[PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[Jan Beulich]

[-- Attachment #1: Type: text/plain, Size: 947 bytes --]

Instead saturate the value if the input can't be represented in the
respective struct domain field.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

--- a/xen/common/domctl.c
+++ b/xen/common/domctl.c
@@ -944,7 +944,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
 
     case XEN_DOMCTL_max_mem:
     {
-        unsigned long new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
+        uint64_t new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
 
         spin_lock(&d->page_alloc_lock);
         /*
@@ -952,7 +952,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
          * that the domain will now be allowed to "ratchet" down to new_max. In
          * the meantime, while tot > max, all new allocations are disallowed.
          */
-        d->max_pages = new_max;
+        d->max_pages = min(new_max, (uint64_t)(typeof(d->max_pages))-1);
         spin_unlock(&d->page_alloc_lock);
         break;
     }




[-- Attachment #2: domctl-max-mem-saturation.patch --]
[-- Type: text/plain, Size: 995 bytes --]

domctl: don't truncate XEN_DOMCTL_max_mem requests

Instead saturate the value if the input can't be represented in the
respective struct domain field.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

--- a/xen/common/domctl.c
+++ b/xen/common/domctl.c
@@ -944,7 +944,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
 
     case XEN_DOMCTL_max_mem:
     {
-        unsigned long new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
+        uint64_t new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
 
         spin_lock(&d->page_alloc_lock);
         /*
@@ -952,7 +952,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
          * that the domain will now be allowed to "ratchet" down to new_max. In
          * the meantime, while tot > max, all new allocations are disallowed.
          */
-        d->max_pages = new_max;
+        d->max_pages = min(new_max, (uint64_t)(typeof(d->max_pages))-1);
         spin_unlock(&d->page_alloc_lock);
         break;
     }

[-- Attachment #3: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[Andrew Cooper]

[-- Attachment #1.1: Type: text/plain, Size: 1249 bytes --]

On 22/04/15 12:56, Jan Beulich wrote:
> Instead saturate the value if the input can't be represented in the
> respective struct domain field.
>
> Signed-off-by: Jan Beulich <jbeulich@suse.com>

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>

Do I take it you have a 16TB guest to hand?

>
> --- a/xen/common/domctl.c
> +++ b/xen/common/domctl.c
> @@ -944,7 +944,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
>  
>      case XEN_DOMCTL_max_mem:
>      {
> -        unsigned long new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
> +        uint64_t new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
>  
>          spin_lock(&d->page_alloc_lock);
>          /*
> @@ -952,7 +952,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
>           * that the domain will now be allowed to "ratchet" down to new_max. In
>           * the meantime, while tot > max, all new allocations are disallowed.
>           */
> -        d->max_pages = new_max;
> +        d->max_pages = min(new_max, (uint64_t)(typeof(d->max_pages))-1);
>          spin_unlock(&d->page_alloc_lock);
>          break;
>      }
>
>
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel


[-- Attachment #1.2: Type: text/html, Size: 2185 bytes --]

[-- Attachment #2: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[Jan Beulich]

>>> On 22.04.15 at 14:03, <andrew.cooper3@citrix.com> wrote:
> On 22/04/15 12:56, Jan Beulich wrote:
>> Instead saturate the value if the input can't be represented in the
>> respective struct domain field.
>>
>> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> 
> Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
> 
> Do I take it you have a 16TB guest to hand?

No, but see the subsequently sent libx{l,c} patch for why this is
needed anyway.

Jan

Ping: [PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[Jan Beulich]

While I got a review from Andrew, I'm still waiting for an ack (or nak).

Thanks, Jan

>>> On 22.04.15 at 13:56, <JBeulich@suse.com> wrote:
> Instead saturate the value if the input can't be represented in the
> respective struct domain field.
> 
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> 
> --- a/xen/common/domctl.c
> +++ b/xen/common/domctl.c
> @@ -944,7 +944,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
>  
>      case XEN_DOMCTL_max_mem:
>      {
> -        unsigned long new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
> +        uint64_t new_max = op->u.max_mem.max_memkb >> (PAGE_SHIFT - 10);
>  
>          spin_lock(&d->page_alloc_lock);
>          /*
> @@ -952,7 +952,7 @@ long do_domctl(XEN_GUEST_HANDLE_PARAM(xe
>           * that the domain will now be allowed to "ratchet" down to new_max. In
>           * the meantime, while tot > max, all new allocations are disallowed.
>           */
> -        d->max_pages = new_max;
> +        d->max_pages = min(new_max, (uint64_t)(typeof(d->max_pages))-1);
>          spin_unlock(&d->page_alloc_lock);
>          break;
>      }

Re: Ping: [PATCH] domctl: don't truncate XEN_DOMCTL_max_mem requests

[Tim Deegan]

At 08:49 +0100 on 04 May (1430729386), Jan Beulich wrote:
> While I got a review from Andrew, I'm still waiting for an ack (or nak).

Acked-by: Tim Deegan <tim@xen.org>

I expect Coverity will complain on architectures where this isn't
necessary, but acked nonetheless.

Tim.