I have a small two node network where one machine is running 386bsd and the
other is running SVR4.0.3.6. I have been NFS mounting a partition from the
386bsd machine to the SVR4.0.3.6 machine. With this mount, I have tried compiling a large application (i.e. gcc.2.3.1) for use in the SVR4 environment.
However, I have run into what looks like a memory leak in 386bsd within the
NFS (or wd8003) subsystem.
Basically, what happens is that the NFS mount works fine when the NFS traffic
is fairly light. However, during heavy NFS I/O, I'll sometimes panic the
386bsd kernel. The panic complains:
kmem_malloc: kmem_map too small
I have increased the define MAX_KMAPENT from 1000 to 1500 (I believe that
one of the 58 patches increased this define from 500 to 1000). This only
extends the life of the kernel but does not solve the problem. I managed
to have the NFS up and flying for a day but it still crashed with the same
problem.
I basically want to know if anyone else has encountered this.. If so, I'll
begin snooping around to find any memory leaks. I have a virgin kernel patched
to pl58 with Terry's patches plus some added patches (locore.s alignment,
"ping" memory leak problem,... that were posted to this group. None of the
patches involed NFS and the "ping" patch fixed a memory leak in the ICMP level
I believe). The machine running 386BSD is a Compaq 386/20 with 5Mb ram. The
standard install was used.
On another note, I noticed a posting regarding 0.2 of 386bsd.. I
havent heard hide nor hare from the Jolitz's in a couple of months..
With the 58+ patches for 0.1 I agree that it is time to change the version.
Regarding 0.2, does anybody know that features, added value, etc... 0.2 will
contain? Maybe the Jolitz's (without any added dates or other added stress...)
could give a brief list of the differences between 0.1 and 0.2 as they
they currently stand. I would personally be interested in knowing where 0.2
is going eventhough I am quite satisfied with 0.1.
