On 2014-11-27 at 16:09, Stefan Hajnoczi wrote:
On Thu, Nov 20, 2014 at 06:06:22PM +0100, Max Reitz wrote:
+/**
+ * Reallocates *array so that it can hold new_size entries. *size must contain
+ * the current number of entries in *array. If the reallocation fails, *array
+ * and *size will not be modified and -errno will be returned. If the
+ * reallocation is successful, *array will be set to the new buffer and *size
+ * will be set to new_size. The size of the reallocated refcount array buffer
+ * will be aligned to a cluster boundary, and the newly allocated area will be
+ * zeroed.
+ */
+static int realloc_refcount_array(BDRVQcowState *s, uint16_t **array,
+ int64_t *size, int64_t new_size)
+{
+ /* Round to clusters so the array can be directly written to disk */
+ size_t old_byte_size = ROUND_UP(refcount_array_byte_size(s, *size),
+ s->cluster_size);
+ size_t new_byte_size = ROUND_UP(refcount_array_byte_size(s, new_size),
+ s->cluster_size);
+ uint16_t *new_ptr;
+
+ if (new_byte_size <= old_byte_size) {
+ *size = new_size;
+ return 0;
+ }
Why not realloc the array to the new smaller size? ...
Because such a call will actually never happen. I could replace this if
() by assert(new_byte_size >= old_byte_size); if (new_byte_size ==
old_byte_size), but as I said before, I'm not a friend of assertions
when the code can deal perfectly well with the "unsupported" case.
Max
+
+ assert(new_byte_size > 0);
+
+ new_ptr = g_try_realloc(*array, new_byte_size);
+ if (!new_ptr) {
+ return -ENOMEM;
+ }
+
+ memset((void *)((uintptr_t)new_ptr + old_byte_size), 0,
+ new_byte_size - old_byte_size);
...we just need to skip the memset in when new_byte_size is smaller
than old_byte_size.