Re: [PATCH 3/3] dax/kmem: Always enroll hotplugged memory for memmap_on_memory

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 13.07.23 17:15, Verma, Vishal L wrote:
On Thu, 2023-07-13 at 09:23 +0200, David Hildenbrand wrote:
On 13.07.23 08:45, Verma, Vishal L wrote:

I'm taking a shot at implementing the splitting internally in
memory_hotplug.c. The caller (kmem) side does become trivial with this
approach, but there's a slight complication if I don't have the module
param override (patch 1 of this series).

The kmem diff now looks like:

     diff --git a/drivers/dax/kmem.c b/drivers/dax/kmem.c
     index 898ca9505754..8be932f63f90 100644
     --- a/drivers/dax/kmem.c
     +++ b/drivers/dax/kmem.c
     @@ -105,6 +105,8 @@ static int dev_dax_kmem_probe(struct dev_dax *dev_dax)
             data->mgid = rc;
            for (i = 0; i < dev_dax->nr_range; i++) {
     +               mhp_t mhp_flags = MHP_NID_IS_MGID | MHP_MEMMAP_ON_MEMORY |
     +                                 MHP_SPLIT_MEMBLOCKS;
                     struct resource *res;
                     struct range range;
    @@ -141,7 +143,7 @@ static int dev_dax_kmem_probe(struct dev_dax *dev_dax)
                      * this as RAM automatically.
                      */
                     rc = add_memory_driver_managed(data->mgid, range.start,
     -                               range_len(&range), kmem_name, MHP_NID_IS_MGID);
     +                               range_len(&range), kmem_name, mhp_flags);
                    if (rc) {
                             dev_warn(dev, "mapping%d: %#llx-%#llx memory add failed\n",

Why do we need the MHP_SPLIT_MEMBLOCKS?

I thought we still wanted either an opt-in or opt-out for the kmem
driver to be able to do memmap_on_memory, in case there were
performance implications or the lack of 1GiB PUDs. I haven't
implemented that yet, but I was thinking along the lines of a sysfs
knob exposed by kmem, that controls setting of this new
MHP_SPLIT_MEMBLOCKS flag.

Why is MHP_MEMMAP_ON_MEMORY not sufficient for that?



In add_memory_driver_managed(), if memmap_on_memory = 1 AND is effective for a
single memory block, you can simply split up internally, no?

Essentially in add_memory_resource() something like

if (mhp_flags & MHP_MEMMAP_ON_MEMORY &&
      mhp_supports_memmap_on_memory(memory_block_size_bytes())) {
         for (cur_start = start, cur_start < start + size;
              cur_start += memory_block_size_bytes()) {
                 mhp_altmap.free = PHYS_PFN(memory_block_size_bytes());
                 mhp_altmap.base_pfn = PHYS_PFN(start);
                 params.altmap = &mhp_altmap;

                 ret = arch_add_memory(nid, start,
                                       memory_block_size_bytes(), &params);
                 if (ret < 0) ...

                 ret = create_memory_block_devices(start, memory_block_size_bytes(),
                                                   mhp_altmap.alloc, group);
                 if (ret) ...
        }
} else {
         /* old boring stuff */
}

Of course, doing it a bit cleaner, factoring out adding of mem+creating devices into
a helper so we can use it on the other path, avoiding repeating memory_block_size_bytes()
...

My current approach was looping a level higher, on the call to
add_memory_resource, but this looks reasonable too, and I can switch to
this. In fact it is better as in my case I had to loop twice, once for
the regular add_memory() path and once for the _driver_managed() path.
Yours should avoid that.

As we only care about the altmap here, handling it for arch_add_memory() + create_memory_block_devices() should be sufficient.

--
Cheers,

David / dhildenb





[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux