Message-ID: <CA+CK2bCpg9ND7FbbA7R_BK5FjR+rMA6DUoaCXmre70xNQXZoLw@mail.gmail.com>
Date: Mon, 18 Mar 2024 11:13:30 -0400
From: Pasha Tatashin <pasha.tatashin@...een.com>
To: David Laight <David.Laight@...lab.com>
Cc: "H. Peter Anvin" <hpa@...or.com>, Matthew Wilcox <willy@...radead.org>, 
	Kent Overstreet <kent.overstreet@...ux.dev>, 
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>, "linux-mm@...ck.org" <linux-mm@...ck.org>, 
	"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>, "x86@...nel.org" <x86@...nel.org>, 
	"bp@...en8.de" <bp@...en8.de>, "brauner@...nel.org" <brauner@...nel.org>, 
	"bristot@...hat.com" <bristot@...hat.com>, "bsegall@...gle.com" <bsegall@...gle.com>, 
	"dave.hansen@...ux.intel.com" <dave.hansen@...ux.intel.com>, 
	"dianders@...omium.org" <dianders@...omium.org>, 
	"dietmar.eggemann@....com" <dietmar.eggemann@....com>, 
	"eric.devolder@...cle.com" <eric.devolder@...cle.com>, "hca@...ux.ibm.com" <hca@...ux.ibm.com>, 
	"hch@...radead.org" <hch@...radead.org>, 
	"jacob.jun.pan@...ux.intel.com" <jacob.jun.pan@...ux.intel.com>, "jgg@...pe.ca" <jgg@...pe.ca>, 
	"jpoimboe@...nel.org" <jpoimboe@...nel.org>, "jroedel@...e.de" <jroedel@...e.de>, 
	"juri.lelli@...hat.com" <juri.lelli@...hat.com>, "kinseyho@...gle.com" <kinseyho@...gle.com>, 
	"kirill.shutemov@...ux.intel.com" <kirill.shutemov@...ux.intel.com>, "lstoakes@...il.com" <lstoakes@...il.com>, 
	"luto@...nel.org" <luto@...nel.org>, "mgorman@...e.de" <mgorman@...e.de>, "mic@...ikod.net" <mic@...ikod.net>, 
	"michael.christie@...cle.com" <michael.christie@...cle.com>, "mingo@...hat.com" <mingo@...hat.com>, 
	"mjguzik@...il.com" <mjguzik@...il.com>, "mst@...hat.com" <mst@...hat.com>, 
	"npiggin@...il.com" <npiggin@...il.com>, "peterz@...radead.org" <peterz@...radead.org>, 
	"pmladek@...e.com" <pmladek@...e.com>, 
	"rick.p.edgecombe@...el.com" <rick.p.edgecombe@...el.com>, "rostedt@...dmis.org" <rostedt@...dmis.org>, 
	"surenb@...gle.com" <surenb@...gle.com>, "tglx@...utronix.de" <tglx@...utronix.de>, 
	"urezki@...il.com" <urezki@...il.com>, 
	"vincent.guittot@...aro.org" <vincent.guittot@...aro.org>, "vschneid@...hat.com" <vschneid@...hat.com>
Subject: Re: [RFC 00/14] Dynamic Kernel Stacks

On Mon, Mar 18, 2024 at 11:09 AM Pasha Tatashin
<pasha.tatashin@...een.com> wrote:
>
> On Sun, Mar 17, 2024 at 2:58 PM David Laight <David.Laight@...labcom> wrote:
> >
> > From: Pasha Tatashin
> > > Sent: 16 March 2024 19:18
> > ...
> > > Expanding on Mathew's idea of an interface for dynamic kernel stack
> > > sizes, here's what I'm thinking:
> > >
> > > - Kernel Threads: Create all kernel threads with a fully populated
> > > THREAD_SIZE stack.  (i.e. 16K)
> > > - User Threads: Create all user threads with THREAD_SIZE kernel stack
> > > but only the top page mapped. (i.e. 4K)
> > > - In enter_from_user_mode(): Expand the thread stack to 16K by mapping
> > > three additional pages from the per-CPU stack cache. This function is
> > > called early in kernel entry points.
> > > - exit_to_user_mode(): Unmap the extra three pages and return them to
> > > the per-CPU cache. This function is called late in the kernel exit
> > > path.
> >
> > Isn't that entirely horrid for TLB use and so will require a lot of IPI?
>
> The TLB load is going to be exactly the same as today, we already use
> small pages for VMA mapped stacks. We won't need to have extra
> flushing either, the mappings are in the kernel space, and once pages
> are removed from the page table, no one is going to access that VA
> space until that thread enters the kernel again. We will need to
> invalidate the VA range only when the pages are mapped, and only on
> the local cpu.

The TLB miss rate is going to slightly increase, but very slightly,
because stacks are small 4-pages with only 3-dynamic pages, and
therefore only up-to 2-3 new misses per syscalls, and that is only for
the complicated deep syscalls, therefore, I suspect it won't affect
the real world performance.

> > Remember, if a thread sleeps in 'extra stack' and is then resheduled
> > on a different cpu the extra pages get 'pumped' from one cpu to
> > another.
>
> Yes, the per-cpu cache can get unbalanced this way, we can remember
> the original CPU where we acquired the pages to return to the same
> place.
>
> > I also suspect a stack_probe() is likely to end up being a cache miss
> > and also slow???
>
> Can you please elaborate on this point. I am not aware of
> stack_probe() and how it is used.
>
> > So you wouldn't want one on all calls.
> > I'm not sure you'd want a conditional branch either.
> >
> > The explicit request for 'more stack' can be required to be allowed
> > to sleep - removing a lot of issues.
> > It would also be portable to all architectures.
> > I'd also suspect that any thread that needs extra stack is likely
> > to need to again.
> > So while the memory could be recovered, I'd bet is isn't worth
> > doing except under memory pressure.
> > The call could also return 'no' - perhaps useful for (broken) code
> > that insists on being recursive.
>
> The current approach discussed is somewhat different from explicit
> more stack requests API. I am investigating how feasible it is to use
> kernel stack multiplexing, so the same pages can be re-used by many
> threads when they are actually used. If the multiplexing approach
> won't work, I will come back to the explicit more stack API.
>
> > Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
> > Registration No: 1397386 (Wales)

Powered by blists - more mailing lists