Re: [PATCH] rrd_update: fix memory leak during CDEF processing

Hi Frederik,

makes sense ... (I guess CDEF DSes are not used a lot)
thanks
tobi

Yesterday Frederik Kriewitz wrote:

> I think rpnp in process_pdp_st (allocated in rpn_expand) in is never freed.
>
> Index: src/rrd_update.c
> ===================================================================
> --- src/rrd_update.c    (revision 2072)
> +++ src/rrd_update.c    (working copy)
> @@ -1354,6 +1354,10 @@
>
>          rpnp =
>              rpn_expand((rpn_cdefds_t *) &(rrd->ds_def[ds_idx].par[DS_cdef]));
> +        if(rpnp == NULL) {
> +          rpnstack_free(&rpnstack);
> +          return -1;
> +        }
>          /* substitute data values for OP_VARIABLE nodes */
>          for (i = 0; rpnp[i].op != OP_END; i++) {
>              if (rpnp[i].op == OP_VARIABLE) {
> @@ -1367,6 +1371,7 @@
>              rpnstack_free(&rpnstack);
>              return -1;
>          }
> +        free(rpnp);
>      }
>
>      /* make pdp_prep ready for the next run */
> Index: src/rrd_rpncalc.c
> ===================================================================
> --- src/rrd_rpncalc.c   (revision 2072)
> +++ src/rrd_rpncalc.c   (working copy)
> @@ -72,8 +72,10 @@
>      /* DS_CDEF_MAX_RPN_NODES is small, so at the expense of some wasted
>       * memory we avoid any reallocs */
>      rpnp = (rpnp_t *) calloc(DS_CDEF_MAX_RPN_NODES, sizeof(rpnp_t));
> -    if (rpnp == NULL)
> +    if (rpnp == NULL) {
> +        rrd_set_error("failed allocating rpnp array");
>          return NULL;
> +    }
>      for (i = 0; rpnc[i].op != OP_END; ++i) {
>          rpnp[i].op = rpnc[i].op;
>          if (rpnp[i].op == OP_NUMBER) {
>
> _______________________________________________
> rrd-developers mailing list
> [email protected]
> https://lists.oetiker.ch/cgi-bin/listinfo/rrd-developers
>
>

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
http://it.oetiker.ch [email protected] ++41 62 775 9902 / sb: -9900