Hi Lihong,

On 2019/10/16 18:11, Lihong Kou wrote:
> Introdce a new fio cache for IPU.
> After commit 8648de2c581(f2fs: add bio cache for IPU)
> in the mainline, we still have the problem in SQLite. So I
> reuse the fio data structs to cache the mergeale write IO in
> wirtepages().

I have discussed with Jaegeuk about this in below thread,

https://lore.kernel.org/linux-f2fs-devel/20190219081529.5106-1-yuch...@huawei.com/T/#md5d6dcf60052201994e6df4a915a07a04869d427

IMO, we need a per data type global list to cache all bios to enhance its
scalability, as single bio cache may suffer performance regression due to race
of multiple threads, I've sent a patch [1] for that previously.

For any thoughts, we can discuss in above thread.

[1]
https://lore.kernel.org/linux-f2fs-devel/20190930105325.42870-1-yuch...@huawei.com/T/#u

Thanks,

> 
> Lihong Kou (1):
>   Revert "f2fs: add bio cache for IPU"
> 
> Lihong Kou (1):
>   f2fs: introduce a new fio cache for IPU.
> 
>  fs/f2fs/checkpoint.c |  1 +
>  fs/f2fs/data.c       | 93 
> ++++++++--------------------------------------------
>  fs/f2fs/f2fs.h       |  4 +--
>  fs/f2fs/segment.c    | 22 ++++++-------
>  fs/f2fs/super.c      |  2 ++
>  5 files changed, 29 insertions(+), 93 deletions(-)
> 


_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

Reply via email to