Repository navigation
Merge chips-it contributions to pulp cluster - #124
RiccardoGandolfi wants to merge 16 commits into
Conversation
* Fixes for QuestaOne simulation without gui * Updateed README.md with QuestaOne flow * Assiging undriven signals to 0 * Minor udpate on iDMA version --> cleaning some Xs * Added axi_burst_splitter module * Updated iDMA commit --> removed useless registers + remapped registers in pulp-runtime * Revert "Added axi_burst_splitter module" This reverts commit 670057e. * Added Burst_len parameter to iDMA + update Bender entry for iDMA * Updated cluster tb with new value for DMAs burst length parameter * Updated hci commit + updated runtime commit (fix on pi_l2_malloc) + aligned memory sizes to cluster instance in pulp-open * Updated iDMA reference --------- Co-authored-by: RiccardoGandolfi <riccardo.gandolfiu@chips.it>
…p-runtime + updated regression tests * [IDMA_CLOCK_GATING] First version of explicit clock gating in idma_wrap.sv * [IDMA_CLOCK_GATING] Fixed clock gating (works also for enqueued mode) * [IDMA_CLOCK_GATING] Updated rtl + updated entry for cluster_peripherals Chips-It fork * [IDMA_CLOCK_GATING] Clean-up on signal names + removed datapath clock gating + updated cluster_peripherals entry * [IDMA CLOCK GATING] Updated dmac_wrap.sv with split clock gating: datapath & frontend * [IDMA CLOCK GATING] Fix for hw-controlled datapath clock gating * [IDMA CLOCK GATING] Moved sw-controlled iDMA clock gating cell from wrap to cluster + clean-up on clock gating logic * [MAKEFILE] Updated commits for regression tests and pulp runtime --> updates for iDMA clock gating * [MAKEFILE][IDMA CLOCK GATING] Updated commits for regression tests and pulp runtime after updating iDMA with clock gating features
…e in pulp repository.
…d to 0 from idma wrap + update Makefile rule to generate iDMA rtl
05d26c8 to
9e64761
Compare
9e64761 to
c8e8721
Compare
FrancescoConti
left a comment
There was a problem hiding this comment.
A couple of minor comments to solve but I will none the less put approve
| ## Clone pulp-runtime as SW stack | ||
| PULP_RUNTIME_REMOTE ?= https://github.com/pulp-platform/pulp-runtime.git | ||
| PULP_RUNTIME_COMMIT ?= 3b48b0c6872cc01ba169a2c9c886ebb815d22cdc | ||
| PULP_RUNTIME_REMOTE ?= https://github.com/FondazioneChipsIT/pulp-runtime.git |
There was a problem hiding this comment.
is it willing? shouldn't the pulp_runtime contrib also be merged?
There was a problem hiding this comment.
Good point, the fact is that we're dropping the pulp-runtime support in favor of the pulp-sdk (which now supports the standalone pulp cluster as well on this PR: pulp-platform/pulp-sdk#171).
So actually I'd be for removing the references to pulp-runtime, which anyway is no longer being supported by us.
| cd $(shell bender path idma) && \ | ||
| curl -LsSf https://astral.sh/uv/install.sh | sh && \ | ||
| uv sync --locked | ||
| idma_venv: |
There was a problem hiding this comment.
why rolling back from uv to venv? Feels like a regression
There was a problem hiding this comment.
For this I think the reason is that we need to bump iDMA to the latest release first. But I think that should be done in a separate PR.
|
Superseded by #125 |
No description provided.