[otbn, dv] Fixes failures in otbn_imem_err when secure wipe is enabled
When secure wipe is enabled, wiping takes extra 98 cycles and if an imem
error is injected in this period, it will be ignored.
This commit increases the delay fron 3 ro 100 cycles to allow for secure
wiping to complete and then check for idle status.
Signed-off-by: Prajwala Puttappa <prajwalaputtappa@lowrisc.org>
diff --git a/hw/ip/otbn/dv/uvm/env/seq_lib/otbn_imem_err_vseq.sv b/hw/ip/otbn/dv/uvm/env/seq_lib/otbn_imem_err_vseq.sv
index 9f2bc94..e1429c4 100644
--- a/hw/ip/otbn/dv/uvm/env/seq_lib/otbn_imem_err_vseq.sv
+++ b/hw/ip/otbn/dv/uvm/env/seq_lib/otbn_imem_err_vseq.sv
@@ -45,11 +45,9 @@
cfg.model_agent_cfg.vif.invalidate_imem();
// If we were unlucky, we might have injected the errors while OTBN was executing an instruction
- // that was already causing it to stop. To allow for this specific case, we wait three cycles to
- // let that instruction flush through. We need this much time to handle things like branches
- // that might take until the second cycle before realising that something failed and then one
- // more cycle to clean up.
- repeat (3) @(cfg.clk_rst_vif.cbn);
+ // that was already causing it to stop. To allow for this specific case, we wait 100 cycles to
+ // let that instruction flush through and secure wiping to finish.
+ repeat (100) @(cfg.clk_rst_vif.cbn);
// If OTBN is now idle, we hit this exact window and the test didn't do anything useful. Ho
// hum... Note that we don't need to apply a reset in this case.
if (cfg.model_agent_cfg.vif.status == otbn_pkg::StatusIdle) begin
