On Thu, Apr 22, 2004 at 01:36:20PM -0700, lawalsh wrote:
> >Blocksize cannot be larger than pagesize - the kernel is right
> >to fail this mount request.
> For those of us losing or gaining a bit in pagesize now and then,
> perhaps it would be good for mkfs.xfs to either deny create or warn
> on creating fs's on the system architecture it is being created on?
> How often does one create a fs under linux that they can't copy
> anything onto -- just mkfs it, then move the filesystem to another
> architecture where it is supported?
> I'd be tempted to add an "-ff" switch to mkfs to really force creating
> a file system that will be unusable on the system creating it.
well, thats tricky on ia64 right? (where pagesize is configurable).
I guess people just need to know to read dmesg when mount fails;
not much more we can really do from an xfs point of view. as long
as we're putting a sensible message there that explains the issue,
we should be ok.