Re: [PATCH 78/81] x86/mm: Check if PUD is large when validating akernel address

[Fio Cattaneo]

Hello,

[ apologies for not responding to the original thread ]

Wrt kernel paging fault in kernel_addr_valid(), we have been able to
get a consistent repro here at Twitter using "crash" on the live
kernel by using /proc/kcore, either by:

* having the kernel built with CONFIG_STRICT_DEVMEM enabled
* or running crash as "crash vmlinux_image /proc/kcore"

Our stack signature is exactly the same. Once we apply Mel Gorman's
patch, the problem is gone.


Kind Regards,



fio@xxxxxxxxxxx
fcattaneo@xxxxxxxxxxx
@fiorenzo1963




>>>>>>>>>> in reply to <<<<<<<<<<<<<<<<

List:       linux-kernel
Subject:    [PATCH 78/81] x86/mm: Check if PUD is large when
validating a kernel address
From:       Herton Ronaldo Krzesinski <herton.krzesinski () canonical ! com>
Date:       2013-02-19 18:49:41
Message-ID: 1361299784-8830-79-git-send-email-herton.krzesinski ()
canonical ! com
[Download message RAW]

3.5.7.6 -stable review patch.  If anyone has any objections, please let me know.

------------------

From: Mel Gorman <mgorman@xxxxxxx>

commit 0ee364eb316348ddf3e0dfcd986f5f13f528f821 upstream.

A user reported the following oops when a backup process reads
/proc/kcore:

 BUG: unable to handle kernel paging request at ffffbb00ff33b000
 IP: [<ffffffff8103157e>] kern_addr_valid+0xbe/0x110
 [...]

 Call Trace:
  [<ffffffff811b8aaa>] read_kcore+0x17a/0x370
  [<ffffffff811ad847>] proc_reg_read+0x77/0xc0
  [<ffffffff81151687>] vfs_read+0xc7/0x130
  [<ffffffff811517f3>] sys_read+0x53/0xa0
  [<ffffffff81449692>] system_call_fastpath+0x16/0x1b

Investigation determined that the bug triggered when reading
system RAM at the 4G mark. On this system, that was the first
address using 1G pages for the virt->phys direct mapping so the
PUD is pointing to a physical address, not a PMD page.

The problem is that the page table walker in kern_addr_valid() is
not checking pud_large() and treats the physical address as if
it was a PMD.  If it happens to look like pmd_none then it'll
silently fail, probably returning zeros instead of real data. If
the data happens to look like a present PMD though, it will be
walked resulting in the oops above.

This patch adds the necessary pud_large() check.

Unfortunately the problem was not readily reproducible and now
they are running the backup program without accessing
/proc/kcore so the patch has not been validated but I think it
makes sense.

Signed-off-by: Mel Gorman <mgorman@xxxxxxx>
Reviewed-by: Rik van Riel <riel@xxxxxxxxxx>
Reviewed-by: Michal Hocko <mhocko@xxxxxxx>
Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: linux-mm@xxxxxxxxx
Link: http://lkml.kernel.org/r/20130211145236.GX21389@xxxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
Signed-off-by: Herton Ronaldo Krzesinski <herton.krzesinski@xxxxxxxxxxxxx>
---
 arch/x86/include/asm/pgtable.h |    5 +++++
 arch/x86/mm/init_64.c          |    3 +++
 2 files changed, 8 insertions(+)

diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
index c3520d7..3f3dd52 100644
--- a/arch/x86/include/asm/pgtable.h
+++ b/arch/x86/include/asm/pgtable.h
@@ -142,6 +142,11 @@ static inline unsigned long pmd_pfn(pmd_t pmd)
  return (pmd_val(pmd) & PTE_PFN_MASK) >> PAGE_SHIFT;
 }

+static inline unsigned long pud_pfn(pud_t pud)
+{
+ return (pud_val(pud) & PTE_PFN_MASK) >> PAGE_SHIFT;
+}
+
 #define pte_page(pte) pfn_to_page(pte_pfn(pte))

 static inline int pmd_large(pmd_t pte)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 3baff25..ce42da7 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -829,6 +829,9 @@ int kern_addr_valid(unsigned long addr)
  if (pud_none(*pud))
  return 0;

+ if (pud_large(*pud))
+ return pfn_valid(pud_pfn(*pud));
+
  pmd = pmd_offset(pud, addr);
  if (pmd_none(*pmd))
  return 0;
--


--
-- Fio Cattaneo

@fiorenzo1963

Universal AC, how can entropy be reversed? "THERE IS AS YET
INSUFFICIENT DATA FOR A MEANINGFUL ANSWER".
--
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/