Merge upstream/master into master

Resolve llama-context.cpp: keep compute lock on the new llama_batch_ext
encode/decode, and reset both gf_res_prev slots in the shared compute path.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

common/common.cpp

@@ -2220,9 +2220,28 @@ bool common_replay_last_token(struct llama_context * ctx, llama_token last_token
     return true;
 }
 
+llama_batch_ext_ptr common_batch_ext_get_one(llama_context * ctx, const llama_tokens & tokens) {
+    llama_batch_ext_ptr batch(llama_batch_ext_init(ctx));
+
+    auto mem = llama_get_memory(ctx);
+    llama_pos pos = mem ? llama_memory_seq_pos_max(mem, 0) + 1 : 0;
+
+    for (size_t i = 0; i < tokens.size(); ++i) {
+        const int32_t idx = llama_batch_ext_add_token(batch.get(), 0, tokens[i]);
+        llama_batch_ext_set_pos(batch.get(), idx, &pos);
+        pos++;
+    }
+
+    if (!tokens.empty()) {
+        llama_batch_ext_set_output_logits(batch.get(), (int32_t) tokens.size() - 1, true);
+    }
+
+    return batch;
+}
+
 bool common_prompt_batch_decode(
               struct llama_context * ctx,
-    const std::vector<llama_token> & all_tokens,
+                const llama_tokens & all_tokens,
                                int   n_new,
                                int & n_past,
                                int   n_batch,
@@ -2243,7 +2262,9 @@ bool common_prompt_batch_decode(
         // Memory implementations in recurrent/hybrid models don't support removing tokens from their
         // memory, so we can't just remove the last token from the memory and replay the last token which
         // is the reason for this logic.
-        if (llama_decode(ctx, llama_batch_get_one(const_cast<llama_token*>(all_tokens.data() + offset), n_tokens_before_last))) {
+        llama_tokens prefix_tokens(all_tokens.begin() + offset, all_tokens.begin() + offset + n_tokens_before_last);
+        llama_batch_ext_ptr batch_prefix = common_batch_ext_get_one(ctx, prefix_tokens);
+        if (llama_process(ctx, LLAMA_PROCESS_TYPE_DECODE, batch_prefix.get())) {
             COM_ERR("%s", "failed to eval\n");
             return false;
         }
@@ -2253,17 +2274,19 @@ bool common_prompt_batch_decode(
         COM_INF("saved session before last token to %s, n_new = %zu\n", state_path.data(), all_tokens.size());
 
         llama_token last_token = all_tokens.back();
-        llama_batch batch = llama_batch_get_one(&last_token, 1);
-        int32_t pos = n_past;
-        batch.pos = &pos;
+        llama_batch_ext_ptr batch_last = common_batch_ext_get_one(ctx, { last_token });
+        llama_pos pos = n_past;
+        llama_batch_ext_set_pos(batch_last.get(), 0, &pos);
 
-        if (llama_decode(ctx, batch)) {
+        if (llama_process(ctx, LLAMA_PROCESS_TYPE_DECODE, batch_last.get())) {
             COM_ERR("%s", "failed to eval last token\n");
             return false;
         }
         n_past++;
     } else {
-        if (llama_decode(ctx, llama_batch_get_one(const_cast<llama_token*>(all_tokens.data() + offset), n_new))) {
+        llama_tokens new_tokens(all_tokens.begin() + offset, all_tokens.begin() + offset + n_new);
+        llama_batch_ext_ptr batch = common_batch_ext_get_one(ctx, new_tokens);
+        if (llama_process(ctx, LLAMA_PROCESS_TYPE_DECODE, batch.get())) {
             COM_ERR("%s", "failed to eval\n");
             return false;
         }