From: Usama Arif <usama.arif@linux•dev>
To: Muchun Song <songmuchun@bytedance•com>
Cc: Usama Arif <usama.arif@linux•dev>,
Oscar Salvador <osalvador@suse•de>,
David Hildenbrand <david@kernel•org>,
Andrew Morton <akpm@linux-foundation•org>,
Madhavan Srinivasan <maddy@linux•ibm.com>,
Michael Ellerman <mpe@ellerman•id.au>,
Muchun Song <muchun.song@linux•dev>,
Mike Rapoport <rppt@kernel•org>, Lorenzo Stoakes <ljs@kernel•org>,
"Liam R. Howlett" <liam@infradead•org>,
Vlastimil Babka <vbabka@kernel•org>,
linux-mm@kvack•org, linux-kernel@vger•kernel.org,
Nicholas Piggin <npiggin@gmail•com>,
"Christophe Leroy (CS GROUP)" <chleroy@kernel•org>,
"Ritesh Harjani (IBM)" <ritesh.list@gmail•com>,
"Aneesh Kumar K.V" <aneesh.kumar@linux•ibm.com>,
linuxppc-dev@lists•ozlabs.org,
Mike Kravetz <mike.kravetz@oracle•com>
Subject: Re: [PATCH v3 15/19] mm/hugetlb_vmemmap: Move bootmem HVO setup to early init
Date: Wed, 3 Jun 2026 05:02:44 -0700 [thread overview]
Message-ID: <20260603120246.1572177-1-usama.arif@linux.dev> (raw)
In-Reply-To: <20260602101039.1867613-16-songmuchun@bytedance.com>
On Tue, 2 Jun 2026 18:10:35 +0800 Muchun Song <songmuchun@bytedance•com> wrote:
> Bootmem HugeTLB pages currently defer HVO setup to
> hugetlb_vmemmap_init_late(), because the optimization needs zone
> information.
>
> Now that zone initialization is available earlier, the bootmem HVO setup
> can be done directly from hugetlb_vmemmap_init_early(). This lets
> gigantic HugeTLB pages apply HVO as soon as they are allocated.
>
> Bootmem gigantic pages that span multiple zones are now filtered out
> when they are allocated, so the remaining bootmem gigantic pages seen by
> later hugetlb initialization are already zone-valid. As a result,
> hugetlb_vmemmap_init_late() no longer needs to handle bootmem HVO setup.
>
> Signed-off-by: Muchun Song <songmuchun@bytedance•com>
> ---
> mm/hugetlb_vmemmap.c | 67 +++++++++-----------------------------------
> 1 file changed, 13 insertions(+), 54 deletions(-)
>
> diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c
> index ea6af85bfec1..464578ee246e 100644
> --- a/mm/hugetlb_vmemmap.c
> +++ b/mm/hugetlb_vmemmap.c
> @@ -745,6 +745,8 @@ static bool vmemmap_should_optimize_bootmem_page(struct huge_bootmem_page *m)
> return true;
> }
>
> +static struct zone *pfn_to_zone(unsigned nid, unsigned long pfn);
> +
> /*
> * Initialize memmap section for a gigantic page, HVO-style.
> */
> @@ -752,6 +754,7 @@ void __init hugetlb_vmemmap_init_early(int nid)
> {
> unsigned long psize, paddr, section_size;
> unsigned long ns, i, pnum, pfn, nr_pages;
> + unsigned long start, end;
> struct huge_bootmem_page *m = NULL;
> void *map;
>
> @@ -761,6 +764,8 @@ void __init hugetlb_vmemmap_init_early(int nid)
> section_size = (1UL << PA_SECTION_SHIFT);
>
> list_for_each_entry(m, &huge_boot_pages[nid], list) {
> + struct zone *zone;
> +
> if (!vmemmap_should_optimize_bootmem_page(m))
> continue;
>
> @@ -769,6 +774,14 @@ void __init hugetlb_vmemmap_init_early(int nid)
> paddr = virt_to_phys(m);
> pfn = PHYS_PFN(paddr);
> map = pfn_to_page(pfn);
> + start = (unsigned long)map;
> + end = start + hugetlb_vmemmap_size(m->hstate);
> + zone = pfn_to_zone(nid, pfn);
> +
> + if (vmemmap_populate_hvo(start, end, huge_page_order(m->hstate),
> + zone, HUGETLB_VMEMMAP_RESERVE_SIZE))
> + panic("Failed to allocate memmap for HugeTLB page\n");
The replaced hugetlb_vmemmap_init_late() path used to fall back to
vmemmap_populate() if vmemmap_populate_hvo() returned an error and
just lost the HVO optimization for that page.
The new path panics on any non-zero return. Is the panic intended,
given that vmemmap_populate_hvo() returns -ENOMEM on allocation
failure and HVO is normally treated as an optimization rather than a
hard requirement?
> + memmap_boot_pages_add(DIV_ROUND_UP(HUGETLB_VMEMMAP_RESERVE_SIZE, PAGE_SIZE));
>
> pnum = pfn_to_section_nr(pfn);
> ns = psize / section_size;
> @@ -800,60 +813,6 @@ static struct zone *pfn_to_zone(unsigned nid, unsigned long pfn)
>
> void __init hugetlb_vmemmap_init_late(int nid)
> {
> - struct huge_bootmem_page *m, *tm;
> - unsigned long phys, nr_pages, start, end;
> - unsigned long pfn, nr_mmap;
> - struct zone *zone = NULL;
> - struct hstate *h;
> - void *map;
> -
> - if (!READ_ONCE(vmemmap_optimize_enabled))
> - return;
> -
> - list_for_each_entry_safe(m, tm, &huge_boot_pages[nid], list) {
> - if (!(m->flags & HUGE_BOOTMEM_HVO))
> - continue;
> -
> - phys = virt_to_phys(m);
> - h = m->hstate;
> - pfn = PHYS_PFN(phys);
> - nr_pages = pages_per_huge_page(h);
> - map = pfn_to_page(pfn);
> - start = (unsigned long)map;
> - end = start + nr_pages * sizeof(struct page);
> -
> - if (!hugetlb_bootmem_page_zones_valid(nid, m)) {
> - /*
> - * Oops, the hugetlb page spans multiple zones.
> - * Remove it from the list, and populate it normally.
> - */
> - list_del(&m->list);
> -
> - vmemmap_populate(start, end, nid, NULL);
> - nr_mmap = end - start;
> - memmap_boot_pages_add(DIV_ROUND_UP(nr_mmap, PAGE_SIZE));
> -
> - memblock_phys_free(phys, huge_page_size(h));
> - continue;
> - }
> -
> - if (!zone || !zone_spans_pfn(zone, pfn))
> - zone = pfn_to_zone(nid, pfn);
> - if (WARN_ON_ONCE(!zone))
> - continue;
> -
> - if (vmemmap_populate_hvo(start, end, huge_page_order(h), zone,
> - HUGETLB_VMEMMAP_RESERVE_SIZE) < 0) {
> - /* Fallback if HVO population fails */
> - vmemmap_populate(start, end, nid, NULL);
> - nr_mmap = end - start;
> - } else {
> - m->flags |= HUGE_BOOTMEM_ZONES_VALID;
> - nr_mmap = HUGETLB_VMEMMAP_RESERVE_SIZE;
> - }
> -
> - memmap_boot_pages_add(DIV_ROUND_UP(nr_mmap, PAGE_SIZE));
> - }
> }
> #endif
>
> --
> 2.54.0
>
>
next prev parent reply other threads:[~2026-06-03 12:03 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-02 10:10 [PATCH v3 00/19] mm: Refactor bootmem gigantic hugepage allocation Muchun Song
2026-06-02 10:10 ` [PATCH v3 01/19] mm/hugetlb: Fix boot panic with CONFIG_DEBUG_VM and HVO bootmem pages Muchun Song
2026-06-02 10:10 ` [PATCH v3 02/19] mm/hugetlb_vmemmap: Fix __hugetlb_vmemmap_optimize_folios() Muchun Song
2026-06-02 10:10 ` [PATCH v3 03/19] powerpc/mm: Fix wrong addr_pfn tracking in compound vmemmap population Muchun Song
2026-06-03 14:36 ` Ritesh Harjani
2026-06-04 2:09 ` Muchun Song
2026-06-02 10:10 ` [PATCH v3 04/19] mm/hugetlb: Initialize gigantic bootmem hugepage struct pages earlier Muchun Song
2026-06-02 10:10 ` [PATCH v3 05/19] mm/mm_init: Simplify deferred_free_pages() migratetype init Muchun Song
2026-06-02 10:10 ` [PATCH v3 06/19] mm/sparse: Panic on memmap and usemap allocation failure Muchun Song
2026-06-02 10:10 ` [PATCH v3 07/19] mm/sparse: Move subsection_map_init() into sparse_init() Muchun Song
2026-06-02 10:10 ` [PATCH v3 08/19] mm/mm_init: Defer sparse_init() until after zone initialization Muchun Song
2026-06-02 10:10 ` [PATCH v3 09/19] mm/mm_init: Defer hugetlb reservation " Muchun Song
2026-06-02 10:10 ` [PATCH v3 10/19] mm/mm_init: Remove set_pageblock_order() call from sparse_init() Muchun Song
2026-06-02 10:10 ` [PATCH v3 11/19] mm/sparse: Move sparse_vmemmap_init_nid_late() into sparse_init_nid() Muchun Song
2026-06-02 10:10 ` [PATCH v3 12/19] mm/hugetlb_cma: Validate hugetlb CMA range by zone at reserve time Muchun Song
2026-06-02 10:10 ` [PATCH v3 13/19] mm/hugetlb: Refactor early boot gigantic hugepage allocation Muchun Song
2026-06-02 10:10 ` [PATCH v3 14/19] mm/hugetlb: Free cross-zone bootmem gigantic pages after allocation Muchun Song
2026-06-02 15:41 ` Mike Rapoport
2026-06-03 2:53 ` Muchun Song
2026-06-02 10:10 ` [PATCH v3 15/19] mm/hugetlb_vmemmap: Move bootmem HVO setup to early init Muchun Song
2026-06-02 15:41 ` Mike Rapoport
2026-06-03 2:42 ` Muchun Song
2026-06-03 12:02 ` Usama Arif [this message]
2026-06-03 12:24 ` Muchun Song
2026-06-03 12:35 ` Usama Arif
2026-06-02 10:10 ` [PATCH v3 16/19] mm/hugetlb: Remove obsolete bootmem cross-zone checks Muchun Song
2026-06-02 15:41 ` Mike Rapoport
2026-06-02 10:10 ` [PATCH v3 17/19] mm/sparse-vmemmap: Remove sparse_vmemmap_init_nid_late() Muchun Song
2026-06-02 15:41 ` Mike Rapoport
2026-06-02 10:10 ` [PATCH v3 18/19] mm/hugetlb: Remove unused bootmem cma field Muchun Song
2026-06-02 15:41 ` Mike Rapoport
2026-06-03 2:41 ` Muchun Song
2026-06-02 10:10 ` [PATCH v3 19/19] mm/mm_init: Fold __init_page_from_nid() into __init_deferred_page() Muchun Song
2026-06-02 14:46 ` Mike Rapoport
2026-06-02 15:41 ` Mike Rapoport
2026-06-03 2:39 ` Muchun Song
2026-06-02 10:34 ` [PATCH v3 00/19] mm: Refactor bootmem gigantic hugepage allocation Oscar Salvador (SUSE)
2026-06-02 12:01 ` Muchun Song
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260603120246.1572177-1-usama.arif@linux.dev \
--to=usama.arif@linux$(echo .)dev \
--cc=akpm@linux-foundation$(echo .)org \
--cc=aneesh.kumar@linux$(echo .)ibm.com \
--cc=chleroy@kernel$(echo .)org \
--cc=david@kernel$(echo .)org \
--cc=liam@infradead$(echo .)org \
--cc=linux-kernel@vger$(echo .)kernel.org \
--cc=linux-mm@kvack$(echo .)org \
--cc=linuxppc-dev@lists$(echo .)ozlabs.org \
--cc=ljs@kernel$(echo .)org \
--cc=maddy@linux$(echo .)ibm.com \
--cc=mike.kravetz@oracle$(echo .)com \
--cc=mpe@ellerman$(echo .)id.au \
--cc=muchun.song@linux$(echo .)dev \
--cc=npiggin@gmail$(echo .)com \
--cc=osalvador@suse$(echo .)de \
--cc=ritesh.list@gmail$(echo .)com \
--cc=rppt@kernel$(echo .)org \
--cc=songmuchun@bytedance$(echo .)com \
--cc=vbabka@kernel$(echo .)org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox