Re: [PATCH 0/2] dax: Fix Device DAX range allocation validation

[Muchun Song <muchun.song@linux.dev>]

> On Sep 22, 2026, at 02:58, Dave Jiang <dave.jiang@intel.com> wrote:
> 
> 
> 
> On 9/18/26 10:03 PM, Muchun Song wrote:
>> 
>> 
>>> On Sep 19, 2026, at 01:02, Dave Jiang <dave.jiang@intel.com> wrote:
>>> 
>>> 
>>> 
>>> On 9/15/26 2:56 AM, Muchun Song wrote:
>>>> While working on the HugeTLB Vmemmap Optimization (HVO) generalization,
>>>> Sashiko reported that memmap_init_zone_device() could be asked to
>>>> initialize struct pages with a large-page order for a physical range
>>>> that was not naturally aligned to that order.
>>>> 
>>>> Following the Device DAX call path confirmed that the condition was
>>>> reachable. Device DAX range validation checks the range size, but not
>>>> its start address. An unaligned range can therefore reach
>>>> memmap_init_zone_device() after the device binds, and a subsequent write
>>>> to a userspace mapping may trigger a kernel panic. With the help of an
>>>> LLM, I was able to reproduce the failure.
>>>> 
>>>> The automatic resize path has a related problem. It can split an aligned
>>>> size request across arbitrary free gaps, consume an unaligned fragment,
>>>> and leave the resize partially applied when a later allocation fails.
>>>> 
>>>> This series validates both the start and size of every Device DAX range
>>>> and makes automatic resize account only for usable aligned space. It
>>>> also fixes the mapping sysfs attribute to propagate validation failures
>>>> instead of reporting success without allocating the requested range.
>>>> 
>>>> Muchun Song (2):
>>>> dax/bus: fix Device DAX range alignment validation
>>>> dax/bus: fix mapping attribute error reporting
>>>> 
>>>> drivers/dax/bus.c | 144 +++++++++++++++++++++++++++++++++++-----------
>>>> 1 file changed, 109 insertions(+), 35 deletions(-)
>>>> 
>>>> 
>>>> base-commit: 1a1de54f7369cd2b5bac0f265910e60ad3a6b4c3
>>> 
>>> This commit does not exist in Linus's tree. Can you please base the series on top of the latest 7.3-rc tag so sashiko can look over it? Thanks!
>> 
>> Hi,
>> 
>> The series is based on linux-next 20260914, and drivers/dax/bus.c is identical
>> between that branch and the latest 7.3-rc. Sashiko has also already reviewed it
>> here: [1].
>> 
>> Therefore, I didn't quite catch your point. Are you expecting Sashiko to conduct
>> another round of review based on the 7.3-rc code? (Are you concerned about issues
>> introduced by the additional commits in linux-next compared to 7.3-rc?)
> 
> Ok that's fine. But next time please use a tag that's valid from Linus tree and not linux-next. While there's a sashiko link, I'd like to see the sashiko response on the mailing list for easier review. I'm not sure if it's due to the base commit or something else for it to not show up. Sometimes sashiko responds with no regression found and sometimes it just does not respond. I'm not sure why the varied behavior.

I previously worked based on the mm development model, and this is my
first time learning about the development rules for DAX. I will develop
DAX based on Linus's tree going forward.

Thanks,
Muchun

> 
> DJ
> 
> 
>> 
>> [1] https://sashiko.dev/#/patchset/20260915095621.3744167-1-songmuchun%40bytedance.com
>> 
>> Thanks,
>> Muchun