2015-05-04 09:56:38

by Hui Zhu

[permalink] [raw]
Subject: [PATCH] CMA: page_isolation: check buddy before access it

I got a issue:
[ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
[ 214.303013] pgd = cc970000
[ 214.305721] [0000082a] *pgd=00000000
[ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
[ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
[ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
[ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
[ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
[ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
[ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
[ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
[ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
[ 214.949720] Backtrace:
[ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
[ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
[ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
[ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
[ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
r4:c7029db4
[ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
[ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)

This issue is because when call unset_migratetype_isolate to unset a part
of CMA memory, it try to access the buddy page to get its status:
if (order >= pageblock_order) {
page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

if (!is_migrate_isolate_page(buddy)) {
But the begin addr of this part of CMA memory is very close to a part of
memory that is reserved in the boot time (not in buddy system).
So add a check before access it.

Signed-off-by: Hui Zhu <[email protected]>
---
mm/page_isolation.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 755a42c..434730b 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

- if (!is_migrate_isolate_page(buddy)) {
+ if (!pfn_present(page_to_pfn(buddy))
+ || !is_migrate_isolate_page(buddy)) {
__isolate_free_page(page, order);
kernel_map_pages(page, (1 << order), 1);
set_page_refcounted(page);
--
1.9.1


2015-05-04 18:34:52

by Laura Abbott

[permalink] [raw]
Subject: Re: [PATCH] CMA: page_isolation: check buddy before access it

On 05/04/2015 02:41 AM, Hui Zhu wrote:
> I got a issue:
> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
> [ 214.303013] pgd = cc970000
> [ 214.305721] [0000082a] *pgd=00000000
> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
> [ 214.949720] Backtrace:
> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
> r4:c7029db4
> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>
> This issue is because when call unset_migratetype_isolate to unset a part
> of CMA memory, it try to access the buddy page to get its status:
> if (order >= pageblock_order) {
> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> if (!is_migrate_isolate_page(buddy)) {
> But the begin addr of this part of CMA memory is very close to a part of
> memory that is reserved in the boot time (not in buddy system).
> So add a check before access it.
>
> Signed-off-by: Hui Zhu <[email protected]>
> ---
> mm/page_isolation.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 755a42c..434730b 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> - if (!is_migrate_isolate_page(buddy)) {
> + if (!pfn_present(page_to_pfn(buddy))
> + || !is_migrate_isolate_page(buddy)) {
> __isolate_free_page(page, order);
> kernel_map_pages(page, (1 << order), 1);
> set_page_refcounted(page);
>

I think you want to use pfn_valid_within instead of pfn_present.

Thanks,
Laura

2015-05-05 03:18:03

by Hui Zhu

[permalink] [raw]
Subject: Re: [PATCH] CMA: page_isolation: check buddy before access it

On Tue, May 5, 2015 at 2:34 AM, Laura Abbott <[email protected]> wrote:
> On 05/04/2015 02:41 AM, Hui Zhu wrote:
>>
>> I got a issue:
>> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual
>> address 0000082a
>> [ 214.303013] pgd = cc970000
>> [ 214.305721] [0000082a] *pgd=00000000
>> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
>> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
>> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
>> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
>> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
>> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
>> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
>> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
>> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM
>> Segment user
>> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
>> [ 214.949720] Backtrace:
>> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from
>> [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
>> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
>> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from
>> [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
>> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from
>> [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
>> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00
>> r5:00000005
>> r4:c7029db4
>> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from
>> [<c00d0a88>] (alloc_contig_range+0x14/0x18)
>> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from
>> [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>>
>> This issue is because when call unset_migratetype_isolate to unset a part
>> of CMA memory, it try to access the buddy page to get its status:
>> if (order >= pageblock_order) {
>> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) -
>> 1);
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> if (!is_migrate_isolate_page(buddy)) {
>> But the begin addr of this part of CMA memory is very close to a part of
>> memory that is reserved in the boot time (not in buddy system).
>> So add a check before access it.
>>
>> Signed-off-by: Hui Zhu <[email protected]>
>> ---
>> mm/page_isolation.c | 3 ++-
>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
>> index 755a42c..434730b 100644
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page,
>> unsigned migratetype)
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> - if (!is_migrate_isolate_page(buddy)) {
>> + if (!pfn_present(page_to_pfn(buddy))
>> + || !is_migrate_isolate_page(buddy)) {
>> __isolate_free_page(page, order);
>> kernel_map_pages(page, (1 << order), 1);
>> set_page_refcounted(page);
>>
>
> I think you want to use pfn_valid_within instead of pfn_present.

Thanks. I will post a new version for it.

Best,
Hui

>
> Thanks,
> Laura
>

2015-05-05 03:23:25

by Hui Zhu

[permalink] [raw]
Subject: [PATCH v2] CMA: page_isolation: check buddy before access it

Change pfn_present to pfn_valid_within according to the review of Laura.

I got a issue:
[ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
[ 214.303013] pgd = cc970000
[ 214.305721] [0000082a] *pgd=00000000
[ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
[ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
[ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
[ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
[ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
[ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
[ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
[ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
[ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
[ 214.949720] Backtrace:
[ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
[ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
[ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
[ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
[ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
r4:c7029db4
[ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
[ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)

This issue is because when call unset_migratetype_isolate to unset a part
of CMA memory, it try to access the buddy page to get its status:
if (order >= pageblock_order) {
page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

if (!is_migrate_isolate_page(buddy)) {
But the begin addr of this part of CMA memory is very close to a part of
memory that is reserved in the boot time (not in buddy system).
So add a check before access it.

Signed-off-by: Hui Zhu <[email protected]>
---
mm/page_isolation.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 755a42c..eb22d1f 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

- if (!is_migrate_isolate_page(buddy)) {
+ if (!pfn_valid_within(page_to_pfn(buddy))
+ || !is_migrate_isolate_page(buddy)) {
__isolate_free_page(page, order);
kernel_map_pages(page, (1 << order), 1);
set_page_refcounted(page);
--
1.9.1

2015-05-05 08:43:56

by Vlastimil Babka

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On 05/05/2015 05:22 AM, Hui Zhu wrote:
> Change pfn_present to pfn_valid_within according to the review of Laura.
>
> I got a issue:
> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
> [ 214.303013] pgd = cc970000
> [ 214.305721] [0000082a] *pgd=00000000
> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
> [ 214.949720] Backtrace:
> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
> r4:c7029db4
> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>
> This issue is because when call unset_migratetype_isolate to unset a part
> of CMA memory, it try to access the buddy page to get its status:
> if (order >= pageblock_order) {
> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> if (!is_migrate_isolate_page(buddy)) {
> But the begin addr of this part of CMA memory is very close to a part of
> memory that is reserved in the boot time (not in buddy system).
> So add a check before access it.
>
> Signed-off-by: Hui Zhu <[email protected]>

Acked-by: Vlastimil Babka <[email protected]>

> ---
> mm/page_isolation.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 755a42c..eb22d1f 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> - if (!is_migrate_isolate_page(buddy)) {
> + if (!pfn_valid_within(page_to_pfn(buddy))
> + || !is_migrate_isolate_page(buddy)) {
> __isolate_free_page(page, order);
> kernel_map_pages(page, (1 << order), 1);
> set_page_refcounted(page);
>

2015-05-05 21:29:51

by Andrew Morton

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On Tue, 5 May 2015 11:22:59 +0800 Hui Zhu <[email protected]> wrote:

> Change pfn_present to pfn_valid_within according to the review of Laura.
>
> I got a issue:
> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
> [ 214.303013] pgd = cc970000
> [ 214.305721] [0000082a] *pgd=00000000
> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
> [ 214.949720] Backtrace:
> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
> r4:c7029db4
> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>
> This issue is because when call unset_migratetype_isolate to unset a part
> of CMA memory, it try to access the buddy page to get its status:
> if (order >= pageblock_order) {
> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> if (!is_migrate_isolate_page(buddy)) {
> But the begin addr of this part of CMA memory is very close to a part of
> memory that is reserved in the boot time (not in buddy system).
> So add a check before access it.
>
> ...
>
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> - if (!is_migrate_isolate_page(buddy)) {
> + if (!pfn_valid_within(page_to_pfn(buddy))
> + || !is_migrate_isolate_page(buddy)) {
> __isolate_free_page(page, order);
> kernel_map_pages(page, (1 << order), 1);
> set_page_refcounted(page);

This fix is needed in kernel versions 4.0.x isn't it?

2015-05-06 01:40:10

by Hui Zhu

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On Wed, May 6, 2015 at 5:29 AM, Andrew Morton <[email protected]> wrote:
> On Tue, 5 May 2015 11:22:59 +0800 Hui Zhu <[email protected]> wrote:
>
>> Change pfn_present to pfn_valid_within according to the review of Laura.
>>
>> I got a issue:
>> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
>> [ 214.303013] pgd = cc970000
>> [ 214.305721] [0000082a] *pgd=00000000
>> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
>> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
>> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
>> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
>> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
>> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
>> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
>> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
>> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
>> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
>> [ 214.949720] Backtrace:
>> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
>> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
>> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
>> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
>> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
>> r4:c7029db4
>> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
>> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>>
>> This issue is because when call unset_migratetype_isolate to unset a part
>> of CMA memory, it try to access the buddy page to get its status:
>> if (order >= pageblock_order) {
>> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> if (!is_migrate_isolate_page(buddy)) {
>> But the begin addr of this part of CMA memory is very close to a part of
>> memory that is reserved in the boot time (not in buddy system).
>> So add a check before access it.
>>
>> ...
>>
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> - if (!is_migrate_isolate_page(buddy)) {
>> + if (!pfn_valid_within(page_to_pfn(buddy))
>> + || !is_migrate_isolate_page(buddy)) {
>> __isolate_free_page(page, order);
>> kernel_map_pages(page, (1 << order), 1);
>> set_page_refcounted(page);
>
> This fix is needed in kernel versions 4.0.x isn't it?

I think it need it.

Thanks,
Hui

2015-05-06 06:25:28

by Joonsoo Kim

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On Tue, May 05, 2015 at 11:22:59AM +0800, Hui Zhu wrote:
> Change pfn_present to pfn_valid_within according to the review of Laura.
>
> I got a issue:
> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
> [ 214.303013] pgd = cc970000
> [ 214.305721] [0000082a] *pgd=00000000
> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
> [ 214.949720] Backtrace:
> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
> r4:c7029db4
> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>
> This issue is because when call unset_migratetype_isolate to unset a part
> of CMA memory, it try to access the buddy page to get its status:
> if (order >= pageblock_order) {
> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> if (!is_migrate_isolate_page(buddy)) {
> But the begin addr of this part of CMA memory is very close to a part of
> memory that is reserved in the boot time (not in buddy system).
> So add a check before access it.
>
> Signed-off-by: Hui Zhu <[email protected]>
> ---
> mm/page_isolation.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index 755a42c..eb22d1f 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
> buddy_idx = __find_buddy_index(page_idx, order);
> buddy = page + (buddy_idx - page_idx);
>
> - if (!is_migrate_isolate_page(buddy)) {
> + if (!pfn_valid_within(page_to_pfn(buddy))
> + || !is_migrate_isolate_page(buddy)) {
> __isolate_free_page(page, order);
> kernel_map_pages(page, (1 << order), 1);
> set_page_refcounted(page);

Hello,

This isolation is for merging buddy pages. If buddy is not valid, we
don't need to isolate page, because we can't merge them.
I think that correct code would be:

pfn_valid_within(page_to_pfn(buddy)) &&
!is_migrate_isolate_page(buddy)

But, isolation and free here is safe operation so your code will work
fine.

Thanks.

2015-05-06 07:08:32

by Hui Zhu

[permalink] [raw]
Subject: [PATCH v3] CMA: page_isolation: check buddy before access it

Changelog:
v3, Change the behavior according to the review of Joonsoo.
v2, Change pfn_present to pfn_valid_within according to the review of Laura.

I got a issue:
[ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
[ 214.303013] pgd = cc970000
[ 214.305721] [0000082a] *pgd=00000000
[ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
[ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
[ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
[ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
[ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
[ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
[ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
[ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
[ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
[ 214.949720] Backtrace:
[ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
[ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
[ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
[ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
[ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
r4:c7029db4
[ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
[ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)

This issue is because when call unset_migratetype_isolate to unset a part
of CMA memory, it try to access the buddy page to get its status:
if (order >= pageblock_order) {
page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

if (!is_migrate_isolate_page(buddy)) {
But the begin addr of this part of CMA memory is very close to a part of
memory that is reserved in the boot time (not in buddy system).
So add a check before access it.

Suggested-by: Laura Abbott <[email protected]>
Suggested-by: Joonsoo Kim <[email protected]>
Signed-off-by: Hui Zhu <[email protected]>
---
mm/page_isolation.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 755a42c..4a5624c 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
buddy_idx = __find_buddy_index(page_idx, order);
buddy = page + (buddy_idx - page_idx);

- if (!is_migrate_isolate_page(buddy)) {
+ if (pfn_valid_within(page_to_pfn(buddy))
+ && !is_migrate_isolate_page(buddy)) {
__isolate_free_page(page, order);
kernel_map_pages(page, (1 << order), 1);
set_page_refcounted(page);
--
1.9.1

2015-05-06 07:10:22

by Hui Zhu

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On Wed, May 6, 2015 at 2:28 PM, Joonsoo Kim <[email protected]> wrote:
> On Tue, May 05, 2015 at 11:22:59AM +0800, Hui Zhu wrote:
>> Change pfn_present to pfn_valid_within according to the review of Laura.
>>
>> I got a issue:
>> [ 214.294917] Unable to handle kernel NULL pointer dereference at virtual address 0000082a
>> [ 214.303013] pgd = cc970000
>> [ 214.305721] [0000082a] *pgd=00000000
>> [ 214.309316] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
>> [ 214.335704] PC is at get_pageblock_flags_group+0x5c/0xb0
>> [ 214.341030] LR is at unset_migratetype_isolate+0x148/0x1b0
>> [ 214.346523] pc : [<c00cc9a0>] lr : [<c0109874>] psr: 80000093
>> [ 214.346523] sp : c7029d00 ip : 00000105 fp : c7029d1c
>> [ 214.358005] r10: 00000001 r9 : 0000000a r8 : 00000004
>> [ 214.363231] r7 : 60000013 r6 : 000000a4 r5 : c0a357e4 r4 : 00000000
>> [ 214.369761] r3 : 00000826 r2 : 00000002 r1 : 00000000 r0 : 0000003f
>> [ 214.376291] Flags: Nzcv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user
>> [ 214.383516] Control: 10c5387d Table: 2cb7006a DAC: 00000015
>> [ 214.949720] Backtrace:
>> [ 214.952192] [<c00cc944>] (get_pageblock_flags_group+0x0/0xb0) from [<c0109874>] (unset_migratetype_isolate+0x148/0x1b0)
>> [ 214.962978] r7:60000013 r6:c0a357c0 r5:c0a357e4 r4:c1555000
>> [ 214.968693] [<c010972c>] (unset_migratetype_isolate+0x0/0x1b0) from [<c0109adc>] (undo_isolate_page_range+0xd0/0xdc)
>> [ 214.979222] [<c0109a0c>] (undo_isolate_page_range+0x0/0xdc) from [<c00d097c>] (__alloc_contig_range+0x254/0x34c)
>> [ 214.989398] r9:000abc00 r8:c7028000 r7:000b1f53 r6:000b3e00 r5:00000005
>> r4:c7029db4
>> [ 214.997308] [<c00d0728>] (__alloc_contig_range+0x0/0x34c) from [<c00d0a88>] (alloc_contig_range+0x14/0x18)
>> [ 215.006973] [<c00d0a74>] (alloc_contig_range+0x0/0x18) from [<c0398148>] (dma_alloc_from_contiguous_addr+0x1ac/0x304)
>>
>> This issue is because when call unset_migratetype_isolate to unset a part
>> of CMA memory, it try to access the buddy page to get its status:
>> if (order >= pageblock_order) {
>> page_idx = page_to_pfn(page) & ((1 << MAX_ORDER) - 1);
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> if (!is_migrate_isolate_page(buddy)) {
>> But the begin addr of this part of CMA memory is very close to a part of
>> memory that is reserved in the boot time (not in buddy system).
>> So add a check before access it.
>>
>> Signed-off-by: Hui Zhu <[email protected]>
>> ---
>> mm/page_isolation.c | 3 ++-
>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
>> index 755a42c..eb22d1f 100644
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> - if (!is_migrate_isolate_page(buddy)) {
>> + if (!pfn_valid_within(page_to_pfn(buddy))
>> + || !is_migrate_isolate_page(buddy)) {
>> __isolate_free_page(page, order);
>> kernel_map_pages(page, (1 << order), 1);
>> set_page_refcounted(page);
>
> Hello,
>
> This isolation is for merging buddy pages. If buddy is not valid, we
> don't need to isolate page, because we can't merge them.
> I think that correct code would be:
>
> pfn_valid_within(page_to_pfn(buddy)) &&
> !is_migrate_isolate_page(buddy)
>
> But, isolation and free here is safe operation so your code will work
> fine.
>

Oops! I posted a new version for the patch.

Thanks,
Hui

> Thanks.

2015-05-06 07:55:40

by Vlastimil Babka

[permalink] [raw]
Subject: Re: [PATCH v2] CMA: page_isolation: check buddy before access it

On 6.5.2015 8:28, Joonsoo Kim wrote:
> On Tue, May 05, 2015 at 11:22:59AM +0800, Hui Zhu wrote:
>>
>> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
>> index 755a42c..eb22d1f 100644
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -101,7 +101,8 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>> buddy_idx = __find_buddy_index(page_idx, order);
>> buddy = page + (buddy_idx - page_idx);
>>
>> - if (!is_migrate_isolate_page(buddy)) {
>> + if (!pfn_valid_within(page_to_pfn(buddy))
>> + || !is_migrate_isolate_page(buddy)) {
>> __isolate_free_page(page, order);
>> kernel_map_pages(page, (1 << order), 1);
>> set_page_refcounted(page);
>
> Hello,
>
> This isolation is for merging buddy pages. If buddy is not valid, we
> don't need to isolate page, because we can't merge them.
> I think that correct code would be:
>
> pfn_valid_within(page_to_pfn(buddy)) &&
> !is_migrate_isolate_page(buddy)
>
> But, isolation and free here is safe operation so your code will work
> fine.

Ah damnit, you're right. But now you got me thinking about it more, and
paranoid... I thought I saw more bugs since the buddy might be in different zone
and we are not locking that zone, but then again it's probably fine, just very
tricky. Then I thought it could be simplified but then not again. Guess I'll
just run away fast :)

> Thanks.
>