On 01.11.19 16:42, Kevin Wolf wrote: > Am 01.11.2019 um 15:01 hat Max Reitz geschrieben: >> On 01.11.19 13:40, Eric Blake wrote: >>> On 11/1/19 11:00 AM, Max Reitz wrote: >>>> This reverts commit c8bb23cbdbe32f5c326365e0a82e1b0e68cdcd8a. >>>> >>>> This commit causes fundamental performance problems on XFS (because >>>> fallocate() stalls the AIO pipeline), and as such it is not clear that >>>> we should unconditionally enable this behavior. >>>> >>>> We expect subclusters to alleviate the performance penalty of small >>>> writes to newly allocated clusters, so when we get them, the originally >>>> intended performance gain may actually no longer be significant. >>>> >>>> If we want to reintroduce something similar to c8bb23cbdbe, it will >>>> require extensive benchmarking on various systems with subclusters >>>> enabled. >>>> >>>> Cc: qemu-sta...@nongnu.org >>>> Signed-off-by: Max Reitz <mre...@redhat.com> >>>> --- >>> >>>> +++ b/qapi/block-core.json >>>> @@ -3304,8 +3304,6 @@ >>>> # >>>> # @cor_write: a write due to copy-on-read (since 2.11) >>>> # >>>> -# @cluster_alloc_space: an allocation of file space for a cluster >>>> (since 4.1) >>>> -# >>>> # @none: triggers once at creation of the blkdebug node (since 4.1) >>> >>> Deleting released qapi is not backwards-compatible. >> >> Right. :-/ >> >> I’ll just not make changes to the QAPI schema. It doesn’t hurt to leave >> this in. > > Why would it be incompatible to drop an enum value that is only ever > used in output and that QEMU doesn't generate?
This isn’t output at all, it’s input for blockdev-add for blkdebug nodes. Max
signature.asc
Description: OpenPGP digital signature
