Re: [Patch] vmbus: Simply hv_get_next_write_location() function

From: Tianyu Lan
Date: Thu Jan 25 2018 - 03:33:43 EST


On Thu, Jan 25, 2018 at 12:37 AM, Stephen Hemminger
<stephen@xxxxxxxxxxxxxxxxxx> wrote:
> On Wed, 24 Jan 2018 14:21:30 +0800
> lantianyu1986@xxxxxxxxx wrote:
>
>> From: Tianyu Lan <Tianyu.Lan@xxxxxxxxxxxxx>
>>
>> The "next" variable is redundant in hv_get_next_write_location().
>> This patch is to remove it and return write_index directly.
>>
>> Signed-off-by: Tianyu Lan <Tianyu.Lan@xxxxxxxxxxxxx>
>> ---
>> drivers/hv/ring_buffer.c | 4 +---
>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c
>> index 12eb8ca..71558e7 100644
>> --- a/drivers/hv/ring_buffer.c
>> +++ b/drivers/hv/ring_buffer.c
>> @@ -82,9 +82,7 @@ static void hv_signal_on_write(u32 old_write, struct vmbus_channel *channel)
>> static inline u32
>> hv_get_next_write_location(struct hv_ring_buffer_info *ring_info)
>> {
>> - u32 next = ring_info->ring_buffer->write_index;
>> -
>> - return next;
>> + return ring_info->ring_buffer->write_index;
>> }
>>
>> /* Set the next write location for the specified ring buffer. */
>
> Looks good.
> But let's go farther since function is only used in one location in the file
> just eliminate it completely and do simple variable references.
>
> The get/set functions in this file are unnecessary.

Yes, agree and will update patch.

>
> Better still it is possible to replace the lock based ring structure
> with a compare-exchange solution.

There are several read/write operations of ring structure in the
hv_ringbuffer_write()
and these operations should be under protection. Especially for
ring_buffer->write_index,
we need to read it to calculate available write buffer, determine
write position and then update it after
writing buffer. This sequence should be under protection, right?

--
Best regards
Tianyu Lan