From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Jan Beulich <JBeulich@suse.com>
Cc: xen-devel <xen-devel@lists.xenproject.org>,
Keir Fraser <keir@xen.org>, Tim Deegan <tim@xen.org>
Subject: Re: [PATCH 2/2] x86: make hypercall preemption checks consistent
Date: Tue, 4 Mar 2014 11:35:46 +0000 [thread overview]
Message-ID: <5315BA92.8080902@citrix.com> (raw)
In-Reply-To: <5315C65E0200007800120CF7@nat28.tlf.novell.com>
[-- Attachment #1.1: Type: text/plain, Size: 5052 bytes --]
On 04/03/14 11:26, Jan Beulich wrote:
> - never preempt on the first iteration (ensure forward progress)
> - never preempt on the last iteration (pointless/wasteful)
>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
>
> --- a/xen/arch/x86/mm.c
> +++ b/xen/arch/x86/mm.c
> @@ -2934,7 +2934,7 @@ long do_mmuext_op(
>
> for ( i = 0; i < count; i++ )
> {
> - if ( curr->arch.old_guest_table || hypercall_preempt_check() )
> + if ( curr->arch.old_guest_table || (i && hypercall_preempt_check()) )
> {
> rc = -EAGAIN;
> break;
> @@ -3481,7 +3481,7 @@ long do_mmu_update(
>
> for ( i = 0; i < count; i++ )
> {
> - if ( curr->arch.old_guest_table || hypercall_preempt_check() )
> + if ( curr->arch.old_guest_table || (i && hypercall_preempt_check()) )
> {
> rc = -EAGAIN;
> break;
> --- a/xen/arch/x86/mm/hap/hap.c
> +++ b/xen/arch/x86/mm/hap/hap.c
> @@ -326,7 +326,7 @@ hap_set_allocation(struct domain *d, uns
> else
> pages -= d->arch.paging.hap.p2m_pages;
>
> - while ( d->arch.paging.hap.total_pages != pages )
> + for ( ; ; )
> {
> if ( d->arch.paging.hap.total_pages < pages )
> {
> @@ -355,6 +355,8 @@ hap_set_allocation(struct domain *d, uns
> d->arch.paging.hap.total_pages--;
> free_domheap_page(pg);
> }
> + else
> + break;
>
> /* Check to see if we need to yield and try again */
> if ( preempted && hypercall_preempt_check() )
> --- a/xen/arch/x86/mm/p2m-pod.c
> +++ b/xen/arch/x86/mm/p2m-pod.c
> @@ -242,7 +242,8 @@ p2m_pod_set_cache_target(struct p2m_doma
>
> p2m_pod_cache_add(p2m, page, order);
>
> - if ( hypercall_preempt_check() && preemptible )
> + if ( preemptible && pod_target != p2m->pod.count &&
> + hypercall_preempt_check() )
> {
> ret = -EAGAIN;
> goto out;
> @@ -286,7 +287,8 @@ p2m_pod_set_cache_target(struct p2m_doma
>
> put_page(page+i);
>
> - if ( hypercall_preempt_check() && preemptible )
> + if ( preemptible && pod_target != p2m->pod.count &&
> + hypercall_preempt_check() )
> {
> ret = -EAGAIN;
> goto out;
> --- a/xen/arch/x86/mm/shadow/common.c
> +++ b/xen/arch/x86/mm/shadow/common.c
> @@ -1672,7 +1672,7 @@ static unsigned int sh_set_allocation(st
> SHADOW_PRINTK("current %i target %i\n",
> d->arch.paging.shadow.total_pages, pages);
>
> - while ( d->arch.paging.shadow.total_pages != pages )
> + for ( ; ; )
> {
> if ( d->arch.paging.shadow.total_pages < pages )
> {
> @@ -1707,6 +1707,8 @@ static unsigned int sh_set_allocation(st
> d->arch.paging.shadow.total_pages--;
> free_domheap_page(sp);
> }
> + else
> + break;
>
> /* Check to see if we need to yield and try again */
> if ( preempted && hypercall_preempt_check() )
> --- a/xen/arch/x86/traps.c
> +++ b/xen/arch/x86/traps.c
> @@ -3596,13 +3596,6 @@ long do_set_trap_table(XEN_GUEST_HANDLE_
>
> for ( ; ; )
> {
> - if ( hypercall_preempt_check() )
> - {
> - rc = hypercall_create_continuation(
> - __HYPERVISOR_set_trap_table, "h", traps);
> - break;
> - }
> -
> if ( copy_from_guest(&cur, traps, 1) )
> {
> rc = -EFAULT;
> @@ -3623,6 +3616,13 @@ long do_set_trap_table(XEN_GUEST_HANDLE_
> init_int80_direct_trap(curr);
>
> guest_handle_add_offset(traps, 1);
> +
> + if ( hypercall_preempt_check() )
> + {
> + rc = hypercall_create_continuation(
> + __HYPERVISOR_set_trap_table, "h", traps);
> + break;
> + }
> }
>
> return rc;
> --- a/xen/arch/x86/x86_64/compat/traps.c
> +++ b/xen/arch/x86/x86_64/compat/traps.c
> @@ -329,13 +329,6 @@ int compat_set_trap_table(XEN_GUEST_HAND
>
> for ( ; ; )
> {
> - if ( hypercall_preempt_check() )
> - {
> - rc = hypercall_create_continuation(
> - __HYPERVISOR_set_trap_table, "h", traps);
> - break;
> - }
> -
> if ( copy_from_guest(&cur, traps, 1) )
> {
> rc = -EFAULT;
> @@ -353,6 +346,13 @@ int compat_set_trap_table(XEN_GUEST_HAND
> init_int80_direct_trap(current);
>
> guest_handle_add_offset(traps, 1);
> +
> + if ( hypercall_preempt_check() )
> + {
> + rc = hypercall_create_continuation(
> + __HYPERVISOR_set_trap_table, "h", traps);
> + break;
> + }
> }
>
> return rc;
>
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
[-- Attachment #1.2: Type: text/html, Size: 5825 bytes --]
[-- Attachment #2: Type: text/plain, Size: 126 bytes --]
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
next prev parent reply other threads:[~2014-03-04 11:35 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-03-04 11:21 [PATCH 0/2] make hypercall preemption checks consistent Jan Beulich
2014-03-04 11:24 ` [PATCH 1/2] common: " Jan Beulich
2014-03-04 11:29 ` Andrew Cooper
2014-03-04 11:26 ` [PATCH 2/2] x86: " Jan Beulich
2014-03-04 11:35 ` Andrew Cooper [this message]
2014-03-04 11:30 ` [PATCH 0/2] " Jan Beulich
2014-03-11 11:38 ` Ian Campbell
2014-03-04 11:46 ` Tim Deegan
2014-03-04 11:52 ` Andrew Cooper
2014-03-04 12:00 ` Jan Beulich
2014-03-04 12:10 ` David Vrabel
2014-03-04 13:06 ` Jan Beulich
2014-03-04 14:07 ` David Vrabel
2014-03-12 9:35 ` Ping: " Jan Beulich
2014-03-13 10:26 ` Keir Fraser
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=5315BA92.8080902@citrix.com \
--to=andrew.cooper3@citrix.com \
--cc=JBeulich@suse.com \
--cc=keir@xen.org \
--cc=tim@xen.org \
--cc=xen-devel@lists.xenproject.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;
as well as URLs for NNTP newsgroup(s).