From: Thomas Gleixner <tglx@linutronix.de>
To: Dave Hansen <dave.hansen@intel.com>
Cc: Qiaowei Ren <qiaowei.ren@intel.com>,
"H. Peter Anvin" <hpa@zytor.com>, Ingo Molnar <mingo@redhat.com>,
x86@kernel.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org,
linux-ia64@vger.kernel.org, linux-mips@linux-mips.org
Subject: Re: [PATCH v9 05/12] x86, mpx: on-demand kernel allocation of bounds tables
Date: Tue, 28 Oct 2014 18:57:46 +0100 (CET) [thread overview]
Message-ID: <alpine.DEB.2.11.1410281851390.5308@nanos> (raw)
In-Reply-To: <544FD5D4.4090404@intel.com>
On Tue, 28 Oct 2014, Dave Hansen wrote:
> On 10/24/2014 05:08 AM, Thomas Gleixner wrote:
> > On Sun, 12 Oct 2014, Qiaowei Ren wrote:
> >> + /*
> >> + * Go poke the address of the new bounds table in to the
> >> + * bounds directory entry out in userspace memory. Note:
> >> + * we may race with another CPU instantiating the same table.
> >> + * In that case the cmpxchg will see an unexpected
> >> + * 'actual_old_val'.
> >> + */
> >> + ret = user_atomic_cmpxchg_inatomic(&actual_old_val, bd_entry,
> >> + expected_old_val, bt_addr);
> >
> > This is fully preemptible non-atomic context, right?
> >
> > So this wants a proper comment, why using
> > user_atomic_cmpxchg_inatomic() is the right thing to do here.
>
> Hey Thomas,
>
> How's this for a new comment? Does this cover the points you think need
> clarified?
>
> ====
>
> The kernel has allocated a bounds table and needs to point the
> (userspace-allocated) directory to it. The directory entry is the
> *only* place we track that this table was allocated, so we essentially
> use it instead of an kernel data structure for synchronization. A
> copy_to_user()-style function would not give us the atomicity that we need.
>
> If two threads race to instantiate a table, the cmpxchg ensures we know
> which one lost the race and that the loser frees the table that they
> just allocated.
Yup. That explains the cmpxchg.
The other thing which puzzled me was that it calls
user_atomic_cmpxchg_inatomic() but the context is not atomic at
all. Its fully preemptible and actually we want it to be able to
handle the fault. The implementation does that, just the function
itself suggest something different.
Thanks,
tglx
--
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:[~2014-10-28 17:57 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-12 4:41 [PATCH v9 00/12] Intel MPX support Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 01/12] x86, mpx: introduce VM_MPX to indicate that a VMA is MPX specific Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 02/12] x86, mpx: rename cfg_reg_u and status_reg Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 03/12] x86, mpx: add MPX specific mmap interface Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 04/12] x86, mpx: add MPX to disaabled features Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 05/12] x86, mpx: on-demand kernel allocation of bounds tables Qiaowei Ren
2014-10-24 12:08 ` Thomas Gleixner
2014-10-27 3:20 ` Ren Qiaowei
2014-10-28 17:43 ` Dave Hansen
2014-10-28 17:57 ` Thomas Gleixner [this message]
2014-10-12 4:41 ` [PATCH v9 06/12] mpx: extend siginfo structure to include bound violation information Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 07/12] mips: sync struct siginfo with general version Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 08/12] ia64: " Qiaowei Ren
2014-10-12 4:41 ` [PATCH v9 09/12] x86, mpx: decode MPX instruction to get bound violation information Qiaowei Ren
2014-10-24 12:36 ` Thomas Gleixner
2014-10-27 1:43 ` Ren, Qiaowei
2014-10-27 20:36 ` Thomas Gleixner
2014-10-28 5:58 ` Ren Qiaowei
2014-10-31 20:16 ` Dave Hansen
2014-10-31 20:33 ` Thomas Gleixner
2014-10-30 22:38 ` Dave Hansen
2014-10-31 2:12 ` Ren Qiaowei
2014-10-31 9:09 ` Thomas Gleixner
2014-10-12 4:41 ` [PATCH v9 10/12] x86, mpx: add prctl commands PR_MPX_ENABLE_MANAGEMENT, PR_MPX_DISABLE_MANAGEMENT Qiaowei Ren
2014-10-24 12:49 ` Thomas Gleixner
2014-10-24 15:10 ` Thomas Gleixner
2014-10-27 2:17 ` Ren, Qiaowei
2014-10-27 20:38 ` Thomas Gleixner
2014-10-28 5:57 ` Ren Qiaowei
2014-10-12 4:41 ` [PATCH v9 11/12] x86, mpx: cleanup unused bound tables Qiaowei Ren
2014-10-24 14:40 ` Thomas Gleixner
2014-10-27 3:13 ` Ren Qiaowei
2014-10-27 20:49 ` Thomas Gleixner
2014-10-28 5:56 ` Ren Qiaowei
2014-10-28 10:42 ` Thomas Gleixner
2014-11-03 20:53 ` Dave Hansen
2014-11-03 21:29 ` Thomas Gleixner
2014-11-04 16:00 ` Dave Hansen
2014-11-04 17:02 ` Thomas Gleixner
2014-11-06 21:50 ` Dave Hansen
2014-11-11 18:27 ` Thomas Gleixner
2014-11-11 20:44 ` Dave Hansen
2014-11-11 21:36 ` Thomas Gleixner
2014-10-12 4:41 ` [PATCH v9 12/12] x86, mpx: add documentation on Intel MPX Qiaowei Ren
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=alpine.DEB.2.11.1410281851390.5308@nanos \
--to=tglx@linutronix.de \
--cc=dave.hansen@intel.com \
--cc=hpa@zytor.com \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@linux-mips.org \
--cc=linux-mm@kvack.org \
--cc=mingo@redhat.com \
--cc=qiaowei.ren@intel.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