>Forcing new log zone size:2048
>ll_rw_block: device 28:00: only 1024-char blocks implemented (2048)
>unable to read i-node block
>
>Anyone know what the problem could be? I'm using a driver I developed for
>the funai doublespin cdrom it also uses the generic cdrom driver. The
>last time I tried the driver it worked fine, though it might have been the
>2.0.x kernels.
I had the same problem here on a driver I'm developing. It seems that
ll_rw_block() doesn't like this kind of code (in your driver):
set_blocksize (dev, 1024);
ll_rw_block(dev, blocks, 2048); /* Or bread(dev, block, 2048 */
set_blocksize must set the logical block size (I'm right?). But actually,
ll_rw_block() tests this value as if it was the hard sector size?
Is this a bug or a limitation so we don't get buffers of different sizes
associated to the same device or partition?
A+
Rani
-- The Truth is Out there...