On 2/17/06, Alan Cox [off-list ref] wrote:

On Gwe, 2006-02-17 at 09:57 +0100, Bartlomiej Zolnierkiewicz wrote:

quoted

Could we get a decent description of the problem and of the patch?

Audit the lock state on all entries to the tune function and it
certainly was the case that the old IDE layer could call it with the
lock either already held or not.

Thank you but this is not a patch description, this is a recipe
for me to spend nice friday's evening staring all over IDE code
and making patch description myself...

http://www.zip.com.au/~akpm/linux/patches/stuff/tpp.txt ?

Bartlomiej