Messages in this thread |  | | From | (really ) | Date | 16 Apr 1996 19:12:29 +0400 | Subject | Re: Steve's crashing 1.3 machine, cured? |
| |
Linus Torvalds (torvalds@cs.helsinki.FI) wrote:
: Sounds like a bad kerneld/module interaction. Maybe the sound module wa= : s=20 : busy being loaded when kerneld started loading it again, or something=20 : like this. Bj=F6rn, is there anything that protects against that kind o= : f=20 : "re-entrancy" problem?
No, current module stuff does no locking and even do not try to resolve any race conditions. I encountered this problem very long ago. F.e. when amd starts, it mounts a lot of NFS file systems, and it starts happily only in 50% cases, is NFS is module. Moreover, I will eat my hat, if someone will able to make kerneld truly reliable without complete rewrite of module.c.
It was the main reason, why I rewrote all the module stuff from the scratch. Unfortunately, it is orthogonal to standard module implementation.
But anybody may get ftp.inr.ac.ru:/PATCH-89.gz and enjoy (well, if he will succeed in patching and compiling 8)8)
Alexey Kuznetsov.
|  |