Hi,
In preparation of 2.4.20-pre, here's an updated version of the
block-highme patch. This enables block drivers to do I/O to high memory
pages instead of reverting to low memory bounce buffers.
Changes:
- (scsi) Fix critical SCpnt->request_buffer -> SCpnt->request.buffer
error
- (scsi) Backport __init_io() changes from 2.5, that enable us to
politely back off on queueing more I/O with any failure. We use this
now instead of reverting to single segment requests on sgtable
allocation failures. For a highmem host this got nasty, because we had
to map the buffer_head and use a bounce buffer for that single segment
anyway. That is now eliminated. Even for non-highmem I/O the single
segment fall back is slower because of the over head associated with
command setup etc.
*.kernel.org/pub/linux/kernel/people/axboe/patches/v2.4/2.4.19-pre10/block-highmem-all-19.bz2
Enjoy,
-- Jens Axboe- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Jun 23 2002 - 22:00:16 EST