Check context switch pointer for eth cores before resetting #17212
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
aliuTT
requested review from
abhullar-tt,
pgkeller,
tt-aho,
tt-dma,
tt-asaigal,
ubcheema,
davorchap,
cfjchu and
omilyutin-tt
as code owners
tt-dma
reviewed
Jan 28, 2025
| (tt::Cluster::instance().is_ethernet_core(virtual_core, this->id())), | ||
| "Invalid core type for context switch check"); | ||
| auto core_type_idx = hal.get_programmable_core_type_index(HalProgrammableCoreType::ACTIVE_ETH); | ||
| std::uint32_t launch_erisc_addr = tt::tt_metal::hal.get_jit_build_config(core_type_idx, 0, 0).fw_launch_addr; |
There was a problem hiding this comment.
Can we just drop a comment here describing why it's this specific address and what value we're looking for? Also, just to confirm - this flag is 0 after a reset, but upon running it at some point becomes nonzero?
There was a problem hiding this comment.
Yes I can add a comment. Flag is 0 after reset, when we launch erisc.cc app FW we write 1 to it.
abhullar-tt
reviewed
Jan 28, 2025
tt_metal/impl/device/device.cpp
Outdated
|
|
||
| // We should be able to remove the check for WORMHOLE_B0 once we assert eth cores for BH | ||
| if ((is_cooperative_eth and erisc_app_still_running(virtual_core) or | ||
| (not is_cooperative_eth and kernel_still_running(launch_msg, go_signal)))) { |
There was a problem hiding this comment.
I don't think we need to do any of this logic for BH eth cores
There was a problem hiding this comment.
I removed the logic here. Added a TODO to add BH eth cores to the assert reset in this function.
aliuTT
force-pushed
the
aliu/check-cs-flag-for-eth-reset
branch
from
ea1464a to
eb29cdd
Compare
tt-dma
approved these changes
Jan 28, 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.
Ticket
Link to Github Issue
Problem description
Today, we're assuming that tt-smi will randomize L1 for ethernet cores, so generally if we read that the launch msg isn't
RUN_MSG_GO, it should tell us that the machine has been reset and is safe to assume erisc app FW is not live. But tt-smi doesn't guarantee randomization for future systems, and reboot does not do anything to L1, so I want to move the "just reset/reboot" check to read a flag from eth L1. On WH we can use theLAUNCH_ERISC_APP_FLAG, which is always set to zero on reset.What's changed
Change
reset_coresto check the erisc app flag, which is always cleared on reset/reboot.Checklist