From: Linus Torvalds <torvalds@linux-foundation.org>
To: Borislav Petkov <bp@alien8.de>
Cc: Markus Trippelsdorf <markus@trippelsdorf.de>,
Andy Lutomirski <luto@kernel.org>, Ingo Molnar <mingo@kernel.org>,
Thomas Gleixner <tglx@linutronix.de>,
Peter Zijlstra <peterz@infradead.org>,
LKML <linux-kernel@vger.kernel.org>,
Ingo Molnar <mingo@redhat.com>,
Tom Lendacky <thomas.lendacky@amd.com>
Subject: Re: Current mainline git (24e700e291d52bd2) hangs when building e.g. perf
Date: Sat, 9 Sep 2017 11:47:33 -0700 [thread overview]
Message-ID: <CA+55aFzxsKzuYKMCXQA+54GUhRzEMCtwYWdj+RTDv7LHUhVj4Q@mail.gmail.com> (raw)
In-Reply-To: <20170909182952.itqad4ryngjwrgqf@pd.tnic>
On Sat, Sep 9, 2017 at 11:29 AM, Borislav Petkov <bp@alien8.de> wrote:
> On Sat, Sep 09, 2017 at 11:26:27AM -0700, Linus Torvalds wrote:
>> But the fact that that fixes it for you does indicate that it's not
>> just a stale TLB entry or something, it really is some CPU using page
>> tables after they have been free'd and been re-allocated to something
>> else (and *then* they may point to garbage).
>
> Cool, I was trying to think of a good use case how we'd hit that. I
> guess you just gave one. :)
The thing is, even with the delayed TLB flushing, I don't think it
should be *so* delayed that we should be seeing a TLB fill from
garbage page tables.
But the part in Andy's patch that worries me the most is that
+ cpumask_clear_cpu(cpu, mm_cpumask(mm));
in enter_lazy_tlb(). It means that we won't be notified by peopel
invalidating the page tables, and while we then do re-validate the TLB
when we switch back from lazy mode, I still worry. I'm not entirely
convinced by that tlb_gen logic.
I can't actually see anything *wrong* in the tlb_gen logic, but it worries me.
Linus
next prev parent reply other threads:[~2017-09-09 18:47 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-05 7:27 Current mainline git (24e700e291d52bd2) hangs when building e.g. perf Markus Trippelsdorf
2017-09-05 8:53 ` Peter Zijlstra
2017-09-05 9:55 ` Markus Trippelsdorf
2017-09-06 12:52 ` Thomas Gleixner
2017-09-06 13:15 ` Markus Trippelsdorf
2017-09-07 6:28 ` Markus Trippelsdorf
2017-09-08 5:35 ` Markus Trippelsdorf
2017-09-08 6:26 ` Thomas Gleixner
2017-09-08 8:05 ` Borislav Petkov
2017-09-08 9:16 ` Borislav Petkov
2017-09-08 9:48 ` Markus Trippelsdorf
2017-09-08 10:35 ` Ingo Molnar
2017-09-08 10:39 ` Markus Trippelsdorf
2017-09-08 11:30 ` Markus Trippelsdorf
2017-09-08 16:12 ` Andy Lutomirski
2017-09-08 17:16 ` Markus Trippelsdorf
2017-09-08 21:47 ` Andy Lutomirski
2017-09-08 21:56 ` Borislav Petkov
2017-09-08 23:07 ` Andy Lutomirski
2017-09-08 23:23 ` Linus Torvalds
2017-09-09 0:00 ` Andy Lutomirski
2017-09-09 1:05 ` Linus Torvalds
2017-09-09 1:39 ` Andy Lutomirski
2017-09-09 17:49 ` Andy Lutomirski
2017-09-09 18:02 ` Linus Torvalds
2017-09-09 6:39 ` Markus Trippelsdorf
2017-09-09 10:18 ` Borislav Petkov
2017-09-09 11:07 ` Markus Trippelsdorf
2017-09-09 13:07 ` Borislav Petkov
2017-09-09 13:37 ` Markus Trippelsdorf
2017-09-09 13:39 ` Markus Trippelsdorf
2017-09-09 14:07 ` Borislav Petkov
2017-09-09 14:20 ` Markus Trippelsdorf
2017-09-09 14:33 ` Borislav Petkov
2017-09-09 14:43 ` Markus Trippelsdorf
2017-09-09 16:32 ` Markus Trippelsdorf
2017-09-09 17:05 ` Borislav Petkov
2017-09-09 17:23 ` Markus Trippelsdorf
2017-09-09 17:36 ` Borislav Petkov
2017-09-09 18:14 ` Markus Trippelsdorf
2017-09-09 18:26 ` Borislav Petkov
2017-09-09 18:46 ` Markus Trippelsdorf
2017-09-09 19:11 ` Borislav Petkov
2017-09-09 19:19 ` Borislav Petkov
2017-09-09 18:26 ` Linus Torvalds
2017-09-09 18:29 ` Borislav Petkov
2017-09-09 18:47 ` Linus Torvalds [this message]
2017-09-09 19:09 ` Borislav Petkov
2017-09-09 19:21 ` Linus Torvalds
2017-09-09 19:28 ` Andy Lutomirski
2017-09-09 19:37 ` Borislav Petkov
2017-09-10 4:42 ` Andy Lutomirski
2017-09-10 20:22 ` Peter Zijlstra
2017-09-10 20:25 ` Andy Lutomirski
2017-09-17 17:04 ` Ingo Molnar
2017-09-11 1:12 ` Rik van Riel
2017-09-11 1:46 ` Andy Lutomirski
2017-09-11 15:08 ` Rik van Riel
2017-09-12 7:14 ` Markus Trippelsdorf
2017-09-09 8:13 ` Markus Trippelsdorf
2017-09-08 14:51 ` Borislav Petkov
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+55aFzxsKzuYKMCXQA+54GUhRzEMCtwYWdj+RTDv7LHUhVj4Q@mail.gmail.com \
--to=torvalds@linux-foundation.org \
--cc=bp@alien8.de \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=markus@trippelsdorf.de \
--cc=mingo@kernel.org \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=thomas.lendacky@amd.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;
as well as URLs for NNTP newsgroup(s).