qwen4exp: fix seq_cp, block position keying, mtmd input, cuda abort, add tests (#27941)

* qwen4exp: follow up fixes

* -kvu NaN collapse fix

Assisted-by: Claude

* indexer cache ext.x/ext.y restore fix

Assisted-by: Claude

* kv-cells: rename seq_set to seq_get_all

seq_get is already taken by the single-id getter, so the suggested name
cannot be overloaded on return type alone.

Assisted-by: Claude

* memory-hybrid-idx: implement set_input_qsa on the memory class

The context held the whole implementation, where the pattern elsewhere is a
thin context forwarding to the memory class, as llama_kv_cache_context does
for set_input_kq_mask. The body reads no context state, so it moves unchanged
and the context keeps a forwarder.

Also shortens the seq_get_all comment as suggested.

* tests: check that a sequence state survives a save/restore round-trip

Saves seq 0, erases it, restores the blob and saves again, requiring the two
blobs to match. Compares blobs rather than generated text, which cannot see a
field dropped on the way back in.

Note this passes on master for qwen4exp, so it does not demonstrate the
ext.x/ext.y drop this PR fixes; reaching that needs 2D mrope content.

* tests: give the synthetic qwen4exp a PLE so the state test bites

has_cell_ext() is n_pos_per_embd() > 1 || ple_n_heads > 0, and the indexer
cache sets rope_type = NONE, so without a PLE it serializes no cell ext at
all and the round-trip test cannot see a dropped ext.x/ext.y. With one,
removing the ext_set restore in state_read_meta fails the test: 198 of
335692 bytes differ, first at offset 282092.

Loading such a model needed two fixes:

- the row count of per_layer_token_embd came from require_weight(), which a
  model synthesised from metadata alone has no file to answer. Derive it
  from the head ranges and prefer the file's padded count where there is one.
- the PLE conv history is a row of the recurrent cache, so a PLE on a full
  attention layer dereferenced a null p_l. Reject it at load time instead.

The meta mirror is skipped for qwen4exp. It returned NaN logits before this
fixture carried a PLE, which the nmse check passes since a NaN comparison is
false, and aborts with one. -sm tensor on real devices works.

Assisted-by: Claude

* llama: disable -sm tensor for qwen4exp

test-llama-archs skipped the tensor split for this arch from inside the
test, so the arch still advertised support it does not have. Declare it in
llm_arch_supports_sm_tensor instead and drop the test-side exception; the
existing llm_arch_supports_sm_tensor branch then does the skipping.

Assisted-by: Claude

src/llama-kv-cache.cpp

@@ -2387,6 +2387,12 @@ bool llama_kv_cache::state_read_meta(llama_io_read_i & io, uint32_t strm, uint32
 
         ubatch.seq_id_unq[0] = dest_seq_id;
 
+        // the ext as it was saved, to put back after apply_ubatch()
+        std::vector<llama_kv_cell_ext> exts;
+        if (has_cell_ext()) {
+            exts.resize(cell_count);
+        }
+
         for (uint32_t i = 0; i < cell_count; ++i) {
             llama_pos pos;
             uint32_t n_seq_id;
@@ -2410,6 +2416,8 @@ bool llama_kv_cache::state_read_meta(llama_io_read_i & io, uint32_t strm, uint32
 
                 // apply_ubatch() below restores ext.tok from the ubatch tokens
                 ubatch.token[i] = ext.tok;
+
+                exts[i] = ext;
             }
 
             // read the sequence id, but directly discard it - we will use dest_seq_id instead
@@ -2461,6 +2469,14 @@ bool llama_kv_cache::state_read_meta(llama_io_read_i & io, uint32_t strm, uint32
         //       see: https://github.com/ggml-org/llama.cpp/pull/16825#issuecomment-3460868350
         apply_ubatch(sinfo, ubatch);
 
+        // apply_ubatch() takes the 2D position from the ubatch, and that ubatch is built with this
+        // cache's own n_pos_per_embd. a cache that does not use M-RoPE itself but mirrors one that
+        // does (the qwen4exp QSA indexer) would drop x and y. put the saved ext back instead, which
+        // is what the whole-context path below already does.
+        for (uint32_t i = 0; i < (uint32_t) exts.size(); ++i) {
+            cells.ext_set(sinfo.idxs[0][i], exts[i]);
+        }
+
         LLAMA_LOG_DEBUG("%s: cell_count = %d, dest_seq_id = %d\n", __func__, cell_count, dest_seq_id);
 
         // DEBUG CHECK: verify that all cells were allocated and have correct seq_id and pos values