From: Linus Torvalds <torvalds@linux-foundation.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org,
linux-mm@kvack.org, Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@elte.hu>,
akpm@linux-foundation.org, Rik van Riel <riel@redhat.com>,
Hugh Dickins <hugh.dickins@tiscali.co.uk>,
Mel Gorman <mel@csn.ul.ie>, Nick Piggin <npiggin@kernel.dk>,
Alex Shi <alex.shi@intel.com>,
"Nikunj A. Dadhania" <nikunj@linux.vnet.ibm.com>,
Konrad Rzeszutek Wilk <konrad@darnok.org>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
David Miller <davem@davemloft.net>,
Russell King <rmk@arm.linux.org.uk>,
Catalin Marinas <catalin.marinas@arm.com>,
Chris Metcalf <cmetcalf@tilera.com>,
Martin Schwidefsky <schwidefsky@de.ibm.com>,
Tony Luck <tony.luck@intel.com>, Paul Mundt <lethal@linux-sh.org>,
Jeff Dike <jdike@addtoit.com>,
Richard Weinberger <richard@nod.at>,
Ralf Baechle <ralf@linux-mips.org>,
Kyle McMartin <kyle@mcmartin.ca>,
James Bottomley <jejb@parisc-linux.org>,
Chris Zankel <chris@zankel.net>
Subject: Re: [PATCH 08/20] mm: Optimize fullmm TLB flushing
Date: Wed, 27 Jun 2012 16:23:30 -0700 [thread overview]
Message-ID: <CA+55aFy6m967fMxyBsRoXVecdpGtSphXi_XdhwS0DB81Qaocdw@mail.gmail.com> (raw)
In-Reply-To: <1340838807.10063.90.camel@twins>
On Wed, Jun 27, 2012 at 4:13 PM, Peter Zijlstra <peterz@infradead.org> wrote:
>
> That triggered.. is this a problem though, at this point userspace is
> very dead so it shouldn't matter, right?
It still matters. Even if user space is dead, kernel space accesses
can result in TLB fills in user space. Exactly because of things like
speculative fills etc.
So what can happen - for example - is that the kernel does a indirect
jump, and the CPU predicts the destination of the jump that using the
branch prediction tables.
But the branch prediction tables are obviously just predictions, and
they easily contain user addresses etc in them. So the kernel may well
end up speculatively doing a TLB fill on a user access.
And your whole optimization depends on this not happening, unless I
read the logic wrong. The whole "invalidate the TLB just once
up-front" approach is *only* valid if you know that nothing is going
to ever fill that TLB again. But see above - if we're still running
within that TLB context, we have no idea what speculative execution
may or may not end up filling.
That said, maybe I misread your patch?
Linus
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2012-06-27 23:23 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-06-27 21:15 [PATCH 00/20] Unify TLB gather implementations -v3 Peter Zijlstra
2012-06-27 21:15 ` [PATCH 01/20] mm, x86: Add HAVE_RCU_TABLE_FREE support Peter Zijlstra
2012-06-27 21:15 ` [PATCH 02/20] mm: Add optional TLB flush to generic RCU page-table freeing Peter Zijlstra
2012-06-27 22:23 ` Linus Torvalds
2012-06-27 23:01 ` Peter Zijlstra
2012-06-27 23:42 ` Linus Torvalds
2012-06-28 7:09 ` Benjamin Herrenschmidt
2012-06-28 11:05 ` Peter Zijlstra
2012-06-28 12:00 ` Benjamin Herrenschmidt
2012-07-24 5:12 ` Nikunj A Dadhania
2012-06-27 21:15 ` [PATCH 03/20] mm, tlb: Remove a few #ifdefs Peter Zijlstra
2012-06-27 21:15 ` [PATCH 04/20] mm, s390: use generic RCU page-table freeing code Peter Zijlstra
2012-06-27 21:15 ` [PATCH 05/20] mm, powerpc: Dont use tlb_flush for external tlb flushes Peter Zijlstra
2012-06-27 21:15 ` [PATCH 06/20] mm, sparc64: " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 07/20] mm, arch: Remove tlb_flush() Peter Zijlstra
2012-06-27 21:15 ` [PATCH 08/20] mm: Optimize fullmm TLB flushing Peter Zijlstra
2012-06-27 22:26 ` Linus Torvalds
2012-06-27 23:02 ` Peter Zijlstra
2012-06-27 23:13 ` Peter Zijlstra
2012-06-27 23:23 ` Linus Torvalds [this message]
2012-06-27 23:33 ` Linus Torvalds
2012-06-28 9:16 ` Catalin Marinas
2012-06-28 10:39 ` Benjamin Herrenschmidt
2012-06-28 10:59 ` Peter Zijlstra
2012-06-28 14:53 ` Catalin Marinas
2012-06-28 16:20 ` Peter Zijlstra
2012-06-28 16:38 ` Peter Zijlstra
2012-06-28 16:45 ` Linus Torvalds
2012-06-28 16:52 ` Peter Zijlstra
2012-06-28 21:57 ` Benjamin Herrenschmidt
2012-06-28 21:58 ` Benjamin Herrenschmidt
2012-06-29 8:49 ` Peter Zijlstra
2012-06-29 15:26 ` Catalin Marinas
2012-06-29 22:11 ` Benjamin Herrenschmidt
2012-06-28 10:55 ` Peter Zijlstra
2012-06-28 11:19 ` Martin Schwidefsky
2012-06-28 11:30 ` Peter Zijlstra
2012-06-28 16:00 ` Avi Kivity
2012-06-27 21:15 ` [PATCH 09/20] mm, arch: Add end argument to p??_free_tlb() Peter Zijlstra
2012-06-27 21:15 ` [PATCH 10/20] mm: Provide generic range tracking and flushing Peter Zijlstra
2012-06-27 21:15 ` [PATCH 11/20] mm, s390: Convert to use generic mmu_gather Peter Zijlstra
2012-06-27 22:13 ` Peter Zijlstra
2012-06-28 7:13 ` Martin Schwidefsky
2012-06-27 21:15 ` [PATCH 12/20] mm, arm: Convert arm to generic tlb Peter Zijlstra
2012-06-27 21:15 ` [PATCH 13/20] mm, ia64: Convert ia64 " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 14/20] mm, sh: Convert sh " Peter Zijlstra
2012-06-28 18:32 ` Paul Mundt
2012-06-28 20:27 ` Peter Zijlstra
2012-06-27 21:15 ` [PATCH 15/20] mm, um: Convert um " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 16/20] mm, avr32: Convert avr32 " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 17/20] mm, mips: Convert mips " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 18/20] mm, parisc: Convert parisc " Peter Zijlstra
2012-06-27 21:15 ` [PATCH 19/20] mm, sparc32: Convert sparc32 " Peter Zijlstra
2012-06-27 21:16 ` [PATCH 20/20] mm, xtensa: Convert xtensa " Peter Zijlstra
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=CA+55aFy6m967fMxyBsRoXVecdpGtSphXi_XdhwS0DB81Qaocdw@mail.gmail.com \
--to=torvalds@linux-foundation.org \
--cc=akpm@linux-foundation.org \
--cc=alex.shi@intel.com \
--cc=benh@kernel.crashing.org \
--cc=catalin.marinas@arm.com \
--cc=chris@zankel.net \
--cc=cmetcalf@tilera.com \
--cc=davem@davemloft.net \
--cc=hugh.dickins@tiscali.co.uk \
--cc=jdike@addtoit.com \
--cc=jejb@parisc-linux.org \
--cc=konrad@darnok.org \
--cc=kyle@mcmartin.ca \
--cc=lethal@linux-sh.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mel@csn.ul.ie \
--cc=mingo@elte.hu \
--cc=nikunj@linux.vnet.ibm.com \
--cc=npiggin@kernel.dk \
--cc=peterz@infradead.org \
--cc=ralf@linux-mips.org \
--cc=richard@nod.at \
--cc=riel@redhat.com \
--cc=rmk@arm.linux.org.uk \
--cc=schwidefsky@de.ibm.com \
--cc=tglx@linutronix.de \
--cc=tony.luck@intel.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