0x25443df8 : 0x6a72e3 (0x3c61c00 0x2 0x25443e28 0x1e)
0x25443e18 : 0x6aee4a (0x523e0000 0x0 0xfd8e30e2 0xc93)
0x25443e78 : 0x546b55 (0x3c61c00 0x25443eac 0x2 0x1a26ce)
0x25443ec8 : 0x54a1d0 (0x3a21e00 0x3 0xfcf5a332 0xc93)
0x25443ee8 : 0x54bc8d (0x3a21e00 0x1917 0x4336be0 0x4336b20)
0x25443f28 : 0x39c536 (0x3a21e00 0x3a40280 0x8 0x2)
0x25443f78 : 0x13d7d9 (0x3a40280 0x3a39ff8 0x0 0x3303)
0x25443fc8 : 0x19ac1c (0x0 0x0 0x19e0b5 0x5cb2144) Backtrace terminated-invalid frame pointer 0x0
Kernel loadable modules in backtrace (with dependencies):
com.apple.NVDAResman(4.5.2)@0x6c9000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
dependency: com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
dependency: com.apple.iokit.IONDRVSupport(1.4.5)@0x6ba000
com.apple.iokit.IONDRVSupport(1.4.5)@0x6ba000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
dependency: com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
com.apple.iokit.IOGraphicsFamily(1.4.5)@0x69f000
dependency: com.apple.iokit.IOPCIFamily(2.2)@0x5e0000
com.apple.iokit.IOHIDFamily(1.4.10)@0x531000

Kernel version:
Darwin Kernel Version 8.9.1: Thu Feb 22 20:55:00 PST 2007; root:xnu-792.18.15~1/RELEASE_I386

 

[AlexUK]

ok so i experimented a little with the hardware and here what i have to the moment:

1. i thought that the usb kb could be the case for mac os to crash so i disconnected the cable from the KVM and plugged normal apple kb to the mac.

2. however when i tried switching back to mac, the machine crashed again.

from what i noticed for the last few days 'working' with this kvm switch was the screen resolution changing while i was switching. meaning the screen flickered for couple of times prior to showing 1280x1024 that was set up in the system. during flickering the resoution became 1024x768. i 've seen it pretty clearly, just it was for half of a second.

also when powering up both machines, they were starting with 1024x768 and not 1280x1024 as was set up. i had to manually change the screen resolution on both machines (pc and mac) every time after successful start up.

so by far my understanding is that kvm switch forces monitor for 1024x768 during switches and on start up. it also causes the OSes (win2K and MacOSX) to lose 1280x1024 and accept 1024x768 as if the monitor was changed and couldn't operate at a higher resolution.

having said that i uploaded this info to kvm mfg (ATEN) and couple more mac forums so that this info may help someone else when ordering a kvm for mac. at the same time i am still open for any thoughts, advices and crazy ideas to get round this bug.

alex

 

[AlexUK]

good stuff!!! i finally have this piece of kit working with my mac pro!

many thanks to Dale @ kvmchoice.com tech support for the brilliant idea to increase the scan interval from 5 secs (default) to 20 (out of 99 possible). it seems like kvm hasn't had enough time to "ping" the video card (nvidia 7300) and read the resolution settings every time i moved the focus to mac pro. being unable to get this data kvm simply forced the video card to switch to kvm's default 1024x768.. this caused mac pro to crash with graphics error in the panic log.

well, i am still testing the system, but by far i switch back and forth without any problems whatsoever and it makes me very happy!!

good luck to all and don't forget to check the scan interval when using your dvi/usb kvm with the mac pro!

 

[brooker]

good stuff!!! i finally have this piece of kit working with my mac pro!

many thanks to Dale @ kvmchoice.com tech support for the brilliant idea to increase the scan interval from 5 secs (default) to 20 (out of 99 possible). it seems like kvm hasn't had enough time to "ping" the video card (nvidia 7300) and read the resolution settings every time i moved the focus to mac pro. being unable to get this data kvm simply forced the video card to switch to kvm's default 1024x768.. this caused mac pro to crash with graphics error in the panic log.

well, i am still testing the system, but by far i switch back and forth without any problems whatsoever and it makes me very happy!!

good luck to all and don't forget to check the scan interval when using your dvi/usb kvm with the mac pro!

Geeez, good bug hunting. I had some ideas, but reading that i wouldn't have come up with anything good for a long time.

Thanks for sharing your experiences!

 

[Spanky Deluxe]

good stuff!!! i finally have this piece of kit working with my mac pro!

many thanks to Dale @ kvmchoice.com tech support for the brilliant idea to increase the scan interval from 5 secs (default) to 20 (out of 99 possible). it seems like kvm hasn't had enough time to "ping" the video card (nvidia 7300) and read the resolution settings every time i moved the focus to mac pro. being unable to get this data kvm simply forced the video card to switch to kvm's default 1024x768.. this caused mac pro to crash with graphics error in the panic log.

well, i am still testing the system, but by far i switch back and forth without any problems whatsoever and it makes me very happy!!

good luck to all and don't forget to check the scan interval when using your dvi/usb kvm with the mac pro!

Thank you for coming back and posting your solution! Chances are some other poor soul will stumble across this with the exact same problem. We don't get enough nerw people posting their solutions.