37 lines
1.5 KiB
Plaintext
37 lines
1.5 KiB
Plaintext
Maybe you want to look at LIMITATIONS in README. Some might consider
|
|
them to be bugs.
|
|
|
|
But there are really problems that might be fixed in the future.
|
|
|
|
-------------------------------------------------------------------------------
|
|
|
|
You might experience lockups of ncpfs volumes. It happens especially
|
|
under high network traffic, not necessarily only IPX traffic. I am not
|
|
able to reproduce this problem on my machine, so I'm sorry I can not
|
|
do anything about that. When such a lockup happens, you have to shut
|
|
down the complete ipx subsystem by deleting all ipx interfaces,
|
|
unmounting all ncpfs volumes (in this order!) and restarting all
|
|
again.
|
|
|
|
For the kernel hackers who want to look at the problem: The routine
|
|
ipx_sendmsg in net/ipx/af_ipx.c sometimes locks forever if called with
|
|
nonblock=0. I DO NOT KNOW WHY!!! HELP ME, PLEASE!
|
|
|
|
-------------------------------------------------------------------------------
|
|
|
|
'df' returns 0: Free disk space is distributed among the volumes in
|
|
NetWare. df is only able to report one number per mounted
|
|
filesystem. As connections are quite expensive for NetWare (with
|
|
mars_nwe and lwared that might change ...), I rejected the alternative
|
|
to mount only a single volume for a unix mount point. So I simply
|
|
return 0.
|
|
|
|
-------------------------------------------------------------------------------
|
|
|
|
If you use Linux 1.2.x, In your kernel log there will appear messages
|
|
like
|
|
|
|
Nov 25 16:09:08 lx01 kernel: alloc_skb called nonatomically from interrupt 0000002e
|
|
|
|
These are a bit annoying, but completely harmless.
|