From: Kalesh Singh <kaleshsingh@google.com>
To: Yang Shi <shy828301@gmail.com>
Cc: akpm@linux-foundation.org, vbabka@suse.cz,
yang@os.amperecomputing.com, riel@surriel.com, david@redhat.com,
linux@armlinux.org.uk, tsbogend@alpha.franken.de,
James.Bottomley@hansenpartnership.com,
ysato@users.sourceforge.jp, dalias@libc.org,
glaubitz@physik.fu-berlin.de, davem@davemloft.net,
andreas@gaisler.com, tglx@linutronix.de, bp@alien8.de,
dave.hansen@linux.intel.com, x86@kernel.org, chris@zankel.net,
jcmvbkbc@gmail.com, bhelgaas@google.com, jason.andryuk@amd.com,
leitao@debian.org, linux-alpha@vger.kernel.org,
linux-kernel@vger.kernel.org,
linux-snps-arc@lists.infradead.org,
linux-arm-kernel@lists.infradead.org,
linux-csky@vger.kernel.org, loongarch@lists.linux.dev,
linux-mips@vger.kernel.org, linux-parisc@vger.kernel.org,
linuxppc-dev@lists.ozlabs.org, linux-s390@vger.kernel.org,
linux-sh@vger.kernel.org, sparclinux@vger.kernel.org,
linux-mm@kvack.org, kernel-team@android.com,
android-mm@google.com
Subject: Re: [PATCH mm-unstable 17/17] mm: Respect mmap hint before THP alignment if allocation is possible
Date: Tue, 10 Dec 2024 09:34:01 -0800 [thread overview]
Message-ID: <CAC_TJvdReRHzBSgg2iqOw3Kw6BBOtwGE=8nB2Hsw-nsmkxN0+g@mail.gmail.com> (raw)
In-Reply-To: <CAHbLzkq2SNaqzx4d981H2QfQvtObS3X0pPL8=oqFsFbMditWPA@mail.gmail.com>
On Mon, Dec 9, 2024 at 7:37 PM Yang Shi <shy828301@gmail.com> wrote:
>
> On Mon, Dec 9, 2024 at 6:45 PM Kalesh Singh <kaleshsingh@google.com> wrote:
> >
> > Commit 249608ee4713 ("mm: respect mmap hint address when aligning for THP")
> > fallsback to PAGE_SIZE alignment instead of THP alignment
> > for anonymous mapping as long as a hint address is provided by the user
> > -- even if we weren't able to allocate the unmapped area at the hint
> > address in the end.
> >
> > This was done to address the immediate regression in anonymous mappings
> > where the hint address were being ignored in some cases; due to commit
> > efa7df3e3bb5 ("mm: align larger anonymous mappings on THP boundaries").
> >
> > It was later pointed out that this issue also existed for file-backed
> > mappings from file systems that use thp_get_unmapped_area() for their
> > .get_unmapped_area() file operation.
> >
> > The same fix was not applied for file-backed mappings since it would
> > mean any mmap requests that provide a hint address would be only
> > PAGE_SIZE-aligned regardless of whether allocation was successful at
> > the hint address or not.
> >
> > Instead, use arch_mmap_hint() to first attempt allocation at the hint
> > address and fallback to THP alignment if that fails.
>
> Thanks for taking time to try to fix this.
>
> >
> > Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
> > ---
> > mm/huge_memory.c | 15 ++++++++-------
> > mm/mmap.c | 1 -
> > 2 files changed, 8 insertions(+), 8 deletions(-)
> >
> > diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> > index 137abeda8602..f070c89dafc9 100644
> > --- a/mm/huge_memory.c
> > +++ b/mm/huge_memory.c
> > @@ -1097,6 +1097,14 @@ static unsigned long __thp_get_unmapped_area(struct file *filp,
> > loff_t off_align = round_up(off, size);
> > unsigned long len_pad, ret, off_sub;
> >
> > + /*
> > + * If allocation at the address hint succeeds; respect the hint and
> > + * don't try to align to THP boundary.
> > + */
> > + addr = arch_mmap_hint(filp, addr, len, off, flags);
> > + if (addr)
> > + return addr;
> > +
Hi Yang,
Thanks for the comments.
>
> IIUC, arch_mmap_hint() will be called in arch_get_unmapped_area() and
> arch_get_unmapped_area_topdown() again. So we will actually look up
> maple tree twice. It sounds like the second hint address search is
> pointless. You should be able to set addr to 0 before calling
> mm_get_unmapped_area_vmflags() in order to skip the second hint
> address search.
You are right that it would call into arch_mmap_hint() twice but it
only attempts the lookup once since on the second attempt addr == 0.
Thanks,
Kalesh
>
> > if (!IS_ENABLED(CONFIG_64BIT) || in_compat_syscall())
> > return 0;
> >
> > @@ -1117,13 +1125,6 @@ static unsigned long __thp_get_unmapped_area(struct file *filp,
> > if (IS_ERR_VALUE(ret))
> > return 0;
> >
> > - /*
> > - * Do not try to align to THP boundary if allocation at the address
> > - * hint succeeds.
> > - */
> > - if (ret == addr)
> > - return addr;
> > -
> > off_sub = (off - ret) & (size - 1);
> >
> > if (test_bit(MMF_TOPDOWN, ¤t->mm->flags) && !off_sub)
> > diff --git a/mm/mmap.c b/mm/mmap.c
> > index 59bf7d127aa1..6bfeec80152a 100644
> > --- a/mm/mmap.c
> > +++ b/mm/mmap.c
> > @@ -807,7 +807,6 @@ __get_unmapped_area(struct file *file, unsigned long addr, unsigned long len,
> > if (get_area) {
> > addr = get_area(file, addr, len, pgoff, flags);
> > } else if (IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE) && !file
> > - && !addr /* no hint */
> > && IS_ALIGNED(len, PMD_SIZE)) {
> > /* Ensures that larger anonymous mappings are THP aligned. */
> > addr = thp_get_unmapped_area_vmflags(file, addr, len,
> > --
> > 2.47.0.338.g60cca15819-goog
> >
> >
next prev parent reply other threads:[~2024-12-10 17:34 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-12-10 2:41 [PATCH mm-unstable 00/17] mm: Introduce arch_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 01/17] mm: Introduce generic_mmap_hint() Kalesh Singh
2024-12-10 3:27 ` Yang Shi
2024-12-10 17:55 ` Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 02/17] mm: x86: Introduce arch_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 03/17] mm: arm: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 04/17] mm: alpha: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 05/17] mm: arc: Use generic_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 06/17] mm: csky: Introduce arch_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 07/17] mm: loongarch: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 08/17] mm: mips: Introduce arch_align_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 09/17] mm: parisc: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 10/17] mm: s390: Introduce arch_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 11/17] mm: sh: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 12/17] mm: sparc32: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 13/17] mm: sparc64: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 14/17] mm: xtensa: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 15/17] mm: powerpc: " Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 16/17] mm: Fallback to generic_mmap_hint() Kalesh Singh
2024-12-10 2:41 ` [PATCH mm-unstable 17/17] mm: Respect mmap hint before THP alignment if allocation is possible Kalesh Singh
2024-12-10 3:37 ` Yang Shi
2024-12-10 17:34 ` Kalesh Singh [this message]
2024-12-10 19:39 ` Yang Shi
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='CAC_TJvdReRHzBSgg2iqOw3Kw6BBOtwGE=8nB2Hsw-nsmkxN0+g@mail.gmail.com' \
--to=kaleshsingh@google.com \
--cc=James.Bottomley@hansenpartnership.com \
--cc=akpm@linux-foundation.org \
--cc=andreas@gaisler.com \
--cc=android-mm@google.com \
--cc=bhelgaas@google.com \
--cc=bp@alien8.de \
--cc=chris@zankel.net \
--cc=dalias@libc.org \
--cc=dave.hansen@linux.intel.com \
--cc=davem@davemloft.net \
--cc=david@redhat.com \
--cc=glaubitz@physik.fu-berlin.de \
--cc=jason.andryuk@amd.com \
--cc=jcmvbkbc@gmail.com \
--cc=kernel-team@android.com \
--cc=leitao@debian.org \
--cc=linux-alpha@vger.kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-csky@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-parisc@vger.kernel.org \
--cc=linux-s390@vger.kernel.org \
--cc=linux-sh@vger.kernel.org \
--cc=linux-snps-arc@lists.infradead.org \
--cc=linux@armlinux.org.uk \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=loongarch@lists.linux.dev \
--cc=riel@surriel.com \
--cc=shy828301@gmail.com \
--cc=sparclinux@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=tsbogend@alpha.franken.de \
--cc=vbabka@suse.cz \
--cc=x86@kernel.org \
--cc=yang@os.amperecomputing.com \
--cc=ysato@users.sourceforge.jp \
/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