On Wed, 7 Jun 2000, Alexander Viro wrote:
>
> New stuff:
> * Fix for races between iput() and ->link() - ->i_count has to be
> atomic_t. Race had been there since long - at least for a couple of
> months. Yes, it means changes to filesystems. Too bad, but it's either
> that or exploitable races able to corrupt the on-disk fs. Submitted to
> Alan.
Aiiee... Idiot me didn't notice that
a) atomic_read() is signed
b) 0 > (1<<31) on 32bit platforms
c) looking into logs is useful, even if flamage is actually
harmless.
Sorry, folks. Fix (compared to ac10-M) is to change the line 820 in
fs/inode.c to
if ((unsigned)atomic_read(&inode->i_count)>(1U<<31)) {
> Please, try it. Thing is on ftp.math.psu.edu/pub/viro/dcache-ac10-M.gz.
Fixed patch (the only difference from the -M being that line) placed into
the same place as dcache-ac10-M1.gz
Alan, the same change is needed in stuff I've sent to you.
Apologies,
Al
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 21:00:28 EST