From: Kiryl Shutsemau <kas@kernel.org>
To: Andrew Morton <akpm@linux-foundation.org>,
Muchun Song <muchun.song@linux.dev>,
David Hildenbrand <david@redhat.com>,
Matthew Wilcox <willy@infradead.org>,
Usama Arif <usamaarif642@gmail.com>,
Frank van der Linden <fvdl@google.com>
Cc: Oscar Salvador <osalvador@suse.de>,
Mike Rapoport <rppt@kernel.org>, Vlastimil Babka <vbabka@suse.cz>,
Lorenzo Stoakes <lorenzo.stoakes@oracle.com>,
Zi Yan <ziy@nvidia.com>, Baoquan He <bhe@redhat.com>,
Michal Hocko <mhocko@suse.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Jonathan Corbet <corbet@lwn.net>,
Huacai Chen <chenhuacai@kernel.org>,
WANG Xuerui <kernel@xen0n.name>,
Palmer Dabbelt <palmer@dabbelt.com>,
Paul Walmsley <paul.walmsley@sifive.com>,
Albert Ou <aou@eecs.berkeley.edu>,
Alexandre Ghiti <alex@ghiti.fr>,
kernel-team@meta.com, linux-mm@kvack.org,
linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
loongarch@lists.linux.dev, linux-riscv@lists.infradead.org,
Kiryl Shutsemau <kas@kernel.org>
Subject: [PATCHv6 15/17] mm: Remove the branch from compound_head()
Date: Mon, 2 Feb 2026 15:56:31 +0000 [thread overview]
Message-ID: <20260202155634.650837-16-kas@kernel.org> (raw)
In-Reply-To: <20260202155634.650837-1-kas@kernel.org>
The compound_head() function is a hot path. For example, the zap path
calls it for every leaf page table entry.
Rewrite the helper function in a branchless manner to eliminate the risk
of CPU branch misprediction.
Signed-off-by: Kiryl Shutsemau <kas@kernel.org>
Reviewed-by: Muchun Song <muchun.song@linux.dev>
Reviewed-by: Zi Yan <ziy@nvidia.com>
---
include/linux/page-flags.h | 27 +++++++++++++++++----------
1 file changed, 17 insertions(+), 10 deletions(-)
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 1aaa604f4b9b..16384cb6f962 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -224,25 +224,32 @@ static __always_inline bool compound_info_has_mask(void)
static __always_inline unsigned long _compound_head(const struct page *page)
{
unsigned long info = READ_ONCE(page->compound_info);
+ unsigned long mask;
+
+ if (!compound_info_has_mask()) {
+ /* Bit 0 encodes PageTail() */
+ if (info & 1)
+ return info - 1;
- /* Bit 0 encodes PageTail() */
- if (!(info & 1))
return (unsigned long)page;
-
- /*
- * If compound_info_has_mask() is false, the rest of compound_info is
- * the pointer to the head page.
- */
- if (!compound_info_has_mask())
- return info - 1;
+ }
/*
* If compoun_info_has_mask() is true the rest of the info encodes
* the mask that converts the address of the tail page to the head page.
*
* No need to clear bit 0 in the mask as 'page' always has it clear.
+ *
+ * Let's do it in a branchless manner.
*/
- return (unsigned long)page & info;
+
+ /* Non-tail: -1UL, Tail: 0 */
+ mask = (info & 1) - 1;
+
+ /* Non-tail: -1UL, Tail: info */
+ mask |= info;
+
+ return (unsigned long)page & mask;
}
#define compound_head(page) ((typeof(page))_compound_head(page))
--
2.51.2
next prev parent reply other threads:[~2026-02-02 15:58 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-02-02 15:56 [PATCHv6 00/17] mm: Eliminate fake head pages from vmemmap optimization Kiryl Shutsemau
2026-02-02 15:56 ` [PATCHv6 01/17] mm: Move MAX_FOLIO_ORDER definition to mmzone.h Kiryl Shutsemau
2026-02-07 20:20 ` Usama Arif
2026-02-10 15:01 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 02/17] mm: Change the interface of prep_compound_tail() Kiryl Shutsemau
2026-02-04 16:14 ` David Hildenbrand (arm)
2026-02-05 11:35 ` Kiryl Shutsemau
2026-02-05 11:58 ` David Hildenbrand (arm)
2026-02-10 15:06 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 03/17] mm: Rename the 'compound_head' field in the 'struct page' to 'compound_info' Kiryl Shutsemau
2026-02-04 16:14 ` David Hildenbrand (arm)
2026-02-10 15:09 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 04/17] mm: Move set/clear_compound_head() next to compound_head() Kiryl Shutsemau
2026-02-04 16:35 ` David Hildenbrand (arm)
2026-02-10 15:10 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 05/17] riscv/mm: Align vmemmap to maximal folio size Kiryl Shutsemau
2026-02-04 16:50 ` David Hildenbrand (arm)
2026-02-05 13:50 ` Kiryl Shutsemau
2026-02-05 13:54 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` [PATCHv6 06/17] LoongArch/mm: " Kiryl Shutsemau
2026-02-04 16:56 ` David Hildenbrand (arm)
2026-02-05 12:56 ` David Hildenbrand (Arm)
2026-02-05 13:43 ` Kiryl Shutsemau
2026-02-05 13:52 ` David Hildenbrand (Arm)
2026-02-05 13:52 ` Kiryl Shutsemau
2026-02-05 13:57 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` [PATCHv6 07/17] mm: Rework compound_head() for power-of-2 sizeof(struct page) Kiryl Shutsemau
2026-02-05 14:09 ` David Hildenbrand (Arm)
2026-02-07 20:19 ` Usama Arif
2026-02-10 15:40 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 08/17] mm: Make page_zonenum() use head page Kiryl Shutsemau
2026-02-04 3:40 ` Muchun Song
2026-02-05 13:10 ` David Hildenbrand (Arm)
2026-02-09 11:52 ` Kiryl Shutsemau
2026-02-10 15:57 ` Vlastimil Babka
2026-02-16 11:30 ` Kiryl Shutsemau
2026-02-15 23:13 ` Matthew Wilcox
2026-02-16 9:06 ` David Hildenbrand (Arm)
2026-02-16 11:20 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 09/17] mm/sparse: Check memmap alignment for compound_info_has_mask() Kiryl Shutsemau
2026-02-03 3:35 ` Muchun Song
2026-02-05 13:31 ` David Hildenbrand (Arm)
2026-02-05 13:58 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` [PATCHv6 10/17] mm/hugetlb: Refactor code around vmemmap_walk Kiryl Shutsemau
2026-02-02 15:56 ` [PATCHv6 11/17] mm/hugetlb: Remove fake head pages Kiryl Shutsemau
2026-02-03 9:50 ` Muchun Song
2026-02-06 9:14 ` David Hildenbrand (Arm)
2026-02-06 9:36 ` David Hildenbrand (Arm)
2026-02-07 20:16 ` Usama Arif
2026-02-07 21:25 ` David Hildenbrand (Arm)
2026-02-07 22:50 ` Usama Arif
2026-02-02 15:56 ` [PATCHv6 12/17] mm: Drop fake head checks Kiryl Shutsemau
2026-02-06 9:41 ` David Hildenbrand (Arm)
2026-02-10 16:18 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 13/17] hugetlb: Remove VMEMMAP_SYNCHRONIZE_RCU Kiryl Shutsemau
2026-02-06 9:42 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` [PATCHv6 14/17] mm/hugetlb: Remove hugetlb_optimize_vmemmap_key static key Kiryl Shutsemau
2026-02-06 9:42 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` Kiryl Shutsemau [this message]
2026-02-06 10:23 ` [PATCHv6 15/17] mm: Remove the branch from compound_head() David Hildenbrand (Arm)
2026-02-10 16:42 ` Vlastimil Babka
2026-02-02 15:56 ` [PATCHv6 16/17] hugetlb: Update vmemmap_dedup.rst Kiryl Shutsemau
2026-02-06 10:35 ` David Hildenbrand (Arm)
2026-02-02 15:56 ` [PATCHv6 17/17] mm/slab: Use compound_head() in page_slab() Kiryl Shutsemau
2026-02-04 3:39 ` Muchun Song
2026-02-06 10:42 ` David Hildenbrand (Arm)
2026-02-10 16:45 ` Vlastimil Babka
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=20260202155634.650837-16-kas@kernel.org \
--to=kas@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=alex@ghiti.fr \
--cc=aou@eecs.berkeley.edu \
--cc=bhe@redhat.com \
--cc=chenhuacai@kernel.org \
--cc=corbet@lwn.net \
--cc=david@redhat.com \
--cc=fvdl@google.com \
--cc=hannes@cmpxchg.org \
--cc=kernel-team@meta.com \
--cc=kernel@xen0n.name \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-riscv@lists.infradead.org \
--cc=loongarch@lists.linux.dev \
--cc=lorenzo.stoakes@oracle.com \
--cc=mhocko@suse.com \
--cc=muchun.song@linux.dev \
--cc=osalvador@suse.de \
--cc=palmer@dabbelt.com \
--cc=paul.walmsley@sifive.com \
--cc=rppt@kernel.org \
--cc=usamaarif642@gmail.com \
--cc=vbabka@suse.cz \
--cc=willy@infradead.org \
--cc=ziy@nvidia.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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