struct file_operations xyz_fops = {
open:xyz_open,
read:xyz_read,
write:xyz_write,
};
if I only implement those 3 methods. This is insensitive to members
being moved around and it seems to me that it solves the problem which
Doug explained which is if you don't carefully look at the structure
declaration beforehand, you're stuffed.
Is this the way you'd like things to be done? This appears to be quite
maintainable.
If so, it seems to me it would make sense to change all the drivers
over to this method. Would you accept a patch that does this?
Linus Torvalds writes:
>
>
> On Sat, 29 Aug 1998, Richard Gooch wrote:
> >
> > Let me preface this my saying that Doug is the first person who has
> > taken the trouble to respond to technical issues instead of dumb
> > flames and personal insults.
>
> Let me just say that I answered your complaint two _days_ ago. You never
> answered that, just continued to flame on. I'm tired of you. Go away.
No, I *didn't* flame. I simply didn't see what the actual problem you
were trying to solve was. Doug is the first person who made it clear.
> The reason you keep on getting flames back is that you have been
> rather irritating on this subject.
You seem to equate "irritating" with persistent. If I don't see you're
reasoning, it's fair enough for me to pursue the issue when I see the
pain that the breakage causes. Why are you guys so impatient?
Why not take the trouble to explain it clearly if someone doesn't get
it?
> I put it in the middle very much on purpose - to force people who use
> open/release to at least be aware of it.
So announcing the new method wasn't sufficient?
> For Gods sake, when somebody notices the compiler warnings when compiling
> a new module, if he doesn't take them seriously he has only himself to
> blame.
I didn't talk about ignoring compiler warnings. My concerns are that a
lot of drivers not distributed with the kernel will break.
> I end up maintaining the end result, and I categorically refuse to make
> bad decisions just because there's going to be a very limited amount of
> pain for a week or two - I end up having to maintain it for _years_. As
> such, my priorities are slightly different than other peoples.
I appreciate your concerns over maintenance. But this didn't really
explain what the problems were. Anyway, Doug has enlightened me.
Regards,
Richard....
-
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.altern.org/andrebalsa/doc/lkml-faq.html