Hi,
在 2023/05/13 9:09, Chaitanya Kulkarni 写道:
On 5/12/23 06:00, Christoph Hellwig wrote:
Hell no. This is just obsfucation. We can look into actually exposing
REQ_OP_FLUSH at the bio level, but not something like this.
and that's why I made it RFC, thanks for the can you please elaborate
o
> 2023年5月12日 10:07，Chaitanya Kulkarni 写道：
>
> Flush requests are implemented as REQ_OP_WRITE + REQ_OP_PREFLUSH
> combination and not REQ_OP_FLUSH + REQ_PREFLUSH combination.
>
> This unclear nature has lead to the confusion and bugs in the code for
> block drivers causing more work for testing
On 5/12/23 06:00, Christoph Hellwig wrote:
> Hell no. This is just obsfucation. We can look into actually exposing
> REQ_OP_FLUSH at the bio level, but not something like this.
>
and that's why I made it RFC, thanks for the can you please elaborate
on "exposing REQ_OP_FLUSH at the bio level" ?
Flush requests are implemented as REQ_OP_WRITE + REQ_OP_PREFLUSH
combination and not REQ_OP_FLUSH + REQ_PREFLUSH combination.
This unclear nature has lead to the confusion and bugs in the code for
block drivers causing more work for testing, reviews and fixes :-
1. https://lore.kernel.org/all/zfh
Hell no. This is just obsfucation. We can look into actually exposing
REQ_OP_FLUSH at the bio level, but not something like this.
--
dm-devel mailing list
dm-devel@redhat.com
https://listman.redhat.com/mailman/listinfo/dm-devel
