From: Kent Overstreet <kent.overstreet@linux.dev>
To: Pasha Tatashin <pasha.tatashin@soleen.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
akpm@linux-foundation.org, x86@kernel.org, bp@alien8.de,
brauner@kernel.org, bristot@redhat.com, bsegall@google.com,
dave.hansen@linux.intel.com, dianders@chromium.org,
dietmar.eggemann@arm.com, eric.devolder@oracle.com,
hca@linux.ibm.com, hch@infradead.org,
jacob.jun.pan@linux.intel.com, jgg@ziepe.ca,
jpoimboe@kernel.org, jroedel@suse.de, juri.lelli@redhat.com,
kinseyho@google.com, kirill.shutemov@linux.intel.com,
lstoakes@gmail.com, luto@kernel.org, mgorman@suse.de,
mic@digikod.net, michael.christie@oracle.com, mingo@redhat.com,
mjguzik@gmail.com, mst@redhat.com, npiggin@gmail.com,
peterz@infradead.org, pmladek@suse.com,
rick.p.edgecombe@intel.com, rostedt@goodmis.org,
surenb@google.com, tglx@linutronix.de, urezki@gmail.com,
vincent.guittot@linaro.org, vschneid@redhat.com
Subject: Re: [RFC 00/14] Dynamic Kernel Stacks
Date: Thu, 14 Mar 2024 15:49:22 -0400 [thread overview]
Message-ID: <mvkjn34r3t3qz5qna72gbzp6ampgshemuvgqucjbxlckimh2fc@jqp63wy3v5ag> (raw)
In-Reply-To: <CA+CK2bDQu0jh55H=FQtA-5R-5Qy2=nqzcybEuN=RzVR75NBLRA@mail.gmail.com>
On Thu, Mar 14, 2024 at 03:34:03PM -0400, Pasha Tatashin wrote:
> On Thu, Mar 14, 2024 at 3:29 PM Kent Overstreet
> <kent.overstreet@linux.dev> wrote:
> >
> > On Thu, Mar 14, 2024 at 03:23:08PM -0400, Pasha Tatashin wrote:
> > > > >
> > > > > My point is that what matters is total memory use, not just memory used in
> > > > > the kernel. Amdahl's law.
> > > >
> > > > If userspace is running a few processes with many threads and the
> > > > userspace stacks are small, kernel stacks could end up dominating.
> > > >
> > > > I'd like to see some numbers though.
> > >
> > > The unused kernel stack pages occupy petabytes of memory across the fleet [1].
> >
> > Raw number doesn't mean much here (I know how many machines Google has,
> > of course it's going to be petabytes ;), percentage of system memory
> > would be better.
> >
> > What I'd _really_ like to see is raw output from memory allocation
> > profiling, so we can see how much memory is going to kernel stacks vs.
> > other kernel allocations.
>
> I've heard there is memory profiling working that can help with that...
I heard you've tried it out, too :)
> While I do not have the data you are asking for, the other kernel
> allocations might be useful, but this particular project is targeted
> to help with reducing overhead where the memory is not used, or used
> in very extreme rare cases.
Well, do you think you could gather it? We shouldn't be blindly applying
performance optimizations; we need to know where to focus our efforts.
e.g. on my laptop I've currently got 356 processes for < 6M of kernel
stack out of 32G total ram, so clearly this isn't much use to me. If the
ratio is similar on your servers - nah, don't want it. I expect the
ratio is not similar and you are burning proportially more memory on
kernel stacks, but we still need to gather the data and do the math :)
>
> > Number of kernel threads vs. number of user threads would also be good
> > to know - I've been seeing ps output lately where we've got a lot more
> > workqueue workers than we should, perhaps that's something that could be
> > addressed.
>
> Yes, doing other optimizations make sense, reducing the total number
> kernel threads if possible might help as well. I will look into this
> as well to see how many user threads vs kernel threads we have.
Great, that will help too.
next prev parent reply other threads:[~2024-03-14 19:49 UTC|newest]
Thread overview: 98+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-11 16:46 [RFC 00/14] Dynamic Kernel Stacks Pasha Tatashin
2024-03-11 16:46 ` [RFC 01/14] task_stack.h: remove obsolete __HAVE_ARCH_KSTACK_END check Pasha Tatashin
2024-03-17 14:36 ` Christophe JAILLET
2024-03-17 15:13 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 02/14] fork: Clean-up ifdef logic around stack allocation Pasha Tatashin
2024-03-11 16:46 ` [RFC 03/14] fork: Clean-up naming of vm_strack/vm_struct variables in vmap stacks code Pasha Tatashin
2024-03-17 14:42 ` Christophe JAILLET
2024-03-19 16:32 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 04/14] fork: Remove assumption that vm_area->nr_pages equals to THREAD_SIZE Pasha Tatashin
2024-03-17 14:45 ` Christophe JAILLET
2024-03-17 15:14 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 05/14] fork: check charging success before zeroing stack Pasha Tatashin
2024-03-12 15:57 ` Kirill A. Shutemov
2024-03-12 16:52 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 06/14] fork: zero vmap stack using clear_page() instead of memset() Pasha Tatashin
2024-03-12 7:15 ` Nikolay Borisov
2024-03-12 16:53 ` Pasha Tatashin
2024-03-14 7:55 ` Christophe Leroy
2024-03-14 13:52 ` Pasha Tatashin
2024-03-17 14:48 ` Christophe JAILLET
2024-03-17 15:15 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 07/14] fork: use the first page in stack to store vm_stack in cached_stacks Pasha Tatashin
2024-03-11 16:46 ` [RFC 08/14] fork: separate vmap stack alloction and free calls Pasha Tatashin
2024-03-14 15:18 ` Jeff Xie
2024-03-14 17:14 ` Pasha Tatashin
2024-03-17 14:51 ` Christophe JAILLET
2024-03-17 15:15 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 09/14] mm/vmalloc: Add a get_vm_area_node() and vmap_pages_range_noflush() public functions Pasha Tatashin
2024-03-11 16:46 ` [RFC 10/14] fork: Dynamic Kernel Stacks Pasha Tatashin
2024-03-11 19:32 ` Randy Dunlap
2024-03-11 19:55 ` Pasha Tatashin
2024-03-11 16:46 ` [RFC 11/14] x86: add support for " Pasha Tatashin
2024-03-11 22:17 ` Andy Lutomirski
2024-03-11 23:10 ` Pasha Tatashin
2024-03-11 23:33 ` Thomas Gleixner
2024-03-11 23:34 ` Andy Lutomirski
2024-03-12 0:08 ` Pasha Tatashin
2024-03-12 0:23 ` Pasha Tatashin
2024-03-11 23:34 ` Dave Hansen
2024-03-11 23:41 ` Andy Lutomirski
2024-03-11 23:56 ` Nadav Amit
2024-03-12 0:02 ` Andy Lutomirski
2024-03-12 7:20 ` Nadav Amit
2024-03-12 0:53 ` Dave Hansen
2024-03-12 1:25 ` H. Peter Anvin
2024-03-12 2:16 ` Andy Lutomirski
2024-03-12 2:20 ` H. Peter Anvin
2024-03-12 21:58 ` Andi Kleen
2024-03-13 10:23 ` Thomas Gleixner
2024-03-13 13:43 ` Pasha Tatashin
2024-03-13 15:28 ` Pasha Tatashin
2024-03-13 16:12 ` Thomas Gleixner
2024-03-14 14:03 ` Pasha Tatashin
2024-03-14 18:26 ` Thomas Gleixner
2024-03-11 16:46 ` [RFC 12/14] task_stack.h: Clean-up stack_not_used() implementation Pasha Tatashin
2024-03-11 16:46 ` [RFC 13/14] task_stack.h: Add stack_not_used() support for dynamic stack Pasha Tatashin
2024-03-11 16:46 ` [RFC 14/14] fork: Dynamic Kernel Stack accounting Pasha Tatashin
2024-03-11 17:09 ` [RFC 00/14] Dynamic Kernel Stacks Mateusz Guzik
2024-03-11 18:58 ` Pasha Tatashin
2024-03-11 19:21 ` Mateusz Guzik
2024-03-11 19:55 ` Pasha Tatashin
2024-03-12 17:18 ` H. Peter Anvin
2024-03-12 19:45 ` Pasha Tatashin
2024-03-12 21:36 ` H. Peter Anvin
2024-03-14 19:05 ` Kent Overstreet
2024-03-14 19:23 ` Pasha Tatashin
2024-03-14 19:28 ` Kent Overstreet
2024-03-14 19:34 ` Pasha Tatashin
2024-03-14 19:49 ` Kent Overstreet [this message]
2024-03-12 22:18 ` David Laight
2024-03-14 19:43 ` Matthew Wilcox
2024-03-14 19:53 ` Kent Overstreet
2024-03-14 19:57 ` Matthew Wilcox
2024-03-14 19:58 ` Kent Overstreet
2024-03-15 3:13 ` Pasha Tatashin
2024-03-15 3:39 ` H. Peter Anvin
2024-03-16 19:17 ` Pasha Tatashin
2024-03-17 0:41 ` Matthew Wilcox
2024-03-17 1:32 ` Kent Overstreet
2024-03-17 14:19 ` Pasha Tatashin
2024-03-17 14:43 ` Brian Gerst
2024-03-17 16:15 ` Pasha Tatashin
2024-03-17 21:30 ` Brian Gerst
2024-03-18 14:59 ` Pasha Tatashin
2024-03-18 21:02 ` Brian Gerst
2024-03-19 14:56 ` Pasha Tatashin
2024-03-17 18:57 ` David Laight
2024-03-18 15:09 ` Pasha Tatashin
2024-03-18 15:13 ` Pasha Tatashin
2024-03-18 15:19 ` Matthew Wilcox
2024-03-18 15:30 ` Pasha Tatashin
2024-03-18 15:53 ` David Laight
2024-03-18 16:57 ` Pasha Tatashin
2024-03-18 15:38 ` David Laight
2024-03-18 17:00 ` Pasha Tatashin
2024-03-18 17:37 ` Pasha Tatashin
2024-03-15 4:17 ` H. Peter Anvin
2024-03-17 0:47 ` H. Peter Anvin
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=mvkjn34r3t3qz5qna72gbzp6ampgshemuvgqucjbxlckimh2fc@jqp63wy3v5ag \
--to=kent.overstreet@linux.dev \
--cc=akpm@linux-foundation.org \
--cc=bp@alien8.de \
--cc=brauner@kernel.org \
--cc=bristot@redhat.com \
--cc=bsegall@google.com \
--cc=dave.hansen@linux.intel.com \
--cc=dianders@chromium.org \
--cc=dietmar.eggemann@arm.com \
--cc=eric.devolder@oracle.com \
--cc=hca@linux.ibm.com \
--cc=hch@infradead.org \
--cc=hpa@zytor.com \
--cc=jacob.jun.pan@linux.intel.com \
--cc=jgg@ziepe.ca \
--cc=jpoimboe@kernel.org \
--cc=jroedel@suse.de \
--cc=juri.lelli@redhat.com \
--cc=kinseyho@google.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lstoakes@gmail.com \
--cc=luto@kernel.org \
--cc=mgorman@suse.de \
--cc=mic@digikod.net \
--cc=michael.christie@oracle.com \
--cc=mingo@redhat.com \
--cc=mjguzik@gmail.com \
--cc=mst@redhat.com \
--cc=npiggin@gmail.com \
--cc=pasha.tatashin@soleen.com \
--cc=peterz@infradead.org \
--cc=pmladek@suse.com \
--cc=rick.p.edgecombe@intel.com \
--cc=rostedt@goodmis.org \
--cc=surenb@google.com \
--cc=tglx@linutronix.de \
--cc=urezki@gmail.com \
--cc=vincent.guittot@linaro.org \
--cc=vschneid@redhat.com \
--cc=x86@kernel.org \
/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;
as well as URLs for NNTP newsgroup(s).