linux-kernel - [v3 0/4] mm: hugetlb: Skip initialization of gigantic tail struct pages if freed by HVO

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20230825111836.1715308-1-usama.arif@bytedance.com>
Date:   Fri, 25 Aug 2023 12:18:32 +0100
From:   Usama Arif <usama.arif@...edance.com>
To:     linux-mm@...ck.org, muchun.song@...ux.dev, mike.kravetz@...cle.com,
        rppt@...nel.org
Cc:     linux-kernel@...r.kernel.org, songmuchun@...edance.com,
        fam.zheng@...edance.com, liangma@...ngbit.com,
        punit.agrawal@...edance.com, Usama Arif <usama.arif@...edance.com>
Subject: [v3 0/4] mm: hugetlb: Skip initialization of gigantic tail struct pages if freed by HVO

This series moves the boot time initialization of tail struct pages of a
gigantic page to after HVO is attempted. If HVO is successful, only
HUGETLB_VMEMMAP_RESERVE_SIZE / sizeof(struct page) struct pages need to
be initialized. For a 1G hugepage, this series avoid initialization of
262144 - 63 = 262081 struct pages per hugepage.

When tested on a 512G system (which can allocate max 500 1G hugepages), the
kexec-boot time with HVO and DEFERRED_STRUCT_PAGE_INIT enabled without this
patchseries to running init is 3.9 seconds. With this patch it is 1.2 seconds.
This represents an approximately 70% reduction in boot time and will
significantly reduce server downtime when using a large number of
gigantic pages.

Thanks,
Usama

[v2->v3]:
- (Muchun Song) skip prep of struct pages backing gigantic hugepages
at boot time only.
- (Muchun Song) move initialization of tail struct pages to after
HVO is attempted. This also means that the hugetlb command line parsing
does not need to be changed.

[v1->v2]:
- (Mike Rapoport) Code quality improvements (function names, arguments,
comments).

[RFC->v1]:
- (Mike Rapoport) Change from passing hugepage_size in
memblock_alloc_try_nid_raw for skipping struct page initialization to
using MEMBLOCK_RSRV_NOINIT flag

Usama Arif (4):
  mm: hugetlb_vmemmap: Use nid of the head page to reallocate it
  memblock: pass memblock_type to memblock_setclr_flag
  memblock: introduce MEMBLOCK_RSRV_NOINIT_VMEMMAP flag
  mm: hugetlb: Skip initialization of gigantic tail struct pages if
    freed by HVO

 include/linux/memblock.h | 10 ++++++++
 mm/hugetlb.c             | 52 ++++++++++++++++++++++++++++++++--------
 mm/hugetlb_vmemmap.c     |  2 +-
 mm/hugetlb_vmemmap.h     |  8 +++----
 mm/internal.h            |  3 +++
 mm/memblock.c            | 47 ++++++++++++++++++++++++++----------
 mm/mm_init.c             |  2 +-
 7 files changed, 95 insertions(+), 29 deletions(-)

-- 
2.25.1