From: Georgi Gerganov <redacted>
Date: Tue, 7 Oct 2025 05:22:35 +0000 (+0300)
Subject: tests : add -INF blocks to the KQ mask in the FA tests (llama/16380)
X-Git-Tag: upstream/1.8.2~39
X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=6cf0c21b094771237e9ba9da7853d6f7bfca90f9;p=pkg%2Fggml%2Fsources%2Fwhisper.cpp

tests : add -INF blocks to the KQ mask in the FA tests (llama/16380)

* tests : add -INF blocks to the KQ mask in the FA tests

* cont : bump -INF block size to 64

Co-authored-by: Jeff Bolz <redacted>

* ggml : prevent division by zero in FA CPU op

---------

Co-authored-by: Jeff Bolz <redacted>
---

diff --git a/ggml/src/ggml-cpu/ops.cpp b/ggml/src/ggml-cpu/ops.cpp
index 6275c830..8e1a2de1 100644
--- a/ggml/src/ggml-cpu/ops.cpp
+++ b/ggml/src/ggml-cpu/ops.cpp
@@ -8135,7 +8135,7 @@ static void ggml_compute_forward_flash_attn_ext_f16(
         }
 
         // V /= S
-        const float S_inv = 1.0f/S;
+        const float S_inv = S == 0.0f ? 0.0f : 1.0f/S;
         ggml_vec_scale_f32(DV, VKQ32, S_inv);
 
         // dst indices