-----Original Message-----
From: Ferruh Yigit [mailto:ferruh.yigit@intel.com]
Sent: Tuesday, June 22, 2021 8:28 PM
To: wangyunjian <redacted>; dev@dpdk.org
Cc: thomas@monjalon.net; gowrishankar.m@linux.vnet.ibm.com;
dingxiaoxiong [off-list ref]; stable@dpdk.org; liucheng (J)
[off-list ref]
Subject: Re: [PATCH v2] kni: fix wrong mbuf alloc count in kni_allocate_mbufs

On 6/22/2021 11:57 AM, wangyunjian wrote:

quoted

From: Yunjian Wang <redacted>

In kni_allocate_mbufs(), we alloc mbuf for alloc_q as this code.
allocq_free = (kni->alloc_q->read - kni->alloc_q->write - 1) \
		& (MAX_MBUF_BURST_NUM - 1);
The value of allocq_free maybe zero, for example :
The ring size is 1024. After init, write = read = 0. Then we fill
kni->alloc_q to full. At this time, write = 1023, read = 0.

Then the kernel send 32 packets to userspace. At this time, write =
1023, read = 32. And then the userspace receive this 32 packets.
Then fill the kni->alloc_q, (32 - 1023 - 1) & 31 = 0, fill nothing.
...
Then the kernel send 32 packets to userspace. At this time, write =
1023, read = 992. And then the userspace receive this 32 packets.
Then fill the kni->alloc_q, (992 - 1023 - 1) & 31 = 0, fill nothing.

Then the kernel send 32 packets to userspace. The kni->alloc_q only
has 31 mbufs and will drop one packet.

Absolutely, this is a special scene. Normally, it will fill some mbufs
everytime, but may not enough for the kernel to use.

In this patch, we always keep the kni->alloc_q to full for the kernel
to use.

Fixes: 49da4e82cf94 ("kni: allocate no more mbuf than empty slots in
queue")
Cc: stable@dpdk.org

Signed-off-by: Cheng Liu <redacted>
Signed-off-by: Yunjian Wang <redacted>

Acked-by: Ferruh Yigit <redacted>

What do you think to change patch title to something like:
kni: fix mbuf allocation for alloc FIFO

OK, I will change patch title later.

Thanks