Patchwork [v3] qemu/xendisk: set maximum number of grants to be used

login
register
mail settings
Submitter Jan Beulich
Date May 31, 2012, 11:23 a.m.
Message ID <4FC770E20200007800087298@nat28.tlf.novell.com>
Download mbox | patch
Permalink /patch/162140/
State New
Headers show

Comments

Jan Beulich - May 31, 2012, 11:23 a.m.
Legacy (non-pvops) gntdev drivers may require this to be done when the
number of grants intended to be used simultaneously exceeds a certain
driver specific default limit.

Change in v2: Double the number requested, as we need to account for
the allocations needing to happen in contiguous chunks. The worst case
number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
but in order to keep things simple just use 2 * max_req * max_seg.

Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
definition.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
qemu/xendisk: set maximum number of grants to be used

Legacy (non-pvops) gntdev drivers may require this to be done when the
number of grants intended to be used simultaneously exceeds a certain
driver specific default limit.

Change in v2: Double the number requested, as we need to account for
the allocations needing to happen in contiguous chunks. The worst case
number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
but in order to keep things simple just use 2 * max_req * max_seg.

Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
definition.

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

--- a/hw/xen_disk.c
+++ b/hw/xen_disk.c
@@ -537,6 +537,15 @@ static void blk_bh(void *opaque)
     blk_handle_requests(blkdev);
 }
 
+/*
+ * We need to account for the grant allocations requiring contiguous
+ * chunks; the worst case number would be
+ *     max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
+ * but in order to keep things simple just use
+ *     2 * max_req * max_seg.
+ */
+#define MAX_GRANTS(max_req, max_seg) (2 * (max_req) * (max_seg))
+
 static void blk_alloc(struct XenDevice *xendev)
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
@@ -548,6 +557,11 @@ static void blk_alloc(struct XenDevice *
     if (xen_mode != XEN_EMULATE) {
         batch_maps = 1;
     }
+    if (xc_gnttab_set_max_grants(xendev->gnttabdev,
+            MAX_GRANTS(max_requests, BLKIF_MAX_SEGMENTS_PER_REQUEST)) < 0) {
+        xen_be_printf(xendev, 0, "xc_gnttab_set_max_grants failed: %s\n",
+                      strerror(errno));
+    }
 }
 
 static int blk_init(struct XenDevice *xendev)
Stefano Stabellini - May 31, 2012, 11:27 a.m.
On Thu, 31 May 2012, Jan Beulich wrote:
> Legacy (non-pvops) gntdev drivers may require this to be done when the
> number of grants intended to be used simultaneously exceeds a certain
> driver specific default limit.
> 
> Change in v2: Double the number requested, as we need to account for
> the allocations needing to happen in contiguous chunks. The worst case
> number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
> but in order to keep things simple just use 2 * max_req * max_seg.
> 
> Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
> definition.
> 
> Signed-off-by: Jan Beulich <jbeulich@suse.com>

I think it is OK, I'll submit it as a bug fix for QEMU 1.1.

It should be applied to qemu-xen-traditional too.


> --- a/hw/xen_disk.c
> +++ b/hw/xen_disk.c
> @@ -537,6 +537,15 @@ static void blk_bh(void *opaque)
>      blk_handle_requests(blkdev);
>  }
>  
> +/*
> + * We need to account for the grant allocations requiring contiguous
> + * chunks; the worst case number would be
> + *     max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
> + * but in order to keep things simple just use
> + *     2 * max_req * max_seg.
> + */
> +#define MAX_GRANTS(max_req, max_seg) (2 * (max_req) * (max_seg))
> +
>  static void blk_alloc(struct XenDevice *xendev)
>  {
>      struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
> @@ -548,6 +557,11 @@ static void blk_alloc(struct XenDevice *
>      if (xen_mode != XEN_EMULATE) {
>          batch_maps = 1;
>      }
> +    if (xc_gnttab_set_max_grants(xendev->gnttabdev,
> +            MAX_GRANTS(max_requests, BLKIF_MAX_SEGMENTS_PER_REQUEST)) < 0) {
> +        xen_be_printf(xendev, 0, "xc_gnttab_set_max_grants failed: %s\n",
> +                      strerror(errno));
> +    }
>  }
>  
>  static int blk_init(struct XenDevice *xendev)
> 
> 
> 
>
Jan Beulich - May 31, 2012, 1:25 p.m.
>>> On 31.05.12 at 13:27, Stefano Stabellini <stefano.stabellini@eu.citrix.com>
wrote:
> On Thu, 31 May 2012, Jan Beulich wrote:
>> Legacy (non-pvops) gntdev drivers may require this to be done when the
>> number of grants intended to be used simultaneously exceeds a certain
>> driver specific default limit.
>> 
>> Change in v2: Double the number requested, as we need to account for
>> the allocations needing to happen in contiguous chunks. The worst case
>> number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
>> but in order to keep things simple just use 2 * max_req * max_seg.
>> 
>> Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
>> definition.
>> 
>> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> 
> I think it is OK, I'll submit it as a bug fix for QEMU 1.1.

Thanks!

> It should be applied to qemu-xen-traditional too.

And the other one ("qemu/xendisk: properly update stats in
ioreq_release()") probably too.

Jan
Stefano Stabellini - May 31, 2012, 1:32 p.m.
On Thu, 31 May 2012, Jan Beulich wrote:
> >>> On 31.05.12 at 13:27, Stefano Stabellini <stefano.stabellini@eu.citrix.com>
> wrote:
> > On Thu, 31 May 2012, Jan Beulich wrote:
> >> Legacy (non-pvops) gntdev drivers may require this to be done when the
> >> number of grants intended to be used simultaneously exceeds a certain
> >> driver specific default limit.
> >> 
> >> Change in v2: Double the number requested, as we need to account for
> >> the allocations needing to happen in contiguous chunks. The worst case
> >> number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
> >> but in order to keep things simple just use 2 * max_req * max_seg.
> >> 
> >> Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
> >> definition.
> >> 
> >> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> > 
> > I think it is OK, I'll submit it as a bug fix for QEMU 1.1.
> 
> Thanks!
> 
> > It should be applied to qemu-xen-traditional too.
> 
> And the other one ("qemu/xendisk: properly update stats in
> ioreq_release()") probably too.

Right.
That patch is already in upstream QEMU and qemu-xen-upstream but it is
missing in qemu-xen-traditional.
Ian Jackson - June 8, 2012, 1:52 p.m.
Stefano Stabellini writes ("Re: [Xen-devel] [PATCH, v3] qemu/xendisk: set maximum number of grants to be used"):
> On Thu, 31 May 2012, Jan Beulich wrote:
> > > It should be applied to qemu-xen-traditional too.
> > 
> > And the other one ("qemu/xendisk: properly update stats in
> > ioreq_release()") probably too.
> 
> Right.
> That patch is already in upstream QEMU and qemu-xen-upstream but it is
> missing in qemu-xen-traditional.

Can you tell me the commit id in qemu-xen-upstream so I can put it in
the commit message in qemu-xen-traditional ?

Thanks,
Ian.
Stefano Stabellini - June 11, 2012, 10:15 a.m.
On Fri, 8 Jun 2012, Ian Jackson wrote:
> Stefano Stabellini writes ("Re: [Xen-devel] [PATCH, v3] qemu/xendisk: set maximum number of grants to be used"):
> > On Thu, 31 May 2012, Jan Beulich wrote:
> > > > It should be applied to qemu-xen-traditional too.
> > > 
> > > And the other one ("qemu/xendisk: properly update stats in
> > > ioreq_release()") probably too.
> > 
> > Right.
> > That patch is already in upstream QEMU and qemu-xen-upstream but it is
> > missing in qemu-xen-traditional.
> 
> Can you tell me the commit id in qemu-xen-upstream so I can put it in
> the commit message in qemu-xen-traditional ?

I'll let you know when I have one.
Stefano Stabellini - June 11, 2012, 10:18 a.m.
On Mon, 11 Jun 2012, Stefano Stabellini wrote:
> On Fri, 8 Jun 2012, Ian Jackson wrote:
> > Stefano Stabellini writes ("Re: [Xen-devel] [PATCH, v3] qemu/xendisk: set maximum number of grants to be used"):
> > > On Thu, 31 May 2012, Jan Beulich wrote:
> > > > > It should be applied to qemu-xen-traditional too.
> > > > 
> > > > And the other one ("qemu/xendisk: properly update stats in
> > > > ioreq_release()") probably too.
> > > 
> > > Right.
> > > That patch is already in upstream QEMU and qemu-xen-upstream but it is
> > > missing in qemu-xen-traditional.
> > 
> > Can you tell me the commit id in qemu-xen-upstream so I can put it in
> > the commit message in qemu-xen-traditional ?
> 
> I'll let you know when I have one.

Reading more carefully, you are actually talking about a different
patch from the one in the subject line.
Here is what you are looking for:

commit ed5477664369c1e9de23b0e7e8f16a418573bd2a
Author: Jan Beulich <JBeulich@suse.com>
Date:   Mon May 14 16:46:33 2012 +0000

    xen_disk: properly update stats in ioreq_release()
    
    While for the "normal" case (called from blk_send_response_all())
    decrementing requests_finished is correct, doing so in the parse error
    case is wrong; requests_inflight needs to be decremented instead.
    
    Signed-off-by: Jan Beulich <jbeulich@suse.com>
    Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
    Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Jan Beulich - June 12, 2012, 7 a.m.
>>> On 31.05.12 at 13:27, Stefano Stabellini <stefano.stabellini@eu.citrix.com>
wrote:
> On Thu, 31 May 2012, Jan Beulich wrote:
>> Legacy (non-pvops) gntdev drivers may require this to be done when the
>> number of grants intended to be used simultaneously exceeds a certain
>> driver specific default limit.
>> 
>> Change in v2: Double the number requested, as we need to account for
>> the allocations needing to happen in contiguous chunks. The worst case
>> number would be max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
>> but in order to keep things simple just use 2 * max_req * max_seg.
>> 
>> Change in v3: introduce MAX_GRANTS(), and add a comment explaining its
>> definition.
>> 
>> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> 
> I think it is OK, I'll submit it as a bug fix for QEMU 1.1.
> 
> It should be applied to qemu-xen-traditional too.

This is commit 64c27e5b1fdb6d94bdc0bda3b1869d7383a35c65 in
git.qemu.org/?p=qemu.git.

Jan

>> --- a/hw/xen_disk.c
>> +++ b/hw/xen_disk.c
>> @@ -537,6 +537,15 @@ static void blk_bh(void *opaque)
>>      blk_handle_requests(blkdev);
>>  }
>>  
>> +/*
>> + * We need to account for the grant allocations requiring contiguous
>> + * chunks; the worst case number would be
>> + *     max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
>> + * but in order to keep things simple just use
>> + *     2 * max_req * max_seg.
>> + */
>> +#define MAX_GRANTS(max_req, max_seg) (2 * (max_req) * (max_seg))
>> +
>>  static void blk_alloc(struct XenDevice *xendev)
>>  {
>>      struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, 
> xendev);
>> @@ -548,6 +557,11 @@ static void blk_alloc(struct XenDevice *
>>      if (xen_mode != XEN_EMULATE) {
>>          batch_maps = 1;
>>      }
>> +    if (xc_gnttab_set_max_grants(xendev->gnttabdev,
>> +            MAX_GRANTS(max_requests, BLKIF_MAX_SEGMENTS_PER_REQUEST)) < 0) {
>> +        xen_be_printf(xendev, 0, "xc_gnttab_set_max_grants failed: %s\n",
>> +                      strerror(errno));
>> +    }
>>  }
>>  
>>  static int blk_init(struct XenDevice *xendev)
>> 
>> 
>> 
>>

Patch

--- a/hw/xen_disk.c
+++ b/hw/xen_disk.c
@@ -537,6 +537,15 @@  static void blk_bh(void *opaque)
     blk_handle_requests(blkdev);
 }
 
+/*
+ * We need to account for the grant allocations requiring contiguous
+ * chunks; the worst case number would be
+ *     max_req * max_seg + (max_req - 1) * (max_seg - 1) + 1,
+ * but in order to keep things simple just use
+ *     2 * max_req * max_seg.
+ */
+#define MAX_GRANTS(max_req, max_seg) (2 * (max_req) * (max_seg))
+
 static void blk_alloc(struct XenDevice *xendev)
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
@@ -548,6 +557,11 @@  static void blk_alloc(struct XenDevice *
     if (xen_mode != XEN_EMULATE) {
         batch_maps = 1;
     }
+    if (xc_gnttab_set_max_grants(xendev->gnttabdev,
+            MAX_GRANTS(max_requests, BLKIF_MAX_SEGMENTS_PER_REQUEST)) < 0) {
+        xen_be_printf(xendev, 0, "xc_gnttab_set_max_grants failed: %s\n",
+                      strerror(errno));
+    }
 }
 
 static int blk_init(struct XenDevice *xendev)