On Thu, Apr 14, 2016 at 12:21:50PM +0200, Lukas Slebodnik wrote: > On (14/04/16 10:39), Pavel Reichl wrote: > >On 04/14/2016 10:28 AM, Lukas Slebodnik wrote: > >>ehlo, > >> > >>@see commit message in attached trivial patch. > >> > >>LS > >> > >> > >Hello, > > > >patch does not apply, code in patch doesn't look like recent master to me...? > > > Patch was from latest 1.13 branch. > This issue is fixed in master branch in similar way. > > The main dfference is that secondary_name is not checked for NULL > before releasing memory. But checking for NULL is not done everywhere > in the idmap code. So it needn't be changed.
I assumed that modern free()-like functions will be able to handle NULL, but you are right, since it is configurable this expectation should be spelled out. Do you think a patch like diff --git a/src/lib/idmap/sss_idmap.h b/src/lib/idmap/sss_idmap.h index 483241e..8a09296 100644 --- a/src/lib/idmap/sss_idmap.h +++ b/src/lib/idmap/sss_idmap.h @@ -135,10 +135,11 @@ struct dom_sid; * @brief Initialize idmap context * * @param[in] alloc_func Function to allocate memory for the context, if - * NULL malloc() id used + * NULL malloc() is used * @param[in] alloc_pvt Private data for allocation routine - * @param[in] free_func Function to free the memory the context, if - * NULL free() id used + * @param[in] free_func Function to free the memory allocated by alloc_func, if + * NULL free() is used, free_func() must handle NULL as + * input gracefully * @param[out] ctx idmap context * * @return would be sufficient? bye, Sumit > > I will remove patch from patchwork. > > LS > _______________________________________________ > sssd-devel mailing list > [email protected] > https://lists.fedorahosted.org/admin/lists/[email protected] _______________________________________________ sssd-devel mailing list [email protected] https://lists.fedorahosted.org/admin/lists/[email protected]
