Am Montag, 17. September 2001 11:24 schrieb Pavel Machek:
> Hi!
>
> > Why is that? The USB 2.0 disks seem like bandwidth won't
> > be an issue. Is it the iffy error handling going through SCSI?
> >
> > > Suggestions for how to fix it?
> >
> > And precisely what is the issue -- is usb-storage ignoring
> > information in the SCSI requests that blocking on memory
> > is forbidden for particular requests, or isn't SCSI passing that
> > kind of requirement down for swap requests?
> >
> > Seems to me that if GFP_KERNEL allocation is an issue,
> > every (!) I/O request path is going to need to pass memory
> > allocation flags down. That capability isn't in the usbcore
>
> I believe GFP_KERNEL is the problem (GFP_KERNEL in function for
> performing I/O seems bad to me); I'm not 100% sure, yet.
Yes GFP_KERNEL must be replaced by GFP_NOIO, not only
for swap but for _any_ memory allocation in the IO path of a block device or
any service a block device uses. I've posted a patch for usb storage itself,
but a patch to add a flag to the usb core and the host controller drivers is
still needed. I'll have a look at it. I propose the flag to be called
URB_MEM_NOIO.
Regards
Oliver
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
