[Model] M3 MSA indexer: align decode path with rebased index API

The MSA builder/impl were written against the pre-#45743 indexer decode
API. After rebasing onto main, MiniMaxM3IndexerDecodeMetadata gained a
required max_decode_query_len field and minimax_m3_index_decode dropped
its sm_scale parameter, so the MSA path crashed at engine init during
cudagraph profiling (missing max_decode_query_len).

- Pass max_decode_query_len when building decode metadata.
- Drop the stale sm_scale arg and pass max_decode_query_len in the decode
  kernel call, matching the Triton builder/impl.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
(cherry picked from commit 0b70ba38a1)

vllm/models/minimax_m3/nvidia/indexer_msa.py

@@ -121,6 +121,7 @@ class MiniMaxM3IndexerMSAMetadataBuilder(MiniMaxM3IndexerMetadataBuilder):
                 block_table=block_table[:num_decodes],
                 max_seq_len=common_attn_metadata.max_seq_len,
                 decode_query_len=decode_query_len,
+                max_decode_query_len=self.max_decode_query_len,
             )
 
         prefill: MiniMaxM3IndexerMSAPrefillMetadata | None = None
@@ -211,8 +212,8 @@ class MiniMaxM3IndexerMSAImpl(MiniMaxM3IndexerImpl):
                 self.init_blocks,
                 self.local_blocks,
                 self.num_kv_heads,
-                self.scale,
                 d.decode_query_len,
+                d.max_decode_query_len,
                 out=buf,
             )