The following reply was made to PR kern/30100; it has been noted by GNATS.

From: Piotr Meyer <aniou@smutek.pl>
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: kern/30100: large number of mounts (mount_null) causes panic
Date: Sun, 1 May 2005 10:03:56 +0200

 Strange. Looks like backtrace and 'how to repeat' has been deleted.
 So, I try again:
 
 >Description:
 After making about 960 (values between 960 and 980, depends of machine
 or kernel) mounts like:
 mount_null /usr/src /mnt/test/001 
 mount_null /usr/src /mnt/test/002
 mount_null /usr/src /mnt/test/003
 
 and so on... causes (in this case on 976 mount):
 
 panic: malloc: out of space in kmem_map
 Stopped in pid 8383.1 (mount_null) at   netbsd:cpu_Debugger+0x4:
 popl   %
 ebp
 db> bt
 cpu_Debugger(c0668fd8,c762dcf8,0,1727,c0d7708c) at
 netbsd:cpu_Debugger+0x4
 panic(c0694620,0,10000,0,ffffffff) at netbsd:panic+0x12b
 malloc(10000,c06cdda0,0,9,9) at netbsd:malloc+0x19d
 hashinit(3fd3,0,c06cdda0,0,c4bf112c) at netbsd:hashinit+0x53
 nullfs_mount(c4bf1200,bfbffa0c,bfbff860,c762de54,c758fb30) at
 netbsd:nullfs_moun
 t+0x1fc
 sys_mount(c6ddcdec,c762df64,c762df5c,c758fb30,c762df74) at
 netbsd:sys_mount+0x25
 a
 syscall_plain() at netbsd:syscall_plain+0x85
 --- syscall (number 21) ---
 0x480839b7:
 db> 
 
 
 Tested on 2.0(official release), 2.0_STABLE(<28.04.2005), 
 3.0_BETA (<27.03.2005). I made some test about jails and
 shared '/usr'... :)
 
 How-To-Repeat:
 
 Try (on 3.0_BETA):
 
 mkdir -p /mnt/test
 for a in `seq -w 1 999`; do mkdir /mnt/test/$a; done
 for a in `seq -w 1 999`; do echo $a; mount_null /usr/src /mnt/test/$a; done
 
 
 -- 
 Piotr 'aniou' Meyer
 http://smutek.pl/~aniou/