From 5a6eb4351d2971edd27ae7d566d9488a4e7e7a94 Mon Sep 17 00:00:00 2001 From: claude-noether Date: Thu, 14 May 2026 08:59:33 +0000 Subject: [PATCH] =?UTF-8?q?iter18=20=CE=B1-22=20TEST:=20skip=20DECODE=5FPA?= =?UTF-8?q?RAMS=20to=20isolate=20validation=20failure?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If removing DECODE_PARAMS from libva's S_EXT_CTRLS batch lets the other 4 controls stage, rkvdec_hevc_run printk will show w=1280 h=720 etc. That confirms DECODE_PARAMS specifically is failing kernel validation and rolling back the whole batch. Co-Authored-By: Claude Opus 4.7 (1M context) --- src/h265.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/src/h265.c b/src/h265.c index dd5104e..81daeae 100644 --- a/src/h265.c +++ b/src/h265.c @@ -635,11 +635,21 @@ int h265_set_controls(struct request_data *driver_data, .ptr = &scaling_matrix, .size = sizeof(scaling_matrix), }; + /* + * iter18 α-22 TEST: skip DECODE_PARAMS submission. + * error_idx=4 from S_EXT_CTRLS showed DECODE_PARAMS fails validation. + * If removing it from the batch lets SPS/PPS/SLICE/SCALING_MATRIX + * stage correctly, the bug is isolated to DECODE_PARAMS validation. + */ +#if 0 controls[n++] = (struct v4l2_ext_control){ .id = V4L2_CID_STATELESS_HEVC_DECODE_PARAMS, .ptr = &decode_params, .size = sizeof(decode_params), }; +#else + (void)decode_params; +#endif rc = v4l2_set_controls(driver_data->video_fd, surface_object->request_fd,