search

bispawel / macfuse

Automatically exported from code.google.com/p/macfuse
Other
0 stars 0 forks source link

Kernel panic with sshfs #52

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
I don't know how easy it is to reproduce this.

What steps caused the problem?
1. mount an sshfs filesystem
2. use it heavily (rsyncing lots of stuff around)
3. try to eject in the finder, which doesn't work
4. try to umount on the command line and get "device is busy" failures
5. do best to ensure nothing is using any of the files on the filesystem
5. try to umount -f and get a kernel panic

panic(cpu 0 caller 0x35B8DDD0): vnode being reclaimed but fufh (type=0) is 
valid
Latest stack backtrace for cpu 0:
      Backtrace:
         0x00095138 0x00095650 0x00026898 0x35B8DDD0 0x000FD718 0x000E5BBC 
0x000E5ECC 0x000E8328 
         0x000E5914 0x000EBE44 0x000EBC74 0x002AAE28 0x000AB930 0xD12E7810 
      Kernel loadable modules in backtrace (with dependencies):
         com.google.filesystems.fusefs(0.1.7)@0x35b87000
Proceeding back via exception chain:
   Exception state (sv=0x35BA6280)
      PC=0x90063B8C; MSR=0x0000D030; DAR=0xA000C56C; DSISR=0x40000000; 
LR=0x00002D94; R1=0xBFFFE330; XCP=0x00000030 (0xC00 - System call)

Kernel version:
Darwin Kernel Version 8.8.0: Fri Sep  8 17:18:57 PDT 2006; root:xnu-
792.12.6.obj~1/RELEASE_PPCModel: PowerBook5,2, BootROM 4.7.1f1, 1 
processors, PowerPC G4  (1.1), 1.25 GHz, 1.25 GB
Graphics: ATI Mobility Radeon 9600, ATY,RV350M10, AGP, 64 MB
Memory Module: SODIMM0/J25LOWER, 256 MB, DDR SDRAM, PC2700U-25330
Memory Module: SODIMM1/J25UPPER, 1 GB, DDR SDRAM, PC2700U-25330
AirPort: AirPort Extreme, 405.1 (3.90.34.0.p18)
Modem: LastDash, UCJ, V.92, 4.0, APPLE VERSION 2.6.6
Bluetooth: Version 1.7.8f2, 2 service, 0 devices, 1 incoming serial ports
Network Service: AirPort, AirPort, en1
PCI Card: TXN,PCIXXXX-00, cardbus, PC Card
Parallel ATA Device: MATSHITADVD-R   UJ-816, 
Parallel ATA Device: Hitachi IC25N080ATMR04-0, 74.53 GB
USB Device: Bluetooth HCI, , Up to 12 Mb/sec, 500 mA

Original issue reported on code.google.com by n8gray@gmail.com on 24 Jan 2007 at 10:46

GoogleCodeExporter commented 8 years ago 
Oh, heh. That's trivial to reproduce.

If you see in the source, I'm causing this panic programmatically if the vnode 
is going away but the refcounts 
don't add up. Of course, in the case of a forced unmount, the refcounts *will* 
be off if some files were open, etc.

I'll fix this in the next release.

That said, if you get a device busy error, try to see what lsof says about 
files that might be open.

Original comment by si...@gmail.com on 25 Jan 2007 at 1:36

GoogleCodeExporter commented 8 years ago 
Fixed in the source tree.

Original comment by si...@gmail.com on 25 Jan 2007 at 5:24