Re: [RFC PATCH rdma-next 01/10] RDMA: mr: Introduce is_pmem

["Li, Zhijian" <lizhijian@cn.fujitsu.com>]

Copied to nvdimm list

Thanks

Zhijian


on 2022/1/6 14:12, Li Zhijian wrote:
>
> Add Dan to the party :)
>
> May i know whether there is any existing APIs to check whether
> a va/page backs to a nvdimm/pmem ?
>
>
>
> On 06/01/2022 08:21, Jason Gunthorpe wrote:
>> On Tue, Dec 28, 2021 at 04:07:08PM +0800, Li Zhijian wrote:
>>> We can use it to indicate whether the registering mr is associated with
>>> a pmem/nvdimm or not.
>>>
>>> Currently, we only assign it in rxe driver, for other device/drivers,
>>> they should implement it if needed.
>>>
>>> RDMA FLUSH will support the persistence feature for a pmem/nvdimm.
>>>
>>> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
>>>   drivers/infiniband/sw/rxe/rxe_mr.c | 47 
>>> ++++++++++++++++++++++++++++++
>>>   include/rdma/ib_verbs.h            |  1 +
>>>   2 files changed, 48 insertions(+)
>>>
>>> diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c 
>>> b/drivers/infiniband/sw/rxe/rxe_mr.c
>>> index 7c4cd19a9db2..bcd5e7afa475 100644
>>> +++ b/drivers/infiniband/sw/rxe/rxe_mr.c
>>> @@ -162,6 +162,50 @@ void rxe_mr_init_dma(struct rxe_pd *pd, int 
>>> access, struct rxe_mr *mr)
>>>       mr->type = IB_MR_TYPE_DMA;
>>>   }
>>>   +// XXX: the logic is similar with mm/memory-failure.c
>>> +static bool page_in_dev_pagemap(struct page *page)
>>> +{
>>> +    unsigned long pfn;
>>> +    struct page *p;
>>> +    struct dev_pagemap *pgmap = NULL;
>>> +
>>> +    pfn = page_to_pfn(page);
>>> +    if (!pfn) {
>>> +        pr_err("no such pfn for page %p\n", page);
>>> +        return false;
>>> +    }
>>> +
>>> +    p = pfn_to_online_page(pfn);
>>> +    if (!p) {
>>> +        if (pfn_valid(pfn)) {
>>> +            pgmap = get_dev_pagemap(pfn, NULL);
>>> +            if (pgmap)
>>> +                put_dev_pagemap(pgmap);
>>> +        }
>>> +    }
>>> +
>>> +    return !!pgmap;
>> You need to get Dan to check this out, but I'm pretty sure this should
>> be more like this:
>>
>> if (is_zone_device_page(page) && page->pgmap->type == 
>> MEMORY_DEVICE_FS_DAX)
>
> Great, i have added him.
>
>
>
>>
>>
>>> +static bool iova_in_pmem(struct rxe_mr *mr, u64 iova, int length)
>>> +{
>>> +    struct page *page = NULL;
>>> +    char *vaddr = iova_to_vaddr(mr, iova, length);
>>> +
>>> +    if (!vaddr) {
>>> +        pr_err("not a valid iova %llu\n", iova);
>>> +        return false;
>>> +    }
>>> +
>>> +    page = virt_to_page(vaddr);
>> And obviously this isn't uniform for the entire umem, so I don't even
>> know what this is supposed to mean.
>
> My intention is to check if a memory region belongs to a nvdimm/pmem.
> The approach is like that:
> iova(user space)-+                     +-> page -> page_in_dev_pagemap()
>                  |                     |
>                  +-> va(kernel space) -+
> Since current MR's va is associated with map_set where it record the 
> relations
> between iova and va and page. Do do you mean we should travel map_set to
> get its page ? or by any other ways.
>
>
>
>
>
>
>
>>> diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
>>> index 6e9ad656ecb7..822ebb3425dc 100644
>>> +++ b/include/rdma/ib_verbs.h
>>> @@ -1807,6 +1807,7 @@ struct ib_mr {
>>>       unsigned int       page_size;
>>>       enum ib_mr_type       type;
>>>       bool           need_inval;
>>> +    bool           is_pmem;
>> Or why it is being stored in the global struct?
>
> Indeed, it's not strong necessary. but i think is_pmem should belongs 
> to a ib_mr
> so that it can be checked by other general code when they needed even 
> though no
> one does such checking so far.
>
>
> Thanks
> Zhijian
>
>
>
>>
>> Jason
>