In message <20070920214447.GB8496@xxxxxxxxxxxxxxxxxxxxxx>, Michael Halcrow writes:
> On Wed, Sep 19, 2007 at 10:46:26PM -0700, Andrew Morton wrote:
> (from ecryptfs_encrypt_page()):
> > > + enc_extent_virt = kmalloc(PAGE_CACHE_SIZE, GFP_USER);
> >
> > I'd have thought that alloc_page() would be nicer. After all, we _are_
> > treating it as a page, and not as a random piece of memry.
> >
> > > + if (!enc_extent_virt) {
> > > + rc = -ENOMEM;
> > > + ecryptfs_printk(KERN_ERR, "Error allocating memory for "
> > > + "encrypted extent\n");
> > > + goto out;
> > > + }
> > > + enc_extent_page = virt_to_page(enc_extent_virt);
> >
> > And then we don't need this.
>
> If neither kmap() nor kmap_atomic() can be safely used to get a
> virtual address to pass to vfs_write(), then I do not know what my
> other options are here.
kmap_atomic is intended for short-lived code sections, where you may not
sleep, so you can't do a vfs_read/write in b/t kmap/kunmap_atomic.
Erez.
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
