When using overlayfs I pretty quickly run into a "page allocation failure: order:4, mode:0x40d0" that looks like this: Call Trace: [<ffffffff8171c149>] dump_stack+0x45/0x56 [<ffffffff811898af>] warn_alloc_failed+0xff/0x170 [<ffffffff8118da65>] __alloc_pages_nodemask+0x6e5/0xbf0 [<ffffffff811d3ee1>] alloc_pages_current+0xb1/0x160 [<ffffffff81189a4b>] alloc_kmem_pages+0x3b/0x110 [<ffffffff811a8418>] kmalloc_order+0x18/0x50 [<ffffffff811a8474>] kmalloc_order_trace+0x24/0xa0 [<ffffffffa064040a>] ? ovl_copy_xattr+0x6a/0x170 [overlayf [<ffffffffa064042b>] ovl_copy_xattr+0x8b/0x170 [overlayfs] [<ffffffffa063de28>] ? ovl_create_real+0x108/0x240 [overla [<ffffffffa064094e>] ovl_copy_up_one+0x33e/0x7f0 [overlayf [<ffffffff813041ed>] ? selinux_sem_alloc_security+0x7d/0xd [<ffffffffa0640ee9>] ovl_copy_up+0xe9/0x120 [overlayfs] [<ffffffffa063ec01>] ovl_do_remove+0x81/0x4a0 [overlayfs] [<ffffffff81303223>] ? may_link.isra.31+0x93/0x120 [<ffffffff812089da>] ? __inode_permission+0x4a/0xc0 [<ffffffffa063f053>] ovl_unlink+0x13/0x20 [overlayfs] [<ffffffff812096d6>] vfs_unlink+0xe6/0x160 [<ffffffff8120f51e>] do_unlinkat+0x27e/0x2f0 [<ffffffff81210646>] SyS_unlink+0x16/0x20 [<ffffffff81722fe9>] system_call_fastpath+0x16/0x1b This seems to be from value = kmalloc(XATTR_SIZE_MAX, GFP_KERNEL); in ovl_copy_xattr(). XATTR_SIZE_MAX is 64k, so its probably a bad idea to always allocate this GFP_KERNEL. We should probably start with a small stack-allocated buffer and then only allocate what is required. Note: The reason I hit this is probably due to selinux, which ensure that most files have an xattr (the selinux context). -- 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
