[Mesa-dev] [PATCH 4/7] util: Port nir_array functionality to u_dynarray

[Kenneth Graunke]

On Saturday, June 3, 2017 11:11:39 AM PDT Thomas Helland wrote:
[snip]
> @@ -87,10 +102,18 @@ util_dynarray_trim(struct util_dynarray *buf)
>  {
>     if (buf->size != buf->capacity) {
>        if (buf->size) {
> -         buf->data = realloc(buf->data, buf->size);
> +         if (buf->mem_ctx) {
> +            reralloc_size(buf->mem_ctx, buf->data, buf->size);
> +         } else {
> +            buf->data = realloc(buf->data, buf->size);
> +         }

Hi Thomas,

Would it make sense to do:

      buf->data = reralloc_size(...)

It's probably safe as is given that you're trying to trim down an
existing allocation to be smaller, but it seems safer to use the return
value nonetheless.  Coverity grumbled after you pushed the patch:

** CID 1412568:  Resource leaks  (RESOURCE_LEAK)
/src/util/u_dynarray.h: 110 in util_dynarray_trim()

______________________________________________________________
*** CID 1412568:  Resource leaks  (RESOURCE_LEAK)
/src/util/u_dynarray.h: 110 in util_dynarray_trim()
104     static inline void
105     util_dynarray_trim(struct util_dynarray *buf)
106     {
107        if (buf->size != buf->capacity) {
108           if (buf->size) {
109              if (buf->mem_ctx) {
>>> CID 1412568:  Resource leaks  (RESOURCE_LEAK)
>>> Ignoring storage allocated by "reralloc_size(buf->mem_ctx, buf->data, buf->size)" leaks it.
110                 reralloc_size(buf->mem_ctx, buf->data, buf->size);
111              } else {
112                 buf->data = realloc(buf->data, buf->size);
113              }
114              buf->capacity = buf->size;
115           } else {

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20170611/054d74f9/attachment-0001.sig>