x86 warnings in arch/x86/mm/pageattr.c:560

From: Nick Piggin
Date: Tue Jan 20 2009 - 04:19:03 EST


Hi,

I've been getting this warning intermittently since the merge window
opened. Considering that track_pfn_vma_new is called before the ptes
are set up in remap_pfn_range (and remap_pfn_range contains a warning
if the ptes *are* nonzero), I don't know how this is supposed to work?

Anyway, I can of course test patches.

------------[ cut here ]------------
WARNING: at arch/x86/mm/pageattr.c:560 __change_page_attr_set_clr+0x9ef/0xa00()
Hardware name: empty
CPA: called for zero pte. vaddr = ffff8800cff4e000 cpa->vaddr = ffff8800cff4e000
Modules linked in:
Pid: 2311, comm: mcelog Not tainted 2.6.29-rc2-00013-gf3b8436 #75
Call Trace:
[<ffffffff8024053a>] warn_slowpath+0xea/0x160
[<ffffffff803ab692>] mix_pool_bytes_extract+0x52/0x160
[<ffffffff803abc63>] extract_buf+0x83/0xf0
[<ffffffff8028c65d>] zone_statistics+0x7d/0x80
[<ffffffff80280686>] get_page_from_freelist+0x4b6/0x690
[<ffffffff80280a57>] __alloc_pages_internal+0xe7/0x4d0
[<ffffffff8022e9ff>] __change_page_attr_set_clr+0x9ef/0xa00
[<ffffffff8029bad4>] __purge_vmap_area_lazy+0x104/0x180
[<ffffffff8029d3ce>] vm_unmap_aliases+0x17e/0x190
[<ffffffff8022eaea>] change_page_attr_set_clr+0xda/0x380
[<ffffffff8022eeed>] _set_memory_wb+0x1d/0x30
[<ffffffff802305a6>] reserve_pfn_range+0x126/0x300
[<ffffffff8022eaea>] change_page_attr_set_clr+0xda/0x380
[<ffffffff80292510>] remap_pfn_range+0xa0/0x3d0
[<ffffffff80230c33>] phys_mem_access_prot_allowed+0xf3/0x1e0
[<ffffffff8028be48>] vma_prio_tree_insert+0x28/0x110
[<ffffffff803ab1f6>] mmap_mem+0x96/0xc0
[<ffffffff80297887>] mmap_region+0x247/0x670
[<ffffffff80297f94>] do_mmap_pgoff+0x2e4/0x3f0
[<ffffffff80210905>] sys_mmap+0x115/0x140
[<ffffffff8020b91b>] system_call_fastpath+0x16/0x1b
---[ end trace 488160a0a7b56966 ]---
mcelog:2311 reserve_pfn_range ioremap_change_attr failed write-back for cff4e000-cff50000

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/