[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Masahiro Yamada]

The current rounding of "size" is wrong.

    size -= start & ~PAGE_MASK;

should be

    size -= PAGE_SIZE - (start & ~PAGE_MASK);

Or more simply

    size -= aligned_start - start;

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
---
 arch/arm/kernel/setup.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
index 50e198c..84508c0 100644
--- a/arch/arm/kernel/setup.c
+++ b/arch/arm/kernel/setup.c
@@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
 	 * Ensure that start/size are aligned to a page boundary.
 	 * Size is appropriately rounded down, start is rounded up.
 	 */
-	size -= start & ~PAGE_MASK;
 	aligned_start = PAGE_ALIGN(start);
+	size -= aligned_start - start;
 
 #ifndef CONFIG_ARCH_PHYS_ADDR_T_64BIT
 	if (aligned_start > ULONG_MAX) {
-- 
1.9.1

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Masahiro Yamada]

Hi.

Any comments for this patch?


Thanks

Best Regards
Masahiro Yamada



On Fri,  6 Jun 2014 12:32:47 +0900
Masahiro Yamada <yamada.m@jp.panasonic.com> wrote:

> The current rounding of "size" is wrong.
> 
>     size -= start & ~PAGE_MASK;
> 
> should be
> 
>     size -= PAGE_SIZE - (start & ~PAGE_MASK);
> 
> Or more simply
> 
>     size -= aligned_start - start;
> 
> Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
> ---
>  arch/arm/kernel/setup.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
> index 50e198c..84508c0 100644
> --- a/arch/arm/kernel/setup.c
> +++ b/arch/arm/kernel/setup.c
> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
>  	 * Ensure that start/size are aligned to a page boundary.
>  	 * Size is appropriately rounded down, start is rounded up.
>  	 */
> -	size -= start & ~PAGE_MASK;
>  	aligned_start = PAGE_ALIGN(start);
> +	size -= aligned_start - start;
>  
>  #ifndef CONFIG_ARCH_PHYS_ADDR_T_64BIT
>  	if (aligned_start > ULONG_MAX) {
> -- 
> 1.9.1

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Geert Uytterhoeven]

Hi Yamada-san,

On Fri, Aug 1, 2014 at 2:37 PM, Masahiro Yamada
<yamada.m@jp.panasonic.com> wrote:
> Any comments for this patch?

Thanks for your patch!

I had a slightly different fix for the same issue, but apparently it got stuck
on the freshly installed machine that didn't have email set up yet ;-(

Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>

early_init_dt_add_memory_arch() has the same bug, I'll resend my (also stuck)
fix for that.

> Thanks
>
> Best Regards
> Masahiro Yamada
>
>
>
> On Fri,  6 Jun 2014 12:32:47 +0900
> Masahiro Yamada <yamada.m@jp.panasonic.com> wrote:
>
>> The current rounding of "size" is wrong.
>>
>>     size -= start & ~PAGE_MASK;
>>
>> should be
>>
>>     size -= PAGE_SIZE - (start & ~PAGE_MASK);
>>
>> Or more simply
>>
>>     size -= aligned_start - start;
>>
>> Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
>> ---
>>  arch/arm/kernel/setup.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
>> index 50e198c..84508c0 100644
>> --- a/arch/arm/kernel/setup.c
>> +++ b/arch/arm/kernel/setup.c
>> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
>>        * Ensure that start/size are aligned to a page boundary.
>>        * Size is appropriately rounded down, start is rounded up.
>>        */
>> -     size -= start & ~PAGE_MASK;
>>       aligned_start = PAGE_ALIGN(start);
>> +     size -= aligned_start - start;
>>
>>  #ifndef CONFIG_ARCH_PHYS_ADDR_T_64BIT
>>       if (aligned_start > ULONG_MAX) {
>> --
>> 1.9.1

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Geert Uytterhoeven]

Any chance we can get this fix applied?

On Wed, Aug 20, 2014 at 5:08 PM, Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
> Hi Yamada-san,
>
> On Fri, Aug 1, 2014 at 2:37 PM, Masahiro Yamada
> <yamada.m@jp.panasonic.com> wrote:
>> Any comments for this patch?
>
> Thanks for your patch!
>
> I had a slightly different fix for the same issue, but apparently it got stuck
> on the freshly installed machine that didn't have email set up yet ;-(
>
> Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
>
> early_init_dt_add_memory_arch() has the same bug, I'll resend my (also stuck)
> fix for that.
>
>> Thanks
>>
>> Best Regards
>> Masahiro Yamada
>>
>>
>>
>> On Fri,  6 Jun 2014 12:32:47 +0900
>> Masahiro Yamada <yamada.m@jp.panasonic.com> wrote:
>>
>>> The current rounding of "size" is wrong.
>>>
>>>     size -= start & ~PAGE_MASK;
>>>
>>> should be
>>>
>>>     size -= PAGE_SIZE - (start & ~PAGE_MASK);
>>>
>>> Or more simply
>>>
>>>     size -= aligned_start - start;
>>>
>>> Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
>>> ---
>>>  arch/arm/kernel/setup.c | 2 +-
>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
>>> index 50e198c..84508c0 100644
>>> --- a/arch/arm/kernel/setup.c
>>> +++ b/arch/arm/kernel/setup.c
>>> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
>>>        * Ensure that start/size are aligned to a page boundary.
>>>        * Size is appropriately rounded down, start is rounded up.
>>>        */
>>> -     size -= start & ~PAGE_MASK;
>>>       aligned_start = PAGE_ALIGN(start);
>>> +     size -= aligned_start - start;
>>>
>>>  #ifndef CONFIG_ARCH_PHYS_ADDR_T_64BIT
>>>       if (aligned_start > ULONG_MAX) {
>>> --
>>> 1.9.1

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Russell King - ARM Linux]

On Thu, Nov 27, 2014 at 11:36:31AM +0100, Geert Uytterhoeven wrote:
> Any chance we can get this fix applied?

As ever, if it ends up in the patch system, then I'll apply it, otherwise
it very quickly falls off of my radar due to the shere volume of email I
receive.

-- 
FTTC broadband for 0.8mile line: currently at 9.5Mbps down 400kbps up
according to speedtest.net.

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Ard Biesheuvel]

On 27 November 2014 at 11:36, Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> Any chance we can get this fix applied?
>
> On Wed, Aug 20, 2014 at 5:08 PM, Geert Uytterhoeven
> <geert@linux-m68k.org> wrote:
>> Hi Yamada-san,
>>
>> On Fri, Aug 1, 2014 at 2:37 PM, Masahiro Yamada
>> <yamada.m@jp.panasonic.com> wrote:
>>> Any comments for this patch?
>>
>> Thanks for your patch!
>>
>> I had a slightly different fix for the same issue, but apparently it got stuck
>> on the freshly installed machine that didn't have email set up yet ;-(
>>
>> Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
>>
>> early_init_dt_add_memory_arch() has the same bug, I'll resend my (also stuck)
>> fix for that.
>>
>>> Thanks
>>>
>>> Best Regards
>>> Masahiro Yamada
>>>
>>>
>>>
>>> On Fri,  6 Jun 2014 12:32:47 +0900
>>> Masahiro Yamada <yamada.m@jp.panasonic.com> wrote:
>>>
>>>> The current rounding of "size" is wrong.
>>>>
>>>>     size -= start & ~PAGE_MASK;
>>>>
>>>> should be
>>>>
>>>>     size -= PAGE_SIZE - (start & ~PAGE_MASK);
>>>>
>>>> Or more simply
>>>>
>>>>     size -= aligned_start - start;
>>>>
>>>> Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
>>>> ---
>>>>  arch/arm/kernel/setup.c | 2 +-
>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
>>>> index 50e198c..84508c0 100644
>>>> --- a/arch/arm/kernel/setup.c
>>>> +++ b/arch/arm/kernel/setup.c
>>>> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
>>>>        * Ensure that start/size are aligned to a page boundary.
>>>>        * Size is appropriately rounded down, start is rounded up.
>>>>        */
>>>> -     size -= start & ~PAGE_MASK;
>>>>       aligned_start = PAGE_ALIGN(start);
>>>> +     size -= aligned_start - start;

Beware that if size if sufficiently small, (i.e., smaller than the
unaligned leading fraction @ start), this wraps around and produces a
bogus value.

-- 
Ard.


>>>>
>>>>  #ifndef CONFIG_ARCH_PHYS_ADDR_T_64BIT
>>>>       if (aligned_start > ULONG_MAX) {
>>>> --
>>>> 1.9.1
>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Geert Uytterhoeven]

On Thu, Nov 27, 2014 at 11:50 AM, Ard Biesheuvel
<ard.biesheuvel@linaro.org> wrote:
>>>>> --- a/arch/arm/kernel/setup.c
>>>>> +++ b/arch/arm/kernel/setup.c
>>>>> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
>>>>>        * Ensure that start/size are aligned to a page boundary.
>>>>>        * Size is appropriately rounded down, start is rounded up.
>>>>>        */
>>>>> -     size -= start & ~PAGE_MASK;
>>>>>       aligned_start = PAGE_ALIGN(start);
>>>>> +     size -= aligned_start - start;
>
> Beware that if size if sufficiently small, (i.e., smaller than the
> unaligned leading fraction @ start), this wraps around and produces a
> bogus value.

That could have happened before, couldn't it?

Nevertheless, fixing that can be a follow-up patch?
Unless Yamada-san wants to incorporate the fix in this patch?

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[PATCH] ARM: mm: fix size rounding-down of arm_add_memory()

[Masahiro Yamada]

Hi.


On Thu, 27 Nov 2014 12:04:47 +0100
Geert Uytterhoeven <geert@linux-m68k.org> wrote:

> On Thu, Nov 27, 2014 at 11:50 AM, Ard Biesheuvel
> <ard.biesheuvel@linaro.org> wrote:
> >>>>> --- a/arch/arm/kernel/setup.c
> >>>>> +++ b/arch/arm/kernel/setup.c
> >>>>> @@ -641,8 +641,8 @@ int __init arm_add_memory(u64 start, u64 size)
> >>>>>        * Ensure that start/size are aligned to a page boundary.
> >>>>>        * Size is appropriately rounded down, start is rounded up.
> >>>>>        */
> >>>>> -     size -= start & ~PAGE_MASK;
> >>>>>       aligned_start = PAGE_ALIGN(start);
> >>>>> +     size -= aligned_start - start;
> >
> > Beware that if size if sufficiently small, (i.e., smaller than the
> > unaligned leading fraction @ start), this wraps around and produces a
> > bogus value.

Uh-oh, yes. I missed this.


> 
> That could have happened before, couldn't it?
> 
> Nevertheless, fixing that can be a follow-up patch?
> Unless Yamada-san wants to incorporate the fix in this patch?

We don't two patches to fix one place.

If you are willing to fix it,
please feel free to discard mine and send a new patch.


(If nobody would fix that, I am happy to contribute
for fixing this issue, of course.)



Best Regards
Masahiro Yamada