From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ferruh Yigit <ferruh.yigit@intel.com>
Subject: Re: [PATCH] eal: fix API to get error string
Date: Wed, 31 Oct 2018 18:26:43 +0000
Message-ID: <cd17cee8-0f77-6a29-25dd-728174e3132e@intel.com>
References: <20181031171928.61110-1-ferruh.yigit@intel.com>
 <5469782.njCRi2Lb0J@xps> <20d0a785-ea11-ca9f-8bed-7e5b184adbd9@intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Cc: dev@dpdk.org, Bruce Richardson <bruce.richardson@intel.com>,
 stable@dpdk.org
To: Thomas Monjalon <thomas@monjalon.net>
Return-path: <dev-bounces@dpdk.org>
In-Reply-To: <20d0a785-ea11-ca9f-8bed-7e5b184adbd9@intel.com>
Content-Language: en-US
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

On 10/31/2018 6:26 PM, Ferruh Yigit wrote:
> On 10/31/2018 5:16 PM, Thomas Monjalon wrote:
>> 31/10/2018 18:19, Ferruh Yigit:
>>> rte_strerror uses strerror_r(), and strerror_r() has two version of it.
>>> - XSI-compliant version, (_POSIX_C_SOURCE >= 200112L) && !  _GNU_SOURCE
>>> - GNU-specific version
>>>
>>> Those two has different return types, so the exiting return type check
>>> is not correct for GNU-specific version.
>>>
>>> And this is causing failure in errno_autotest unit test.
>>>
>>> Adding different implementation for FreeBSD and Linux.
>>>
>>> Fixes: 016c32bd3e3d ("eal: cleanup strerror function")
>>> Cc: stable@dpdk.org
>>>
>>> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
>>> ---
>>> --- a/lib/librte_eal/common/eal_common_errno.c
>>> +++ b/lib/librte_eal/common/eal_common_errno.c
>>>  		default:
>>> +#ifdef RTE_EXEC_ENV_BSDAPP
>>>  			if (strerror_r(errnum, ret, RETVAL_SZ) != 0)
>>>  				snprintf(ret, RETVAL_SZ, "Unknown error%s %d",
>>>  						sep, errnum);
>>> +#else
>>> +			/*
>>> +			 * _GNU_SOURCE version, error string is not always
>>> +			 * strored in "ret" buffer, need to use return value
>>> +			 */
>>> +			ret = strerror_r(errnum, ret, RETVAL_SZ);
>>> +#endif
>>
>> Why not use the return value in both cases?
>>
>> Why not writing an error message in Linux case?
> 
> "man strerror_r" has more details, but briefly,
> 
> The XSI-compliant strerror_r() function returns 0 on success. GNU one returns
> the pointer to string.
> 
> The XSI-compliant can return an empty buffer, GNU one always return a string,
> either proper error string or "Unknown .." one.

strerror_r() not portable. An alternative can be not using it at all...