fusedav hangs without any output

Bug #152710 reported by Michael Blakeley
2
Affects Status Importance Assigned to Milestone
fusedav (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

Binary package hint: fusedav

$ uname -a
Linux mblakele-lt 2.6.22-14-generic #1 SMP Tue Oct 9 09:24:55 GMT 2007 x86_64 GNU/Linux
$ apt-show-versions fusedav
fusedav/gutsy uptodate 0.2-2
$ groups
mblakele adm disk dialout fax cdrom floppy audio dip video plugdev lpadmin scanner admin fuse kvm nvram
$ ls -l /dev/fuse
crw-rw---- 1 root fuse 10, 229 2007-10-09 09:29 /dev/fuse
$ mkdir /tmp/foo
$ sudo netstat -nlp | grep 1234
$ fusedav http://localhost:1234/ /tmp/foo
[ hangs... ]

There is no webdav server at port 1234 (as a debugging measure, since fusedav wouldn't seem to talk to my webdav server). I've also eliminated the username and password options, for simplicity (their presence doesn't seem to change any of the behavior I'm describing).

But the interesting thing is that fusedav doesn't appear to get as far as trying to open any tcp/ip ports. So what *is* fusedav doing?

$ netstat -an | grep 1234
$ pstack `pgrep fusedav`
Thread 3 (Thread 1082132816 (LWP 13169)):
#0 0x00002b8a256ecefb in ?? () from /lib/libpthread.so.0
#1 0x00002b8a25b2c3da in ?? () from /lib/libfuse.so.2
#2 0x00002b8a25b2c764 in ?? () from /lib/libfuse.so.2
#3 0x00002b8a256e6317 in start_thread () from /lib/libpthread.so.0
#4 0x00002b8a2601ad5d in clone () from /lib/libc.so.6
#5 0x0000000000000000 in ?? ()
Thread 2 (Thread 1090525520 (LWP 13170)):
#0 0x00002b8a256ecefb in ?? () from /lib/libpthread.so.0
#1 0x00002b8a25b2c3da in ?? () from /lib/libfuse.so.2
#2 0x00002b8a25b2c764 in ?? () from /lib/libfuse.so.2
#3 0x00002b8a256e6317 in start_thread () from /lib/libpthread.so.0
#4 0x00002b8a2601ad5d in clone () from /lib/libc.so.6
#5 0x0000000000000000 in ?? ()
Thread 1 (Thread 47872381224240 (LWP 13166)):
#0 0x00002b8a256ec39d in sem_wait () from /lib/libpthread.so.0
#1 0x00002b8a25b2c965 in fuse_session_loop_mt () from /lib/libfuse.so.2
#2 0x00000000004055cd in ?? ()
#3 0x00002b8a25f64b44 in __libc_start_main () from /lib/libc.so.6
#4 0x0000000000402aa9 in ?? ()
#5 0x00007fff855e6aa8 in ?? ()
#6 0x0000000000000000 in ?? ()
#0 0x00002b8a256ec39d in sem_wait () from /lib/libpthread.so.0
$ strace -f -p `pgrep fusedav`
Process 13170 attached with 3 threads - interrupt to quit
[pid 13166] futex(0x7fff855e57a0, FUTEX_WAIT, 0, NULL <unfinished ...>
[pid 13169] read(4, <unfinished ...>
[pid 13170] read(4,

I'd guess that fd 4 is probably /dev/fuse. Note that strace and fusedav don't play well together. After attaching with strace, I have to kill -9 the fusedav process to get it to let go.

On the same system, ntfsmount works fine.

$ apt-show-versions ntfsprogs
ntfsprogs/gutsy uptodate 1.13.1-6

Revision history for this message
Michael Blakeley (mike+ubuntu) wrote :

Not a bug - I didn't realize that fusedav doesn't background itself, and doesn't perform any network operations until the first filesystem access.

Changed in fusedav:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.