[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH 5/7] vdi: leave bounce buffering to block layer
|
From: |
Kevin Wolf |
|
Subject: |
Re: [Qemu-devel] [PATCH 5/7] vdi: leave bounce buffering to block layer |
|
Date: |
Fri, 16 Mar 2012 10:54:13 +0100 |
|
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:10.0.1) Gecko/20120209 Thunderbird/10.0.1 |
Am 05.03.2012 18:40, schrieb Paolo Bonzini:
> vdi.c really works as if it implemented bdrv_read and bdrv_write.
> However, because only vector I/O is supported by the asynchronous
> callbacks, it went through extra pain to bounce-buffer the I/O.
> With the conversion to coroutines bdrv_read and bdrv_write are now
> asynchronous, so they can be handled by the block layer now that the
> format is coroutine-based.
>
> Signed-off-by: Paolo Bonzini <address@hidden>
In the long run, the right thing to do would be to convert VDI to deal
with vectored I/O, of course. Stefan, maybe you want to do that on top.
It's not that hard, I did it for qcow2 a while ago.
Kevin
- [Qemu-devel] [PATCH 0/7] block: convert VDI image format to coroutines, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 5/7] vdi: leave bounce buffering to block layer, Paolo Bonzini, 2012/03/05
- Re: [Qemu-devel] [PATCH 5/7] vdi: leave bounce buffering to block layer,
Kevin Wolf <=
- [Qemu-devel] [PATCH 1/7] vdi: basic conversion to coroutines, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 7/7] vdi: change goto to loop, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 6/7] vdi: do not create useless iovecs, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 3/7] vdi: merge aio_read_cb and aio_write_cb into callers, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 4/7] vdi: move aiocb fields to locals, Paolo Bonzini, 2012/03/05
- [Qemu-devel] [PATCH 2/7] vdi: move end-of-I/O handling at the end, Paolo Bonzini, 2012/03/05
- Re: [Qemu-devel] [PATCH 0/7] block: convert VDI image format to coroutines, Stefan Weil, 2012/03/15