On Zynq, we haven't been reserving the correct amount of DMA-incapable
RAM to keep DMA away from it (per the Zynq TRM Section 4.1, it should be
the first 512k). In older kernels, this was masked by the
memblock_reserve call in arm_memblock_init(). Now, reserve the correct
amount excplicitly rather than relying on swapper_pg_dir, which is an
address and not a size anyway.
Fixes: 46f5b96 ("ARM: zynq: Reserve not DMAable space in front of the
kernel")
Signed-off-by: Kyle Roeschley <[email protected]>
---
Found when migrating from 4.1 to 4.6.
arch/arm/mach-zynq/common.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/arm/mach-zynq/common.c b/arch/arm/mach-zynq/common.c
index 6cefdb8..75885bc 100644
--- a/arch/arm/mach-zynq/common.c
+++ b/arch/arm/mach-zynq/common.c
@@ -59,7 +59,7 @@ void __iomem *zynq_scu_base;
static void __init zynq_memory_init(void)
{
if (!__pa(PAGE_OFFSET))
- memblock_reserve(__pa(PAGE_OFFSET), __pa(swapper_pg_dir));
+ memblock_reserve(__pa(PAGE_OFFSET), 0x80000);
}
static struct platform_device zynq_cpuidle_device = {
--
2.9.3
On 10 November 2016 at 18:41, Michal Simek <[email protected]> wrote:
> + Nathan
>
> 2016-10-31 17:26 GMT+01:00 Kyle Roeschley <[email protected]>:
>>
>> On Zynq, we haven't been reserving the correct amount of DMA-incapable
>> RAM to keep DMA away from it (per the Zynq TRM Section 4.1, it should be
>> the first 512k). In older kernels, this was masked by the
>> memblock_reserve call in arm_memblock_init(). Now, reserve the correct
>> amount excplicitly rather than relying on swapper_pg_dir, which is an
>> address and not a size anyway.
>>
>> Fixes: 46f5b96 ("ARM: zynq: Reserve not DMAable space in front of the
>> kernel")
>>
>> Signed-off-by: Kyle Roeschley <[email protected]>
Tested-by: Nathan Rossi <[email protected]>
For reference this causes problems with DEBUG_RODATA (which changed to
default yes for CPU_V7 in v4.6) due to padding memory between
.head.text and .text, allowing memory below 0x80000 to be available
for allocation as non-reserved memory.
Regards,
Nathan
>> ---
>> Found when migrating from 4.1 to 4.6.
>>
>> arch/arm/mach-zynq/common.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/arch/arm/mach-zynq/common.c b/arch/arm/mach-zynq/common.c
>> index 6cefdb8..75885bc 100644
>> --- a/arch/arm/mach-zynq/common.c
>> +++ b/arch/arm/mach-zynq/common.c
>> @@ -59,7 +59,7 @@ void __iomem *zynq_scu_base;
>> static void __init zynq_memory_init(void)
>> {
>> if (!__pa(PAGE_OFFSET))
>> - memblock_reserve(__pa(PAGE_OFFSET), __pa(swapper_pg_dir));
>> + memblock_reserve(__pa(PAGE_OFFSET), 0x80000);
>> }
>>
>> static struct platform_device zynq_cpuidle_device = {
>> --
>> 2.9.3
>>
>>
>> _______________________________________________
>> linux-arm-kernel mailing list
>> [email protected]
>> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>
>
>
>
> --
> Michal Simek, Ing. (M.Eng), OpenPGP -> KeyID: FE3D1F91
> w: http://www.monstr.eu p: +42-0-721842854
> Maintainer of Linux kernel - Microblaze cpu - http://www.monstr.eu/fdt/
> Maintainer of Linux kernel - Xilinx Zynq ARM architecture
> Microblaze U-BOOT custodian and responsible for u-boot arm zynq platform
On 10 November 2016 at 19:33, Nathan Rossi <[email protected]> wrote:
> On 10 November 2016 at 18:41, Michal Simek <[email protected]> wrote:
>> + Nathan
>>
>> 2016-10-31 17:26 GMT+01:00 Kyle Roeschley <[email protected]>:
>>>
>>> On Zynq, we haven't been reserving the correct amount of DMA-incapable
>>> RAM to keep DMA away from it (per the Zynq TRM Section 4.1, it should be
>>> the first 512k). In older kernels, this was masked by the
>>> memblock_reserve call in arm_memblock_init(). Now, reserve the correct
>>> amount excplicitly rather than relying on swapper_pg_dir, which is an
>>> address and not a size anyway.
>>>
>>> Fixes: 46f5b96 ("ARM: zynq: Reserve not DMAable space in front of the
>>> kernel")
>>>
>>> Signed-off-by: Kyle Roeschley <[email protected]>
>
> Tested-by: Nathan Rossi <[email protected]>
>
> For reference this causes problems with DEBUG_RODATA (which changed to
Sorry typo -> s/causes/caused/, as in "... this [incorrect reserving
of the lower 512K] caused ...".
Regards,
Nathan
> default yes for CPU_V7 in v4.6) due to padding memory between
> .head.text and .text, allowing memory below 0x80000 to be available
> for allocation as non-reserved memory.
>
> Regards,
> Nathan
>
>>> ---
>>> Found when migrating from 4.1 to 4.6.
>>>
>>> arch/arm/mach-zynq/common.c | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm/mach-zynq/common.c b/arch/arm/mach-zynq/common.c
>>> index 6cefdb8..75885bc 100644
>>> --- a/arch/arm/mach-zynq/common.c
>>> +++ b/arch/arm/mach-zynq/common.c
>>> @@ -59,7 +59,7 @@ void __iomem *zynq_scu_base;
>>> static void __init zynq_memory_init(void)
>>> {
>>> if (!__pa(PAGE_OFFSET))
>>> - memblock_reserve(__pa(PAGE_OFFSET), __pa(swapper_pg_dir));
>>> + memblock_reserve(__pa(PAGE_OFFSET), 0x80000);
>>> }
>>>
>>> static struct platform_device zynq_cpuidle_device = {
>>> --
>>> 2.9.3
>>>
>>>
>>> _______________________________________________
>>> linux-arm-kernel mailing list
>>> [email protected]
>>> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>>
>>
>>
>>
>> --
>> Michal Simek, Ing. (M.Eng), OpenPGP -> KeyID: FE3D1F91
>> w: http://www.monstr.eu p: +42-0-721842854
>> Maintainer of Linux kernel - Microblaze cpu - http://www.monstr.eu/fdt/
>> Maintainer of Linux kernel - Xilinx Zynq ARM architecture
>> Microblaze U-BOOT custodian and responsible for u-boot arm zynq platform
On 10.11.2016 10:57, Nathan Rossi wrote:
> On 10 November 2016 at 19:33, Nathan Rossi <[email protected]> wrote:
>> On 10 November 2016 at 18:41, Michal Simek <[email protected]> wrote:
>>> + Nathan
>>>
>>> 2016-10-31 17:26 GMT+01:00 Kyle Roeschley <[email protected]>:
>>>>
>>>> On Zynq, we haven't been reserving the correct amount of DMA-incapable
>>>> RAM to keep DMA away from it (per the Zynq TRM Section 4.1, it should be
>>>> the first 512k). In older kernels, this was masked by the
>>>> memblock_reserve call in arm_memblock_init(). Now, reserve the correct
>>>> amount excplicitly rather than relying on swapper_pg_dir, which is an
>>>> address and not a size anyway.
>>>>
>>>> Fixes: 46f5b96 ("ARM: zynq: Reserve not DMAable space in front of the
>>>> kernel")
>>>>
>>>> Signed-off-by: Kyle Roeschley <[email protected]>
>>
>> Tested-by: Nathan Rossi <[email protected]>
>>
>> For reference this causes problems with DEBUG_RODATA (which changed to
>
> Sorry typo -> s/causes/caused/, as in "... this [incorrect reserving
> of the lower 512K] caused ...".
>
> Regards,
> Nathan
>
>> default yes for CPU_V7 in v4.6) due to padding memory between
>> .head.text and .text, allowing memory below 0x80000 to be available
>> for allocation as non-reserved memory.
Applied to zynq/soc
https://github.com/Xilinx/linux-xlnx/commits/zynq/soc
Thanks,
Michal