From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Juergen Gross <jgross@suse.com>,
David Vrabel <david.vrabel@citrix.com>,
<linux-kernel@vger.kernel.org>, <xen-devel@lists.xensource.com>,
<konrad.wilk@oracle.com>, <boris.ostrovsky@oracle.com>
Subject: Re: [Xen-devel] [PATCH 02/13] xen: anchor linear p2m list in shared info structure
Date: Wed, 18 Feb 2015 10:50:09 +0000 [thread overview]
Message-ID: <54E46E61.8030805@citrix.com> (raw)
In-Reply-To: <54E46CB1.5010702@suse.com>
On 18/02/15 10:42, Juergen Gross wrote:
>
>>> /* Set up p2m_top to point to the domain-builder provided p2m
>>> pages */
>>> @@ -469,8 +473,10 @@ static pte_t *alloc_p2m_pmd(unsigned long addr,
>>> pte_t *pte_pg)
>>>
>>> ptechk = lookup_address(vaddr, &level);
>>> if (ptechk == pte_pg) {
>>> + HYPERVISOR_shared_info->arch.p2m_generation++;
>>> set_pmd(pmdp,
>>> __pmd(__pa(pte_newpg[i]) | _KERNPG_TABLE));
>>> + HYPERVISOR_shared_info->arch.p2m_generation++;
>>
>> Do these increments of p2m_generation need to be atomic?
>
> Hmm, they are done under lock. I don't think the compiler is allowed to
> reorder the writes to p2m_generation across set_pmd().
They do need smp_wmb() to guarantee that the increment is visible before
the update occurs, just as the toolstack will need smp_rmb() to read.
They also need to be protected from concurrent update inside the kernel,
for which a lock should appear to suffice.
~Andrew
WARNING: multiple messages have this Message-ID (diff)
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Juergen Gross <jgross@suse.com>,
David Vrabel <david.vrabel@citrix.com>,
linux-kernel@vger.kernel.org, xen-devel@lists.xensource.com,
konrad.wilk@oracle.com, boris.ostrovsky@oracle.com
Subject: Re: [Xen-devel] [PATCH 02/13] xen: anchor linear p2m list in shared info structure
Date: Wed, 18 Feb 2015 10:50:09 +0000 [thread overview]
Message-ID: <54E46E61.8030805@citrix.com> (raw)
In-Reply-To: <54E46CB1.5010702@suse.com>
On 18/02/15 10:42, Juergen Gross wrote:
>
>>> /* Set up p2m_top to point to the domain-builder provided p2m
>>> pages */
>>> @@ -469,8 +473,10 @@ static pte_t *alloc_p2m_pmd(unsigned long addr,
>>> pte_t *pte_pg)
>>>
>>> ptechk = lookup_address(vaddr, &level);
>>> if (ptechk == pte_pg) {
>>> + HYPERVISOR_shared_info->arch.p2m_generation++;
>>> set_pmd(pmdp,
>>> __pmd(__pa(pte_newpg[i]) | _KERNPG_TABLE));
>>> + HYPERVISOR_shared_info->arch.p2m_generation++;
>>
>> Do these increments of p2m_generation need to be atomic?
>
> Hmm, they are done under lock. I don't think the compiler is allowed to
> reorder the writes to p2m_generation across set_pmd().
They do need smp_wmb() to guarantee that the increment is visible before
the update occurs, just as the toolstack will need smp_rmb() to read.
They also need to be protected from concurrent update inside the kernel,
for which a lock should appear to suffice.
~Andrew
next prev parent reply other threads:[~2015-02-18 10:50 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-18 6:51 [PATCH 00/13] xen: support pv-domains larger than 512GB Juergen Gross
2015-02-18 6:51 ` [PATCH 01/13] xen: sync with xen header Juergen Gross
2015-02-18 6:51 ` [PATCH 02/13] xen: anchor linear p2m list in shared info structure Juergen Gross
2015-02-18 10:32 ` [Xen-devel] " David Vrabel
2015-02-18 10:32 ` David Vrabel
2015-02-18 10:42 ` Juergen Gross
2015-02-18 10:50 ` Andrew Cooper [this message]
2015-02-18 10:50 ` Andrew Cooper
2015-02-18 10:54 ` David Vrabel
2015-02-18 10:54 ` David Vrabel
2015-02-18 10:57 ` Andrew Cooper
2015-02-18 10:57 ` Andrew Cooper
2015-02-18 10:57 ` Juergen Gross
2015-02-18 10:56 ` Juergen Gross
2015-02-18 10:51 ` David Vrabel
2015-02-18 10:51 ` David Vrabel
2015-02-18 6:51 ` [PATCH 03/13] xen: eliminate scalability issues from initial mapping setup Juergen Gross
2015-02-18 6:51 ` [PATCH 04/13] xen: move static e820 map to global scope Juergen Gross
2015-02-19 17:27 ` [Xen-devel] " David Vrabel
2015-02-18 6:51 ` [PATCH 05/13] xen: simplify xen_set_identity_and_remap() by using global variables Juergen Gross
2015-02-19 18:10 ` [Xen-devel] " David Vrabel
2015-02-24 6:15 ` Juergen Gross
2015-02-18 6:51 ` [PATCH 06/13] xen: detect pre-allocated memory interfering with e820 map Juergen Gross
2015-02-19 18:07 ` [Xen-devel] " David Vrabel
2015-02-24 6:27 ` Juergen Gross
2015-02-25 14:24 ` David Vrabel
2015-02-25 14:24 ` David Vrabel
2015-02-25 16:00 ` Juergen Gross
2015-03-30 10:00 ` Juergen Gross
2015-02-18 6:52 ` [PATCH 07/13] xen: find unused contiguous memory area Juergen Gross
2015-02-19 17:31 ` [Xen-devel] " David Vrabel
2015-02-18 6:52 ` [PATCH 08/13] xen: add service function to copy physical memory areas Juergen Gross
2015-02-19 17:35 ` [Xen-devel] " David Vrabel
2015-02-24 6:34 ` Juergen Gross
2015-02-18 6:52 ` [PATCH 09/13] xen: check for kernel memory conflicting with memory layout Juergen Gross
2015-02-19 17:36 ` [Xen-devel] " David Vrabel
2015-02-24 6:45 ` Juergen Gross
2015-02-18 6:52 ` [PATCH 10/13] xen: check pre-allocated page tables for conflict with memory map Juergen Gross
2015-02-19 17:37 ` [Xen-devel] " David Vrabel
2015-02-24 6:45 ` Juergen Gross
2015-02-18 6:52 ` [PATCH 11/13] xen: move initrd away from e820 non-ram area Juergen Gross
2015-02-19 17:42 ` [Xen-devel] " David Vrabel
2015-02-18 6:52 ` [PATCH 12/13] xen: if p2m list located in to be remapped region delay remapping Juergen Gross
2015-02-19 17:44 ` [Xen-devel] " David Vrabel
2015-02-24 7:01 ` Juergen Gross
2015-02-18 6:52 ` [PATCH 13/13] xen: allow more than 512 GB of RAM for 64 bit pv-domains Juergen Gross
2015-02-18 9:21 ` Paul Bolle
2015-02-18 9:37 ` Juergen Gross
2015-02-18 9:49 ` [Xen-devel] " Jan Beulich
2015-02-18 9:49 ` Jan Beulich
[not found] ` <54E46E3C0200007800060F98@suse.com>
2015-02-18 9:59 ` Juergen Gross
2015-02-18 10:35 ` Paul Bolle
2015-02-18 11:18 ` [Xen-devel] " David Vrabel
2015-02-18 11:18 ` David Vrabel
2015-02-18 11:51 ` Juergen Gross
2015-02-18 11:56 ` Andrew Cooper
2015-02-19 17:51 ` David Vrabel
2015-02-24 7:28 ` Juergen Gross
2015-02-18 11:53 ` Juergen Gross
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=54E46E61.8030805@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=boris.ostrovsky@oracle.com \
--cc=david.vrabel@citrix.com \
--cc=jgross@suse.com \
--cc=konrad.wilk@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=xen-devel@lists.xensource.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.