On Thu, Feb 13, 2025 at 02:26:37PM +0100, Daniel Gomez wrote:
> That's a good question. The stx_blksize field description indicates the value
> should be referring to the fs block size that avoids RMW.
One that is optimal, the RMW is an example. This what Posix says:
blksize_t st_blksize A file system-specific preferred I/O block size
for this object. In some file system types, this
may vary from file to file.
> So I think, if devices report high values in stx_blksize, it is either because
> smaller values than the reported one cause RMW or they are incorrectly reporting
> a value in the wrong statx field.
Or it is just more efficient. E.g. on NFS or XFS you'll get fairly
big values.
> The commit I refer in the commit message maps the minimum_io_size reported by
> the block layer with stx_blksize.
Yes, and that was my question again - minimum_io_size for block
devices is specified even more handwavy.
In other words I'm really sceptical this is going to be a net win.
To win me over you'll need to show a improvement over a wide variety
of devіces and workloads.
