qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v5 3/6] libvhost-user: Support tracking inflight


From: Jason Wang
Subject: Re: [Qemu-devel] [PATCH v5 3/6] libvhost-user: Support tracking inflight I/O in shared memory
Date: Wed, 30 Jan 2019 17:52:52 +0800
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.4.0


On 2019/1/30 上午11:14, Michael S. Tsirkin wrote:
On Wed, Jan 30, 2019 at 10:31:49AM +0800, Jason Wang wrote:
On 2019/1/22 下午4:31, address@hidden wrote:
+static int
+vu_queue_inflight_get(VuDev *dev, VuVirtq *vq, int desc_idx)
+{
+    if (!has_feature(dev->protocol_features,
+        VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD)) {
+        return 0;
+    }
+
+    if (unlikely(!vq->inflight)) {
+        return -1;
+    }
+
+    vq->inflight->desc[desc_idx].inuse = 1;
+
+    vq->inflight->desc[desc_idx].avail_idx = vq->last_avail_idx;
+
+    return 0;
+}
+
+static int
+vu_queue_inflight_pre_put(VuDev *dev, VuVirtq *vq, int desc_idx)
+{
+    if (!has_feature(dev->protocol_features,
+        VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD)) {
+        return 0;
+    }
+
+    if (unlikely(!vq->inflight)) {
+        return -1;
+    }
+
+    vq->inflight->desc[desc_idx].used_idx = vq->used_idx;
+
+    barrier();
+
+    vq->inflight->desc[desc_idx].version++;
+
+    return 0;
+}

You probably need WRITE_ONCE() semantic (e.g volatile) to make sure the
value reach memory.

Thanks

WRITE_ONCE is literally volatile + dependency memory barrier.
So unless compiler is a very agressive one, it does not
buy you much.


Well, since version is increased twice, if compiler decide the inline both vu_queue_inflight_pre_put() and vu_queue_inflight_post_put(), can we make sure it always generate instructions that write to memory instead of registers?

Thanks




reply via email to

[Prev in Thread] Current Thread [Next in Thread]