sync : ggml

examples/cli/cli.cpp

@@ -9,7 +9,6 @@
 #include <cstdio>
 #include <string>
 #include <thread>
-#include <utility>
 #include <vector>
 #include <cstring>
 
@@ -1019,10 +1018,12 @@ int main(int argc, char ** argv) {
 
             bool open(const char * ext, const char * function) {
                 if (is_stdout) {
-                    if (std::exchange(used_stdout, true)) {
+                    if (used_stdout) {
                         fprintf(stderr, "warning: Not appending multiple file formats to stdout\n");
                         return false;
                     }
+
+                    used_stdout = true;
 #ifdef _WIN32
                     fout = std::ofstream{"CON"};
 #else
@@ -1032,6 +1033,7 @@ int main(int argc, char ** argv) {
                     // Also assuming /dev is mounted
                     return true;
                 }
+
                 fname_out.resize(basename_length);
                 fname_out += ext;
                 fout = std::ofstream{fname_out};

ggml/src/ggml-alloc.c

@@ -816,7 +816,10 @@ static void ggml_gallocr_init_tensor(ggml_gallocr_t galloc, struct ggml_tensor *
 static bool ggml_gallocr_node_needs_realloc(ggml_gallocr_t galloc, struct ggml_tensor * node, struct tensor_alloc * talloc) {
     size_t node_size = 0;
     if (!node->data && !node->view_src) {
-        GGML_ASSERT(talloc->buffer_id >= 0); // prevent segfault when misusing the API
+        // If we previously had data but don't now then reallocate
+        if (talloc->buffer_id < 0) {
+            return false;
+        }
         node_size = ggml_backend_buft_get_alloc_size(galloc->bufts[talloc->buffer_id], node);
     }
     return talloc->size_max >= node_size;