On 22.02.21 14:51, Oscar Salvador wrote:
v2 -> v3:
- Drop usage of high-level generic helpers in favour of
low-level approach (per Michal)
- Check for the page to be marked as PageHugeFreed
- Add a one-time retry in case someone grabbed the free huge page
from under us
v1 -> v2:
- Adressed feedback by Michal
- Restrict the allocation to a node with __GFP_THISNODE
- Drop PageHuge check in alloc_and_dissolve_huge_page
- Re-order comments in isolate_or_dissolve_huge_page
- Extend comment in isolate_migratepages_block
- Place put_page right after we got the page, otherwise
dissolve_free_huge_page will fail
RFC -> v1:
- Drop RFC
- Addressed feedback from David and Mike
- Fence off gigantic pages as there is a cyclic dependency between
them and alloc_contig_range
- Re-organize the code to make race-window smaller and to put
all details in hugetlb code
- Drop nodemask initialization. First a node will be tried and then we
will back to other nodes containing memory (N_MEMORY). Details in
patch#1's changelog
- Count new page as surplus in case we failed to dissolve the old page
and the new one. Details in patch#1.
Cover letter:
alloc_contig_range lacks the hability for handling HugeTLB pages.
This can be problematic for some users, e.g: CMA and virtio-mem, where those
users will fail the call if alloc_contig_range ever sees a HugeTLB page, even
when those pages lay in ZONE_MOVABLE and are free.
That problem can be easily solved by replacing the page in the free hugepage
pool.
In-use HugeTLB are no exception though, as those can be isolated and migrated
as any other LRU or Movable page.
This patchset aims for improving alloc_contig_range->isolate_migratepages_block,
so HugeTLB pages can be recognized and handled.
Below is an insight from David (thanks), where the problem can clearly be seen:
"Start a VM with 4G. Hotplug 1G via virtio-mem and online it to
ZONE_MOVABLE. Allocate 512 huge pages.
[root@localhost ~]# cat /proc/meminfo
MemTotal: 5061512 kB
MemFree: 3319396 kB
MemAvailable: 3457144 kB
...
HugePages_Total: 512
HugePages_Free: 512
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 2048 kB
The huge pages get partially allocate from ZONE_MOVABLE. Try unplugging
1G via virtio-mem (remember, all ZONE_MOVABLE). Inside the guest:
[ 180.058992] alloc_contig_range: [1b8000, 1c0000) PFNs busy
[ 180.060531] alloc_contig_range: [1b8000, 1c0000) PFNs busy
[ 180.061972] alloc_contig_range: [1b8000, 1c0000) PFNs busy
[ 180.063413] alloc_contig_range: [1b8000, 1c0000) PFNs busy
[ 180.064838] alloc_contig_range: [1b8000, 1c0000) PFNs busy
[ 180.065848] alloc_contig_range: [1bfc00, 1c0000) PFNs busy
[ 180.066794] alloc_contig_range: [1bfc00, 1c0000) PFNs busy
[ 180.067738] alloc_contig_range: [1bfc00, 1c0000) PFNs busy
[ 180.068669] alloc_contig_range: [1bfc00, 1c0000) PFNs busy
[ 180.069598] alloc_contig_range: [1bfc00, 1c0000) PFNs busy"
Same experiment with ZONE_MOVABLE:
a) Free huge pages: all memory can get unplugged again.
b) Allocated/populated but idle huge pages: all memory can get unplugged
again.
c) Allocated/populated but all 512 huge pages are read/written in a
loop: all memory can get unplugged again, but I get a single
[ 121.192345] alloc_contig_range: [180000, 188000) PFNs busy
Most probably because it happened to try migrating a huge page while it
was busy. As virtio-mem retries on ZONE_MOVABLE a couple of times, it
can deal with this temporary failure.
Last but not least, I did something extreme:
]# cat /proc/meminfo
MemTotal: 5061568 kB
MemFree: 186560 kB
MemAvailable: 354524 kB
...
HugePages_Total: 2048
HugePages_Free: 2048
HugePages_Rsvd: 0
HugePages_Surp: 0
Triggering unplug would require to dissolve+alloc - which now fails when
trying to allocate an additional ~512 huge pages (1G).
As expected, I can properly see memory unplug not fully succeeding. + I
get a fairly continuous stream of
[ 226.611584] alloc_contig_range: [19f400, 19f800) PFNs busy
...
But more importantly, the hugepage count remains stable, as configured
by the admin (me):
HugePages_Total: 2048
HugePages_Free: 2048
HugePages_Rsvd: 0
HugePages_Surp: 0
--
Thanks,
David / dhildenb
