[dpdk-dev] [PATCH v2 3/3] vhost: optimize vhost virtqueue struct

[Maxime Coquelin]

On 3/16/21 2:38 PM, David Marchand wrote:
> On Tue, Mar 16, 2021 at 1:42 PM Maxime Coquelin
> <maxime.coquelin at redhat.com> wrote:
>>
>> This patch moves vhost_virtuqueue struct fields in order
> 
> virtqueue
> You might want to add something in your local dictionary :-p.
> 
> 
>> to both optimize packing and move hot fields on the first
>> cachelines.
>>
>> Signed-off-by: Maxime Coquelin <maxime.coquelin at redhat.com>
>> ---
>>  lib/librte_vhost/vhost.h | 52 +++++++++++++++++++++-------------------
>>  1 file changed, 27 insertions(+), 25 deletions(-)
>>
>> diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
>> index 3a71dfeed9..ae45b05dd1 100644
>> --- a/lib/librte_vhost/vhost.h
>> +++ b/lib/librte_vhost/vhost.h
>> @@ -133,7 +133,7 @@ struct vhost_virtqueue {
>>                 struct vring_used       *used;
>>                 struct vring_packed_desc_event *device_event;
>>         };
>> -       uint32_t                size;
>> +       uint16_t                size;
> 
> There is one site in the code that could be problematic:
> 
> vhost_user_set_vring_num()
> ...
> vq->size = msg->payload.state.num;
> ...
> if (vq->size > 32768) {
> 
> payload.state.num is an unsigned int, so we'd better check its value
> as an unsigned int, before storing to vq->size.

Indeed, I will rework the function, so that it checks that
msg->payload.state.num is smaller than 32K to comply with the spec,
and so remove the later check on vq->size.

> 
>>
>>         uint16_t                last_avail_idx;
>>         uint16_t                last_used_idx;
>> @@ -143,29 +143,12 @@ struct vhost_virtqueue {
>>  #define VIRTIO_INVALID_EVENTFD         (-1)
>>  #define VIRTIO_UNINITIALIZED_EVENTFD   (-2)
>>
>> -       int                     enabled;
>> -       int                     access_ok;
>> -       int                     ready;
>> -       int                     notif_enable;
>> -#define VIRTIO_UNINITIALIZED_NOTIF     (-1)
>> +       bool                    enabled;
>> +       bool                    access_ok;
>> +       bool                    ready;
> 
> Changing those types is fine, but it is a bit odd to still see
> boolean_var = 0 or boolean_var = 1 in the rest of the code.

Yes, I fix all the places with assigning boolean values.

> 
>>
>>         rte_spinlock_t          access_lock;
>>
>> -       /* Used to notify the guest (trigger interrupt) */
>> -       int                     callfd;
>> -       /* Currently unused as polling mode is enabled */
>> -       int                     kickfd;
>> -
>> -       /* Physical address of used ring, for logging */
>> -       uint64_t                log_guest_addr;
>> -
>> -       /* inflight share memory info */
>> -       union {
>> -               struct rte_vhost_inflight_info_split *inflight_split;
>> -               struct rte_vhost_inflight_info_packed *inflight_packed;
>> -       };
>> -       struct rte_vhost_resubmit_info *resubmit_inflight;
>> -       uint64_t                global_counter;
>>
>>         union {
>>                 struct vring_used_elem  *shadow_used_split;
>> @@ -176,22 +159,36 @@ struct vhost_virtqueue {
>>         uint16_t                shadow_aligned_idx;
>>         /* Record packed ring first dequeue desc index */
>>         uint16_t                shadow_last_used_idx;
>> -       struct vhost_vring_addr ring_addrs;
>>
>> -       struct batch_copy_elem  *batch_copy_elems;
>>         uint16_t                batch_copy_nb_elems;
>> +       struct batch_copy_elem  *batch_copy_elems;
>>         bool                    used_wrap_counter;
>>         bool                    avail_wrap_counter;
>>
>> -       struct log_cache_entry *log_cache;
>> +       /* Physical address of used ring, for logging */
>>         uint16_t log_cache_nb_elem;
> 
> Indent is broken / not consistent, probably because you cut/pasted lines.

Will fix it.

Thanks,
Maxime