From: Marcelo Tosatti <mtosatti@redhat•com>
To: David Hildenbrand <david@redhat•com>
Cc: geert+renesas@glider•be, Peter Zijlstra <peterz@infradead•org>,
linus.walleij@linaro•org, dave.hansen@linux•intel.com,
Yair Podemsky <ypodemsk@redhat•com>,
sebastian.reichel@collabora•com, linux-mm@kvack•org,
hpa@zytor•com, sparclinux@vger•kernel.org,
agordeev@linux•ibm.com, will@kernel•org, ardb@kernel•org,
linux-arch@vger•kernel.org, linux-s390@vger•kernel.org,
vschneid@redhat•com, arnd@arndb•de, paulmck@kernel•org,
aneesh.kumar@linux•ibm.com, x86@kernel•org,
linux@armlinux•org.uk, tony@atomide•com, mingo@redhat•com,
samitolvanen@google•com, borntraeger@linux•ibm.com,
hca@linux•ibm.com, keescook@chromium•org, gor@linux•ibm.com,
jannh@google•com, Frederic Weisbecker <frederic@kernel•org>,
npiggin@gmail•com, rmk+kernel@armlinux•org.uk,
Yang Shi <shy828301@gmail•com>,
bp@alien8•de, nick.hawkins@hpe•com, tglx@linutronix•de,
jpoimboe@kernel•org, linux-arm-kernel@lists•infradead.org,
alougovs@redhat•com, linux-kernel@vger•kernel.org,
juerg.haefliger@canonical•com, svens@linux•ibm.com ,
dhildenb@redhat•com, akpm@linux-foundation•org,
linuxppc-dev@lists•ozlabs.org, davem@davemloft•net
Subject: Re: [PATCH 3/3] mm/mmu_gather: send tlb_remove_table_smp_sync IPI only to CPUs in kernel mode
Date: Wed, 19 Apr 2023 08:39:48 -0300 [thread overview]
Message-ID: <ZD/TBN5doEex3iag@tpad> (raw)
In-Reply-To: <914e826e-3fab-4540-d3a1-24ca39b1cf0a@redhat.com>
On Wed, Apr 19, 2023 at 01:30:57PM +0200, David Hildenbrand wrote:
> On 06.04.23 20:27, Peter Zijlstra wrote:
> > On Thu, Apr 06, 2023 at 05:51:52PM +0200, David Hildenbrand wrote:
> > > On 06.04.23 17:02, Peter Zijlstra wrote:
> >
> > > > DavidH, what do you thikn about reviving Jann's patches here:
> > > >
> > > > https://bugs.chromium.org/p/project-zero/issues/detail?id=2365#c1
> > > >
> > > > Those are far more invasive, but afaict they seem to do the right thing.
> > > >
> > >
> > > I recall seeing those while discussed on security@kernel•org. What we
> > > currently have was (IMHO for good reasons) deemed better to fix the issue,
> > > especially when caring about backports and getting it right.
> >
> > Yes, and I think that was the right call. However, we can now revisit
> > without having the pressure of a known defect and backport
> > considerations.
> >
> > > The alternative that was discussed in that context IIRC was to simply
> > > allocate a fresh page table, place the fresh page table into the list
> > > instead, and simply free the old page table (then using common machinery).
> > >
> > > TBH, I'd wish (and recently raised) that we could just stop wasting memory
> > > on page tables for THPs that are maybe never going to get PTE-mapped ... and
> > > eventually just allocate on demand (with some caching?) and handle the
> > > places where we're OOM and cannot PTE-map a THP in some descend way.
> > >
> > > ... instead of trying to figure out how to deal with these page tables we
> > > cannot free but have to special-case simply because of GUP-fast.
> >
> > Not keeping them around sounds good to me, but I'm not *that* familiar
> > with the THP code, most of that happened after I stopped tracking mm. So
> > I'm not sure how feasible is it.
> >
> > But it does look entirely feasible to rework this page-table freeing
> > along the lines Jann did.
>
> It's most probably more feasible, although the easiest would be to just
> allocate a fresh page table to deposit and free the old one using the mmu
> gatherer.
>
> This way we can avoid the khugepaged of tlb_remove_table_smp_sync(), but not
> the tlb_remove_table_one() usage. I suspect khugepaged isn't really relevant
> in RT kernels (IIRC, most of RT setups disable THP completely).
People will disable khugepaged because it causes IPIs (and the fact one
has to disable khugepaged is a configuration overhead, and a source of
headache for configuring the realtime system, since one can forget of
doing that, etc).
But people do want to run non-RT applications along with RT applications
(in case you have a single box on a priviledged location, for example).
>
> tlb_remove_table_one() only triggers if __get_free_page(GFP_NOWAIT |
> __GFP_NOWARN); fails. IIUC, that can happen easily under memory pressure
> because it doesn't wait for direct reclaim.
>
> I don't know much about RT workloads (so I'd appreciate some feedback), but
> I guess we can run int memory pressure as well due to some !rt housekeeping
> task on the system?
Yes, exactly (memory for -RT app will be mlocked).
next prev parent reply other threads:[~2023-04-19 11:42 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-04 13:42 [PATCH 0/3] send tlb_remove_table_smp_sync IPI only to necessary CPUs Yair Podemsky
2023-04-04 13:42 ` [PATCH 1/3] arch: Introduce ARCH_HAS_CPUMASK_BITS Yair Podemsky
2023-04-04 13:47 ` David Hildenbrand
2023-04-04 13:42 ` [PATCH 2/3] mm/mmu_gather: send tlb_remove_table_smp_sync IPI only to MM CPUs Yair Podemsky
2023-04-04 14:57 ` Peter Zijlstra
2023-04-04 13:42 ` [PATCH 3/3] mm/mmu_gather: send tlb_remove_table_smp_sync IPI only to CPUs in kernel mode Yair Podemsky
2023-04-04 14:03 ` David Hildenbrand
2023-04-04 15:12 ` Peter Zijlstra
2023-04-04 16:00 ` Peter Zijlstra
2023-04-05 10:43 ` Frederic Weisbecker
2023-04-05 11:10 ` Frederic Weisbecker
2023-04-05 11:41 ` Peter Zijlstra
2023-04-05 12:00 ` David Hildenbrand
2023-04-05 12:05 ` Frederic Weisbecker
2023-04-05 12:31 ` Frederic Weisbecker
2023-04-05 12:45 ` Valentin Schneider
2023-04-06 13:38 ` Peter Zijlstra
2023-04-06 14:11 ` Valentin Schneider
2023-04-06 14:39 ` Peter Zijlstra
2023-04-05 19:45 ` Marcelo Tosatti
2023-04-05 19:52 ` Peter Zijlstra
2023-04-06 12:38 ` Marcelo Tosatti
2023-04-06 13:29 ` Peter Zijlstra
2023-04-06 14:04 ` Peter Zijlstra
2023-04-06 14:42 ` David Hildenbrand
2023-04-06 15:06 ` Peter Zijlstra
2023-04-06 15:02 ` Peter Zijlstra
2023-04-06 15:51 ` David Hildenbrand
2023-04-06 18:27 ` Peter Zijlstra
2023-04-19 11:30 ` David Hildenbrand
2023-04-19 11:39 ` Marcelo Tosatti [this message]
2023-04-05 19:43 ` Marcelo Tosatti
2023-04-05 19:54 ` Peter Zijlstra
2023-04-06 12:49 ` Marcelo Tosatti
2023-04-06 13:32 ` Peter Zijlstra
2023-04-19 11:01 ` Marcelo Tosatti
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=ZD/TBN5doEex3iag@tpad \
--to=mtosatti@redhat$(echo .)com \
--cc=agordeev@linux$(echo .)ibm.com \
--cc=akpm@linux-foundation$(echo .)org \
--cc=alougovs@redhat$(echo .)com \
--cc=aneesh.kumar@linux$(echo .)ibm.com \
--cc=ardb@kernel$(echo .)org \
--cc=arnd@arndb$(echo .)de \
--cc=borntraeger@linux$(echo .)ibm.com \
--cc=bp@alien8$(echo .)de \
--cc=dave.hansen@linux$(echo .)intel.com \
--cc=davem@davemloft$(echo .)net \
--cc=david@redhat$(echo .)com \
--cc=dhildenb@redhat$(echo .)com \
--cc=frederic@kernel$(echo .)org \
--cc=geert+renesas@glider$(echo .)be \
--cc=gor@linux$(echo .)ibm.com \
--cc=hca@linux$(echo .)ibm.com \
--cc=hpa@zytor$(echo .)com \
--cc=jannh@google$(echo .)com \
--cc=jpoimboe@kernel$(echo .)org \
--cc=juerg.haefliger@canonical$(echo .)com \
--cc=keescook@chromium$(echo .)org \
--cc=linus.walleij@linaro$(echo .)org \
--cc=linux-arch@vger$(echo .)kernel.org \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
--cc=linux-kernel@vger$(echo .)kernel.org \
--cc=linux-mm@kvack$(echo .)org \
--cc=linux-s390@vger$(echo .)kernel.org \
--cc=linux@armlinux$(echo .)org.uk \
--cc=linuxppc-dev@lists$(echo .)ozlabs.org \
--cc=mingo@redhat$(echo .)com \
--cc=nick.hawkins@hpe$(echo .)com \
--cc=npiggin@gmail$(echo .)com \
--cc=paulmck@kernel$(echo .)org \
--cc=peterz@infradead$(echo .)org \
--cc=rmk+kernel@armlinux$(echo .)org.uk \
--cc=samitolvanen@google$(echo .)com \
--cc=sebastian.reichel@collabora$(echo .)com \
--cc=shy828301@gmail$(echo .)com \
--cc=sparclinux@vger$(echo .)kernel.org \
--cc=svens@linux$(echo .)ibm.com \
--cc=tglx@linutronix$(echo .)de \
--cc=tony@atomide$(echo .)com \
--cc=vschneid@redhat$(echo .)com \
--cc=will@kernel$(echo .)org \
--cc=x86@kernel$(echo .)org \
--cc=ypodemsk@redhat$(echo .)com \
/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