[dpdk-dev] [PATCH v3] app/testpmd: fix testpmd packets dump overlapping

[Ferruh Yigit]

On 11/23/2020 6:13 AM, Jiawei(Jonny) Wang wrote:
> Hi Ferruh,
> 
>> -----Original Message-----
>> From: Ferruh Yigit <ferruh.yigit at intel.com>
>> Sent: Saturday, November 21, 2020 1:50 AM
>> To: Jiawei(Jonny) Wang <jiaweiw at nvidia.com>; wenzhuo.lu at intel.com;
>> beilei.xing at intel.com; bernard.iremonger at intel.com; Ori Kam
>> <orika at nvidia.com>; Slava Ovsiienko <viacheslavo at nvidia.com>; NBU-
>> Contact-Thomas Monjalon <thomas at monjalon.net>; Raslan Darawsheh
>> <rasland at nvidia.com>
>> Cc: dev at dpdk.org
>> Subject: Re: [PATCH v3] app/testpmd: fix testpmd packets dump overlapping
>>
>> On 11/20/2020 5:35 PM, Jiawei Wang wrote:
>>> When testpmd enabled the verbosity for the received packets, if two
>>> packets were received at the same time, for example, sampling packet
>>> and normal packet, the dump output of these packets may be overlapping
>>> due to multiple core handling the multiple queues simultaneously.
>>>
>>> The patch uses one string buffer that collects all the packet dump
>>> output into this buffer and then printouts it at last, that guarantees
>>> to printout separately the dump output per packet.
>>>
>>> Fixes: d862c45 ("app/testpmd: move dumping packets to a separate
>>> function")
>>>
>>> Signed-off-by: Jiawei Wang <jiaweiw at nvidia.com>
>>
>> <...>
>>
>>> @@ -74,13 +85,16 @@
>>>    	uint32_t vx_vni;
>>>    	const char *reason;
>>>    	int dynf_index;
>>> +	int buf_size = MAX_STRING_LEN;
>>> +	char print_buf[buf_size];
>>> +	int cur_len = 0;
>>>
>>> +	memset(print_buf, 0, sizeof(print_buf));
>>
>> Should 'print_buf' cleaned per each packet below, if not can we drop 'memset'
>> completely?
>>
> Since the 'snprintf' always append the a terminating null character('\0') after the written string,
> and in the code the following character be appended start from previous null character, so only one
> '\0' will be appended in the last 'snprintf' calls,
> 	snprintf(buf + cur_len, buf_size - cur_len ..
> At the last 'printf("%s") will print the string buffer up to a terminating null character ('\0').
> so it's ok even not 'memset' calls to clean 'print_buf' per each packets.

agree, so why not drop the memset completely?