> since we no longer have any concept of temporary, for-IO-only buffer_heads,
> reuse_list, free_async_buffers() and recover_reusable_buffer_heads() become
> redundant. recover...() is a nop, and it's use in create_buffers before
> scheduling is pointless. create_buffers should probably wakeup bdflush and
> schedule after setting policy to SCHED_YIELD.
yep, i'm doing cleanups to buffer.c right now. (there was a
file-corrupting bug wrt. file system holes in there too). David S. Miller
noticed that BH_Protected is redundant as well (can be substitued via
b_count++), he is also fixing a bad bug as we are speaking in
make_request()+end_io_async().
these latest changes have really enabled us to clean up lots of code all
across the filesystem and IO layer - i think we should be really happy
about these 'side effects' - and it shows that we are going in the right
direction, design-wise :)
-- mingo
-
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.tux.org/lkml/