[rom_ctrl,dv] Pass the .prior argument for some mirror calls #25839
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Confusingly, the default priority (-1) seems to mostly work.
Unfortunately, this *doesn't* work properly with the
rom_ctrl_stress_all_with_rand_reset sequence for some seeds for
particular iterations of the stressed sequence.
In this situation, we end up with an item from a call to mirror() from
the end of one iteration racing against some other register item from
the start of the next one. Because there is more than once item, the
sequencer tries to decide which one to grant... and explodes because
they have a negative priority.
The error message is rather confusing:
UVM_FATAL @ 2122662017 ps: uvm_test_top.env.m_tl_agent_rom_ctrl_regs_reg_block.sequencer [SEQDEFPRI] Sequence m_tl_host_base_seq has illegal priority: -1
and the point is that if a sequence item has a priority of -1, this
means "look at the priority of the *sequence* as a fall-back". This
priority happens to be -1 (because the parent_sequence object didn't
have a priority in the first place, and its do_copy function didn't
copy that across either).
But it turns out that the correct fix has nothing to do with the
sequence itself. That took... a LONG time to debug.
Signed-off-by: Rupert Swarbrick <rswarbrick@lowrisc.org>
rswarbrick
requested review from
eshapira and
hcallahan-lowrisc
and removed request for
a team
alees24
approved these changes
Jan 24, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Confusingly, the default priority (-1) seems to mostly work. Unfortunately, this doesn't work properly with the rom_ctrl_stress_all_with_rand_reset sequence for some seeds for particular iterations of the stressed sequence.
In this situation, we end up with an item from a call to mirror() from the end of one iteration racing against some other register item from the start of the next one. Because there is more than once item, the sequencer tries to decide which one to grant... and explodes because they have a negative priority.
The error message is rather confusing:
and the point is that if a sequence item has a priority of -1, this means "look at the priority of the sequence as a fall-back". This priority happens to be -1 (because the parent_sequence object didn't have a priority in the first place, and its do_copy function didn't copy that across either).
But it turns out that the correct fix has nothing to do with the sequence itself. That took... a LONG time to debug.