From 44c3dad69bbf4aeabbbbab5a7a15150c23aefc63 Mon Sep 17 00:00:00 2001 From: "Galantsev, Dmitrii" Date: Mon, 24 Mar 2025 22:46:57 +0000 Subject: [PATCH] README - Add gpu reset known issue Change-Id: I4f9ac6ce807d4d670a19ae84fe553eb3a7484d96 Signed-off-by: Galantsev, Dmitrii --- README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/README.md b/README.md index 65c4e60..7babd23 100644 --- a/README.md +++ b/README.md @@ -576,6 +576,14 @@ The RAS plugin enables monitoring and counting of ECC (Error-Correcting Code) er > - Limited metrics on MI200. > - Consumer GPUs (e.g., RX6800) have fewer supported metrics. > +>#### 🛑 RDC crashes on GPU reset +> +> This is expected behavior. Plugins for RDC (such as RVS and RocProfiler) +> enter an undefined state when GPU resets. Currently there is no way to +> automatically re-initialize those without restarting RDC. +> +> Due to above issues - **Policy** feature cannot monitor for GPU reset. +> >#### 🐍 dmon RocProfiler Fields Return Zeros > >**Solution:**