On Thu, Sep 12, 2019 at 09:03:44PM +1000, Matthew Bobrowski wrote:
> +static bool ext4_dio_checks(struct inode *inode)
> +{
> +#if IS_ENABLED(CONFIG_FS_ENCRYPTION)
> + if (IS_ENCRYPTED(inode))
> + return false;
> +#endif
> + if (ext4_should_journal_data(inode))
> + return false;
> + if (ext4_has_inline_data(inode))
> + return false;
> + return true;
Shouldn't this function be called ext4_dio_supported or similar?
Also bonus points of adding a patch that defines a IS_ENCRYPTED stub
for !CONFIG_FS_ENCRYPTION to make this a little cleaner.
Also the iomap direct I/O code supports inline data, so the above
might not be required (at least with small updates elsewhere).
