qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v4 0/27] write/create for Parallels images with


From: Stefan Hajnoczi
Subject: Re: [Qemu-devel] [PATCH v4 0/27] write/create for Parallels images with reasonable performance
Date: Mon, 18 May 2015 17:45:27 +0100
User-agent: Mutt/1.5.23 (2014-03-12)

On Tue, Apr 28, 2015 at 10:46:33AM +0300, Denis V. Lunev wrote:
> This patchset provides an ability to create of/write to Parallels
> images and some testing of the new code. Readings and writings are
> optimized out and I expect the same or slightly better performance
> as qcow2.
> 
> Changes from v4:
> - parallels format driver marked as bdrv_has_zero_init_1
> - added missed unlocks to parallels_co_readv/writev on error path, locking
>   is shortened and simplified
> - changed test number for created images
> - added check for bdrv_has_zero_init() and availability of bdrv_truncate()
>   in parallels_open() and proper error handling in alloc_cluster
> - some patch comments are improved
> 
> Changes from v3:
> - fixed checkpatch warnings even in just moved code. I am tired of them
> - fixed contingency check in patch 18
> 
> Changes from v2:
> - read performance is almost doubled (up to 360 Mb/sec), write performance
>   is improved by 15-20%
> - bat caching approach changed completely. bat_bitmap now contains the data
>   in on-disk format, which allows to use this data for metadata cache
> - incorrect close detection code is added (inuse field in the header)
> - very basic check consistency code added
> 
> Changes from v1:
> - patches 13-19 added, which boosts performance from 800 KiB/sec to
>   near native performance
> 
> Signed-off-by: Denis V. Lunev <address@hidden>
> CC: Roman Kagan <address@hidden>
> CC: Kevin Wolf <address@hidden>
> CC: Stefan Hajnoczi <address@hidden>

Thanks, applied to my master tree:
https://github.com/stefanha/qemu/commits/master

Stefan

Attachment: pgpYlIUHyvs8K.pgp
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]