ggml-alloc : fix discrepency between measure&eval (#2639)

The GGML memory allocator consistently places a tensor within the
optimal-fit memory block, which is the smallest block capable of
accommodating the tensor's size. During the measurement phase, the final
block is generously sized, ensuring it never qualifies as the
optimal-fit block as long as there exists another block capable of
accommodating the tensor. Nevertheless, in the evaluation phase, the
last block is constrained in size and could potentially qualify as the
optimal-fit block. Consequently, there exists the possibility of a
tensor being allocated to a different region during evaluation, leading
to more memory fragmentation in our scratch buffer.

This recent commit guarantees uniform behavior of the allocator across
both the measurement and evaluation phases, eliminating discrepancies
between the two.

Author: lshzh-ww

ggml-alloc.c

@@ -113,10 +113,10 @@ void ggml_allocr_alloc(struct ggml_allocr * alloc, struct ggml_tensor * tensor)
 
     size_t max_avail = 0;
 
-    // find the best fitting free block
+    // find the best fitting free block besides the last block
     int best_fit_block = -1;
     size_t best_fit_size = SIZE_MAX;
-    for (int i = 0; i < alloc->n_free_blocks; i++) {
+    for (int i = 0; i < alloc->n_free_blocks - 1; i++) {
         struct free_block * block = &alloc->free_blocks[i];
         max_avail = MAX(max_avail, block->size);
         if (block->size >= size && block->size <= best_fit_size) {
@@ -128,10 +128,17 @@ void ggml_allocr_alloc(struct ggml_allocr * alloc, struct ggml_tensor * tensor)
     AT_PRINTF("block %d\n", best_fit_block);
 
     if (best_fit_block == -1) {
-        fprintf(stderr, "%s: not enough space in the buffer (needed %zu, largest block available %zu)\n",
-                __func__, size, max_avail);
-        GGML_ASSERT(!"not enough space in the buffer");
+        // the last block is our last resort
+        struct free_block * block = &alloc->free_blocks[alloc->n_free_blocks - 1];
+        if (block->size >= size) {
+            best_fit_block = alloc->n_free_blocks - 1;
+            max_avail = MAX(max_avail, block->size);
+        } else {
+            fprintf(stderr, "%s: not enough space in the buffer (needed %zu, largest block available %zu)\n",
+                    __func__, size, max_avail);
+            GGML_ASSERT(!"not enough space in the buffer");
         return;
+        }
     }
     struct free_block * block = &alloc->free_blocks[best_fit_block];
     void * addr = block->addr;