Jakub> This issue really is not matter of filesystem, but of the
Jakub> architecture. I wrote support for large files in ext2,
Jakub> hopefully it should make it soon into 2.1.9x. Only for 64bit
Jakub> platforms though. If you want to convert all variables in the
Jakub> kernel to long long on 32bit machines, you'd get unbelievably
Jakub> bad performance. ...
I can imagine. But what if you change only those variables that have
something to do with the fact that files are currently 32bit limited?
I.e., size, current position, a few things like that.
It seems fairly logical that adding a few cycles overhead to system
services that do I/O (i.e., spend lots of time either waiting for
devices, or moving bytes to/from cache blocks) is a non-issue.
I can understand an argument that this sort of change is tricky, but I
can't see a good reason why it should be a performance problem.
paul
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu