Re: QUOTA [with patch this time]

Adam D. Bradley (bradley@cs.unca.edu)
Mon, 16 Dec 1996 22:08:21 -0500 (EST)

> There seems to be a bug in kernel quotas. I've verified that it is present
> in the first 1.3.x kernels to support quotas.
>
[snip]
>
> It happens on roughly every 40th UID, but not exactly. I've
> verified this on four different machines with various kernels
> from 1.3.x to 2.1.15.

Hmm...dquot.c uses a hash function for some data lookup, table is of
size NR_DQHASH which is defined as 43 (see linux/include/linux/quota.h).
Coincidence? I wish the source had better documented parameters...but the
stuff in linux/fs/dquot.c is where we should start looking...

Adam

--
He feeds on ashes; a deluded mind has led him    Adam Bradley, UNCA Senior
astray, and he cannot deliver himself or say,             Computer Science
"Is there not a lie in my right hand?"   Isaiah 44:20
        bradley@cs.unca.edu       http://www.cs.unca.edu/~bradley      <><