Re: [PATCH v6 5/6] mm/vmalloc: map contiguous pages in batches for vmap() if possible

[Dev Jain]

On 14/07/26 10:46 am, Wen Jiang wrote:
> On Mon, 13 Jul 2026 at 23:19, Dev Jain <dev.jain@xxxxxxx> wrote:
>>
>>
>>
>> On 09/07/26 1:08 pm, Wen Jiang wrote:
>>> From: "Barry Song (Xiaomi)" <baohua@xxxxxxxxxx>
>>>
>>> In many cases, the pages passed to vmap() may include high-order
>>> pages. For example, the systemheap often allocates pages in descending
>>> order: order 8, then 4, then 0. Currently, vmap() iterates over every
>>> page individually—even pages inside a high-order block are handled
>>> one by one.
>>>
>>> This patch detects physically contiguous pages (regardless of whether
>>> they are compound or non-compound) by scanning with
>>> num_pages_contiguous(), and maps them as a single contiguous block
>>> whenever possible. The mapping order is determined by taking the
>>> minimum of the contiguous page count and the pfn alignment, allowing
>>> graceful degradation when pfn alignment is less than the contiguous
>>> range.
>>>
>>> Pages with the same page_shift are coalesced and mapped via
>>> vmap_pages_range_noflush_walk() to avoid page table rewalk.
>>>
>>> As users typically allocate memory in descending orders (e.g.
>>> 8 → 4 → 0), once an order-0 page is encountered, we stop scanning
>>> for contiguous pages since subsequent pages are likely order-0 as well.
>>>
>>> Signed-off-by: Barry Song (Xiaomi) <baohua@xxxxxxxxxx>
>>> Co-developed-by: Dev Jain <dev.jain@xxxxxxx>
>>> Signed-off-by: Dev Jain <dev.jain@xxxxxxx>
>>> Signed-off-by: Wen Jiang <jiangwen6@xxxxxxxxxx>
>>> Tested-by: Xueyuan Chen <xueyuan.chen21@xxxxxxxxx>
>>> Tested-by: Leo Yan <leo.yan@xxxxxxx>
>>> ---
>>>  mm/vmalloc.c | 87 ++++++++++++++++++++++++++++++++++++++++++++++++++--
>>>  1 file changed, 85 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
>>> index d2a4d649af549..db0492151ad08 100644
>>> --- a/mm/vmalloc.c
>>> +++ b/mm/vmalloc.c
>>> @@ -3543,6 +3543,89 @@ void vunmap(const void *addr)
>>>  }
>>>  EXPORT_SYMBOL(vunmap);
>>>
>>> +static inline unsigned int vm_shift(pgprot_t prot, unsigned long size)
>>> +{
>>> +     if (arch_vmap_pmd_supported(prot) && size >= PMD_SIZE)
>>> +             return PMD_SHIFT;
>>> +
>>> +     return arch_vmap_pte_supported_shift(size);
>>> +}
>>
>> Need to throw in a preparatory patch for this, which will (in addition to
>> introduction of the vm_shift() function) do:
>>
> 
> Hi Dev,
> 
> Your suggestion is to extract "vm_shift" into a separate patch, say
> Patch 5, to make it a common utility function just like Patch 3. And
> move the existing code that can use vm_shift into this new patch as
> well, correct?


Yes it will look like the current patch 3. So before the current patch 5,
you can have a patch introducing vm_shift and making the code chunk
in __vmalloc_node_range_noprof() call it.



> 
> Thanks,
> Wen
> 
>> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
>> index afaa14ebf17bb..dac87e1cd484b 100644
>> --- a/mm/vmalloc.c
>> +++ b/mm/vmalloc.c
>> @@ -4175,10 +4175,7 @@ void *__vmalloc_node_range_noprof(unsigned long size, unsigned long align,
>>                  * supporting them.
>>                  */
>>
>> -               if (arch_vmap_pmd_supported(prot) && size >= PMD_SIZE)
>> -                       shift = PMD_SHIFT;
>> -               else
>> -                       shift = arch_vmap_pte_supported_shift(size);
>> +               shift = vm_shift(prot, size);
>>
>>                 align = max(original_align, 1UL << shift);
>>         }
>>
>>
>>
>>> +
>>> +static inline int get_vmap_batch_order(struct page **pages,
>>> +             pgprot_t prot, unsigned int max_steps, unsigned int idx)
>>> +{
>>> +     unsigned int nr_contig;
>>> +     int order;
>>> +
>>> +     if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP))
>>> +             return 0;
>>> +
>>> +     nr_contig = num_pages_contiguous(&pages[idx], max_steps);
>>> +     if (nr_contig < 2)
>>> +             return 0;
>>> +
>>> +     order = ilog2(nr_contig);
>>> +
>>> +     /* Limit order by pfn alignment */
>>> +     order = min_t(int, order, __ffs(page_to_pfn(pages[idx])));
>>
>> You missed Sashiko's point here :) the pfn may be zero and this
>> will blow up. You can just first derive the pfn and do the clamping
>> only when pfn > 0.
>>
>>
> 
> Will fix this.
> 
> Thanks,
> Wen
>>
>>> +
>>> +     if (vm_shift(prot, PAGE_SIZE << order) == PAGE_SHIFT)
>>> +             return 0;
>>> +
>>> +     return order;
>>> +}
>>> +