> +
> +/*
> + * __vmalloc() will allocate data pages and auxillary structures (e.g.
> + * pagetables) with GFP_KERNEL, yet we may be under GFP_NOFS context here. Hence
> + * we need to tell memory reclaim that we are in such a context via
> + * PF_MEMALLOC_NOFS to prevent memory reclaim re-entering the filesystem here
> + * and potentially deadlocking.
> + */
Btw, I think we should eventually kill off KM_NOFS and just use
PF_MEMALLOC_NOFS in XFS, as the interface makes so much more sense.
But that's something for the future.
> +/*
> + * Same as kmem_alloc_large, except we guarantee a 512 byte aligned buffer is
> + * returned. vmalloc always returns an aligned region.
> + */
> +void *
> +kmem_alloc_io(size_t size, xfs_km_flags_t flags)
> +{
> + void *ptr;
> +
> + trace_kmem_alloc_io(size, flags, _RET_IP_);
> +
> + ptr = kmem_alloc(size, flags | KM_MAYFAIL);
> + if (ptr) {
> + if (!((long)ptr & 511))
Please use unsigned long (or uintptr_t if you want to be fancy),
and (SECTOR_SIZE - 1).
As said elsewhere if we want to be fancy we should probably pass a
request queue or something pointing to it. But then again I don't think
it really matters much, it would save us the reallocation with slub debug
for a bunch of scsi adapters that support dword aligned I/O. But last
least the interface would be a little more obvious.
