On 10/16/23 02:20, Niklas Cassel wrote:
So when you say: "It depends on how we want the user to specify the data lifetime for direct I/O.", do you mean that buffered I/O should use fcntl() to specify data lifetime, but direct I/O should used Linux IO priority API to specify the same? Because, to me that seems to be how the series is currently working. (I am sorry if I am missing something.)
Let's drop support for passing the data lifetime in the I/O priority field from user space as Damien proposed. The remaining question is whether only to restore support for F_SET_RW_HINT (per inode) or also for F_SET_FILE_RW_HINT (per file)? Thanks, Bart.
