On Fri, Jun 17, 2016 at 12:21:50PM +0530, Chandan Rajendra wrote:
> > > t%2dsend%2demail%2dchandan%40linux.vnet.ibm.com) restrict the stripesize
> > > to be either sectorsize or 4096. So I think these restrictions are good
> > > enough to make sure we don't get into the situation explained by you.
> >
> > It's a workaround anyway, I'd rather fix the kernel to not use
> > stripesize and we can remove all checks against super_stripesize.
> >
> > The code has evolved a lot to have free space align well to sectorsize,
> > so stripesize is not as necessary as when it's introduced firstly.
> >
>
> Yes, I mostly agree with what you are suggesting.
>
> But, I am not completely sure about the following ...
>
> static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target)
> {
> /* TODO allow them to set a preferred stripe size */
> return SZ_64K;
> }
>
> Chris, Josef, David ... Do you have any objections to replacing invocations
> of find_raid56_stripe_len() with BTRFS_STRIPE_LEN? Or do we need to retain
> find_raid56_stripe_len() to have the ability for having configurable
> stripesize in the future?
The stripe will be configurable some day, so please keep it for now.
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
