Replace our somewhat fragile code to reuse the bio, which caused a
regression in the past with the block layer bio_reuse helper.

Signed-off-by: Christoph Hellwig <[email protected]>
Reviewed-by: Hans Holmberg <[email protected]>
Reviewed-by: Carlos Maiolino <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
Reviewed-by: "Darrick J. Wong" <[email protected]>

The double buffering where just one scratch area is used at a time does
not efficiently use the available memory.  It was originally implemented
when GC I/O could happen out of order, but that was removed before
upstream submission to avoid fragmentation.  Now that all GC I/Os are
processed in order, just use a number of buffers as a simple ring buffer.

For a synthetic benchmark that fills 256MiB HDD zones and punches out
holes to free half the space this leads to a decrease of GC time by
a little more than 25%.

Thanks to Hans Holmberg <[email protected]> for testing and
benchmarking.

Signed-off-by: Christoph Hellwig <[email protected]>
Reviewed-by: Hans Holmberg <[email protected]>
Reviewed-by: Carlos Maiolino <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
Reviewed-by: "Darrick J. Wong" <[email protected]>