[dpdk-stable] Should we backport or skip e132ee86 "devargs: fix memory leak on parsing failure"

Xueming(Steven) Li xuemingl at nvidia.com
Tue Jun 1 09:01:41 CEST 2021



> -----Original Message-----
> From: Christian Ehrhardt <christian.ehrhardt at canonical.com>
> Sent: Tuesday, June 1, 2021 2:07 PM
> To: dpdk stable <stable at dpdk.org>; Xueming(Steven) Li <xuemingl at nvidia.com>
> Cc: Luca Boccassi <bluca at debian.org>
> Subject: Should we backport or skip e132ee86 "devargs: fix memory leak on parsing failure"
> 
> Hi Xueming,
> 
> I wanted to let you know that while backporting:
>   Commit e132ee8690474fa985e26f1d4db75823024748c8
>   Author: Xueming Li <xuemingl at nvidia.com>
>   Date:   Tue Apr 13 03:14:09 2021 +0000
>       devargs: fix memory leak on parsing failure
> 
> I've found
> 
> [  288s] ../lib/librte_eal/common/eal_common_devargs.c:171:16: error:
> passing argument 1 of ‘free’ discards ‘const’ qualifier from pointer target type [-Werror=discarded-qualifiers]
> [  288s]   171 |    free(devargs->data);
> [  288s]       |         ~~~~~~~^~~~~~
> [  288s] In file included from ../lib/librte_eal/common/include/rte_common.h:20,
> [  288s]                  from ../lib/librte_eal/common/include/rte_log.h:25,
> [  288s]                  from ../lib/librte_eal/common/include/rte_bus.h:24,
> [  288s]                  from ../lib/librte_eal/common/eal_common_devargs.c:13:
> [  288s] /usr/include/stdlib.h:565:25: note: expected ‘void *’ but argument is of type ‘const char *’
> [  288s]   565 | extern void free (void *__ptr) __THROW;
> [  288s]       |                   ~~~~~~^~~~~
> 
> I've seen that to make it work it would require some more from the same series, but that was not targeted at dpgk-stable
> 
>   Commit 64051bb1f144c418f3fc76e6d0973337b05d5886
>   Author: Xueming Li <xuemingl at nvidia.com>
>   Date:   Tue Apr 13 03:14:08 2021 +0000
>       devargs: unify scratch buffer storage
> 
> Then I've checked and seen that it is also not present in lib/librte_eal/common/eal_common_devargs.c of 20.11.2 that you are
> preparing so I have also dropped it from 19.11.9 as well for now.
> 
> Do you think you'd want to create backports of this for 19.11/20.11 or is it not important enough to bother about it?

Yes, need another patch to make this one work. But I don't think it important enough to have it.
> 
> --
> Christian Ehrhardt
> Staff Engineer, Ubuntu Server
> Canonical Ltd


More information about the stable mailing list