From: Mircea CIRJALIU - MELIU <mcirjaliu@bitdefender.com>
To: "linux-mm@kvack.org" <linux-mm@kvack.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
Jerome Glisse <jglisse@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>,
"aarcange@redhat.com" <aarcange@redhat.com>
Subject: [RFC PATCH v1 4/4] mm/gup: flag to limit follow_page() to transhuge pages
Date: Wed, 11 Dec 2019 09:29:19 +0000 [thread overview]
Message-ID: <DB7PR02MB3979ECF6271070E49D062EAFBB5A0@DB7PR02MB3979.eurprd02.prod.outlook.com> (raw)
Sometimes the user needs to look up a transhuge page mapped at a given address.
So instead of being given a normal page and having to test it, save some cycles
by filtering out PTE mappings.
Signed-off-by: Mircea Cirjaliu <mcirjaliu@bitdefender.com>
---
include/linux/mm.h | 1 +
mm/gup.c | 13 +++++++++++--
2 files changed, 12 insertions(+), 2 deletions(-)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index c97ea3b..64bbf83 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2579,6 +2579,7 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
#define FOLL_ANON 0x8000 /* don't do file mappings */
#define FOLL_LONGTERM 0x10000 /* mapping lifetime is indefinite: see below */
#define FOLL_SPLIT_PMD 0x20000 /* split huge pmd before returning */
+#define FOLL_HUGE 0x40000 /* only return huge mappings */
/*
* NOTE on FOLL_LONGTERM:
diff --git a/mm/gup.c b/mm/gup.c
index 7646bf9..a776bdc 100644
--- a/mm/gup.c
+++ b/mm/gup.c
@@ -361,9 +361,11 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma,
if (page)
return page;
}
- if (likely(!pmd_trans_huge(pmdval)))
+ if (likely(!pmd_trans_huge(pmdval))) {
+ if (flags & FOLL_HUGE)
+ return ERR_PTR(-EFAULT);
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
-
+ }
if ((flags & FOLL_NUMA) && pmd_protnone(pmdval))
return no_page_table(vma, flags);
@@ -382,6 +384,8 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma,
}
if (unlikely(!pmd_trans_huge(*pmd))) {
spin_unlock(ptl);
+ if (flags & FOLL_HUGE)
+ return ERR_PTR(-EFAULT);
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
}
if (flags & (FOLL_SPLIT | FOLL_SPLIT_PMD)) {
@@ -513,6 +517,8 @@ static struct page *follow_page_mask(struct vm_area_struct *vma,
struct page *page;
struct mm_struct *mm = vma->vm_mm;
+ VM_BUG_ON((flags & (FOLL_SPLIT | FOLL_HUGE)) == (FOLL_SPLIT | FOLL_HUGE));
+
ctx->page_mask = 0;
/* make this handle hugepd */
@@ -685,6 +691,9 @@ static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags)
if (gup_flags & FOLL_ANON && !vma_is_anonymous(vma))
return -EFAULT;
+ if (gup_flags & FOLL_HUGE && !transparent_hugepage_enabled(vma))
+ return -EFAULT;
+
if (write) {
if (!(vm_flags & VM_WRITE)) {
if (!(gup_flags & FOLL_FORCE))
next reply other threads:[~2019-12-11 9:29 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-11 9:29 Mircea CIRJALIU - MELIU [this message]
2019-12-11 15:56 ` Kirill A. Shutemov
2019-12-11 17:13 ` Mircea CIRJALIU - MELIU
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=DB7PR02MB3979ECF6271070E49D062EAFBB5A0@DB7PR02MB3979.eurprd02.prod.outlook.com \
--to=mcirjaliu@bitdefender.com \
--cc=aarcange@redhat.com \
--cc=jglisse@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=pbonzini@redhat.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