> Let's see how quickly we can cut the list down, shall we?
That's why I posted the list. =)
> Can we do the rest with awk? Put the line
> "struct inode *inode = filp->f_dentry->d_inode;" in and remove it iff gcc complains that it's not used? :)
I'm uneasy about automating the process because I don't own hardware to
test most of the affected drivers. There have already been lots of 'xyz
broke in 2.1.60 and I don't know why' messages that are traceable to the
VFS changes; it seems these problems will be much harder to track down if
the code is grokked to *look* VFS-compliant, possibly leaving some subtle
interaction uncorrected. Things like possible pointer aliasing make
me nervous without human review of all the currently broken code -- am I
being paranoid?
--Scott
@ @
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-oOO-(_)-OOo-=-=-=-=-=
C. Scott Ananian: cananian@lcs.mit.edu / Declare the Truth boldly and
Laboratory for Computer Science/Crypto / without hindrance.
Massachusetts Institute of Technology /META-PARRESIAS AKOLUTOS:Acts 28:31
-.-. .-.. .. ..-. ..-. --- .-. -.. ... -.-. --- - - .- -. .- -. .. .- -.
PGP key available via finger and from http://www.pdos.lcs.mit.edu/~cananian