Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On 4/8/21 9:20 PM, Guenter Roeck wrote: > On 4/8/21 3:53 PM, Frank Rowand wrote: >> On 4/8/21 4:54 PM, Guenter Roeck wrote: >>> On 4/8/21 2:28 PM, Rob Herring wrote: Applying now so this gets into linux-next this week. >>> The patch doesn't apply on top of today's -next; it conflicts >>> with "of: properly check for error returned by fdt_get_name()". >>> >>> I reverted that patch and applied this one, and the DT unittests >>> run with it on openrisc. I do get a single test failure, but I that >>> is a different problem (possibly with the test case itself). >>> >>> ### dt-test ### FAIL of_unittest_dma_ranges_one():923 of_dma_get_range: >>> wrong DMA addr 0x >>> (expecting 1) on node >>> /testcase-data/address-tests/bus@8000/device@1000 >> >> That is a known regression on the target that I use for testing (and >> has been since 5.10-rc1) - the 8074 dragonboard, arm 32. No >> one else has reported it on the list, so even though I want to debug >> and fix it "promptly", other tasks have had higher priority. In my >> notes I list two suspect commits: >> >> e0d072782c73 dma-mapping: introduce DMA range map, supplanting >> dma_pfn_offset >> 0a0f0d8be76d dma-mapping: split >> >> I think that was purely based on looking at the list of commits that >> may have touched OF dma. I have not done a bisect. >> > > Here you are: > > # bad: [2c85ebc57b3e1817b6ce1a6b703928e113a90442] Linux 5.10 > # good: [bbf5c979011a099af5dc76498918ed7df445635b] Linux 5.9 > git bisect start 'v5.10' 'v5.9' > # bad: [4d0e9df5e43dba52d38b251e3b909df8fa1110be] lib, uaccess: add failure > injection to usercopy functions > git bisect bad 4d0e9df5e43dba52d38b251e3b909df8fa1110be > # good: [f888bdf9823c85fe945c4eb3ba353f749dec3856] Merge tag > 'devicetree-for-5.10' of > git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux > git bisect good f888bdf9823c85fe945c4eb3ba353f749dec3856 > # good: [640eee067d9aae0bb98d8706001976ff1affaf00] Merge tag > 'drm-misc-next-fixes-2020-10-13' of > git://anongit.freedesktop.org/drm/drm-misc into drm-next > git bisect good 640eee067d9aae0bb98d8706001976ff1affaf00 > # good: [c6dbef7307629cce855aa6b482b60cbfed88] Merge tag 'usb-5.10-rc1' > of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb > git bisect good c6dbef7307629cce855aa6b482b60cbfed88 > # good: [ce1558c285f9ad04c03b46833a028230771cc0a7] ALSA: hda/hdmi: fix > incorrect locking in hdmi_pcm_close > git bisect good ce1558c285f9ad04c03b46833a028230771cc0a7 > # good: [c48b75b7271db23c1b2d1204d6e8496d91f27711] Merge tag 'sound-5.10-rc1' > of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound > git bisect good c48b75b7271db23c1b2d1204d6e8496d91f27711 > # bad: [0cd7d9795fa82226e7516d38b474bddae8b1ff26] Merge branch 'for-linus' of > git://git.kernel.org/pub/scm/linux/kernel/git/livepatching/livepatching > git bisect bad 0cd7d9795fa82226e7516d38b474bddae8b1ff26 > # good: [b1839e7c2a42ccd9a0587c0092e880c7a213ee2a] dmaengine: xilinx: dpdma: > convert tasklets to use new tasklet_setup() API > git bisect good b1839e7c2a42ccd9a0587c0092e880c7a213ee2a > # bad: [0de327969b61a245e3a47b60009eae73fe513cef] cma: decrease CMA_ALIGNMENT > lower limit to 2 > git bisect bad 0de327969b61a245e3a47b60009eae73fe513cef > # good: [6eb0233ec2d0df288fe8515d5b0b2b15562e05bb] usb: don't inherity DMA > properties for USB devices > git bisect good 6eb0233ec2d0df288fe8515d5b0b2b15562e05bb > # bad: [48d15814dd0fc429e3205b87f1af6cc472018478] lib82596: move DMA > allocation into the callers of i82596_probe > git bisect bad 48d15814dd0fc429e3205b87f1af6cc472018478 > # bad: [eba304c6861613a649ba46cfab835b1eddeacd8e] dma-mapping: better > document dma_addr_t and DMA_MAPPING_ERROR > git bisect bad eba304c6861613a649ba46cfab835b1eddeacd8e > # bad: [b9bb694b9f62f4b31652223ed3ca38cf98bbb370] iommu/io-pgtable-arm: Clean > up faulty sanity check > git bisect bad b9bb694b9f62f4b31652223ed3ca38cf98bbb370 > # bad: [a97740f81874c8063c12c24f34d25f10c4f5e9aa] dma-debug: convert comma to > semicolon > git bisect bad a97740f81874c8063c12c24f34d25f10c4f5e9aa > # bad: [e0d072782c734d27f5af062c62266f2598f68542] dma-mapping: introduce DMA > range map, supplanting dma_pfn_offset > git bisect bad e0d072782c734d27f5af062c62266f2598f68542 > # first bad commit: [e0d072782c734d27f5af062c62266f2598f68542] dma-mapping: > introduce DMA range map, supplanting dma_pfn_offset > > Guenter > Thank you
Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On 4/8/21 3:53 PM, Frank Rowand wrote: > On 4/8/21 4:54 PM, Guenter Roeck wrote: >> On 4/8/21 2:28 PM, Rob Herring wrote: >>> >>> Applying now so this gets into linux-next this week. >>> >> The patch doesn't apply on top of today's -next; it conflicts >> with "of: properly check for error returned by fdt_get_name()". >> >> I reverted that patch and applied this one, and the DT unittests >> run with it on openrisc. I do get a single test failure, but I that >> is a different problem (possibly with the test case itself). >> >> ### dt-test ### FAIL of_unittest_dma_ranges_one():923 of_dma_get_range: >> wrong DMA addr 0x >> (expecting 1) on node >> /testcase-data/address-tests/bus@8000/device@1000 > > That is a known regression on the target that I use for testing (and > has been since 5.10-rc1) - the 8074 dragonboard, arm 32. No > one else has reported it on the list, so even though I want to debug > and fix it "promptly", other tasks have had higher priority. In my > notes I list two suspect commits: > > e0d072782c73 dma-mapping: introduce DMA range map, supplanting > dma_pfn_offset > 0a0f0d8be76d dma-mapping: split > > I think that was purely based on looking at the list of commits that > may have touched OF dma. I have not done a bisect. > Here you are: # bad: [2c85ebc57b3e1817b6ce1a6b703928e113a90442] Linux 5.10 # good: [bbf5c979011a099af5dc76498918ed7df445635b] Linux 5.9 git bisect start 'v5.10' 'v5.9' # bad: [4d0e9df5e43dba52d38b251e3b909df8fa1110be] lib, uaccess: add failure injection to usercopy functions git bisect bad 4d0e9df5e43dba52d38b251e3b909df8fa1110be # good: [f888bdf9823c85fe945c4eb3ba353f749dec3856] Merge tag 'devicetree-for-5.10' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux git bisect good f888bdf9823c85fe945c4eb3ba353f749dec3856 # good: [640eee067d9aae0bb98d8706001976ff1affaf00] Merge tag 'drm-misc-next-fixes-2020-10-13' of git://anongit.freedesktop.org/drm/drm-misc into drm-next git bisect good 640eee067d9aae0bb98d8706001976ff1affaf00 # good: [c6dbef7307629cce855aa6b482b60cbfed88] Merge tag 'usb-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb git bisect good c6dbef7307629cce855aa6b482b60cbfed88 # good: [ce1558c285f9ad04c03b46833a028230771cc0a7] ALSA: hda/hdmi: fix incorrect locking in hdmi_pcm_close git bisect good ce1558c285f9ad04c03b46833a028230771cc0a7 # good: [c48b75b7271db23c1b2d1204d6e8496d91f27711] Merge tag 'sound-5.10-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound git bisect good c48b75b7271db23c1b2d1204d6e8496d91f27711 # bad: [0cd7d9795fa82226e7516d38b474bddae8b1ff26] Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/livepatching/livepatching git bisect bad 0cd7d9795fa82226e7516d38b474bddae8b1ff26 # good: [b1839e7c2a42ccd9a0587c0092e880c7a213ee2a] dmaengine: xilinx: dpdma: convert tasklets to use new tasklet_setup() API git bisect good b1839e7c2a42ccd9a0587c0092e880c7a213ee2a # bad: [0de327969b61a245e3a47b60009eae73fe513cef] cma: decrease CMA_ALIGNMENT lower limit to 2 git bisect bad 0de327969b61a245e3a47b60009eae73fe513cef # good: [6eb0233ec2d0df288fe8515d5b0b2b15562e05bb] usb: don't inherity DMA properties for USB devices git bisect good 6eb0233ec2d0df288fe8515d5b0b2b15562e05bb # bad: [48d15814dd0fc429e3205b87f1af6cc472018478] lib82596: move DMA allocation into the callers of i82596_probe git bisect bad 48d15814dd0fc429e3205b87f1af6cc472018478 # bad: [eba304c6861613a649ba46cfab835b1eddeacd8e] dma-mapping: better document dma_addr_t and DMA_MAPPING_ERROR git bisect bad eba304c6861613a649ba46cfab835b1eddeacd8e # bad: [b9bb694b9f62f4b31652223ed3ca38cf98bbb370] iommu/io-pgtable-arm: Clean up faulty sanity check git bisect bad b9bb694b9f62f4b31652223ed3ca38cf98bbb370 # bad: [a97740f81874c8063c12c24f34d25f10c4f5e9aa] dma-debug: convert comma to semicolon git bisect bad a97740f81874c8063c12c24f34d25f10c4f5e9aa # bad: [e0d072782c734d27f5af062c62266f2598f68542] dma-mapping: introduce DMA range map, supplanting dma_pfn_offset git bisect bad e0d072782c734d27f5af062c62266f2598f68542 # first bad commit: [e0d072782c734d27f5af062c62266f2598f68542] dma-mapping: introduce DMA range map, supplanting dma_pfn_offset Guenter
Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On 4/8/21 4:54 PM, Guenter Roeck wrote: > On 4/8/21 2:28 PM, Rob Herring wrote: >> >> Applying now so this gets into linux-next this week. >> > The patch doesn't apply on top of today's -next; it conflicts > with "of: properly check for error returned by fdt_get_name()". > > I reverted that patch and applied this one, and the DT unittests > run with it on openrisc. I do get a single test failure, but I that > is a different problem (possibly with the test case itself). > > ### dt-test ### FAIL of_unittest_dma_ranges_one():923 of_dma_get_range: wrong > DMA addr 0x > (expecting 1) on node > /testcase-data/address-tests/bus@8000/device@1000 That is a known regression on the target that I use for testing (and has been since 5.10-rc1) - the 8074 dragonboard, arm 32. No one else has reported it on the list, so even though I want to debug and fix it "promptly", other tasks have had higher priority. In my notes I list two suspect commits: e0d072782c73 dma-mapping: introduce DMA range map, supplanting dma_pfn_offset 0a0f0d8be76d dma-mapping: split I think that was purely based on looking at the list of commits that may have touched OF dma. I have not done a bisect. One specific report of not seeing the FAIL was Vireshk on 5.11-rc6 with a Hikey board. > > Tested-by: Guenter Roeck Thanks for testing! > > Guenter >
Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On 4/8/21 4:28 PM, Rob Herring wrote:
> On Thu, Apr 8, 2021 at 3:45 PM wrote:
>>
>> From: Frank Rowand
>>
>> The Devicetree standard specifies an 8 byte alignment of the FDT.
>> Code in libfdt expects this alignment for an FDT image in memory.
>> kmemdup() returns 4 byte alignment on openrisc. Replace kmemdup()
>> with kmalloc(), align pointer, memcpy() to get proper alignment.
>>
>> The 4 byte alignment exposed a related bug which triggered a crash
>> on openrisc with:
>> commit 79edff12060f ("scripts/dtc: Update to upstream version
>> v1.6.0-51-g183df9e9c2b9")
>> as reported in:
>> https://lore.kernel.org/lkml/20210327224116.69309-1-li...@roeck-us.net/
>>
>> Reported-by: Guenter Roeck
>> Signed-off-by: Frank Rowand
>>
>> ---
>>
>> changes since version 1:
>> - use pointer from kmalloc() for kfree() instead of using pointer that
>> has been modified for FDT alignment
>>
>> changes since version 2:
>> - version 1 was a work in progress version, I failed to commit the
>> following
>> final changes
>> - reorder first two arguments of of_overlay_apply()
>>
>> changes since version 3:
>> - size of memory allocation and size of copy after pointer alignment
>> differ, use separate variables with correct values for each case
>> - edit comment to more clearly describe that ovcs->fdt is the allocated
>> memory region, which may be different than where the aligned pointer
>> points
>> - remove unused parameter from of_overlay_apply()
>>
>> drivers/of/of_private.h | 2 ++
>> drivers/of/overlay.c| 27 +--
>> drivers/of/unittest.c | 13 ++---
>> 3 files changed, 29 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/of/of_private.h b/drivers/of/of_private.h
>> index d9e6a324de0a..d717efbd637d 100644
>> --- a/drivers/of/of_private.h
>> +++ b/drivers/of/of_private.h
>> @@ -8,6 +8,8 @@
>> * Copyright (C) 1996-2005 Paul Mackerras.
>> */
>>
>> +#define FDT_ALIGN_SIZE 8
>> +
>> /**
>> * struct alias_prop - Alias property in 'aliases' node
>> * @link: List node to link the structure in aliases_lookup list
>> diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
>> index 50bbe0edf538..ecf967c57900 100644
>> --- a/drivers/of/overlay.c
>> +++ b/drivers/of/overlay.c
>> @@ -57,7 +57,7 @@ struct fragment {
>> * struct overlay_changeset
>> * @id:changeset identifier
>> * @ovcs_list: list on which we are located
>> - * @fdt: FDT that was unflattened to create @overlay_tree
>> + * @fdt: base of memory allocated to hold aligned FDT that
>> was unflattened to create @overlay_tree
>> * @overlay_tree: expanded device tree that contains the fragment nodes
>> * @count: count of fragment structures
>> * @fragments: fragment nodes in the overlay expanded device tree
>> @@ -719,8 +719,8 @@ static struct device_node *find_target(struct
>> device_node *info_node)
>> /**
>> * init_overlay_changeset() - initialize overlay changeset from overlay tree
>> * @ovcs: Overlay changeset to build
>> - * @fdt: the FDT that was unflattened to create @tree
>> - * @tree: Contains all the overlay fragments and overlay fixup nodes
>> + * @fdt: base of memory allocated to hold aligned FDT that was
>> unflattened to create @tree
>> + * @tree: Contains the overlay fragments and overlay fixup nodes
>> *
>> * Initialize @ovcs. Populate @ovcs->fragments with node information from
>> * the top level of @tree. The relevant top level nodes are the fragment
>> @@ -873,7 +873,7 @@ static void free_overlay_changeset(struct
>> overlay_changeset *ovcs)
>> * internal documentation
>> *
>> * of_overlay_apply() - Create and apply an overlay changeset
>> - * @fdt: the FDT that was unflattened to create @tree
>> + * @fdt: base of memory allocated to hold the aligned FDT
>> * @tree: Expanded overlay device tree
>> * @ovcs_id: Pointer to overlay changeset id
>> *
>> @@ -913,7 +913,7 @@ static void free_overlay_changeset(struct
>> overlay_changeset *ovcs)
>> */
>>
>> static int of_overlay_apply(const void *fdt, struct device_node *tree,
>> - int *ovcs_id)
>> + int *ovcs_id)
>> {
>> struct overlay_changeset *ovcs;
>> int ret = 0, ret_revert, ret_tmp;
>> @@ -953,7 +953,9 @@ static int of_overlay_apply(const void *fdt, struct
>> device_node *tree,
>> /*
>> * after overlay_notify(), ovcs->overlay_tree related pointers may
>> have
>> * leaked to drivers, so can not kfree() tree, aka
>> ovcs->overlay_tree;
>> -* and can not free fdt, aka ovcs->fdt
>> +* and can not free memory containing aligned fdt. The aligned fdt
>> +* is contained within the memory at ovcs->fdt, possibly at an offset
>> +* from ovcs->fdt.
>> */
>> ret = overlay_notify(ovcs, OF_OVERLAY_PRE_APPLY);
>> if
Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On 4/8/21 2:28 PM, Rob Herring wrote: > > Applying now so this gets into linux-next this week. > The patch doesn't apply on top of today's -next; it conflicts with "of: properly check for error returned by fdt_get_name()". I reverted that patch and applied this one, and the DT unittests run with it on openrisc. I do get a single test failure, but I that is a different problem (possibly with the test case itself). ### dt-test ### FAIL of_unittest_dma_ranges_one():923 of_dma_get_range: wrong DMA addr 0x (expecting 1) on node /testcase-data/address-tests/bus@8000/device@1000 Tested-by: Guenter Roeck Guenter
Re: [PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
On Thu, Apr 8, 2021 at 3:45 PM wrote:
>
> From: Frank Rowand
>
> The Devicetree standard specifies an 8 byte alignment of the FDT.
> Code in libfdt expects this alignment for an FDT image in memory.
> kmemdup() returns 4 byte alignment on openrisc. Replace kmemdup()
> with kmalloc(), align pointer, memcpy() to get proper alignment.
>
> The 4 byte alignment exposed a related bug which triggered a crash
> on openrisc with:
> commit 79edff12060f ("scripts/dtc: Update to upstream version
> v1.6.0-51-g183df9e9c2b9")
> as reported in:
> https://lore.kernel.org/lkml/20210327224116.69309-1-li...@roeck-us.net/
>
> Reported-by: Guenter Roeck
> Signed-off-by: Frank Rowand
>
> ---
>
> changes since version 1:
> - use pointer from kmalloc() for kfree() instead of using pointer that
> has been modified for FDT alignment
>
> changes since version 2:
> - version 1 was a work in progress version, I failed to commit the following
> final changes
> - reorder first two arguments of of_overlay_apply()
>
> changes since version 3:
> - size of memory allocation and size of copy after pointer alignment
> differ, use separate variables with correct values for each case
> - edit comment to more clearly describe that ovcs->fdt is the allocated
> memory region, which may be different than where the aligned pointer
> points
> - remove unused parameter from of_overlay_apply()
>
> drivers/of/of_private.h | 2 ++
> drivers/of/overlay.c| 27 +--
> drivers/of/unittest.c | 13 ++---
> 3 files changed, 29 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/of/of_private.h b/drivers/of/of_private.h
> index d9e6a324de0a..d717efbd637d 100644
> --- a/drivers/of/of_private.h
> +++ b/drivers/of/of_private.h
> @@ -8,6 +8,8 @@
> * Copyright (C) 1996-2005 Paul Mackerras.
> */
>
> +#define FDT_ALIGN_SIZE 8
> +
> /**
> * struct alias_prop - Alias property in 'aliases' node
> * @link: List node to link the structure in aliases_lookup list
> diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
> index 50bbe0edf538..ecf967c57900 100644
> --- a/drivers/of/overlay.c
> +++ b/drivers/of/overlay.c
> @@ -57,7 +57,7 @@ struct fragment {
> * struct overlay_changeset
> * @id:changeset identifier
> * @ovcs_list: list on which we are located
> - * @fdt: FDT that was unflattened to create @overlay_tree
> + * @fdt: base of memory allocated to hold aligned FDT that was
> unflattened to create @overlay_tree
> * @overlay_tree: expanded device tree that contains the fragment nodes
> * @count: count of fragment structures
> * @fragments: fragment nodes in the overlay expanded device tree
> @@ -719,8 +719,8 @@ static struct device_node *find_target(struct device_node
> *info_node)
> /**
> * init_overlay_changeset() - initialize overlay changeset from overlay tree
> * @ovcs: Overlay changeset to build
> - * @fdt: the FDT that was unflattened to create @tree
> - * @tree: Contains all the overlay fragments and overlay fixup nodes
> + * @fdt: base of memory allocated to hold aligned FDT that was
> unflattened to create @tree
> + * @tree: Contains the overlay fragments and overlay fixup nodes
> *
> * Initialize @ovcs. Populate @ovcs->fragments with node information from
> * the top level of @tree. The relevant top level nodes are the fragment
> @@ -873,7 +873,7 @@ static void free_overlay_changeset(struct
> overlay_changeset *ovcs)
> * internal documentation
> *
> * of_overlay_apply() - Create and apply an overlay changeset
> - * @fdt: the FDT that was unflattened to create @tree
> + * @fdt: base of memory allocated to hold the aligned FDT
> * @tree: Expanded overlay device tree
> * @ovcs_id: Pointer to overlay changeset id
> *
> @@ -913,7 +913,7 @@ static void free_overlay_changeset(struct
> overlay_changeset *ovcs)
> */
>
> static int of_overlay_apply(const void *fdt, struct device_node *tree,
> - int *ovcs_id)
> + int *ovcs_id)
> {
> struct overlay_changeset *ovcs;
> int ret = 0, ret_revert, ret_tmp;
> @@ -953,7 +953,9 @@ static int of_overlay_apply(const void *fdt, struct
> device_node *tree,
> /*
> * after overlay_notify(), ovcs->overlay_tree related pointers may
> have
> * leaked to drivers, so can not kfree() tree, aka ovcs->overlay_tree;
> -* and can not free fdt, aka ovcs->fdt
> +* and can not free memory containing aligned fdt. The aligned fdt
> +* is contained within the memory at ovcs->fdt, possibly at an offset
> +* from ovcs->fdt.
> */
> ret = overlay_notify(ovcs, OF_OVERLAY_PRE_APPLY);
> if (ret) {
> @@ -1014,9 +1016,10 @@ static int of_overlay_apply(const void *fdt, struct
> device_node *tree,
> int of_overlay_fdt_apply(const void *overlay_fdt, u32
[PATCH v4 1/1] of: unittest: overlay: ensure proper alignment of copied FDT
From: Frank Rowand
The Devicetree standard specifies an 8 byte alignment of the FDT.
Code in libfdt expects this alignment for an FDT image in memory.
kmemdup() returns 4 byte alignment on openrisc. Replace kmemdup()
with kmalloc(), align pointer, memcpy() to get proper alignment.
The 4 byte alignment exposed a related bug which triggered a crash
on openrisc with:
commit 79edff12060f ("scripts/dtc: Update to upstream version
v1.6.0-51-g183df9e9c2b9")
as reported in:
https://lore.kernel.org/lkml/20210327224116.69309-1-li...@roeck-us.net/
Reported-by: Guenter Roeck
Signed-off-by: Frank Rowand
---
changes since version 1:
- use pointer from kmalloc() for kfree() instead of using pointer that
has been modified for FDT alignment
changes since version 2:
- version 1 was a work in progress version, I failed to commit the following
final changes
- reorder first two arguments of of_overlay_apply()
changes since version 3:
- size of memory allocation and size of copy after pointer alignment
differ, use separate variables with correct values for each case
- edit comment to more clearly describe that ovcs->fdt is the allocated
memory region, which may be different than where the aligned pointer points
- remove unused parameter from of_overlay_apply()
drivers/of/of_private.h | 2 ++
drivers/of/overlay.c| 27 +--
drivers/of/unittest.c | 13 ++---
3 files changed, 29 insertions(+), 13 deletions(-)
diff --git a/drivers/of/of_private.h b/drivers/of/of_private.h
index d9e6a324de0a..d717efbd637d 100644
--- a/drivers/of/of_private.h
+++ b/drivers/of/of_private.h
@@ -8,6 +8,8 @@
* Copyright (C) 1996-2005 Paul Mackerras.
*/
+#define FDT_ALIGN_SIZE 8
+
/**
* struct alias_prop - Alias property in 'aliases' node
* @link: List node to link the structure in aliases_lookup list
diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
index 50bbe0edf538..ecf967c57900 100644
--- a/drivers/of/overlay.c
+++ b/drivers/of/overlay.c
@@ -57,7 +57,7 @@ struct fragment {
* struct overlay_changeset
* @id:changeset identifier
* @ovcs_list: list on which we are located
- * @fdt: FDT that was unflattened to create @overlay_tree
+ * @fdt: base of memory allocated to hold aligned FDT that was
unflattened to create @overlay_tree
* @overlay_tree: expanded device tree that contains the fragment nodes
* @count: count of fragment structures
* @fragments: fragment nodes in the overlay expanded device tree
@@ -719,8 +719,8 @@ static struct device_node *find_target(struct device_node
*info_node)
/**
* init_overlay_changeset() - initialize overlay changeset from overlay tree
* @ovcs: Overlay changeset to build
- * @fdt: the FDT that was unflattened to create @tree
- * @tree: Contains all the overlay fragments and overlay fixup nodes
+ * @fdt: base of memory allocated to hold aligned FDT that was
unflattened to create @tree
+ * @tree: Contains the overlay fragments and overlay fixup nodes
*
* Initialize @ovcs. Populate @ovcs->fragments with node information from
* the top level of @tree. The relevant top level nodes are the fragment
@@ -873,7 +873,7 @@ static void free_overlay_changeset(struct overlay_changeset
*ovcs)
* internal documentation
*
* of_overlay_apply() - Create and apply an overlay changeset
- * @fdt: the FDT that was unflattened to create @tree
+ * @fdt: base of memory allocated to hold the aligned FDT
* @tree: Expanded overlay device tree
* @ovcs_id: Pointer to overlay changeset id
*
@@ -913,7 +913,7 @@ static void free_overlay_changeset(struct overlay_changeset
*ovcs)
*/
static int of_overlay_apply(const void *fdt, struct device_node *tree,
- int *ovcs_id)
+ int *ovcs_id)
{
struct overlay_changeset *ovcs;
int ret = 0, ret_revert, ret_tmp;
@@ -953,7 +953,9 @@ static int of_overlay_apply(const void *fdt, struct
device_node *tree,
/*
* after overlay_notify(), ovcs->overlay_tree related pointers may have
* leaked to drivers, so can not kfree() tree, aka ovcs->overlay_tree;
-* and can not free fdt, aka ovcs->fdt
+* and can not free memory containing aligned fdt. The aligned fdt
+* is contained within the memory at ovcs->fdt, possibly at an offset
+* from ovcs->fdt.
*/
ret = overlay_notify(ovcs, OF_OVERLAY_PRE_APPLY);
if (ret) {
@@ -1014,9 +1016,10 @@ static int of_overlay_apply(const void *fdt, struct
device_node *tree,
int of_overlay_fdt_apply(const void *overlay_fdt, u32 overlay_fdt_size,
int *ovcs_id)
{
- const void *new_fdt;
+ void *new_fdt;
+ void *new_fdt_align;
int ret;
- u32 size;
+ u32 size, size_alloc;
struct device_node *overlay_root;
*ovcs_id
