Just insist on 1 lp staged work per mining thread on new block.
diff --git a/main.c b/main.c
index 7390646..d440341 100644
--- a/main.c
+++ b/main.c
@@ -1210,10 +1210,7 @@ static void flush_requests(bool longpoll)
/* Temporarily increase the staged count so that get_work thinks there
* is work available instead of making threads reuse existing work */
- if (extra >= mining_threads)
- inc_staged(mining_threads, true);
- else
- inc_staged(extra, true);
+ inc_staged(mining_threads, true);
for (i = 0; i < extra; i++) {
/* Queue a whole batch of new requests */