All of lore.kernel.org
 help / color / mirror / Atom feed
* [Qemu-devel] [PATCH] memory: Don't call memory_region_update_coalesced_range if nothing changed
@ 2014-06-13  6:34 Fam Zheng
  2014-06-13  7:46 ` Paolo Bonzini
  0 siblings, 1 reply; 2+ messages in thread
From: Fam Zheng @ 2014-06-13  6:34 UTC (permalink / raw)
  To: qemu-devel; +Cc: Paolo Bonzini, stefanha

With huge number of PCI devices in the system (for example, 200
virtio-blk-pci), this unconditional call can slow down emulation of
irrelevant PCI operations drastically, such as a BAR update on a device
that has no coalescing region. So avoid it.

Signed-off-by: Fam Zheng <famz@redhat.com>
---
 memory.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/memory.c b/memory.c
index d8ea212..57227f4 100644
--- a/memory.c
+++ b/memory.c
@@ -1324,6 +1324,7 @@ void memory_region_add_coalescing(MemoryRegion *mr,
 void memory_region_clear_coalescing(MemoryRegion *mr)
 {
     CoalescedMemoryRange *cmr;
+    bool updated = false;
 
     qemu_flush_coalesced_mmio_buffer();
     mr->flush_coalesced_mmio = false;
@@ -1332,8 +1333,12 @@ void memory_region_clear_coalescing(MemoryRegion *mr)
         cmr = QTAILQ_FIRST(&mr->coalesced);
         QTAILQ_REMOVE(&mr->coalesced, cmr, link);
         g_free(cmr);
+        updated = true;
+    }
+
+    if (updated) {
+        memory_region_update_coalesced_range(mr);
     }
-    memory_region_update_coalesced_range(mr);
 }
 
 void memory_region_set_flush_coalesced(MemoryRegion *mr)
-- 
2.0.0

^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [Qemu-devel] [PATCH] memory: Don't call memory_region_update_coalesced_range if nothing changed
  2014-06-13  6:34 [Qemu-devel] [PATCH] memory: Don't call memory_region_update_coalesced_range if nothing changed Fam Zheng
@ 2014-06-13  7:46 ` Paolo Bonzini
  0 siblings, 0 replies; 2+ messages in thread
From: Paolo Bonzini @ 2014-06-13  7:46 UTC (permalink / raw)
  To: Fam Zheng, qemu-devel; +Cc: stefanha

Il 13/06/2014 08:34, Fam Zheng ha scritto:
> With huge number of PCI devices in the system (for example, 200
> virtio-blk-pci), this unconditional call can slow down emulation of
> irrelevant PCI operations drastically, such as a BAR update on a device
> that has no coalescing region. So avoid it.
>
> Signed-off-by: Fam Zheng <famz@redhat.com>
> ---
>  memory.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/memory.c b/memory.c
> index d8ea212..57227f4 100644
> --- a/memory.c
> +++ b/memory.c
> @@ -1324,6 +1324,7 @@ void memory_region_add_coalescing(MemoryRegion *mr,
>  void memory_region_clear_coalescing(MemoryRegion *mr)
>  {
>      CoalescedMemoryRange *cmr;
> +    bool updated = false;
>
>      qemu_flush_coalesced_mmio_buffer();
>      mr->flush_coalesced_mmio = false;
> @@ -1332,8 +1333,12 @@ void memory_region_clear_coalescing(MemoryRegion *mr)
>          cmr = QTAILQ_FIRST(&mr->coalesced);
>          QTAILQ_REMOVE(&mr->coalesced, cmr, link);
>          g_free(cmr);
> +        updated = true;
> +    }
> +
> +    if (updated) {
> +        memory_region_update_coalesced_range(mr);
>      }
> -    memory_region_update_coalesced_range(mr);
>  }
>
>  void memory_region_set_flush_coalesced(MemoryRegion *mr)
>

Very nice.  Applying to memory branch, thank you.

Paolo

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2014-06-13  7:46 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-06-13  6:34 [Qemu-devel] [PATCH] memory: Don't call memory_region_update_coalesced_range if nothing changed Fam Zheng
2014-06-13  7:46 ` Paolo Bonzini

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.